Social Norms Marketing Social Networks and Alcohol Consumption

Dublin Institute of Technology
ARROW@DIT
Doctoral
Business
2013-05-01
Social Norms Marketing, Social Networks and

Alcohol Consumption: A Collegiate Context.
Investigating Feasability in Ireland.
Sarah Samdani
Follow this and additional works at: http://arrow.dit.ie/busdoc

Part of the Marketing Commons
Recommended Citation
Samdani, S. Social Norms Marketing, Social Networks and Alcohol Consumption: A Collegiate Context. Investigating Feasability in
Ireland. Submitted in part fulfilment of the requirements of Dublin Institute of Rechnology for the award of Doctor of Philosophy.
This Theses, Ph.D is brought to you for free and open access by the
Business at ARROW@DIT. It has been accepted for inclusion in Doctoral
by an authorized administrator of ARROW@DIT. For more information,
please contact yvonne.desmond@dit.ie, arrow.admin@dit.ie.
This work is licensed under a Creative Commons AttributionNoncommercial-Share Alike 3.0 License
Social Norms Marketing, Social Networks and

Alcohol Consumption: A Collegiate Context
Investigating Feasibility in Ireland
Submitted in partial fulfilment of the requirements of Dublin Institute of Technology for the
award of Doctor of Philosophy
By
Engr. Sarah Samdani (MSc.)

School of Marketing, Faculty of Business, Dublin Institute of Technology
11th May, 2013
_______________________________________________________________________
Supervisors:
Dr. Eddie Rohan 1
Mr. Patrick Kenny 1
Dr. Elisa Bellotti 2
1
School of Marketing, Faculty of Business

Dublin Institute of Technology, Ireland
Sustainable Consumption Institute and Mitchell Centre for Social Network Analysis
University of Manchester, United Kingdom
Declaration
Declaration
I certify that this thesis, which I now submit for examination for the award of PhD degree, is
entirely my own work and has not been taken from the work of others, save and to the extent
that, such work has been cited and acknowledged within the text of my work.
This thesis was prepared according to the regulations for postgraduate study by research of the
Dublin Institute of Technology and has not been submitted in whole or in part for another
award in any Institute.
The work reported on in this thesis conforms to the principles and requirements of the
Institute's guidelines for ethics in research.
The Institute has permission to keep, lend or copy this thesis in whole or in part, on condition
that any such use of the material of the thesis be duly acknowledged.
Signature ____________________________
Candidate
Date _______________
Abstract
Abstract
The current Irish policies have not been adequately effective in reducing alcohol consumption.
There is a need to consider alternative strategies, such as the increasingly popular SN
marketing campaigns, which have been applied successfully in the US college system.
However, the potential of these campaigns has not been evaluated in Ireland. It is also not clear
from the literature if descriptive or injunctive norm types will be more likely to induce
behaviour change. Further, while SN interventions tend to provide friends or typical student
as referent groups, little is understood about how individuals visualize these groups and how
salient these peers are. The present study addressed these issues by combining web based
survey methods with social network analysis. Hierarchical multiple regression was used to
analyse a web survey of 1700 DIT students. Further, 26 ego networks generated via in depth
interviews were examined using network techniques combined with a qualitative analysis to
understand norm salience. The study provides evidence of overestimations of the campus
drinking norm at DIT. It shows that perceived norms impact personal consumption and that
social distance is a key consideration in this regards. Further, the findings demonstrate that
descriptive norms are stronger predictors of personal consumption than injunctive norms. Most
importantly, the study provides evidence that individuals social networks are key determinants
of their drinking behaviours and that the most salient peers for DIT students are embedded in
cohesive sub groups outside college. The study does not support using SN campaigns to reduce
alcohol consumption in DIT. It urges policy makers to address norm salience in intervention
work as it is critical for the applicability, planning and success of SN campaigns.
Keywords: Social norms theory, college drinking, social network analysis, ego networks.
ii
DEDICATION
This is to
My parents
Dr. G.M Samdani and Tahira Samdani
(I am because you are)
And my husband
Dr. Salman Ali Mumtaz
(For none of this would have been possible without you)
iii
Acknowledgements
Acknowledgements
First and foremost, I extend my deepest gratitude to the Almighty whose kindness to me knows
no bounds. Thank You for always being right next to me in everything that I do in life.
I am also sincerely grateful to
Mr. Patrick Kenny, who has supervised me the longest. Thank you for sharing your immense
knowledge on the subject with me, for providing timely advice on every aspect of this
research, for helping me formulate the design of this study, for giving the most thorough
feedback on my draft chapters and for providing the motivation one requires to reach the finish
line.
Dr. Edmund Rohan, also my supervisor and who I have known since the day I joined DIT as an
MSc student. Thank you for your continued interest in this study, for your guidance on many
methodological aspects, for your most useful insights and reflections on my work, for your
time and commitment to this study above and beyond your official obligations and for always
being extremely encouraging.
Dr. Elisa Bellotti, my supervisor at the University of Manchester. Thank you for providing
supervision to this project despite your own commitments, for the countless discussions on
networks, for advising on the design of in depth interviews, for your feedback on my draft
chapters and for your time and guidance on the many occasions that I visited you in
Manchester.
Thank you all 3, for frequently going above and beyond your duties to provide me with
excellent supervision. I could not have asked for a better supervisory team.
Dr. John McAlaney, for providing guidance and advice from his own experiences to help me
finalize my research questionnaire, Dr. Rich Dejordy, Dr. Tom Snidjers and Dr. Alan Reifman
for providing the initial directions and guidance in social network analysis and Dr. Jeremy
Michaels for helping me make sense of regression.
iv
Acknowledgements
Ms. Melda Slattery at the DIT Public Affairs Office, Ms. Rachel O Connor at the DIT
Campus Life Office and Ms. Raffaella Salvante at the DIT Graduate Research School Office
for providing assistance in launching the web survey and making it available to the entire
student body at DIT. Ms. Gillian Dann and Ms. Paula Maguire at the marketing office in the
Faculty of Business at DIT for providing assistance with room allotments for the in depth
interviews, sometimes at a very brief notice. All students who participated in this study.
Salman, my husband who believed before there was anything to believe in. Thank you for
being the most consistent source of motivation throughout this study, for proof reading my
work and for spending days in formatting this thesis. Thank you for the million times youve
reminded me why I want to do this and the million others when you took the reins and made it
possible for me to concentrate on the task at hand.
Dr. G.M. Samdani, my father for always placing his pride in me. If there was anything that
convinced me that I could do this, it was the comfort that I had your genes. I am what I am
because I am your child.
The two mothers in my life; my mom for always supporting me unconditionally and for
minding my baby for several months despite several unavoidable commitments, and my
mother in law, for taking care of my home, my family and me while I spent weeks after weeks
writing up this thesis. I cannot thank the two of you enough neither can I ever match the
generosity and love you extend to me.
Hina Samdani, my sister in law, who has been extremely motivating and supportive. Thank
you for providing me with the push I needed and for being my partner in crime on several
occasions when I needed to escape from the burden of research and just have FUN. Sumaira
Naz, also my sister in law for always being there when I needed her the most and for taking
care of the most important details in my life while I worked.
Shanza Baig, a friend I am glad I found and Zaffaryab Ahmed, a friend who I know will last a
lifetime. Thank you both for proof reading my work at a very short notice despite your own
professional commitments.
Finally, the many friends who made my time in Ireland a memorable one. Thank you all!
v
Abbreviations
Abbreviations
AAI
Alcohol Action Ireland
ABFI
Alcohol Beverage Foundation of Ireland
AMCMB
Alcohol Marketing Communications Monitoring Body
ASAI
Advertising Standards Authority for Ireland
BAC
Blood Alcohol Content
BAI
Broadcasting Authority of Ireland
BCC
Broadcasting Complaints Commission
BCI
Broadcasting Commission of Ireland
CCCI
Central Copy Clearance Ireland
DIGI
Drinks Industry Group of Ireland
DIT
DoHC
Department of Health and Children
ESPAD
European School Survey Project on Alcohol and Other Drugs
GAAG
Government Alcohol Advisory Group
HSB
Health Service Board
HSE
Health Service Executive
vi
Abbreviations
ITA
Intoxicating Liquor Act
MEAS
Mature Enjoyment of Alcohol in Society
NDTRS
National Drug Treatment Reporting System
NDTRS
National Drugs Treatment Reporting System
OECD
Organization for Economic Cooperation Development
OTC
Office of Tobacco Control
PRSI
Pay Related Social Insurance
PULSE
Police Using Leading Systems Effectively
RAND Corporation
Research and Development Corporation
RRAI
Responsible Retailing of Alcohol in Ireland
RSA Training
Responsible Serving of Alcohol Training
RSA
Road Safety Authority
SN
Social Norms
SNA
Social Network Analysis
STFA
Strategic Task Force on Alcohol
UCD
University College Dublin
WHO
World Health Organization
vii
Table of Content
Table of Content
Declaration ...................................................................................................................................i
Abstract ...................................................................................................................................... ii
Acknowledgements ....................................................................................................................iv
Abbreviations .............................................................................................................................vi
Table of Content ..................................................................................................................... viii
List of Figures ..........................................................................................................................xvi
List of Tables .......................................................................................................................... xvii
1
Introduction ...................................................................................................................... 19
1.1
The Social Norms (SN) Theory and SN Marketing Interventions .............................. 19
1.2
Scope and Limitations of the Study ............................................................................ 20
1.2.1
Scope ...................................................................................................................... 20
1.2.2
Limitations .............................................................................................................. 22
1.3
2
Thesis Plan ................................................................................................................. 22
Research Context .............................................................................................................. 25

2.1
Introduction ................................................................................................................ 25
2.2
Alcohol Consumption in Ireland ................................................................................ 25
2.3
Alcohol Related Health and Social Consequences in Ireland .................................... 26
2.4
Irish Policy Interventions ........................................................................................... 28
2.5
Conclusion .................................................................................................................. 29
Literature Review: The Social Norms Approach .......................................................... 30
viii
Table of Content
3.1
Introduction ................................................................................................................ 30
3.2
What are Norms? ........................................................................................................ 31
3.2.1
Descriptive Norms .................................................................................................. 31
3.2.2
Injunctive Norms .................................................................................................... 32
3.2.3
Distinction between Injunctive Norms and Subjective Norms .............................. 33
3.2.4
Characteristics of Norms ........................................................................................ 34
3.3
Theory of Social Ecology and Importance of Norms ................................................. 35
3.4
Social Influence Research .......................................................................................... 36
3.4.1
Direct Peer Influence .............................................................................................. 37
3.4.2
Indirect Peer Influence ........................................................................................... 38
3.4.2.1
Modelling ....................................................................................................... 39
3.4.2.2
Normative Beliefs ........................................................................................... 40
3.5
Students and Susceptibility of Social Norms .............................................................. 42
3.6
Social Norms (SN) Theory .......................................................................................... 43
3.6.1
Causes and Types of Misperceptions ..................................................................... 44
3.6.1.1
Pluralistic Ignorance (I think they differ, when they dont) ........................... 45
3.6.1.2
False Consensus (I think they are as bad as me, so I dont really have a
problem) 45
3.6.1.3
False Uniqueness (Not many are like me and abstain or drink less) .............. 46
3.6.2
Documentation of Misperceptions ......................................................................... 46
3.6.3
Consequences of Misperceptions ........................................................................... 48
3.7
SN Prevention Strategies for Reducing Alcohol Related Harm ................................. 50
3.7.1
Social Norms Marketing Interventions Outside the US ......................................... 54
3.7.2
Expansion beyond Academics ................................................................................ 58
3.7.3
Unsuccessful Interventions ..................................................................................... 58
3.7.4
Issues in the Evaluation of SN Interventions ......................................................... 59
3.8
3.8.1
Criticisms of SN Marketing Interventions .................................................................. 61

Not Publicizing the Problem .................................................................................. 61
ix
Table of Content
3.8.2
Overstating Misperceptions .................................................................................... 62
3.8.3
Offering Encouragement to Abstainers .................................................................. 62
3.8.4
Feasibility in Situations when the Problem Behaviour is the Norm ...................... 63
3.9
3.9.1
Applicability of SN Marketing Interventions in Ireland ........................................ 63
3.9.2
The Conceptualization of Norms............................................................................ 64
3.9.3
Norm Salience ........................................................................................................ 66
3.10
4
Theoretical Gap and Social Need............................................................................... 63
Conclusion .................................................................................................................. 68
Literature Review: Social Network Analysis (SNA) ..................................................... 70

4.1
Introduction ................................................................................................................ 70
4.2
Network Theory .......................................................................................................... 71
4.2.1
What is Social Network Analysis (SNA)? ............................................................. 71
4.2.2
Types of Ties .......................................................................................................... 72
4.2.3
History and the Current State of SNA .................................................................... 73
4.2.4
The Guiding Principles of SNA ............................................................................. 76
4.3
Network Data and Measurement ................................................................................ 79
4.4
Integration of SNA and Norm Salience ...................................................................... 81
4.4.1
Network Visualization and Norm Salience ............................................................ 82
4.4.2
Network Composition and Norm Salience ............................................................. 83
4.4.3
Network Structure and Norm Salience ................................................................... 86
4.5
SNA and Substance Use Research .............................................................................. 92
4.5.1
Birds of a Feather Flock Together .......................................................................... 93
4.5.2
Theories of Social Learning, Differential Association and Reasoned Action........ 94
4.5.3
Influence Vs Selection ............................................................................................ 95
4.5.4
Prevention Interventions using SNA ...................................................................... 99
4.6
Challenges in SNA .................................................................................................... 101
4.6.1
Boundary Specification ........................................................................................ 102
4.6.2
Respondent Burden .............................................................................................. 103

x
Table of Content
4.6.3
4.7
Criticisms of SNA ..................................................................................................... 105
4.7.1
Lacks Theory ........................................................................................................ 105
4.7.2
Lacks Agency ....................................................................................................... 106
4.7.3
Ignores Dynamics of Relationships ...................................................................... 107
4.8
5
Causality ............................................................................................................... 104
Conclusion ................................................................................................................ 107
Methodology .................................................................................................................... 109

5.1
Introduction .............................................................................................................. 109
5.2
Research Objectives ................................................................................................. 109
5.3
Research Philosophy ................................................................................................ 114
5.4
Research Design ....................................................................................................... 117
5.5
Wave 1: The Web Survey .......................................................................................... 121
5.5.1
Questionnaire Design ........................................................................................... 121
5.5.2
Cognitive Interviewing ......................................................................................... 122
5.5.3
The Pilot Study ..................................................................................................... 124
5.5.4
Sampling for the Web Survey .............................................................................. 126
5.5.5
Evaluation of Online Survey Tools ...................................................................... 127
5.5.6
Measures ............................................................................................................... 129
5.5.6.1
Control Variables.......................................................................................... 129
5.5.6.2
Personal Consumption .................................................................................. 134
5.5.6.3
Perceptions of Prevalence (The Descriptive Norms) ................................... 138
5.5.6.4
Perceptions of Approval (The Injunctive Norms) ........................................ 139
5.5.7
5.6
Reliability and Validity ........................................................................................ 140

Wave 2: In Depth Interviews .................................................................................... 144
5.6.1
Sampling and Related Issues ................................................................................ 144
5.6.1.1
Campus Structure at DIT .............................................................................. 144
5.6.1.2
Recruitment of Subjects ............................................................................... 146
5.6.1.3
Sample Size .................................................................................................. 149

xi
Table of Content
5.6.2
The Ego Network Approach ................................................................................. 149
5.6.3
Interview Design .................................................................................................. 150
5.6.3.1
Settings and Protocol .................................................................................... 150
5.6.3.2
Name Generators .......................................................................................... 151
5.6.3.3
Name Interpreters ......................................................................................... 154
5.6.4
Interview Structure ............................................................................................... 155
5.6.4.1
5.6.5
The Main Interviews............................................................................................. 157
5.6.5.1
Examining Interrelationships among Alters ................................................. 160
5.6.5.2
Tie Strength .................................................................................................. 162
5.7
Pilot Interviews ............................................................................................. 156
Ethics of Research .................................................................................................... 165
5.7.1
Informed Consent ................................................................................................. 165
5.7.2
Privacy and Confidentiality .................................................................................. 165
5.7.3
Anonymity ............................................................................................................ 166
5.7.4
DIT Ethics Committee Approval.......................................................................... 167
5.8
Data Transformation and Analysis .......................................................................... 168
5.9
Limitations of Methodological Approach ................................................................. 169
5.9.1
Specific Issues Related to the Web Survey .......................................................... 169
5.9.2
Specific Issues Related to Ego Network Analysis and In-depth Interviews ........ 172
Analysis and Results of Survey Data ............................................................................ 177

6.1
Data Cleaning .......................................................................................................... 177
6.2
Descriptive Statistics ................................................................................................ 178
6.2.1
Personal Alcohol Consumption ............................................................................ 179
6.2.2
Descriptive Norms (DN) Vs. Personal Consumption........................................... 180
6.2.3
Injunctive Norms vs. Personal Consumption ....................................................... 186
6.3
Correlations .............................................................................................................. 186
6.4
Hierarchical Multiple Regression ............................................................................ 190
6.4.1
Data set considerations ......................................................................................... 191

xii
Table of Content
6.4.2
Assumption Testing .............................................................................................. 192
6.4.3
First set of regressions .......................................................................................... 192
6.4.3.1
Model 1 ......................................................................................................... 194
6.4.3.2
Model 2 ......................................................................................................... 196
6.4.4
6.4.4.1
Model 1 ......................................................................................................... 201
6.4.4.2
Model 2 ......................................................................................................... 203
6.4.5
Third set of regressions ........................................................................................ 205
6.4.5.1
Model 1:........................................................................................................ 207
6.4.5.2
Model 2:........................................................................................................ 209
6.5
7
Second set of regressions...................................................................................... 199
Conclusion ................................................................................................................ 211
Network Analysis and Results: Second Wave.............................................................. 215

7.1
Introduction .............................................................................................................. 215
7.2
Sample Overview ...................................................................................................... 216
7.3
Network Analysis ...................................................................................................... 217
7.4
Overview of Network Composition ........................................................................... 219
7.4.1
Size and Effective Size of Ego Networks ............................................................. 219
7.4.2
Referent Group Composition of Ego Networks ................................................... 223
7.4.3
Gender and Age Composition of the Ego Networks ............................................ 226
7.5
Evaluation of Tie Strength ........................................................................................ 227
7.6
Variations in Network Structures ............................................................................. 228
7.6.1
The Small Cliques ................................................................................................ 230
7.6.2
The Overlapping Cliques ...................................................................................... 235
7.6.3
The Group ............................................................................................................. 246
7.6.4
The Core/Periphery .............................................................................................. 253
7.6.5
The Contextualized Network ................................................................................ 263
7.7
Conclusion ................................................................................................................ 268
xiii
Table of Content
Personal Networks, Normative Perceptions and Drinking Behaviours .................... 271

8.1
Introduction .............................................................................................................. 271
8.2
Linking the web survey and interview data .............................................................. 272
8.3
Development and Reinforcement of Drinking Behaviours in Networks ................... 279
8.3.1
The Overlapping Cliques ...................................................................................... 279
8.3.1.1
Sams network .............................................................................................. 279
8.3.1.2
Megs Network ............................................................................................. 285
8.3.1.3
Sues Network .............................................................................................. 291
8.3.1.4
Peters network ............................................................................................. 299
8.3.2
The Group .......................................................................................................... 304
8.3.2.1
Lisas Network ............................................................................................. 304
8.3.2.2
Kens network .............................................................................................. 308
8.3.3
Core/Periphery ...................................................................................................... 313
8.3.3.1
Garys network ............................................................................................. 313
8.3.3.2
Lindas Network ........................................................................................... 319
8.3.4
Contextualized Network ....................................................................................... 328
8.3.4.1
Ruths network ............................................................................................. 328
8.4
Cross Case Analysis ................................................................................................. 331
8.5
Linking the Outcomes of Web Survey and SNA ........................................................ 337
Discussion of Results ...................................................................................................... 342

9.1
Introduction .............................................................................................................. 342
9.2
Summary of Evidence Against Research Objectives ................................................ 342
9.2.1
Research Objectives 1 and 2................................................................................. 342
9.2.2
Research Objective 3 ............................................................................................ 344
9.2.3
Research Objectives 4 and 5................................................................................. 345
9.3
Implications and Theoretical Contributions ............................................................ 347
9.3.1
First Evidence of Alcohol Related Misperceptions in Ireland ............................. 347
9.3.2
Integration of Norm Salience and SNA................................................................ 348

xiv
Table of Content
9.3.3
Descriptive norms vs. Injunctive norms ............................................................... 353
9.4
Other Findings and Their Contributions.................................................................. 354
9.5
Limitations ................................................................................................................ 355
9.6
Recommendations for Future Research ................................................................... 358
9.7
Conclusions .............................................................................................................. 364
References ............................................................................................................................... 365

Appendices...399
xv
List of Figures
List of Figures
Figure 1: Alcohol consumption per Adult in Litres of Pure Alcohol. (Hope, 2007; Foley, 2012) ...........................26
Figure 2: Typology of ties studied in social networks (Borgatti et al., 2009) ..........................................................73
Figure 3: Network visualization showing various positions.....................................................................................90
Figure 4: Location of DIT Campuses around Dublin City .....................................................................................145
Figure 5: Example Adjacency Matrix ....................................................................................................................161
Figure 6: Concentric Circles Diagram ....................................................................................................................164
Figure 7: Frequency of drinking in a typical month Vs the perceived norm ..........................................................183
Figure 8: No. of drinks on a typical occasion Vs the perceived norm ....................................................................184
Figure 9: Frequency of drunkenness in a typical month Vs the perceived norm....................................................185
Figure 10: Example of UCINET matrix .................................................................................................................218
Figure 11: Effective size in a 3 mode network .......................................................................................................220
Figure 12: Sue's important discussants' network: A small clique ...........................................................................231
Figure 13: Sams overall network: An overlapping clique .....................................................................................236
Figure 14: Megs overall network: An overlapping clique ....................................................................................243
Figure 15: Lisas overall network: A group .........................................................................................................248
Figure 16: Enlarged Image of the 'group' in Lisa's network ...................................................................................251
Figure 17: Garys overall network: A core/periphery ............................................................................................254
Figure 18: Lindas overall network: A core/periphery ...........................................................................................259
Figure 19: Ruths overall network: A contextualized network...............................................................................265
Figure 20: Sam's overall network ...........................................................................................................................280
Figure 21: Meg's overall network ...........................................................................................................................286
Figure 22: Sue's overall network ............................................................................................................................292
Figure 23: Peter's overall network ..........................................................................................................................300
Figure 24: Lisa's overall network ...........................................................................................................................305
Figure 25: Ken's overall network ...........................................................................................................................309
Figure 26: Gary's overall network ..........................................................................................................................314
Figure 27: Linda's overall network .........................................................................................................................320
Figure 28: Ruth's overall network ..........................................................................................................................328
xvi
List of Tables
List of Tables
Table 1: Colleges at DIT ........................................................................................................................................145
Table 2: Mean alcohol consumption across DIT and individual campuses............................................................147
Table 3: Drinking Intensity Criteria (Beck and Treiman, 1996) ............................................................................147
Table 4: Drinking Intensity Criteria for the Current Study ....................................................................................148
Table 5: Standard descriptive Statistics ..................................................................................................................178
Table 6: Sample distribution across campuses in percentage .................................................................................178
Table 7: Personal alcohol consumption ..................................................................................................................179
Table 8: Means and St Deviations of personal consumption vs. the perceived descriptive norm ..........................180
Table 9: Means of perceived descriptive norm for abstainers ................................................................................181
Table 10: Pearson's correlation between personal consumption and perceived descriptive norms .......................187
Table 11: Pearson's correlation between personal consumption and perceived injunctive norms .........................189
Table 12: Single set regressions for frequency of drinking in a typical month ......................................................193
Table 13: Change in R2 for model 1 (first set of regressions) .................................................................................195
Table 14: Unstandardized regression coefficients for model 1 (first set of regressions) ........................................195
Table 15: Change in R2 for model 2 (first set of regressions) .................................................................................197
Table 16: Unstandardized regression coefficients for model 2 (first set of regressions) ........................................198
Table 17: Single set regressions for number of drinks on a typical occasion, significant at p<0.05 ......................200
Table 18: Change in R2 for model 1 (Second set of regressions) ...........................................................................201
Table 19: Unstandardized regression coefficients for model 1 ..............................................................................202
Table 20: Change in R2 for model 2 (Second set of regressions) ...........................................................................204
Table 21: Unstandardized regression coefficients for model 2 (Second set of regressions) ..................................204
Table 22: Single set regressions for frequency of drunkenness in a typical month, significant at p<0.05 .............206
Table 23: Change in R2 for model 1 (Third set of regressions) ..............................................................................207
Table 24: Unstandardized regression coefficients for model 1 ..............................................................................208
Table 25: Change in R2 for model 2 (Third set of regressions) ..............................................................................210
Table 26: Unstandardized regression coefficients for model 2 (Third set of regressions) .....................................210
Table 27: Drinking Intensity Criteria for the Current Study ..................................................................................217
Table 28: Size and effective size of the examined ego networks ...........................................................................221
Table 29: Peer group composition (%) based on overall network size...................................................................224
Table 30: Number of people named by gender. .....................................................................................................226
Table 31: Summary of the structures found ...........................................................................................................229
Table 32: Tie strength scores for Sams network ...................................................................................................236
Table 33: Tie strength scores for Meg's network ...................................................................................................242
Table 34: Tie strength scores for Lisas network ...................................................................................................248
xvii
List of Tables
Table 35: Tie strength scores for Garys network ..................................................................................................254

Table 36: Tie strength scores for Linda's network .................................................................................................258
Table 37: Tie strength scores for Ruths network ..................................................................................................264
Table 38: Association between perceived norms and personal networks...............................................................275
Table 39: Association between individual drinking behaviour and personal networks..........................................277
Table 40: Tie strength scores for Peter's network...................................................................................................299
Table 41: Tie strength scores for Ken's network ....................................................................................................309
xviii
Introduction
Introduction
This chapter presents a preface to the thesis. It moves from describing the theoretical
framework to outlining the scope and limitations of this study before culminating in a roadmap
of the thesis.
1.1
The Social Norms (SN) Theory and SN Marketing Interventions
The SN theory focuses on peer influence and the role it plays in individual decision making
around behaviours such as drinking. It argues that our behaviour is influenced by
misperceptions of how other members of our social groups think and act. With regards to
drinking, students often perceive their peers to have more permissive attitudes and behaviours
towards alcohol than they themselves. In efforts to match these exaggerated perceptions, their
own consumption levels increase. Accordingly, the theory also states that correcting
misperceptions of perceived norms can result in a decrease in problem behaviour or an increase
in the desired behaviour.
The specific application of SN theory to college drinking as a prevention strategy was first
recommended by Perkins and Berkowitz (1986). Since then, the use of SN based marketing
campaigns has gained immense popularity and prominence in the US where it has been found
to be successful in reducing drinking rates in college students. These interventions seek to
correct misperceptions of peer drinking norms by providing normative feedback about the
actual drinking rate of an average or typical student on campus (Perkins and Craig, 2003a).
This feedback is often provided by utilizing social marketing techniques and hence the name
SN marketing came forth. The potential of these campaigns in Ireland is yet to be evaluated.
19
Introduction
1.2
Scope and Limitations of the Study
1.2.1 Scope
Coverage of Issues: The study draws on three bodies of literature namely, the SN theory, SN
marketing interventions and social network analysis (SNA). It tests the SN theory in Ireland to
assess the potential of SN marketing campaigns as a strategy to reduce drinking rates in Irish
colleges. It also focuses on deepening our understanding of two key theoretical issues in the
SN literature. The first issue is related to examining the relative impact of descriptive and
injunctive norm types on individuals drinking behaviours. The second issue concerns the need
to understand how norm salience influences drinking behaviour and how it may impact the
effectiveness of a SN based intervention. The study utilized SNA as a methodology to examine
and understand norm salience. In doing so it examined the compositional and structural aspects
of social networks surrounding the participants and the subjective meanings associated with
these relationships.
Research settings and population: The study took Dublin Institute of Technology (DIT) as its
research setting. The population for this study comprises full time Irish undergraduate students
enrolled at DIT during the academic year 2010-11.
Methodological Approach: The study was based around five research objectives which
emerged from the review of literature presented in chapters 3 and 4. These are stated formally
in chapter 5. The data was collected and analysed in two waves.
Establishing misperceptions of peer drinking norms and determining their impact on the
drinking behaviour of population of interest is a key assumption of the SN theory and a
20
Introduction
necessary prerequisite of any SN marketing intervention. In the first wave of data collection, a
campus wide web survey was conducted which primarily assessed the drinking behaviours of
respondents and their perceptions of the behaviour among most students in DIT. In addition,
the survey also assessed the perceived norms for close friends, best friend, mother and father.
Further, the perceptions of two types of norms namely descriptive and injunctive were
examined in the survey. The resulting data comprising responses from a sample of 1700
students was imported into SPSS and hierarchical multiple regression was used for analysis.
Central to the effectiveness of any SN marketing campaign is targeting the most salient norms.
It is a complex issue because it requires knowledge of the individuals environments, their
social networks and the importance they attach to these networks. SNA provides a way to
address this issue. It is a growing field which encompasses theories, models, and applications
to study the relationships between a set of actors (Marsden, 2005). The second wave comprised
unstructured in depth interviews with 26 individuals who were selected from the survey sample
based on a pre defined criteria. These interviews focused on extracting the networks
comprising people with whom the participants discussed important matters and with whom
they liked to socialize or party. In addition, these interviews were based around exploring the
relationships participants shared with their network members, the subjective meanings they
attached with these associations and the development and reinforcement of drinking
behaviours and attitudes in the networks. The in depth interviews generated two types of data.
The network data, which was imported in UCINET software (Borgatti et al., 2002) for the
application of network techniques and the verbatim text of the interviews generated from the
audio recordings which was subjected to a qualitative analysis. The study linked the outcomes
of the two data collection methods which are discussed in chapter 8.
21
Introduction
1.2.2 Limitations
The study did not cover undergraduate student populations in other third level institutes in
Ireland. The quantitative findings therefore may only be generalized to DIT students. The
network results are based on a non representative sample but may be naturalistically and
tentatively generalized to the extent of being a phenomenon of commuter colleges as will be
explained more fully in chapter 5 and chapter 9.
This study was cross sectional in design which is why causal inferences cannot be made with
certainty. The findings presented herein are based on the assumption that normative
perceptions cause personal consumption. This assumption was made on the basis of theoretical
knowledge and past longitudinal research (Fearnow-Kenny et al., 2001).
As is mostly the case with alcohol related survey research, the findings of this study are based
on self reports of personal alcohol use.
Networks examined in this study were based on how individuals viewed their networks rather
than collecting network data from each network member. This is an inherent feature of ego
network analysis. Finally, the network typologies identified in this study are not exhaustive of
all possible typologies.
1.3
Thesis Plan
A chapter by chapter outline of this thesis is presented next.

Chapter 2 provides context to this study. It looks at the current consumption trends in Ireland,
draws attention to the excess of heavy drinking in Ireland and the harms caused by it. The
22
Introduction
discussion highlights that the current policy has not been hugely effective in producing positive
results and that there is a need to consider alternative strategies.
Chapter 3 reviews the research related to SN approach. It moves from describing the
theoretical aspects of the SN approach to examining it as an intervention strategy and
describing its postulates, assumptions, criticisms and limitations. It identifies the theoretical
gap in literature which warrants further investigation.
Chapter 4 introduces SNA as a perspective, summarizes its history and guiding principles and
examines how it can be suitable in addressing norm salience. It also reviews network research
conducted in the domain of substance use, particularly alcohol consumption. It concludes with
a discussion on the limitations of SNA and criticisms raised against it.
Chapter 5 describes the methodology of this research in detail. It formally presents the research
objectives and hypothesis of this study, their origin, rationale and the research philosophy. It
then describes various aspects of research design including the choice of data collection
instruments and specific issues related to them, the development, pre testing, administration
and reliability/validity of these instruments, the sampling strategy for both waves of data
collection, the recruitment of subjects, the ethical concerns in this study and the steps taken to
address them. It concludes with a discussion on the limitations of the underlying
methodological approach.
Chapter 6 provides the descriptive statistics and present the results of hierarchical multiple
regression. In parallel, it also compares the results with past research. The chapter concludes
with a summary of the main findings
23
Introduction
Chapter 7 presents a compositional and structural analysis of the networks combined with a
qualitative interpretation of the interviews. An assessment of tie strength and the subjective
meanings associated with these relationships is presented in parallel. The chapter concludes
with a summary of the main findings.
Chapter 8 integrates the survey and interview data. It explores the relationship of individuals
drinking behaviours and perceived norms of their proximal peers as reported in the web survey
with the perceived norms of their network members. In doing so, the chapter investigates how
drink related norms were formed, reinforced and transmitted through the networks. A cross
case analysis provides an overall summary of the findings. The results are based on a
qualitative interpretation of the interviews and guided by the sense of the participants
descriptions of their networks. The chapter also links the outcomes of the two data collection
methods.
Finally, chapter 9 revisits the research objectives and addresses them one by one in light of the
evidence. It also discusses the theoretical contributions of this study and their practical
implications. It then describes the limitations of this study and concludes with some useful
recommendations for future research.
24
Research Context
Research Context
2.1
Introduction
This chapter provides context to the study by showing that excessive drinking is a serious
concern in Ireland having led to several social and health consequences. It draws attention to
the inadequacy of Irish policies in dealing with this issue and highlights the need to consider
alternative strategies such as the SN approach, which has been successful in other countries.
Additional details about alcohol consumption trends in Ireland, related social and health harm
and an outline of alcohol policy activity in Ireland from 1990-2012 is presented in appendix 1.
Chapter 3 extends the discussion of SN approach and reviews related literature
2.2
Alcohol Consumption in Ireland
Ireland is one of the heaviest consumers of alcohol in the enlarged EU and worldwide with the
Irish people drinking about 20% more than the average European (AAI, 2011). The trend for
alcohol consumption (per adult) for the years 1987- 2011 is presented in Figure 1. The
consumption of alcohol in Ireland increased by 46% between 1987 and 2001 when it hit a
record high of 14.3 litres of pure alcohol per adult. There was a decline of 6% in 2003 for the
first time in sixteen years attributed mainly to a large drop in spirit sales following an increase
in excise duty (DoHC, 2004). Another decline in consumption was observed in 2008-09,
followed by an increase in 2010. This was partially influenced by changing trends in cross
border shopping (Hope, 2007). Since, 19% of the adult population abstain from alcohol
25
Research Context
completely, those who do drink consume much more than the average statistics indicate,
making the effective drinking ra
rates even higher (DoH, 2012).
Figure 1:: Alcohol consumption per Adult in Litres of Pure Alcohol. (Hope, 2007; Foley,
2012)
About a quarter of the Irish population report binge drinking every week (AAI, 2011).
Underage drinking is a serious problem constituting an illegal market of at least 145m (OTC,
2006). Irish teenagers aged 15
15-16 years have been shown to binge
inge drink and abuse alcohol
more than their European counterparts (Hibell et al., 2012).
2.3
Alcohol Related Health and Social Consequences in Ireland
Every seven hours, someone in Ireland dies from an alcohol

alcohol-related
related illness (AAI, 2011).
Between 1992-2002,
2002, more than 14000 people died in Ireland from alcohol related conditions
and suicide (DoHC, 2004)
2004). An increase of 92% was recorded in alcohol related hospital
26
Research Context
discharges between 1995-2002 (Mongan, 2007). A total of 42,333 cases were presented with
alcohol as the main problem substance during 2005-2010 accounting for over half of all cases
treated for problem substance use during this period (Carew et al., 2011). Alcohol contributed
to 36.5% of all fatal road accidents in Ireland in 2003 (Mongan et al., 2009).
At least 271,000 children in Ireland are being exposed to risk as a result of parental hazardous
drinking (Hope, 2011). Divorce rates are twice as high in marriages suffering from issues of
alcohol use compared to those without such problems (Mongan et al., 2009). A study
examining domestic violence reported that 11% of the population experienced severe abuse
and in one quarter of these cases, alcohol was always involved (Watson and Parsons, 2005).
In a study of over 3000 Irish adults, alcohol was found to be involved in almost half of the
cases (53% of men and 45% of women) of sexual abuse that occurred in adulthood (McGee et
al., 2002). Assaults increased by 20% and disorderly conduct by 23% between 2004-2010 and
half of these were alcohol related (AAI, 2011). Analysis of Garda PULSE1 data for the years
2003-07 found that the total number of drunkenness, public order and assault offences2
increased by 30% from 50,948 to 66,406. The 18-24 year old age group accounted for two fifth
offences (Mongan et al., 2009).
Problem alcohol use cost an estimated 3.7 billion to the Irish society in 2007 (Byrne, 2010).
According to the Chief Medical Officer of Ireland, these figures suggest that a 30% reduction
in alcohol-related harm would save taxpayers an estimated 1 billion a year.
The Gardai PULSE (Police Using Leading Systems Effectively) system launched in 1999 records data on crime,
traffic management, progression of criminal cases through the courts, fire arms licensing and drivers licenses
2
Although public order and assault offences are not necessarily alcohol-related, there is sufficient international
evidence suggesting a strong relationship between alcohol consumption and such offences. (Mongan et al, 2009)
27
Research Context
2.4
Irish Policy Interventions
Tax and pricing: International evidence suggests that an increase in alcohol prices decreases
consumption and alcohol related harm (Anderson et al., 2009) and that increasing taxes on
alcohol and minimum pricing3 are effective policy options (Mongan et al., 2009; Purshouse et
al., 2009). Ireland lacks a legislative basis for introducing minimum pricing. Since 2000, there
have been three4 increases in excise duty rates in Ireland followed by a subsequent decrease in
consumption. In 2009, excise duty was reduced followed by a 5.3% increase in consumption in
2010. The Irish drinks industry does not support increasing excise rates on alcohol and
introducing minimum pricing. It points instead to its role in the economic prosperity of Ireland
and cautions that the above measures would threaten Irish jobs, tax revenues and income from
exports. In the mean while, the affordability of alcohol in Ireland has increased by 50%
between 1996-2004 (Rabinovich et al., 2009) indicating that strict policies are required.
Availability: The availability of alcohol has substantially increased through longer opening
hours and an increase in the density of off licensed outlets. The Intoxicating Liquor Act (2008)
strengthened some laws, the most significant of which was restricting the opening hours for off
licenses. However, no evaluation of the impact of these regulations has taken place.
Marketing and Advertising: Alcohol marketing in Ireland is regulated by several voluntary
codes embedded in a self regulation system largely maintained by the drinks industry which
has not been sufficiently adequate in reducing alcohol related harm in Ireland or making
The cost of alcohol is based on the number of units it contains. The lowest price is then set at which an alcohol
product can be sold. The more alcohol content a product has, the more expensive it is (Mongan et al, 2009)
4
Cider in 2001, spirits in 2002 and wine in 2008
28
Research Context
alcohol any less available to the population. It would appear that the drinks industry seeks to
counterbalance reports of alcohol misuse by emphasising responsible marketing of its products.
In the mean while, Irish youth continue to be exposed to alcohol marketing practices through a
variety of channels and strategies including price promotions, appealing adverts and packaging
(Hope and McCrea, 2009). Public health advocates often regard it to be an exercise in public
relations aimed at allaying public concern rather than taking actual measures.
Drink Driving: Random breath testing introduced in 2006 and a new lower BAC limit of 50mg
and 20mg for professional and novice drivers respectively have been effective measures.
Education: The Department of Education and Skills delivers the SPHE (Social, Personal and
Health Education) programme in junior high schools which focuses on substance abuse as part
of a wider curriculum (DoH, 2012). However, the continuation of this programme in senior
cycle remains a challenge. Similarly, prevention efforts such as the SN marketing campaigns
have been largely ignored in Irish policy making despite their success in other countries. In the
mean while, Irish children continue to drink with 1 in 5 being a weekly drinker (GAAG, 2008).
2.5
Conclusion
Alcohol related harm in Ireland far outpaces the economic benefits generated by the drinks
industry and necessitates the consideration of alternative strategies to protect and preserve
public health. One approach in this regards has been the implementation of SN marketing
campaigns in the US and some other countries. Based on the SN theory, these campaigns have
apparently been successful in reducing alcohol consumption rates on college campuses.
Chapter 3 reviews the theoretical basis of this approach and discusses related literature.
29
Literature Review: The Social Norms Approach
3.1
Introduction
Norms have evolved as a popular sociological concept that reflects behavioural rules. They are
cultural phenomenon which have the power to prescribe and proscribe behaviour in specific
situations (Hechter and Opp, 2001) and are presumed by many in the field to represent a
powerful source of influence on human behaviour (Cialdini et al., 1990; Larimer and
Neighbors, 2003; Berkowitz, 2004). From a theoretical point of view, classic research in social
psychology has long argued that friendship affiliation needs and social comparison processes
(Festinger, 1954), pressures towards peer group conformity (Asch, 1951) and the formation
and acquisition of reference group norms (Newcomb, 1943; Newcomb et al., 1966; Sherif,
1972) typically coalesce to encourage (or force upon) individuals to act in accordance with
their peers' expectations and behaviours (Perkins and Wechsler, 1996). The research on social
norms is scattered across several disciplines and substantive topics such as queuing
(MacCormick, 1998), binge eating (Giles et al., 2007), substance use (Maxwell, 2002), crime
(Sampson et al., 1997) and social order (Hechter and Kanazawa, 1993). As discussed in the
forthcoming pages, the literature on substance use particularly that focussing on college
drinking, frequently links alcohol consumption to the normative influence of peers in the peer
intensive environment that characterizes college life (Wood et al., 1992; Turrisi, 1999; Clapp
and McDonnell, 2000; Wood et al., 2001). This forms the basis of a widely implemented social
norms strategy to reduce alcohol consumption on college campuses more commonly known as
the Social Norms Marketing Approach.
30
This chapter begins by specifying the meaning and significance of norms, describing the types
of social norms used specifically in the college drinking literature and examining the literature
on social influence, a widely recognized factor believed to facilitate the dissemination of
drinking norms among young adults. It then introduces the SN theory and describes its
postulates and assumptions. The discussion then focuses on SN marketing interventions which
are based on the SN theory and reviews the success and limitations of this prevention strategy
in light of empirical findings. The chapter concludes with a review of commonly raised
criticisms against the approach and the identification of some substantive theoretical issues in
the literature that call for further examination.
3.2
What are Norms?
The attention that norms and normative beliefs have received over the years especially in the
context of college drinking, has led researchers to explicate further the idea of norms with
several definitions emerging in the literature depending on the focus of the researcher. Put
simply, norms are rules that serve as a guide about how people are supposed to behave or
believe (Francessca, 2009). In order to interpret and understand the research on social norms, it
is crucial to address how norms have been conceptualized in this work. College drinking
literature reflects the use of two closely related terms. These are the descriptive and the
injunctive norms - articulated by Cialdini and colleagues (1990).
3.2.1 Descriptive Norms

One source of reassurance that people look for when trying to determine appropriate social
behaviour, is the descriptive norm relevant to the situation (Cialdini and Trost, 1998).
31
Descriptive norms also known as popular norms, norms of prevalence or norms of is

refer to individuals beliefs about the prevalence of a behaviour (Rimal and Real, 2005). Put
differently, these are derived from the perceived actions of other people in a given situation.
Consequently, such perceptions serve as a guide for behaviour particularly when the situation
is ambiguous (Festinger, 1954; Lapinski and Rimal, 2005). In addition to this, descriptive
norms provide us with consensus information when the majority responds to the same
situation in the same way, we perceive the behaviour to be normative (Cialdini and Trost,
1998) and follow suit. This heuristic of social proof saves us the cognitive effort of choosing
an appropriate course of action (Cialdini, 1993). In the specific context of alcohol
consumption, descriptive norms refer to the perceptions of others quantity and frequency of
drinking (Borsari and Carey, 2001). Largely based on observations and experiences, these
provide information about how people consume alcohol in various drinking situations.
3.2.2 Injunctive Norms

Injunctive norms go beyond simply describing appropriate behaviour to prescribing it as well
as proscribing inappropriate actions (Cialdini and Trost, 1998). Expanding on the above, this
norm construct is commonly used to refer to behaviours which are accompanied by social
acceptance or approval of others. Put formally then, injunctive norms characterize the
perceptions of what most people approve or disapprove (Cialdini et al., 1991). Also known as
prescriptive norms, norms of approval or acceptability or norms of ought, these represent
the perceived moral rules of the group and help people determine, what is acceptable and
unacceptable social behaviour (Borsari and Carey, 2003; Rimal and Real, 2005; Lewis and
32
Neighbors, 2006a). In the context of alcohol consumption, these refer to the perceptions of
social acceptability and unacceptability of drinking and drunkenness among peers.
3.2.3 Distinction between Injunctive Norms and Subjective Norms

The term subjective norm was first used in the theory of reasoned action (TRA) (Ajzen and
Fishbein, 1980) and its successor, the theory of planned behaviour (TPB) (Ajzen, 1991). The
theories conceptualized subjective norms as a combination of perceived expectations from
relevant individuals or groups along with intentions to comply with these expectations.
Since then, research has reflected two contrasting opinions on use of the term subjective
norms. Some researchers regard it as a variation of injunctive norms. For example, Rivis and
Sheeran (2003) and Lapinski and Rimal (2005) describe that the subjective norms construct, as
articulated in the TRA and TPB is a form of injunctive norms in that they are concerned with
the motivation of people to comply with beliefs of important referents. In contrast, other
researchers treat subjective norms as being distinct from injunctive norms and provide more
precise definitions more so in conjunction with TRA and TPB. For example, Finlay and
colleagues (1999) describe subjective norms as an individuals perception about what
important others think the individual should do (Finlay et al., 1999). Park and Smith (2007)
support this and strengthen the distinction by suggesting that subjective norms are measured by
the perceptions of important others expectations (i.e., they think that I should.), and
injunctive norms are measured by the perceptions of important others approval (i.e., they
would approve of my.). It is important to appreciate this distinction to avoid inter
changeable use of the terms. Reflecting on the above, injunctive norms refer to the beliefs of
33
acceptability of behaviour whereas subjective norms refer to the beliefs of expectancy of

behaviour in important referents.
3.2.4 Characteristics of Norms

The aforementioned conceptualization or in fact any definition of social norms has four
principal aspects.
First, norms are not simply rules. In fact norms are thought to be in place if any deviation from
them incurs some kind of punishment (Summer, 1906; Rimal and Real, 2003). The idea of
sanctions is central to the concept of norms because in the absence of some sort of
enforcement mechanism, rules become completely discretionary and have no binding force
(Bendor and Swistak, 2001; Hechter and Opp, 2001). Scholars differ in their views about what
makes norms effective and powerful. Two contrasting viewpoints appear in the literature.
Some researchers argue that norms are internalized, implying that individuals apply sanctions
to their own behaviour and respond to these internally generated rewards or punishments
(Durkheim, 1951; Elster, 1989; Coleman, 1990). A majority of scholars however emphasize
the role of external sanctions. According to this view, norms are generally enforced by
sanctions which materialise either in the form of rewards for complying and conforming with
them or punishments for deviant behaviour (Hechter and Opp, 2001). These sanctions come
from the social networks of individuals rather than the legal system (Cialdini and Trost, 1998;
Marin and Wellman, 2010).
Second, norms are different from laws (Cialdini and Trost, 1998; Rimal and Real, 2003). This
is an important consideration because laws come in force through a deliberate procedure,
34
precisely specified in written texts, linked to particular sanctions and enforced by a specialized
bureaucracy (Bendor and Swistak, 2001). Norms on the other hand are understood through
social interaction. Hechter and Opp (2001) elaborate that social norms are often spontaneous,
unwritten and enforced informally. Sociologists concur that norms propagate through social
influence, the fundamental principle of which is consensus (Cialdini and Trost, 1998; Hechter
and Opp, 2001). Although the amount of acceptance is unspecified, it is generally argued that
at least some level of consensus is necessary among group members regarding the validity of
the rule (Hechter and Opp, 2001).
Third, norms do not exist independently of individuals group identity. The sense of belonging
to the group triggers appropriate behavioural measures in line with group expectations as has
been predicted by theories such as the social identity theory (Tajfel and Turner, 1986).
Finally, norms are shared belief systems which largely disseminate through observation and
communication (Bendor and Swistak, 2001).
3.3
Theory of Social Ecology and Importance of Norms
Social ecological models find their roots in Bronfenbrenners (1977) theory for ecology of
human development. Bronfenbrenner (1977) defines the person-environment interrelationship
in terms of micro, meso, exo and macrosystems. Microsystems refer to the principal and
immediate socialization contexts (family, school and college contexts for adolescents). A
mesosystem is formed from the interrelations among microsystems. For example, the
connection between an adolescents family and his friends represents a mesosystem.
Exosystems comprise more distant social environments for example the neighbourhoods in
35
which the microsystems are embedded. Finally, macrosystems are composed of cultural values,
customs and laws. The theory has been embraced by health specialists who recommend its use
in college alcohol prevention (Perkins and Berkowitz, 1986; Hansen, 1997). Social ecology
theory for alcohol and drug use prevention (Hansen, 1997) postulates that informal rules
(norms) about alcohol consumption and drug use rather than formal ones (policies) influence
individual behaviour. Hansen (1997) also describes 165 social units relevant to college students
which may have varying levels of influence on their alcohol and drug use. The social ecology
theory suggests that the more profoundly a social unit affects interaction among students, the
more likely it will be to influence behaviour and that in order to change a particular behaviour
we must address the effects of social influence. Social influence perspective and related
research is reviewed in the following section.
3.4
Social Influence Research
Norms often operate through the medium of social influence (Cialdini et al., 1990), which is a
fundamental component of all models of adolescent substance use (Graham et al., 1991).
Social influence occurs when people continually compare themselves with others to ascertain if
their own behaviour is appropriate (Maxwell, 2002). In studying risk and protective factors,
many researchers have examined the contribution of social influences in the initiation and
perpetuation of drinking among adolescents (Wood et al., 2004). This literature has
These units in decreasing order of potential for social influence are (1) friends and acquaintances (2) dormitories
and roommates (3) parties (4) cafes, nightspots, stores and hangouts (5) classes and classmates (6) fraternities
and sororities (7) special interest clubs and groups (8) campus sponsored special events (9) worksites (10)
athletic teams (11) the student newspaper (12) religious fellowships (13) the faculty (14) student government
(15) the administration (16) student health services (Hansen, 1997)
36
predominantly focussed on studying peer influences especially among young adults such as
college students (Baer et al., 1991; Baer and Carney, 1993; Borsari and Carey, 2001; Borsari
and Carey, 2003).
Within the context of alcohol consumption, peer influences are those interpersonal factors
which are present in the immediate or potential drinking environment (Borsari and Carey,
2001). The nature of social influence impinging on an adolescent however has been only
broadly described in the early research (Graham et al., 1991) and often theorized under the
concept of peer pressure. Borsari and Carey (2001) argue that peers can exercise social
influence in relation to alcohol consumption in two distinct ways: directly and indirectly.
3.4.1 Direct Peer Influence

Direct (or active) peer influence refers to explicit offers to consume alcohol (Graham et al.,
1991; Wood et al., 2004). Examples of such influences can range from polite gestures such as
buying a round to verbal prompting and encouragement such as encouraging an individual to
drink during drinking games. Borsari and Carey (2001) note that there is indeed relatively little
research documenting if direct offers and active pressures influence personal alcohol
consumption. From research that does exist on the topic, two themes of interest emerge.
First, drinking is an important ingredient of social functions at college and students who do not
drink heavily in college drinking situations are naturally regarded as unusual and become the
targets of active social pressure by being offered drinks frequently and exposed to teasing from
peers (Rabow and Duncan-Schill, 1995).
37
Second, some students can resist peer offerings of alcohol and are thus not susceptible to active
peer pressure. This is mostly associated with maturity, social ease (Shore et al., 1983) and year
in college (Klein, 1992). Intrinsically, younger students who are new to college life may be
more likely to be influenced by overt pressure to consume alcohol than older students. A
gender effect also appears to exist suggesting that women are more resilient to active influence
(Klein, 1992). Based on the findings of a study of 562 under graduate students at a private
American university, Klein (1992) reports that unlike men who demonstrated no significant
changes over the course of their college years, women in this study appeared to mature
throughout, gradually progressing towards an adult like developmental stage at least as far as
their drinking patterns and alcohol related attitudes were concerned.
Preventive interventions designed to address active social influences focus mostly on
reinforcing refusal skills (Wood et al., 2001). The limited studies on active influences make it
difficult to determine the validity and reliability of the measures used in such research and
should therefore be treated with caution (Borsari and Carey, 2001).
3.4.2 Indirect Peer Influence

In contrast, indirect (or passive) peer influences are triggered when peers set examples through
their own actions thus providing an adolescent with information about what is in vogue and
socially normative (Graham et al., 1991; Borsari and Carey, 2001). This provides guidelines to
an adolescent about what behaviours are likely to lead to social acceptance and reinforcement.
Wood et al (2004) summarize indirect peer influences as being related to an individuals
perception and interpretation of alcohol consumption and reinforcement patterns of others.
38
These are further classified to include two dimensions that have been linked to drinking
behaviour, modelling and perceived norms.
3.4.2.1 Modelling
Modelling refers to the temporary and concurrent imitation of anothers behaviour. Interest in
modelling began when an earlier study (Cutler and Storm, 1975) noted the alcohol
consumption of college students to be positively associated with their drinking group size.
Much of the research carried out in relation to social modelling is experimental in nature. A
typical example of this involves placing participants in a situation where they are free to
consume alcohol along with other volunteers, who unknown to participants are actually
confederates of the researcher. A modelling effect is observed if a participants alcohol
consumption matches that of the confederate. Research indicates that this effect is influenced
by the characteristics of both the participant and the confederate (Quigley and Collins, 1999).
Borsari and Carey (2001) note three aspects of modelling research. First, participants only
model concurrent behaviour of the confederate meaning that the effect does not continue to
influence individual consumption once outside the control environment (Cooper et al., 1979).
Second, the modelling process is influenced by the composition of the group, if any. For
example, in a study where two confederates drink at different rates, the participants are likely
to match the fastest (DeRicco and Niemann, 1980), whereas in a larger group, majority
behaviour is modelled (Dericco, 1978). Third, friendliness of the confederate during the
session also appears to have an effect as participant drinking has been noted to converge more
rapidly with a model who appears to be more sociable (Maisto et al., 1999). Research also
notes that heavy social drinkers drink significantly more in the presence of heavy drinking
39
models than do light social drinkers (Lied and Marlatt, 1979). This suggests that regardless of
the modelling conditions, heavy drinkers will drink more than light drinkers. Those with a
family history of drinking problems are believed to match the confederates consumption to a
greater extent than others (Chipperfield and Vogel-Sprott, 1988) and females are believed to
consistently consume less alcohol than males (Cooper et al., 1979; Lied and Marlatt, 1979;
DeRicco and Niemann, 1980).
There is evidence that the modelling effect occurs even after the participants are made fully
aware of the confederate and aims of the study (Dericco and Garlington, 1977). This suggests
that campaigns attempting to raise awareness of modelling and imitation effects may not help
in reducing personal consumption (McAlaney, 2007). Wood et al (2001) suggests that from a
social modelling perspective, preventive interventions to reduce alcohol abuse need to focus on
decreasing the occurrence of heavy drinking in the proximal environment. However, it is
crucial to note that the external validity of the modelling research is limited by highly
controlled experimental interactions, fewer studies with female participants and inability to
compare drinking durations with those commonly encountered in college settings (Borsari and
Carey, 2001).
3.4.2.2 Normative Beliefs
Another source of social influence identified by literature is the normative beliefs or perceived
norms and like modelling, these influence ones alcohol consumption indirectly (Graham et al.,
1991; Borsari and Carey, 2001; Wood et al., 2001; Wood et al., 2004). The literature on
normative beliefs emphasizes that these often stem from adolescents perceptions of peer
behaviour rather than the actual behaviour (Wood et al., 2001). This is in conjunction with a
40
sociological precept and what has come to be known as the Thomas Theorem stating that if
people define situations as real, they are real in their consequences (Thomas and Thomas,
1928). The very idea continues to be accepted and acknowledged by the researchers as a
powerful way of comprehending human behaviour (Perkins, 2002). For example, Perkins
(1997) remarks that indeed the strongest form of peer influence may occur indirectly through
an individuals perception of peers regardless of the accuracy of that perception. This is an
important element of the SN theory the effects of which will be discussed in section 3.6.
The literature suggests that normative beliefs are constructed from three primary sources which
are observable behaviours, direct and indirect communications and knowledge of the self
(Miller and Prentice, 1996). These are the key concepts behind several behavioural theories
and models. The social cognitive theory (Bandura, 1986) for example describes the processes
by which social influence factors contribute to behaviour. Among other things, it emphasizes
the influence of normative perceptions on behaviour over and above the immediate effects of
active social pressure and modelling. Perkins (1997) argues that several factors combine to
move individuals to perceive the world as their group does, to adopt peer attitudes and to act in
accordance with peers expectations and behaviours. Such factors have long been discussed by
classic research in social psychology and form the basis of SN theory. The social comparison
theory (Festinger, 1954) for instance explains individuals need to evaluate their own attitudes
by comparing themselves to similar others. Social Impact theory (Latane, 1981) describes
social influence as a function of strength (how important the referent group is to you),
immediacy (how close the group is to you) and group size. The theory postulates that as
strength and immediacy increase within a group, conformity with its norms also increases.
Further, as the size of a group increases, a single person has less of an effect. Social
41
categorization theory (Turner et al., 1987) assumes that the social norms of an in-group are
perceived by its group members to be objectively true and appropriate.
Even if behaviour such as heavy drinking and drunkenness is viewed as deviant by the larger
society, young people may still socially learn and continue to engage in abusive drinking in
response to peer group norms (Perkins and Wechsler, 1996). As the literature identifies, there
is consistent evidence establishing that normative perceptions of alcohol consumption in others
are predictive of both personal drinking (Wood et al., 1992; Nagoshi, 1999; Clapp and
McDonnell, 2000; Mallett et al., 2009) and drinking problems (Wood et al., 2001). Normative
beliefs are discussed in the literature as being the most researched of all the forms of social
influence particularly in the context of alcohol consumption and arguably the ones most open
to modification. As will be discussed in section 3.7, these are also being used worldwide in an
increasingly popular and apparently effective intervention strategy to reduce alcohol
consumption among college students.
3.5
Students and Susceptibility of Social Norms
College is a time of social anxiety for students characterized by the desire to gain acceptability
among peers and the pressure to fit in with perceived popular behaviour (Borsari and Carey,
2001; Wood et al., 2001). For many students, going away to college is the first experience of
being in an unfamiliar environment requiring knowledge of appropriate and relevant modes of
conduct. This is a time when students experience a great deal of ambiguity as many habitual
behaviours become inapplicable in the new environment (Cialdini and Trost, 1998). The
literature indicates that normative influences such as those outlined in section 3.4.2.2 are
42
heightened under conditions of ambiguity (Cialdini, 1993; Rice, 1993). Normative perceptions
of peer behaviours and attitudes therefore become powerful influencers of behaviour during
this time of distinct vulnerability as students try to mirror peer norms in their efforts to gain
social approval and to be perceived as relaxed and at ease with drinking (Schall et al., 1992;
Perkins, 1997; Wood et al., 2001). Violating these norms can make one appear different, which
is especially undesirable in social situations.
Borsari and Carey (2001) note two aspects that make the college environment a suitable setting
to examine drinking behaviour and relevant norms. First, there is a predominant shift in
influence from parents to peers during this time. Peers become increasingly important as
students become more independent from parental oversight. Second, the increased prevalence
of alcohol based social opportunities during college life adds to the potency of peer influence
and makes normative pressures all the more influential.
3.6
Social Norms (SN) Theory
SN theory is based on the idea that our behaviour is guided by what other members of our
social groups accept and expect and how they behave (Berkowitz, 2004; Berkowitz, 2005). It
means that individuals express or inhabit a behaviour in an attempt to conform to a perceived
norm (Perkins, 2002). According to SN theory, our perceptions of others behaviours and
beliefs are often incorrect. These incorrect beliefs often referred to as misperceptions in SN
literature, have an impact on our own behaviours (Berkowitz, 2004). SN theory argues that
correcting misperceptions can lead to behaviour change for most individuals who will either
reduce their participation in the problem behaviour or be encouraged to engage in protective
43
healthy behaviours (Perkins and Berkowitz, 1986). This has led to the development of a
prevention strategy (SN marketing interventions) discussed in detail in section 3.7.
3.6.1 Causes and Types of Misperceptions

Theoretical explanation of the causes of misperceptions draws attention to processes occurring
at the cognitive, social and cultural levels (Perkins, 1997). At the cognitive level, humans tend
to overly attribute actions of other people to their dispositions rather than to environmental
contexts in which the behaviour occurs (Perkins, 2002). This happens because people often
lack information to make accurate judgement about the causes of other peoples behaviour.
Thus, when students observe an intoxicated peer, they tend to think it is characteristic of that
individual unless they have a concrete basis to think otherwise (Perkins, 1997). At the social
level, extravagant behaviour of an individual or a few people under the influence of alcohol is
easily noticed, remembered and given disproportionate weight in subsequent social
conversations with peers (Perkins, 1997; Perkins, 2003a). This further exaggerates the
perceived drinking norm among students. Finally, at the cultural level, mass media contributes
heavily to the production and reinforcement of alcohol related misperceptions through films,
television shows and advertisements that unrealistically emphasize heavy drinking as part of
youth culture (Perkins, 1997; Perkins, 2003a).
Misperceptions can occur in the form of overestimating or underestimating the norm.
However; research informs us that people often tend to overestimate problem behaviours and
underestimate healthier norms (Perkins and Wechsler, 1996; Perkins et al., 1999; Perkins and
Craig, 2002; Berkowitz, 2004; Berkowitz, 2005). SN theory describes three common types of
misperceptions which are discussed next.
44
3.6.1.1 Pluralistic Ignorance (I think they differ, when they dont)

Pluralistic ignorance is a psychological state which occurs when a majority of individuals
falsely assume that most of their peers behave or think differently from them when in fact their
attitudes and/or behaviours are similar (Miller and McFarland, 1991; Prentice and Miller,
1993). This is the most common type of misperception. SN theory suggests that most college
students drink moderately but incorrectly assume that their peers drink a lot more and do so a
lot more frequently than they do (Berkowitz and Perkins, 1986a; Perkins and Craig, 2002;
Rimal and Real, 2003; Berkowitz, 2004; Berkowitz, 2005; McAlaney and McMahon, 2007).
Pluralistic ignorance encourages individuals to move their attitudes and behaviours closer to
the unhealthy behaviours and attitudes which are perceived incorrectly as being normative. It
also provides encouragement to suppress healthy attitudes and behaviours that are wrongly
thought to be fairly uncommon (Berkowitz, 2005).
3.6.1.2 False Consensus (I think they are as bad as me, so I dont really have a
problem)
False consensus, also known as self-serving bias refers to the tendency of people to assume
that others share their attitudes and behaviours to a greater extent than they really do (Wolfson,
2000). It takes the form of overestimating the frequency with which other people might act or
feel in conjunction with ones self and this is especially relevant to behaviours that are viewed
by an individual with somewhat uncertainty and conflict (Berkowitz, 2004). Drawing from the
above, it represents an individuals denial that his or her attitudes or behaviours are
problematic or unusual therefore motivating him or her to believe in these exaggerated norms.
45
3.6.1.3 False Uniqueness (Not many are like me and abstain or drink less)
False uniqueness refers to the tendency to underestimate the commonality of ones attitudes
and behaviours (Suls and Wan, 1987). For example, those students who abstain from drinking
alcohol can underestimate the prevalence of abstinence and falsely assume that they are more
unique than they really are. This can encourage them to withdraw from the larger community
perceiving it to be more alcohol oriented than it really is.
3.6.2 Documentation of Misperceptions

Following the work done by Perkins and Berkowitz (1986), a body of institutional research
diverse in region, size and student characteristics has reported similar misperceptions related to
overestimations of peer drinking norms (Prentice and Miller, 1993; Beck and Treiman, 1996;
Haines and Spear, 1996; Larimer et al., 1997; Page et al., 1999; Carter and Kahnweiler, 2000;
Clapp and McDonnell, 2000; Peeler et al., 2000; Thombs, 2000; Werch et al., 2000; Glider et
al., 2001; Sher et al., 2001; Fabiano, 2003; Far and Miller, 2003; Haines et al., 2003; Jeffrey et
al., 2003; Linkenbach and Perkins, 2003a; McAlaney and McMahon, 2007). Specifically, these
studies demonstrate that most students perceived substantially greater use of alcohol among
their peers than was the reality.
Further, these misperceptions have been noted for both descriptive and injunctive norms.
Borsari and Carey (2003) from a meta-analysis of 23 studies of normative misperceptions
conducted in the US report that students overestimated both the drinking norms of prevalence
(descriptive) and approval (injunctive) in peer reference groups and that these misperceptions
were greater for injunctive norms compared to descriptive norms. The researchers reflect that
46
one possible explanation for this can be that injunctive norms are based on less direct
information in that they are largely unobservable and hence their estimations require greater
cognitive inference leading to higher misperceptions. Trockel and others (2003) and Larimer
and Neighbors (2003) also report misperceptions of injunctive norms in studies examining
fraternity drinking and college student gambling respectively.
Research also informs us that misperceptions are held by all members of campus communities
inclusive of undergraduate and graduate students, faculty and staff, students and student
leaders (Berkowitz and Perkins, 1986a; Berkowitz, 1997). Likewise, misperceptions have been
documented in a state wide sample of young adults in the US both in college and not in college
(Linkenbach and Perkins, 2003a) as well as among middle and high school students (Beck and
Treiman, 1996; Thombs et al., 1997; Botvin et al., 2001; D Amico et al., 2001; Haines et al.,
2003; Perkins and Craig, 2003b). Similarly, Thombs and others (1997) report misperceptions
relating to DWI (Driving while intoxicated) and RWID (Riding with an intoxicated driver).
There is only one published study which questions the existence of misperceptions (McAlaney
et al., 2011). This study conducted by Wechsler and Kuo (2000) reports that students
accurately perceived campus norms for drinking and that there were no underlying
misperceptions. However, a number of researchers criticize the study and point out
methodological problems (DeJong, 2000; Perkins and Linkenbach, 2003) such as the use of an
arbitrary definition of binge drinking and the use of normative belief items which were
fundamentally different from the measures of personal behaviour a consideration which is
believed to be of particular importance as will be discussed in chapter 5, section 5.5.6. Further,
Wechsler and Kuo (2000) assume ones perceptions of others drinking as being accurate if
47
within +/-10% of the actual campus binge drinking rate. Dejong (2000) argues that this
condition conceals over or under estimates of binge drinking on campus that may be reliable
but less than 10%.
The patterns of exaggerated perceptions have also been reported for other behaviours that
occur frequently in substance use research such as cigarette smoking (Haines et al., 2003;
Hancock and Henry, 2003; Linkenbach and Perkins, 2003b; Perkins and Craig, 2003b),
marijuana and other illegal drug use (Perkins, 1985; Hansen and Graham, 1991; Perkins et al.,
1999; Pollard et al., 2000; Wolfson, 2000; Perkins and Craig, 2003b). Apart from alcohol,
tobacco and other drugs, the evidence of misperceptions has also been documented for a range
of other behaviours such as gambling (Larimer and Neighbors, 2003), bullying (Bigsby, 2002;
Paluck and Shepherd, 2012), homophobia (Dubuque et al., 2002), eating disorders (Kusch,
2002) and sexual assault (Bruce, 2002).
3.6.3 Consequences of Misperceptions

SN theory argues that when misperceptions are perceived as real, they have real consequences.
There is strong and consistent evidence in the literature demonstrating that misperceptions are
positively associated with ones own drinking behaviour, including studies of both cross
sectional (Perkins, 1985; Wood et al., 1992; Perkins and Wechsler, 1996; Thombs et al., 1997;
Nagoshi, 1999; Page et al., 1999; Clapp and McDonnell, 2000; Korcuska and Thombs, 2003;
Page et al., 2008) and longitudinal (Graham et al., 1991; Prentice and Miller, 1993; Botvin et
al., 2001; D Amico et al., 2001; Sher et al., 2001) nature.
48
Perkins and Wechsler (1996) from a nationwide study of 17,592 students at 140 colleges and
universities in the US report that perceptions of campus drinking climate explained more of the
variance in drinking behaviour than any other variable. Similarly, Clapp and McDonnell
(2000) report that perceptions of drinking norms prevalent on campus are predictive of
drinking behaviour and alcohol related problems. Page and others (1999) demonstrate that
overestimations of others binge drinking is directly correlated with an individuals own rate of
binge drinking. A longitudinal study (Sher et al., 2001) reports that perceptions of heavy
drinking in the Greek system were largely responsible for the prevalence of heavy drinking
among fraternity and sorority members.
The research on the association between normative perceptions and drinking behaviour is not
only limited to college campuses but has also been conducted with high school and middle
school populations. For example DAmico and colleagues (2001) in a US based longitudinal
study of over 1500 high school students demonstrate that higher peer perceptions of alcohol
use are associated with subsequent escalations in personal drinking and that only the perceived
intensity of student alcohol use predicts behaviour change.
Just like misperceptions have been documented for both types of norms as has been
commented in section 3.6.2, the impact of misperceived norms in predicting personal
consumption has also been established for both descriptive and injunctive norms. Early
research evaluating the direct effect of injunctive norms on drinking behaviour reports that it is
predictive of alcohol use (Wood et al., 1992; Wood et al., 2001). Larimer and colleagues
49
(2004b) in a US based longitudinal study of fraternity and sorority6 students report similar
associations between injunctive norms and drinking behaviour at two time points. Borsari and
Carey (2003) and Trockel and colleagues (2003) assert that injunctive norms are more likely
than descriptive norms to predict drinking behaviour. Larimer and Neighbours (2003) report
similar findings with regards to student gambling. However, in general, perceptions of
descriptive norms have been found to be more influential than those of injunctive norms,
which is part perhaps reflected in the fact that most of the SN research has focused on the
former. This issue will be addressed in more detail in section 3.9.
In summary, SN theory suggests that misperceptions exist and that they are associated with
ones personal consumption of alcohol. These assumptions of SN theory have been extensively
validated by international research as has been discussed in the preceding discussion.
3.7
SN Prevention Strategies for Reducing Alcohol Related Harm
Problems commonly associated with misuse of alcohol by college students have been shown to
include property damage, poor academic performance, damaged relationships, unprotected
sexual activity, physical injuries, date rape, and suicide (Berkowitz and Perkins, 1986b;
Perkins, 1992; Wechsler et al., 1994; Presley et al., 1996; Wechsler and Kuo, 2000). Beck et al
(2008) report on the relationship between social contexts and alcohol related problems in a
sample of 728 college students. Their findings demonstrate that those students who frequently
drink to enhance general well being and conviviality and to facilitate social interaction are
Fraternities and sororities are club like organizations for under graduate students, common in the American and
the Canadian college system. Collectively these are called the Greek System because their names often
consist of Greek letters. (Sher et al, 2001)
50
more likely to ride with an impaired driver, to drive after drinking themselves and to receive an
alcohol related housing violation. Similarly those who drink in the context of motor vehicles
and emotional pain are likely to suffer from alcohol abuse/dependence and clinical depression
respectively.
The predominant approach in the field of health promotion has been to induce positive
behaviour change by highlighting risk and emphasizing the negative consequences of heavy
drinking often referred to as the scare tactic approach (Berkowitz, 2004). Others have
focussed on the development of skills for dealing with inter and intrapersonal social influences
and enhancement of personal qualities such as self esteem (Hansen, 1997). However, these
strategies have not been effective (Perkins and Craig, 2002).
SN theory was first examined in the context of college drinking by Perkins and Berkowitz
(1986) who found that students at Hobart and William Smith Colleges in New York tended to
overestimate how heavily and how frequently their peers drank alcohol. This overestimation
was found to predict students own drinking. This foundational study recommended the
application of SN theory to reducing drinking on college campuses which led to the
development of SN based prevention strategies. These strategies refer to any variety of
approaches designed to decrease problem behaviour or increase protective behaviour by
reducing misperceptions of healthy norms (Perkins and Craig, 2002). Popularly known as
social norms marketing interventions, these campaigns target the discrepancy between reality
and perception by first exposing and then shrinking the gap between the two. This has an effect
of the target population revising their perceptions downward with their drinking levels (or level
of the behaviour being examined) following. Further, this strategy also leads to a decrease in
51
alcohol related harm effects (Perkins, 2003a). The model of SN based prevention approach
(Perkins, 2003a) is depicted below.
Since the land mark study by Perkins and Berkowitz (1986), numerous interventions have
practically evaluated whether norm education can change both students perceptions of norms
as well as their alcohol consumption. SN based interventions utilize social marketing7
techniques to deliver norm education in several formats to address universal, selective and
indicated prevention or a combination of these (Berkowitz, 2004).
Universal prevention efforts typically combine normative feedback with social marketing
techniques to promote actual healthier norms via campus wide electronic/print media
campaigns. For example, the SN based intervention at the University of Arizona reports a
reduction of 29% in rates of heavy drinking (defined as consuming 5 or more drinks in a
sitting in the last 2 weeks) as well as a decrease in negative consequences8 resulting from
alcohol over the years 1995-98 (Johannessen and Glider, 2003). Some of the other colleges that
Social marketing is the application of commercial marketing techniques to programs designed to influence the
behavior of people in order to improve their personal welfare and that of their society. Andreasen, A. R.
(1995). Marketing social change: Changing behavior to promote health, social development, and the
environment, Jossey-Bass San Francisco, CA.
Getting into a fight or argument (decreased from 32.2% to 20.4%), Getting in trouble with campus police or
other school authorities (decreased from 17.5% to 6.3%), doing something that later regretted (decreased from
41.8% to 31.2%), been taken advantage of sexually (decreased from 14.7% to 8.3%), doing poorly on a test or
important project dropped (decreased from 22.2% to 15.8%), missed class (decreased from 33.5% to 24.7%)
52
have successfully implemented such interventions include the Western Washington University
(Fabiano, 2003), the Northern Illinois University (Haines and Spear, 1996; Haines, 1996;
Haines and Barker, 2003), the Hobbart and William Smith colleges (Perkins and Craig, 2002;
Perkins and Craig, 2003a), the Rowan University (Jeffrey et al., 2003) and the University of
North Caroline Chapel Hill (Foss et al., 2003; Foss et al., 2004). These colleges report a
reduction of 20% or more in high risk drinking rates within 2 years or more through a social
norms marketing campaign. The website of the National Social Norms Center
(www.socialnorm.org) keeps detailed data pertaining to these as well as other studies.
Selective prevention measures also use social marketing techniques specifically targeted at
particular groups such as first year students or fraternity/sorority members (Berkowitz, 2004).
These interventions pioneered at the Washington State University (Barnett et al., 1996; Peeler
et al., 2000; Far and Miller, 2003) typically provide norm education to high risk drinkers in
small interactive group discussions, workshops or academic classes. Due to their small size and
manageable format, many of these interventions assess their effectiveness by randomly
assigning students to experimental and control groups (Barnett et al., 1996; Schroeder and
Prentice, 1998; Steffian, 1999; Peeler et al., 2000; Far and Miller, 2003). For example, Far and
Miller (2003) report a decrease of 29% in students who drink 5 or more drinks on an occasion
and an increase of 8% in abstainers over a nine year period. Another selected prevention
campaign conducted at the University of Virginia (Odahowski and Miller, 2000) reports that
the number of drinks per week for first years went down from 3 to 1, the median number of
drinks per week for fraternity first-year men went down from 15 to 7, and the percentage of
abstainers went up from 35% to 49% over a period of three years. In a subsequent evaluation
of this campaign, Bauerle (2003), reports that the campaign was extended to the entire campus
53
and that negative consequences of alcohol experienced by first-year students continued to trend
downwards.
Indicated prevention endeavours provide personalized feedback to high risk drinkers and
abusers as part of individual counselling interventions. These are based on the idea that sharing
normative data in a motivational interviewing format is a non-judgemental way to catalyze
change. Several studies following this approach report significant reductions of drinking rates
in the target population (Dimeff, 1999; Walters, 2000; Borsari and Carey, 2005; Juarez et al.,
2006; Lewis and Neighbors, 2007; Lewis et al., 2007; Walters et al., 2007). Borsari and Carey
(2005) for example report a 30% reduction in alcohol consumption rates of the participants in a
US based study.
3.7.1 Social Norms Marketing Interventions Outside the US

Much of the SN theory and the methodology behind intervention efforts has been developed
and tested in the US particularly within the American college system with some work also
being done in Canada (Perkins, 2007; Schmidt et al., 2009). However, there are important
cultural and legislative differences between the US and other countries especially those of
Europe which raised questions about the success of the approach in European colleges. First,
the legal drinking age in most European countries ranges between 16 to 18 years (19 in Nordic
countries) while in US, it is 21 (Wicki et al., 2010). This means that alcohol becomes available
to European students at a younger age in contrast to their American counterparts. Second, large
cross-cultural studies among adolescents, such as ESPAD indicate a lower prevalence of
54
regular alcohol consumption and RSOD9 in the US compared to European countries meaning
that the prevalence of alcohol use in European freshmen may be higher than among their
American peers (Hibell et al., 2012). Third, European universities differ considerably from the
American colleges. While American students tend to be members of fraternities, sororities and
special athletic clubs, this type of campus life style and the accompanying drinking culture is
far less common in Europe (Turrisi et al., 2006). Similarly, most American colleges have a
thriving sport culture whereas in Europe, most sporting activities tend to take place in clubs
outside of the education system (Wicki et al., 2010). In order to address the concerns related to
ecological validity of the approach, normative research efforts began in other countries
(McAlaney and McMahon, 2007).
Outside the US, misperceptions of alcohol related norms in peers have been documented in
New Zealand (Kypri and Langley, 2003), Switzerland (Bertholet et al., 2011), Germany (Haug
et al., 2011), Finland (Lintonen and Konu, 2004), Scotland (McAlaney and McMahon, 2007),
England (Bewick et al., 2008), Australia (Hughes et al., 2008), Denmark (Balvig, 2009),
France (Franca et al., 2010), Hungary, Romania, Slovakia and the Czech Republic (Page et al.,
2008). Kypri and Langley (2003) for example report that in a randomly selected sample of
1564 university students in New Zealand, 80% of the women and 73% of the men
overestimated the incidence of heavy drinking among student peers. Similarly, McAlaney and
McMahon (2007) examined heavy episodic drinking and normative misperceptions in a sample
of 500 students at a Scottish university using a web based survey. The study reports significant
correlations between the respondents drinking behaviour and their perceptions of the
Risky Single Occasion Drinking

55
behaviour in others10 with beliefs of the most proximal peers being the most strongly
correlated. In addition, the study also demonstrates overestimations of alcohol use in other
students by the majority of respondents.
Although alcohol related misperceptions have been extensively documented outside of US,
McAlaney and colleagues (2011) note that intervention based studies outside of the US are
more limited in number. Among these studies, the Ringsted Project, an award winning
community based initiative reports a reduction of 39% in alcohol use among a group of Danish
school children over a period of one year (Ringsted., 2010). Balvig and Holmberg (2011)
report on a brief SN intervention targeted at 11-13 year old Danish school children and
demonstrate a ripple effect where the correction of misperceptions regarding one type of risk
behaviour (smoking) also influenced other types of misperceptions and risk behaviours
(alcohol and drug misuse). Following the work conducted in the Ringsted experiment, the
approach is being widely used in Denmark (McAlaney et al., 2011). Though this work has not
been publicized extensively in English language journals to date, information about this and
other Danish endeavours is available on the website of the Danish Centre for SN approach
(www.socialpejling.dk).
SNIPE (Social Norms Intervention for the prevention of Polydrug usE) is a collaborative
research project covering six European countries (Belgium, Denmark, Germany, Slovakia,
Spain and United Kingdom) and Turkey (Pischke et al., 2012). It aims to assess the feasibility
of delivering a web based personalized SN feedback for substance use to a sample of
10
The study included three reference groups namely close friends, others of same age at the university and others
of same age in the UK
56
university students from across Europe. The project is currently in progress with the final
results to be published in 2013. The scope of this initiative will help examine whether
misperceptions of the frequency and perceived acceptability of substance use are shared across
cultures. Similarly, EUDAP (European Drug Addiction Prevention) is a Europe wide alcohol
and drug education endeavour targeting 12-14 year olds which incorporates social norms
marketing (EUDAP, 2010).
In Australia, SNAP (Social Norms Analysis Project) is the first major study to use social norms
for reducing alcohol related harm in high school children in rural areas of Tasmania (Hughes et
al., 2008). The project reports significant reductions in self reported frequency of drunkenness
in the intervention group compared to the control. Another Australian study (Kypri et al.,
2009) randomized heavy drinkers into either a control or intervention group and delivered a
web based personalized feedback intervention. The study reports that heavy drinkers who
received the intervention drank 17% less alcohol than the control participants 1 month after
screening and 11% less alcohol 6 months after screening.
In Britain, Bewick and colleagues (2008) report on a web based personalized feedback
intervention where the participants (university students) were randomly assigned to either a
control or intervention condition. Intervention participants received electronic personalized
feedback and norm education through a website they could access by a secure log on
procedure. The study reports significant reductions in the alcohol consumption rates of
intervention participants.
In contrast to many US based studies the Australian (Hughes et al., 2008; Kypri et al., 2009)
and British (Bewick et al., 2008) studies benefit from the inclusion of control groups. It is also
57
notable that unlike most US based research, studies such as Balvig and Holmberg (2011) and
Hughes et al (2008) target school children instead of college students. Balvig and Holmberg
(2011) emphasize that prevention can be most effective when intervention precedes the risk
behaviour which is a why a younger age group is preferable.
Although alcohol misuse and related harm is a serious concern in Ireland as has been discussed
in chapter 2 sections 2.2 and 2.3, the applicability of SN marketing interventions has not yet
been evaluated in Ireland (More on this in section 3.9.1).
3.7.2 Expansion beyond Academics

McAlaney and colleagues (2011) note that to date most social norms work has focused on
student populations although the field has started to rapidly expand into diverse groups and
research topics. Studies have examined social norms in relation to effective parenting
(Linkenbach et al., 2003), safety behaviours at work (Fugas et al., 2011), community attitudes
towards providing support to women who have experienced intimate partner violence
(McDonnell et al., 2011), cigarette use and religiosity in a national sample of 12-17 year old
American adolescents (Gryczynski and Ward, 2011), biodiversity conservation among Finnish
foresters (Primmer and Karppinen, 2010) and paying taxes (Traxler, 2010).
3.7.3 Unsuccessful Interventions

Studies have also been reported where social norms interventions were unsuccessful (Werch et
al., 2000; Granfield, 2002; Clapp et al., 2003; Thombs et al., 2004). However, researchers note
that most of these studies were flawed by methodological limitations such as the failure to
58
reduce misperceptions which SN theory informs would yield no behaviour change (Perkins et
al., 2005; McAlaney, 2007; McAlaney et al., 2011).
3.7.4 Issues in the Evaluation of SN Interventions

There are certain methodological limitations acknowledged by the researchers to characterize
SN literature.
First, norm education is often combined with other components such as discussion of drinking
problems, suggestions to reduce alcohol use, skills training and strengthening resiliency
(Agostinelli et al., 1995; Barnett et al., 1996; Walters, 2000; Walters and Neighbors, 2005). It
is thus not always possible to evaluate the exclusive impact of normative feedback accurately.
It is important to be wary of the fact that these and not the provision of accurate norm
education may have facilitated the observed reductions in alcohol use in intervention work.
Efforts directed at differentiating the effective components in norm education programs that
are associated with drinking may be useful in addressing this limitation (Borsari and Carey,
2003).
Second, use of several different referent groups ranging from your best friend (Baer and
Carney, 1993) to an average student (Perkins et al., 1999) makes it difficult to compare
different studies. It is likely that these reference groups differ in their degree of familiarity and
specificity to the participants. Research has indicated that students' perceptions become more
distorted for groups they know less well (Baer et al., 1991; Perkins, 1997). Subsequently, some
perceptions may be more factually based (best friend) than others (average student), thus
influencing the accuracy and replication of such estimates.
59
The third limitation relates to the issue of possible question order effects in norm estimation.
Research provides evidence that the order in which participants complete personal
consumption and normative belief items influences norm estimates (Baer and Carney, 1993;
Prentice and Miller, 1993). Borsari and Carey (2001) note that such trends can influence the
interpretation of results. Consequently, most researchers conducting survey research have
preferred to present the questions on personal consumption prior to those on normative
perceptions.
Fourth, most normative research is based on self reported measures of consumption, thus
susceptible to possible response bias. This has been addressed by a number of studies that have
used breathalyzers. Foss and colleagues (2003; 2004) for example, used actual blood alcohol
content measures and demonstrated that the reductions in alcohol use were not due to potential
response bias or the possibility that students were guided by the social norms campaigns to
exaggerate changes in their behaviour.
The fifth limitation concerns the lack of longitudinal studies in normative research compared to
a vast majority of cross sectional research. Cross sectional studies make it difficult to establish
causal relationships (Kypri and Langley, 2003). In contrast to the notion that norms precede
drinking, it has been argued that at least two phenomenon; namely projection and social
selection suggest the opposite (Neighbors et al., 2006). In the former students may simply base
their estimates of others drinking on their own consumption in consistency with the
aforementioned phenomena of false consensus. In the latter heavy drinkers may self-select into
a friendship network with comparable drinking levels to themselves, which can influence
personal consumption as well as their perceptions of drinking by peers. There is evidence of
60
this process occurring in student populations (Kahler et al., 2003; Read et al., 2005). Neighbors
and colleagues (2006) address this by examining the temporal precedence of perceived norms
and behaviour. Whilst a degree of reciprocal causality was identified, there was evidence that
perceived norms were a stronger predictor of drinking behaviour than vice versa.
Finally, the more recent Cochrane review (Moreira et al., 2009) of the SN approach in reducing
alcohol misuse in college or university students, identifies the lack of randomized control trials
as a potential weakness of normative interventions. As commented earlier, in contrast to many
of the existing social norms studies in the US, both Bewick et al. (2008) and Hughes et al.
(2008) benefited from the inclusion of control groups. McAlaney and colleagues (2011) note
that though these studies did not bring about large changes in drinking behaviours, it is due to
the limited time frames of these studies compared to other American studies as mass behaviour
change is believed to occur after several years of sustained SN campaigning.
3.8
Criticisms of SN Marketing Interventions
Normative interventions are often subject to scepticism and criticism in relation to the claims
they make since they deviate substantially from conventional methods of preventive research.
Some of the common concerns that are often raised are reviewed here.
3.8.1 Not Publicizing the Problem

Some critics suggest that SN marketing interventions are doing what they refer to as sugar
coating the problem by not really pointing out the significant number of students who exhibit
problem behaviour thus causing potential harm to themselves and to those around them. The
61
advocates of the approach argue that SN theory does not trivialize the problem of alcohol
misuse among youngsters. In fact, SN research clearly demonstrates that alcohol misuse is a
particular problem in student populations (Gill, 2002; White et al., 2006). However, identifying
a problem must be considered as a different task from prevention. While acknowledging that
problematic behaviour exists, SN interventions focus on accurate presentation of the healthy
majority as a way to combat the real problem that exists among a portion of youth (Perkins,
2003b).
3.8.2 Overstating Misperceptions

There is often a concern whether misperceptions exist in such a consistent manner so as to
justify designing a prevention strategy around reducing them. There is extensive published
literature as has been reviewed in section 3.6.2 which reports substantial misperceptions across
a number of problem behaviours in populations that vary in their characteristics. Further, the
advocates of the theory argue, that in rare instance when a study failed to find misperceptions,
it was mostly because the constructs used to measure personal consumption and perceived
norms were not comparable.
3.8.3 Offering Encouragement to Abstainers

Sometimes the question is raised if normative messages can encourage abstainers to take up
drinking because it is more normative than they initially thought it was. SN researchers
(Perkins, 2003b) argue that those who currently abstain from consuming alcohol do not believe
that abstinence is the norm and therefore putting out accurate drinking figures does not expose
them to new pressures.
62
3.8.4 Feasibility in Situations when the Problem Behaviour is the Norm

SN marketing approach to prevention is often criticised in terms of its applicability to
situations where a majority of the population exhibits problem behaviour thus making it
normative. For example, Ireland is known to have a culture where heavy drinking is seen to be
common as explained in chapter 2. The experts of SN approach respond to this by emphasizing
that even if the norm is at a higher level of risk than we would want, people still believe that
their peers on average engage in a still higher level of risk than the actual norm. Therefore,
addressing the misperception is a useful strategy even in such contexts. As commented in
section 3.7.1, exaggerated perceptions of drinking norms have been documented and targeted
via normative feedback in heavy drinking cultures outside of the US for example the UK
(McAlaney, 2007; McAlaney and McMahon, 2007; Bewick et al., 2008).
3.9
Theoretical Gap and Social Need
3.9.1 Applicability of SN Marketing Interventions in Ireland

As has been discussed in chapter 2, sections 2.2 and 2.3, excessive alcohol consumption is a
serious problem in Ireland which has led to several social and health consequences. The
current policy in this regards has not been highly effective and there is a need to consider
alternative strategies to deal with the issue of alcohol misuse in Ireland. It is clear from the
research reviewed in sections 3.6 and 3.7 that SN marketing campaigns are an effective
strategy to reduce drinking rates especially among college populations. While extensive
international evidence supports these campaigns, their applicability is yet to be examined in
Ireland. Most of the alcohol related research in Ireland has focussed on other aspects of the
63
phenomenon. Delaney and colleagues (2008) for example examine gender and ethnic
differences and report that the Irish students drink more than the foreign students and that
males drink more than females. Another study (Delaney et al., 2007) examines students
perceptions of excessive drinking using statistical vignettes based on an online survey and
several focus groups. The study reveals stark heterogeneity in students perceptions of what is
meant by alcohol excess both in terms of their own self rated drinking behaviour and in terms
of their general conceptions about excessive drinking. While perceived peer drinking has been
found to be an important predictor of personal consumption in a study of 2700 Irish post
primary students (Grube et al., 1989), Ireland lacks published empirical work investigating
misperceptions of drinking norms.
Given, the high prevalence of binge drinking in Ireland, an uncertainty remains about the
applicability of SN approach. As commented in section 3.8, the advocates of SN model argue
that the approach should still theoretically work in an alcohol-tolerant context like Ireland
(Perkins, 2003a). However, as a prerequisite to any SN marketing intervention aimed at
reducing drinking rates, it is necessary to establish that alcohol related misperceptions occur
among the Irish college students and impact their drinking behaviours.
3.9.2 The Conceptualization of Norms

One of the weaknesses of SN literature is that while the theory clearly differentiates between
descriptive and injunctive norms (Borsari and Carey, 2001; Borsari and Carey, 2003), little
attention has been paid to evaluating the relative effectiveness of these norms in influencing
drinking behaviour. Several researchers argue that the two types of norms refer to separate
constructs that uniquely affect behaviour. Cialdini and colleagues (1990)
64
for example
emphasize that descriptive norms which describe what is typical or normal motivate by
providing evidence of what will likely be effective and adaptive action. In contrast, injunctive
norms specify what ought to be done through the threat of social sanctions. This distinction
between the two types of norms and their unique effects on behaviour is emphasized by a
number of theories which describe injunctive norms to play an important role in influencing
behaviour. These include the theory of normative social behaviour (Rimal and Real, 2005), the
focus theory of normative conduct (Cialdini et al., 1990; Cialdini et al., 1991) and the theories
of reasoned action and planned behaviour (Ajzen, 1991).
One of the most consistent works in this regard by Cialdini and colleagues (Cialdini et al.,
1990; Cialdini et al., 1991; Cialdini, 1993; Cialdini and Trost, 1998) establishes the
effectiveness of feedback based on injunctive norms in behaviours such as littering (Reno et
al., 1993) and theft of petrified wood (Cialdini et al., 2006). However, most normative
interventions to date have focused on correcting misperceptions related to descriptive norms.
Only two published interventions consider manipulation of injunctive norms in interventions
aimed at reducing alcohol consumption. The findings are mixed. Barnett and colleagues (1996)
report decreases in the perceived approval of alcohol use of close friends and the typical
student by both dormitory residents and Greek members. In contrast, Schroeder and Prentice
(1998) do not report similar changes at a longer term follow-up of 46 months.
While, addressing descriptive norms in interventions has been effective in reducing drinking
rates, it is not clear if one of these norm types would be more likely to change behaviour than
the other and hence should be the preferred choice in interventions. This is an important
65
question because it impacts the effectiveness of SN based interventions. This is an area which
requires further research.
3.9.3 Norm Salience

A key challenge faced by SN research is the need to understand how and why norms of some
reference groups become more influential and salient than those of others. The interest in
salience is driven by a need to design tailored normative interventions for particular groups
(Berkowitz, 2004). In a comprehensive review of prior research, Borsari and Carey (2001)
report the use of as many as 18 different referent groups in SN research ranging from your
best friend to an average student.
The literature on norm salience is very generic in nature and focuses on norms pertaining to
particular identity groups, based on gender (Lewis and Neighbors, 2004; Lewis and Neighbors,
2006b), fraternity and sorority membership (Baer et al., 1991; Baer, 1994; Larimer et al., 1997;
Carter and Kahnweiler, 2000) and athletes (Thombs, 2000). For example, Lewis and Neighbors
(2004) in a study involving under graduate students, evaluate perceptions of gender specific
versus gender-non-specific drinking norms. The study reports that gender specific norms tend
to be more salient than gender non specific norms. Lewis and Neighbors (2006b) provide
empirical evidence that both men and women tend to think of the typical college student as
male when estimating peer drinking norms. In contrast, McAlaney (2007) reports that female
British students interpret the typical student as being the same gender as themselves. There is
also considerable research demonstrating that fraternity members often overestimate the
drinking of other Greeks but correctly perceive that Greeks drink more than non-Greeks (Baer
et al., 1991; Baer, 1994; Larimer et al., 1997; Carter and Kahnweiler, 2000).
66
It is also well documented in the literature on saliency that misperceptions increase as social
distance increases (Beck and Treiman, 1996; Thombs, 2000; Borsari and Carey, 2003; Kypri
and Langley, 2003; Berkowitz, 2004; Berkowitz, 2005) but social groups that are closer tend to
be more influential in shaping personal behaviour (Borsari and Carey, 2003; Korcuska and
Thombs, 2003). Research also indicates that the influence of a referent group on a person is
also dependent on an individuals perception of similarity between him or herself and that
group (Borsari and Carey, 2003).
Most SN campaigns focus on the norms of a typical student at campus which may or may not
be a salient referent group for the population of interest. In case of latter, a SN based campaign
cannot be effective (Berkowitz, 2004). From a theoretical point of view, while it is clear that
proximal social groups are more influential and that evaluation of salience is imperative to the
success of a campaign, in practice, the identification of these groups and their placement in the
social networks of individuals remains a key challenge. One particular area of importance for
SN research therefore is a better understanding of group identity and how individuals identify
salient peers (McAlaney et al., 2011). When asked to report on the drinking of their close
friends or typical students, how do individuals decide which friendship groups to draw from
and how influential are these groups in determining personal consumption? McAlaney and
colleagues (2011) in a recent review of the international development of SN approach, draw
attention towards the need to address these questions regarding norm salience which represent
an important theoretical weakness.
Social Network Analysis (SNA) which is a set of widely used techniques to map and study
relationships in various contexts, offers a unique methodology which can be used to address
67
the issue of norms salience. However, the utilization of SNA in enhancing our understanding
of norm salience remains an unexplored area which requires further investigation (More on this
in chapter 4).
3.10 Conclusion
It is important to understand and address the drivers of unhealthy drinking habits among
youngsters in order to develop appropriate policy measures to tackle the problem. Social norms
represent a core construct of the research on substance use particularly alcohol consumption.
SN literature demonstrates that misperceived peer drinking norms play a prominent role in
influencing the drinking patterns of college students and argues that correcting these
misperceptions can promote healthy behaviours. This chapter reviewed the theory and practice
of SN approach by examining the empirical literature on normative perceptions and drinking
behaviour. In doing so, some important theoretical and practical issues came to light which
provide the impetus and direction for this research. Specifically, the review highlights three
key gaps in the literature. First, it points out that SN approach is yet to be tested for its practical
application in Ireland where the drinking culture is different and more permissive compared to
the US. Second, it draws attention to the relatively little consideration given to the assessment
of injunctive or attitudinal norms as compared to descriptive or behavioural norms in most SN
studies. Third, the chapter highlights that despite pervasive interest in peer influences and
college drinking; SN literature reflects a lack of clarity on how individuals conceptualize
salient peer groups. While, past research on saliency has focussed on how different referent
groups influence perceived norms, identifying these salient others, locating them in the social
surroundings of individuals and determining their relative influence on individual behaviour
68
remains a challenge. Little consideration is thus given to peer-group structure. It is important

that work is undertaken to accurately identify and target salient reference groups because it is
an important part of the planning process of a SN campaign (Berkowitz, 2004).
Broadly speaking, the present study seeks to address the question how different types of norms
and norm salience influence the drinking behaviours of college students in Ireland. In order to
address these concerns, the social context of individuals has to be taken into account which is
characterized by the social networks they inhabit. As mentioned in section 3.9.3, SNA offers a
unique methodology to study relationships and consequently provide a better understanding of
the issue of norm salience. Chapter 4 extends this discussion by introducing SNA, explaining
its relevance to norm salience and reviewing related substance abuse research.
69
Literature Review: Social Network Analysis (SNA)
4.1
Introduction
As discussed in chapter 3, section 3.9.3, SN researchers have consistently drawn attention to

the importance of understanding the social environments of individuals, so that SN marketing
intervention efforts may benefit from focusing on the most salient peer groups corresponding
to the population of interest (Berkowitz, 2004; Gorman et al., 2004; Mason et al., 2004;
Berkowitz, 2005; McAlaney et al., 2011). SNA offers a way to address this concern because it
facilitates studying the social context of substance use behaviours such as alcohol consumption
(Ennett et al., 2006).
Social Network Analysis (SNA) is based on the idea that individuals are embedded in thick
webs of social relations and that social life is created by these interactions and the patterns
formed by them (Borgatti et al., 2009; Marin and Wellman, 2010). Based on the theoretical
constructs of sociology, mathematical foundations of graph theory and recent developments in
computing technology, it offers a unique methodology for visualizing and examining social
structures and relations. The transdisciplinary nature of network science has increased its
popularity and growth in diverse fields ranging from ecology and epidemiology to social
sciences and business practice (Marin and Wellman, 2010).
This chapter is aimed at introducing SNA and examining how it can be useful in identifying
and exploring salient peer relationships that may influence individuals perceptions of peer
drinking and their own drinking behaviours. The chapter begins by discussing issues related to
defining social networks and summarizing the history and the current state of SNA. It then
70
describes three principles implicit in the network theory that set it apart from attribute or group
based perspectives. A description of network data, its measurement and related issues is
presented next followed by a detailed discussion on the integration of SNA and norm salience.
Network research conducted in the specific domain of substance use is reviewed before the
chapter concludes with a discussion on specific challenges and common criticisms faced by
network science. A glossary of common network terminologies is presented in appendix 9.
4.2
Network Theory
4.2.1 What is Social Network Analysis (SNA)?

A social network can be thought of as a set of socially relevant nodes which are connected by
one or more relations (Borgatti et al., 2009; Marin and Wellman, 2010). Nodes (network
members) are the units that are connected by various relations whose patterns network analysts
study. These can be people, organizations or any units that are meaningfully connected to other
units (Wasserman and Faust, 1994). For example, researchers often examine web pages (Watts,
2003), journal articles (White et al., 2004), countries, neighbourhoods and departments within
organizations (Quan-Haase and Wellman, 2006). The relationships, connections or linkages
between a set of nodes is often referred to as ties in network theory. SNA then is a set of
theories, method and techniques used to understand social relationships and how these
relationships might influence individual and group behaviour (Valente et al., 2004).
71
4.2.2 Types of Ties

Network analysts study various kinds of relationships or ties for example, collaborations,
friendships, trade ties, web links, citations, resource flows, information flows, exchanges of
social support or any other possible connection between the nodes (Wasserman and Faust,
1994). Borgatti et al. (2009) divide these different kinds of relationships into 4 basic types
namely, similarities, social relations, interactions and flows.
Similarities occur when two nodes share similar attributes such as demographic characteristics,
attitudes, locations or group memberships (Marin and Wellman, 2010). Similarities are not
seen as social ties in themselves but rather as conditions that increase the probability of
forming other kinds of ties and therefore shape social networks. For example physical
proximity creates opportunities for interactions which can subsequently create social relations.
Social relations include kinship ties, other types of commonly known role relations (e.g.
workmate, friend); affective ties which are based on network members feelings for one
another (e.g. liking or disliking); or cognitive awareness (e.g. knowing) (Borgatti et al., 2009).
Killworth and others (1990) for example examine the networks of people known by the
respondents in their study.
Interactions refer to discrete events which can be counted over a period of time. These are ties
based on behaviour such as those an individual speaks to, helps, or invites into his/her home.
These usually facilitate and occur in the context of social relations. For example, friends (social
relation) give each other advice (interaction) (Borgatti et al., 2009).
72
Flows are those tangible and intangible things that are transmitted through interactions. These
may include relations in which resources, information or influence flow through networks.
Like interactions, flow based relations often occur in the context of other social relations and
researchers frequently assume or study their co-existence. For example, social relationships
such as kinship or friendship can affect the exchange of different kinds of support and
companionship (Wellman and Wortley, 1990).
Much of network analysis is concerned with examining one or more of these different kinds of
ties among actors or nodes. This typology of ties suggested by Borgatti et al. (2009) is
presented in Figure 2.
Figure 2: Typology of ties studied in social networks (Borgatti et al., 2009)
4.2.3 History and the Current State of SNA

SNA finds its roots in several theoretical perspectives because inherently it is an
interdisciplinary endeavour. The primacy of relations over atomized units is an idea that is
much older than the field which came to be known as SNA. Network theorists find examples
of this notion in the works of influential thinkers such as Einstein, Marx, Durkheim, Weber and
Goffman (Emirbayer, 1997). The primacy of relations is most explicit in the work of Georg
73
Simmel who believes that social world is found in interactions and patterns of relationships
rather than in an aggregation of individuals. Based on this belief, he argues that sociologists
should study patterns among these interactions which he refers to as forms rather than
studying individual motives, emotions, thoughts, feelings and beliefs which he calls content
(Simmel and Wolff, 1950; Simmel, 1964; Simmel and Levine, 1971). His theoretical writings
have inspired major empirical findings (Boorman and White, 1976; White et al., 1976; Burt,
1995; Burt, 2005) and are often referred to as the Simmelian Roots of network analysis
(Marin and Wellman, 2010).
The first person to visualize a social network is Moreno. In the fall of 1932, there was an
epidemic of runaways in New York at the Hudson School for Girls at a rate that was 30 times
higher than the norm. Jacob Moreno a psychiatrist along with his collaborator Helen Jennings
mapped the social network at Hudson using a technique called Sociometry which allows
graphical representation of peoples subjective feelings towards one another. Moreno suggests
that the high rate of runaways had an association with the positions of girls in the underlying
social networks that provided channels for the flow of social influence and ideas among the
girls (Moreno, 1953). Morenos sociograms became hugely popular and the recognition that
they could be utilized to study social structures led to a rapid introduction of analytic
techniques (Wasserman and Faust, 1994).
In the 1940s and 50s, the field of social networks progressed along several fronts (Wasserman
and Faust, 1994). One of these has been the formal use of matrix algebra and graph theory to
understand social cognitive concepts such as groups. At the same time experimental studies
picked up pace (Bavelas, 1950; Leavitt, 1951; Wasserman and Faust, 1994). During this time
74
SNA was also being used by sociologists to study the effect of urbanization on communities
primarily using an approach called ego networks (Fischer, 1982) which will be explained in
detail in section 4.3. A similar study of teenage boys and girls in a Mid Western town indicates
that adolescents behaviour was strongly influenced by the cliques11 or groups to which they
belonged (Hollingshead, 2007). The representation and analysis of community network
structure remains to this day at the forefront of network research. In 1970s, the theory of social
capital (Granovetter, 1973; Coleman, 1988; Putnam, 1993; Nahapiet and Ghoshal, 1998;
Portes, 2000) gained immense popularity. The theory reflects the idea that people to whom
individuals are connected and the pattern of connections between these people, enable
individuals to access resources that ultimately lead them to economic and social benefits (Lin
et al., 2001).
By the 1980s, SNA had become an established field with a professional body (INSNA), an
annual conference (Sunbelt), specialized software (UCINET) and a journal (Social Networks).
Today, it has radiated into a number of disciplines. INSNA members having grown from 177
in 1977 to 1200 in 2009 come from a wide variety of fields such as anthropology,
communications, computer science, education, economics, management science, medicine,
physics, biology, political science, public health, psychology and others (Marin and Wellman,
2010).
11
Refer to appendix 9 for the meaning of cliques in SNA

75
4.2.4 The Guiding Principles of SNA

Network analysts make certain assumptions about how best to describe and explain the social
phenomenon of interest. These often contrast from conventional social science research in a
number of ways (Marin and Hampton, 2007). First, network explanations do not assume that
environment, attributes or circumstances affect people or nodes independently. Second, the
concept of uniformly cohesive and discretely bounded groups does not exist in network
analysis. Finally, network analysts take the context so seriously that every so often, relations
themselves are analyzed in the context of other relations.
Attributes Vs Relations
One of the key distinctions between social science research and network analysis is that while
the former is related to examining attributes, the latter focuses on exploring relations.
Attributes refer to the characteristics, attitudes, opinions and behaviours of individuals and
reflect the properties or qualities which belong to them as individuals or groups (Freeman et
al., 1992). It is common for social scientists to collect attribute data through surveys and
interviews and analyse the associations between variables of interest via statistical procedures.
Network techniques on the other hand are most suitable for relational data which refers to the
contacts, ties and connections relating one individual to another and which cannot be reduced
to the properties of individuals themselves.
Networks are Different from Groups
76
Network analysts must in some way specify the boundaries of networks they wish to study.
However; they do not treat network embeddedness as binary, meaning that the nodes are not
considered as belonging only to sets of mutually exclusive groups (Marin and Wellman, 2010).
Groups are different from networks in a sense that they are discretely bounded whereas
network boundaries are fuzzy. What defines a group is the number of members and their
qualities for example employees in different departments, residents of different city districts or
members of different school clubs. Marin and Wellman (2010) emphasize that group
memberships rarely have a uniform influence on members because some members are more or
less committed, more or less tied to other group members, more or less identified with the
group or more or less identified by other members of the group. When group memberships are
treated as having discretely bounded or mutually exclusive memberships, it overlooks the
importance of differing levels of group membership, membership in multiple groups and cross
cutting ties between groups (Marin and Wellman, 2010).
A network essentially represents an association of members via connections and thrives on
diversity rather than uniformity. A network perspective among other things takes into
consideration the strength and nature of connections, the resources that flow between members
and the pattern these ties form. Marin and Wellman (2010) propose three advantages of
examining groups in this way. First, it allows analysts to consider individuals as being
embedded in groups to varying degrees which subjects them to different opportunities,
constraints and influences created by group membership. Second, it allows researchers to
examine variations in group structure such as which groups are more or less cohesive, which
are clearly bounded and which are more permeable. Third, when the question of boundary
77
definition is left open, it allows researchers to move beyond clearly identifiable groups and
study sets of people or nodes that exist at the intersection of various groups. These people or
nodes often act as brokers12 and facilitate interaction between various groups thus structuring
social relations, an idea that is frequently visited by network analysts (Brieger, 1976; Feld,
1981; Blau, 1994; Bellotti, 2008) and will be discussed further in section 4.4.3.
Relations in a Relational Context
Those involved in network research study patterns of relationships rather than examining just
relations between pairs of people or nodes. Marin and Wellman (2010) reflect that while
relations are measured as existing between pairs of nodes, understanding and interpreting the
effect and meaning of a relationship between two nodes requires knowledge of the broader
pattern of ties within a network. Put simply, network analysts are often interested in examining
how A, who is in touch with B and C is affected by the relation between B and C. This is
because the nature of relationship between two people can vary based on their relations with
others. An example is that of siblings. Wellman and Frank (2001) describe that understanding
relations of support, jealousy and competition between siblings requires taking into account the
relationship of each to their parents. Similarly, identification of people in brokerage positions
who may mediate interactions between different groups requires elicitation of the broader
network (Marin and Wellman, 2010).
12
Refer to appendix 9 for the meaning of brokers in SNA

78
4.3
Network Data and Measurement
Network data collection is primarily based on the type of networks and relations researchers
wish to study. Network literature identifies two different ways of studying networks which
stem from two distinct historical traditions (Marsden, 1990) briefly discussed next.
Whole or sociocentric network approach comes from sociology and is heavily influenced by
the work of Simmel. It provides a birds eye view of social structure focussing on all nodes in
the population of interest (Marin and Wellman, 2010). It usually begins from a list of nodes
and includes data on the presence or absence of relations between each pair of nodes. Well
known examples of whole networks are, a network of all workers in a factory showing who
plays games with whom (Roethlisberger and Dickson, 2003) and a network of actors appearing
on film or television showing who has co-starred with whom (Watts, 2003). Sociocentric
techniques are most often used in small communities, schools and organizations where the
boundary of the network can be defined (Valente et al., 2004). They often involve looking at
an exhaustive map of relationships. Carrying out a whole network study involves challenges
related to obtaining data from every member of the population and having every member
provide relational data on every other member. This increases respondent burden and may be
tedious, time consuming and impractical for many studies (Wasserman and Faust, 2007).
Ego centric or personal network approach arises from anthropology and traces its roots to
Radcliffe-Brown among others. Ego centric network data focuses on the network surrounding
one node or focal actor known as the ego and a set of people who share specific relation(s) of
interest with the ego. These people are generally referred to as the alters. Ego networks are
generally elicited through questions focusing on particular relationships (for example, people
79
with whom an individual plays sports or people with whom an individual shares important
matters). Researchers use these questions called name generators to obtain network data. Name
generators are generally followed by additional questions called name interpreters which seek
other relevant information on the alters. For example, a researcher may ask the respondents to
provide the first names of their closest friends and to state whether each of these friends engage
in certain behaviours and whether the respondent engages in these behaviours with each friend.
(Valente and Vlahov, 2001). Name generators, name interpreters and specific issues associated
with them will be discussed at length in chapter 5, section 5.6.
From an ego network perspective, each individual has his/her own network of relationships cut
across many groups which contribute to their behaviour and attitudes. The strength of this
approach lies in its ability to capture the diversity of social environments surrounding
individuals. Amenable to random sampling techniques, ego centric approach typically involves
interviewing specific people. Researchers cannot possibly interview each respondents alters
and must rely instead on the respondents to report their relationships with network members
and the connections between these network members. These ego networks are then treated as
the units of analyses in standard statistical methods. One of the classic examples of ego centric
approach is the General Social Survey an annual population wide survey conducted in the
US, which has been using ego centric methods for over 25 years to extract the network
members with whom the respondents share important matters (Burt, 1984; Marsden, 1987).
80
4.4
Integration of SNA and Norm Salience
The norm change strategy grounded in the SN marketing approach offers a potentially
powerful method to minimize substance use among adolescents as is apparent from the success
of various campaigns discussed in chapter 3, section 3.7. However; SN interventions are
impeded by insufficient attention paid to the processes of norm formation and to the
identification of salient others whom individuals refer to in developing perceptions about the
prevalence and approval of high risk behaviours such as alcohol consumption (Rhodes et al.,
1996; McAlaney et al., 2011). The research on saliency in the context of college drinking
norms is general and does not provide a comprehensive understanding of the social
environment surrounding college students.
SNA is one method of examining norm formation (Latkin et al., 2003) that is argued to expand
our current understanding of the ways in which the behaviours of network members influence
individual youth (Friedman et al., 1997). This is so because SNA is relationally based and
offers techniques to map and visualize peer networks subsequently allowing for the
documentation, illustration and identification of how individuals networks function and how
members interact with one another. In addition, network analysis allows direct identification of
the structural properties of peer relationships providing a means to examine the patterning of
ties in individuals networks and their relationship with individual and group behaviour
(Valente, 2003; Ennett et al., 2006). Reflecting on the above, network analysts often derive
explanations of the content and patterns of ties based on the analysis of composition and
structure of networks, relevant to their field of interest. The following sections discuss network
visualization and how the knowledge about network composition and structure can aid the
81
identification of salient others in ones social surroundings and evaluate their influence on
individual perceptions and consumption behaviour.
4.4.1 Network Visualization and Norm Salience

Network visualization is a key component of analyzing social networks and consists of
presenting network information in graphic format. Visual imagery has played a key role in
network analysis since its inception. Drawing on the pioneering work of Moreno (1953), most
graphic models of networks are presented as sociograms, displaying the relations among
network members in two-dimensional space. Members of the network are represented as points
or nodes, with lines drawn between pairs of nodes to show a relationship between them. Earlier
researchers have argued that points and lines are the most "natural" way to represent social
networks (Klovdahl, 1981). There are several practical advantages to visualizing networks in
this manner which can improve our understanding of why some norms are more salient than
others and how they diffuse through networks. For example, graphic displays of relationships
between people convey a vivid image of the network and provide an intuitive understanding of
structure and dominant groupings that characterize a network, which might be difficult to
achieve in any other way (Streeter and Gillespie, 1993). This graphic representation then
allows researchers to ask and answer questions about networks which might not be statistically
obvious. Modern network software such as UCINET (Borgatti et al., 2002) incorporates layout
and presentation algorithms that facilitate efficient and accurate interpretation of network
graphs.
82
4.4.2 Network Composition and Norm Salience

The analysis of network composition involves examining information related to the attributes
of ego and alters. The sort of variables and topics assessed in a compositional analysis can be
diverse depending on the nature of research. These might include data on socio demographic
characteristics (such as age, gender, ethnicity, religion and so forth), relational aspects (such as
role relations e.g. kin, friend and workmate etc., duration of a relationship) and behavioural
characteristics (such as drinking behaviour and attitudes). Compositional analysis at the most
basic level provides estimates of the size of a particular network determined by the total
number of contacts one has in a certain context. Though simplistic, estimates of network size
are useful to understanding norm salience in various ways. First, they provide a glimpse of the
extent of normative pressures a person might be exposed to. Second, larger drinking networks
are associated with higher levels of alcohol use in college students (Farrow, 2009). Third, this
knowledge is particularly important for studies examining hidden or hard to reach
populations such as injection drug users so that they may be effectively reached. The
assessment of network size makes it possible to measure gross-level descriptors of individuals
networks, such as the percentage of network members who are male or female, who are family
members or friends and so forth. These characteristics provide useful information about the
content of networks being examined.
A typical analysis of network composition involves examining the functions a network
performs by identifying the type of support or interaction provided by the relationships. This is
synonymous with the idea of transmission which is the most common mechanism employed in
network research and which implies that something flows along a network path from one node
83
or member to another (Borgatti et al., 2009). Network research often treats network ties as
pipelines through which many things flow such as information about jobs (Granovetter, 1973),
social support (Wellman and Wortley, 1990), work place identities (Podolny and Baron, 1997),
disease (Morris et al., 1995), immunity to disease (Cohen et al., 1997) and knowledge of
culture (Erickson, 1996). A compositional analysis can identify various kinds of resources
students have access to in their networks. The network literature describes four kinds of
possible resources namely social (e.g. companionship), emotional (e.g. advice), instrumental
(e.g. borrowing money) and informational support (e.g. assistance with college work) (Burt,
1984; Halgin and DeJordy, 2008). The varying levels of exchange of these resources can
provide an explanation of how some members of a network are more salient and important
than the others.
One of the strengths of network science is its potential to measure the strength of relationships
in a given network (Marsden, 1990). Indeed, all members of an individuals network may not
be equally influential in affecting personal consumption (Reifman et al., 2006). Network
analysts make use of several indicators that describe the strength of a tie such as frequency of
contact, duration of relationship, exchange of support and intensity of a relationship. These
indicators may uniquely or in combination define the most salient ties. These indicators
provide knowledge about who is important in a network and subsequently aids the
identification of salient ties within a network. The Social Ecology Theory (Hansen, 1997)
discussed in chapter 3, section 3.3 also supports this by emphasizing that the social units in
which individuals spend the most time and which foster greater bonding are more likely to
transfer existing normative standards and cause adoption of group norms. This inherent ability
of network techniques can make an important contribution to SN theory by offering a way to
84
evaluate the relative strength of different associations in ones social surroundings and
subsequently tease out the most salient ties. A more detailed discussion of the measures of tie
strength and related issues is presented in chapter 5, section 5.6.5.2 .
The literature on norm salience suggests that perceptions of similarity increase the saliency of a
group to an individual (Berkowitz, 2004). This is referred to as homophily in network terms
which refers to the tendency of individuals to relate to people with similar characteristics.
Network literature identifies two kinds of homophily. Status homophily is based on informal,
formal or ascribed status (such as gender, race, age, occupation, education or behaviour
patterns) and value homophily is based on values attitudes and beliefs (McPherson et al.,
2006). Analysis of network composition allows one to examine the degree of homophily in a
network which provides information about whether a person interacts with others like
him/herself. This generates useful insights about the extent of homogenization in a network
which is believed to affect the persuasive impact of a message (Valente et al., 2004). This
knowledge can have important implications for normative interventions. For example, Visser
and Mirabile (2004) in a research involving 4 studies using both experimental and correlational
designs explore the implications of being embedded in attitudinally homophilous networks
versus attitudinally non-homophilous networks. Their findings demonstrate that people who
were embedded within networks made up of like-minded others were more resistant to attitude
change when they encountered a persuasive message than were people in attitudinally
heterogeneous social networks. Reifman and others (2006) note that similar implications are
likely to exist in context of individual drinking behaviour and that of ones network members.
85
4.4.3 Network Structure and Norm Salience

The structural properties describe the way members fit together to form social networks and a
structural analysis is often aimed at uncovering the patterns of relationships within a network
(McCarty et al., 2007). This can improve our understanding of salient norms by providing
useful information about the connectedness of networks, their structural configurations and the
opportunities and constrains associated with different network positions as will be discussed
shortly. Network structure is typically examined visually as well as quantitatively through
matrix algebra.
The use of matrices has become the dominant and preferred approach in network analysis over
the years typically producing algebraic representations of network relations (Streeter and
Gillespie, 1993). While a matrix does not stimulate the kind of intuitive understanding that a
simple visualization can, it offers important advantages of storing diverse attribute and
relational data. This data is typically pictured in varied ways in network visualizations to aid
the analysis. Moreover, matrix data facilitates extensive quantitative analyses and is especially
appropriate for studying structural properties of relationships. A matrix presents a network in
the form of an array of units arranged in rows and columns (Wasserman and Faust, 2007). In a
typical network matrix, the rows represent network members and the columns represent the
same set of members in identical sequence. Each cell in the matrix contains a number
representing the relationship between two members of a network. Binary methods are typically
used, with 1 representing a relationship between two members and 0 representing the absence
of a relationship. These matrices are then subjected to specific routines in specialized network
software generating a range of measures that describe various structural aspects of a network.
86
As mentioned earlier, an important aspect of structural analysis is to determine the extent of

cohesiveness or connectedness between members of a network. Network techniques allow
calculation of measures that describe the degree to which a network is connected (Valente,
Gallaher and Mouttapa, 2004). For example a dense network indicates that most of a persons
friends know each other. Valente, Gallaher and Mouttapa (2004) note that dense or cohesive
networks can contribute to diffusion of norms through peer modelling and peer influence.
These may also reinforce behavioural norms and increase the effects of prevalence
overestimates. Pearson and West (2003) for example demonstrate that the influence exercised
by an individual in a social network context increases with the cohesiveness of the individuals
social network position and the length of time he or she occupies that position.
Another unique aspect of SNA is its ability to categorize people with regards to their position
in a network and to study network shapes. The significance of network position and the ways
in which individuals are embedded in their networks has been a frequent topic of discussion in
network literature since early days (Granovetter, 1973; Burt, 1995; Burt, 2004; Burt, 2005;
Bellotti, 2008) . Network analysts often look towards positions that individuals occupy within
their social networks because it allows to them to examine the extent of grouping or clustering
in networks. Various terminologies such as cliques, liaisons or brokers and isolates are
common in the network literature to refer to various network shapes based on positions of their
members as will be discussed shortly. Network analysts argue that different network positions
a expose individuals to different norms and conventions (Granovetter, 1973; Burt, 1995;
Krackhardt, 1999) . Reflecting on the above, clique members, liaisons and isolates might differ
in their alcohol consumption and in the influence they exert on others (Ennett et al., 2006).
87
This can have important implications in the design and evaluation of normative marketing
interventions.
Cliques for example are subgroups where all members are connected to one another and are
believed to transmit consistent expectations and clear norms (Coleman, 1990; Podolny and
Baron, 1997).
Isolates are people who are not connected to anyone in a network. Network research on
diffusion of innovation which is primarily concerned with the spread of new ideas in a culture
shows that isolates are later adopters of innovations because their position puts them outside
the flow of information about new ideas (Rogers, 1995; Valente, 1996). Similarly, being an
isolate in a high risk setting offers protection because the individual is metaphorically remote
from the negative influences in the group (Valente, Gallaher and Mouttapa, 2004). On the
other hand, there is always a chance that isolates are connected to another group outside the
boundaries of the network being studied, and this other group may put them at risk (Valente et
al., 2004).
Liaisons or brokers are individuals who connect otherwise unconnected or weakly connected
groups in a network, thus playing important roles of mediation. The structural hole theory
proposed by Burt (1995) argues that an individual who acts as a bridge between people in a
social network (A bridge being an individual who is connected to people who are not
themselves connected to each other) will be able to negotiate better agreements, extract higher
rents and in general have more freedom of choice. This is because he has access to resources
that the rest of the group does not. Based on this notion, researchers often suggest that liaisons
may be more resistant to group peer pressure because they are not embedded within a
88
particular group (Valente, Gallaher and Mouttapa, 2004). Thus, in contrast to a bridging
position, individuals embedded in densely connected groups such as cliques, are exposed to
greater constraints in the form of normative pressures (Valente et al., 2004). Granovetters
work (1973) on the strength of weak ties also points to the advantages of occupying a
bridging position though in contrast to Burt who stresses that the strength of a relationship is
irrelevant in this equation, Granovetter predicts that weaker ties will more likely be such
bridges than stronger ties. A third argument put forth by Krackhardt (1999) differs from Burt in
that Krackhardt focuses on the normative power of groups. He argues that if a person A was
simultaneously a member of two cliques, then person A will be subject to two sets of norms
and though he bridges the two cliques, he will be more constrained by normative pressures
from both groups. In consistency with this argument, liaisons may actually be at a greater risk
of substance use because they are exposed to norms of two different groups either of whom
may support substance use (Ennett and Bauman, 1993). Thus, according to this view the more
cliques one is connected to, the more constrained one is because there are more norms to
adhere to.
89
Figure 3: Network visualization showing various positions13
For illustrative purposes, Figure 3 shows an ego network of a hypothetical individual indicated
by the black node labelled Ego. This individuals network has 8 contacts. The ego is
embedded in 2 cliques, which have been encircled. The members in these cliques know each
other but there are no interconnections between the two groups. In fact, the ego acts as a bridge
between the pink clique, the blue clique and person 7. According to Burts (1995) point of
view, this individual is somewhat constrained because of his membership in cliques however,
he is also empowered as he is in a bridging position. However, according to Krackhardts
(1999) interpretation, this individual is embedded in two cliques and hence is most constrained
by pressures to comply to the norms of both groups. There is no isolate in this visualization
13
This network visualization has been generated with data from the present study
90
because this is an ego network where everyone is naturally connected to the focal individual
(ego). However, if we remove ego from the visualization then person 7 is an isolate because
then he will not be connected to anybody in the network.
Network literature provides some guidance on how to resolve this debate. Krackhardt (1999)
suggests that in case of private behaviours known only to group members, an individual is free
to engage in different behaviours in different groups changing as he/she moves from group to
group. In such cases, Burts argument prevails and being part of different groups is not
constraining. However, if the behaviour in question is a public one, then Krackhardts
argument holds that engaging in such behaviours while maintaining bonding with different
groups is more constraining.
Social scientists use the knowledge on network positions to develop an understanding of how
individuals behave in context of substance use. For example, Ennett and Bauman (1993) apply
the network perspective to determine whether adolescents who fill various social positions
characterizing peer group structure differ in prevalence of current smoking. They asked 1092
9th graders to name their 3 best friends. With this information the peer group structure was
determined by categorizing the adolescents into cliques, liaisons and isolates based on the
extent and pattern of their friendship links with others. The study reports that smoking varied
by social position with significantly higher rates of current smoking among isolates than clique
members or liaisons. A number of other studies focussing on social embeddedness also look at
adolescents social position in the peer networks (Pearson and Michell, 2000; Abel et al., 2002;
Fang et al., 2003; Pearson and West, 2003).
91
Valente (2003) emphasizes that although positions matter because these can lead to persuasive
influence by friends, these processes are dynamic and contextual. Therefore, the situational
factors that affect individual behaviour and attitudes often need to be examined. More recently,
interest in the incorporation of qualitative elements in the design of network studies has grown
with the researchers arguing that such study designs can facilitate uncovering and
understanding the contextual factors that may be associated with an individuals opinions and
behaviour (Martinez et al., 2003). This will be discussed in more detail in chapter 5, section
5.6.
The examination of compositional and structural characteristics of networks and their impact
on normative perceptions and drinking behaviours of individuals is a concept which is highly
relevant to the study of salient norms and which can make a novel contribution to SN
literature. However, to date, it is an area unexplored by normative research.
4.5
SNA and Substance Use Research
As commented in section 4.2.4, most behavioural research in the domain of social science has
focussed by and large on individual attributes and characteristics and how they correlate and
result in certain outcomes. However; more recently researchers and practitioners have begun to
realize the need to study contextual factors such as the physical and the social environment
which can contribute significantly to variation in outcomes (Gorman et al., 2004; Mason et al.,
2004). This is particularly relevant to substance use related behaviours. Over the years, SNA
has emerged as an interesting perspective offering a way to investigate the social context of
substance use related behaviours in a transdisciplinary fashion. This section provides a review
92
of studies that have employed network techniques to study substance use behaviours. Valente,
Gallaher and Mouttapa (2004) note that the underlying theoretical basis for this body of
research is largely provided by the concept of peer influence and the notion of birds of a
feather flock together embedded in theories of Social Learning, Differential Association and
Reasoned Action.
4.5.1 Birds of a Feather Flock Together

This is a simple conception based on day to day observations, meaning that similarity breeds
connection. It is popularly believed to structure network ties of every type including marriage,
friendship, work, advice, support, information transfer, exchange, co membership and other
types of relations (Valente et al., 2004). Within network literature, it is synonymous with the
principle of homophily, a concept frequently studied by network analysts. Most substance use
researchers would concur that people who engage in risky substance use are often surrounded
by friends, family members and associates who either do the same or tacitly approve of doing
so. The idea that youth tend to cluster together based on shared activities is supported by
numerous empirical studies whose findings suggest that an individuals substance use is
associated with that of his or her friends for a variety of behaviours such as smoking (Botvin et
al., 1993; Aloise-Young et al., 1994; Bauman and Ennett, 1994; Flay et al., 1994; Urberg et al.,
1997; Unger and Chen, 1999; Alexander et al., 2001), illicit drug use (Windle, 2000; Rai et al.,
2003) and alcohol use (Windle, 2000). Other studies have examined the association between
the number of friends who use substances (Donato et al., 1994; Meijer et al., 1996; Wang et al.,
1997; Jenkins and Zunguze, 1998) or smoke (Wang et al., 1997) and personal behaviour.
93
4.5.2 Theories of Social Learning, Differential Association and Reasoned Action

Many argue and perhaps rightly so, that birds of a feather flock together is a very simplistic
notion. In order to understand the mechanisms underlying the clustering of similar peers and
provide an adequate explanation, network analysts often invoke the Social Learning Theory
(Bandura, 1986). The theory suggests that youth may develop an interest in substance use
behaviour merely from observing others apparently receiving rewards for use.
Differential Association Theory (Sutherland and Cressey, 1974) offers a slightly different
explanation stating that youngsters learn delinquent behaviour such as substance use from
close friends and family who also use substances themselves and/or have supportive attitudes
towards their usage. Further, the theory also suggests that associations with substance using
peers precede the actual substance use arguing that individuals may not tend to model such
risky behaviours from strangers or from impersonal influences.
Theory of reasoned action (Ajzen, 1991) which is also one of the core theories behind the SN
approach posits that behaviour is influenced in part by perceived peer norms - a key postulate
of the SN approach as has been discussed in chapter 3, section 3.6. TRA is also often referred
to in peer influence related network research (Rice, 1993; Rice et al., 2003; Valente et al.,
2004). In fact, two network studies examine norm perceptions and their findings are consistent
with the SN research documenting misperceptions. Iannotti and Bush (1992) report that though
respondents reports of their friends substance use did not correlate well with those friends
self-reports, yet perceived norms were better predictors of behaviour than friends actual use.
Along similar lines though not related to substance use, Valente et al (1997) report that women
94
in voluntary organizations in Cameroon misjudged their friends contraceptive use and that
these perceptions regardless of their accuracy were associated with their own behaviours.
4.5.3 Influence Vs Selection

The theoretical concepts described above assume that youngsters use substances because they
are influenced by similar behaviour of peers. However, network analysts studying substance
use often argue on the basis of longitudinal research that similarity cannot be equated with peer
influence and that adolescents may just select peers based on similar patterns of delinquent
behaviour (Valente et al., 2004). As commented in chapter 3, section 3.7.4, this issue of
causality is also a matter of concern when evaluating the effectiveness of SN theory which
assumes that personal behaviour is influenced by peers. In fact, much of the earlier substance
use related network research has been focussed on disentangling influence from selection
(Hussong et al., 2001).
In this regards, a number of researchers provide support for peer influence (Bauman and
Ennett, 1994; Friedman et al., 1997; Sieving et al., 2000). For example, Friedman and others
(1997) from a cross sectional study of 767 drug injectors in New York city report that being
connected to a large group of people who used drugs was associated with own drug use.
Sieving and others (2000) examine adolescent friendships longitudinally for three years and
demonstrate that over time, higher levels of friends drug use led to increased alcohol use.
Some analysts have looked into the alcohol consumption and social networks of couples
(Leonard et al., 2000; Homish and Leonard, 2008). These studies provide evidence that a
partners drinking has an influence on ones own drinking during the transition to marriage and
through the early years of marriage (Homish and Leonard, 2008).
95
In contrast, other network research provides evidence of peer selection. For example, Donohew
and colleagues (1999) in a longitudinal study conducted over 2 years find that individuals who
were high on sensation seeking tended to select friends who were also high on sensation
seeking, and were more likely to experiment with alcohol, marijuana, and other substances.
Consistent with this assertion, Pearson and West (2003) report findings from three data
collection points of a longitudinal study of secondary school adolescents at a Scottish school
and suggest that people who became substance users transitioned from belonging to non risktaking groups to risk taking-groups. The research on married couples also suggests that
individuals select social networks which are consistent with their drinking behaviours and that
of their partners (Leonard et al., 2000). The study explains that these social networks which
are moderately stable overtime then shape ones drinking patterns.
The majority of network research argues that both the processes of selection and influence are
responsible for similarity of substance use behaviour among peers (Fisher and Bauman, 1988;
Ennett and Bauman, 1994; Engels et al., 1997; Kirke, 2004; Kirke, 2006; Reifman et al., 2006;
Kirke, 2009; Rosenquist et al., 2010). Kandel (1985) provides one of the earliest evidences. In
a longitudinal network investigation of high school students the study demonstrates that
models which included both selection and influence explained initiation into marijuana use
more fully then either factor alone. Ennett and Bauman (1994) use network techniques to study
the contribution of influence and selection to cigarette smoking homogeneity in adolescent
peer groups. Their work comprises two rounds of data collection and assessment of friendship
links and smoking behaviour of 926 8th graders and that of their friends. The results show equal
contribution of selection and influence towards similarity in smoking behaviour of the
participants and their peer groups. Reifman and colleagues (2006) report similar results in
96
relation to college students heavy drinking based on a 3 wave longitudinal study at a large
South Western University in the US. Having found the evidence for both phenomenon of
interest, the study further addresses specialized issues related to each. It reports that greater
presence of network members whom the focal respondent regarded as drinking buddies was
predictive of ones own later drinking after having controlled for potential confounds. The
study also reports that changes in the overall network drinking over time appeared to be driven
predominantly by the dropping and adding of network members.
A recent network study (Rosenquist et al., 2010) provides the most thorough findings in
relation to selection and influence. The findings are based on a longitudinal investigation
examining person to person spread of alcohol consumption in 12,067 people spanning 32
years. The study uses data from the Framingham Heart Study which is a population based
longitudinal, observational cohort study initiated in 1948 to prospectively investigate risk
factors for cardiovascular disease. The researchers provide evidence for both selection and
interpersonal influence up to three degrees of separation (e.g. a persons friends friends
friends). Further, the study also reports that changes in alcohol consumption behaviour of an
individuals social network had a statistically significant effect on ones subsequent
consumption of alcohol. In addition, the study also reports that female contacts were
significantly more likely to influence the spread of heavy alcohol consumption behaviour than
male contacts.
The most comprehensive work in the domain of social networks and substance use conducted
in Ireland comes from a series of articles by Deirdre Kirke based on a dataset collected in
1987. Kirkes work involves total coverage of all adolescents (298) aged 14-18 in one DED
97
(District Electoral Division) in Dublin County Borough (Kirke, 2004; Kirke, 2006; Kirke,
2009). Kirke (2004) demonstrates from three case studies that youngsters adjacent to each
other in peer networks were likely to be similar in their substance use and to form chains of
users of similar substances which she refers to as chain reactions. Her work shows that the
similarity in substance use among adolescents occurred because of both selection and
influence. Kirke (2006, 2009) further explain how chain reactions drew together those who
were similar with those who were not and resulted in similarity of substance use among
teenagers and their peer groups. The studies also elaborate that gender played an important role
in all aspects of chain reactions which involved peer selection, the patterning of peer ties and
peer influence. Specifically, the findings indicate that teenagers generally selected same sexed
peers, males clustered into large more dense networks than females and male teenagers who
were substance users were predominantly influenced by other males while females were
predominantly influenced by both males and females.
Other network studies focus on HIV risk and prevention among drug users. Several studies
focus primarily on sexual and needle sharing networks and identify risk factors associated with
network characteristics (Curtis et al., 1995; Latkin et al., 1995b; Morris et al., 1995; Friedman
et al., 1997). Network characteristics that have been found to shape injection risk and
behaviour include network size and its cohesiveness (Latkin et al., 1995a; Latkin et al., 2003).
Network research also suggests that Greek membership and frequency of discussing particular
alcohol related issues are positively related to drinking in excess among college students
(Dorsey et al., 1999b).
98
To date, most network research related to substance use has collected sociocentric data, a
(near) complete enumeration of the population of interest as described in section 4.3. This
research as is evident from the above discussion has been conducted in both academic and non
academic settings.
4.5.4 Prevention Interventions using SNA

Most of the prevention research which utilizes social network techniques originates from the
concept of diffusion of innovation (Valente et al., 2004). While diffusion of innovation is
mostly related to the transmission of new ideas, researchers often use it to examine the spread
of substance abuse. Early research on the diffusion of innovations demonstrates that opinion
leaders can be effective health promoters (Valente et al., 2004). Opinion leaders are people
who are believed to influence the opinions, attitudes, beliefs, motivations, and behaviours of
others. Several studies recruit opinion leaders via nominations from network members and use
them to communicate healthier behaviours (Latkin, 1998; Valente et al., 2003; Kelly et al.,
2006) . For example, Latkin (1998) recruited street opinion leaders to communicate safe
injecting practices and reports that these opinion leaders adopted the safe injecting messages
themselves and effectively communicated it to others. In school-based studies, there has been a
long tradition of using peer leaders to assist in program delivery (Perry et al., 2002). The
teacher typically collects the data, and selects those students as peer leaders who receive the
most nominations. Valente and others (1999) expand this methodology to allow students to be
assigned to a leader they nominated, or were closest to structurally. Their study demonstrates
that the use of sociocentric methods of selection is more effective and appropriate than leaders
being popularly chosen but being placed in groups defined randomly.
99
Though not related to alcohol consumption, a recent study combines the norm change strategy
embedded in SN marketing interventions with SNA to reduce problem behaviour. Paluck and
Shepherd (2012) examine peer harassment or bullying norms at a small public high school (N=
291) in Connecticut in an experimental study and demonstrate that changing the public
behaviour of a randomly assigned subset of student social referents changes their peers
perceptions of school norms and related behaviours. To date, this is the first randomized
experiment combining SN theory with SNA which necessitates its discussion. One of the
unique aspects of this experiment is that it assesses both types of norms. A social network and
norms survey was administered to the entire school. The survey assessed individuals
behaviour and their perceptions of prevalence and approval of the behaviour among other
students. In addition, the survey also asked the students several (6) name generator questions
aimed at eliciting their friendship ties at the school and nominations of high status14 peers
which utilized complete social network and norms survey at three time points. A complete
network of the relationships among students at the school was then mapped out. Two types of
social referents (1) those who were nominated as being high status and (2) those who received
many friendship nominations were then identified. These students were randomly assigned to
either a control group or intervention group. The intervention participants were then trained to
communicate the campaign message to the entire school during school assembly through skits
and speeches. The survey was repeated a week later as well as at the end of school year. The
study reports decreases in harassment behaviour and increases in anti-harassment behaviour
accompanied by improvements in perceived norms about harassment among students with
14
Two questions, students who you really respect and who you think are most popular were asked to elicit
high status peers.
100
more ties to intervention social referents. In contrast with past SN studies which typically rely
on self reports to assess personal behaviour (McAlaney and McMahon, 2007), Paluck and
Shephards (2012) study assessed personal behaviour and subsequent change by examining
school records for disciplinary actions, teacher reports and purchase of a wrist band indicating
an individuals support for the anti harassment message. Although this study is experimental in
nature and deviates from past SN research in relation to measurement of personal behaviour, it
indicates that SNA is a novel perspective which can be incorporated in norm change strategies.
To date, no such study has examined the effectiveness of SNA in SN campaigns aimed at
reducing alcohol consumption among youngsters. Some network research in consistency with
SN theory acknowledges the fact that pervasive influence of peers can lead to gross
inaccuracies in ones social perceptions of what is normative subsequently affecting personal
behaviour as described in section 4.5.2 (Iannotti and Bush, 1992; Valente et al., 1997; Valente
et al., 2004). However, substance use related network literature clearly lacks studies that
examine norms at network level addressing specific rather than generalized referenced others.
This brings us back to the fundamental question of which referent groups are salient to an
individual or a group of individuals when it comes to behaviour.
4.6
Challenges in SNA
Some general issues bearing on network measurement and study design are presented next.
101
4.6.1 Boundary Specification

A key issue of importance in the design of network studies concerns the specification of
boundaries on a set of units (respondents or nodes) to be included in a network. Network
analysts (Laumann et al., 1989; Marsden, 1990) caution that care must be taken in specifying
rules of inclusion pertaining to both the selection of actors or nodes for a network and the
choice of types of relationships to be studied. This is a concern for both ego network and whole
network studies. In both approaches, the analysis relies on the interrelationship of nodes, hence
omission of a relevant element or arbitrary delineation of boundaries can lead to misleading
results (Barnes, 1979).
Laumann et al (1989), in a review of boundary specification strategies for whole networks,
distinguishes between the realist and the nominalist approach. The realist approach is based on
the subjective perceptions of the participants and assumes that there is a true network of
relationships out there which researchers should uncover. The nominalist approach offers a
more realistic option of using a-priori conceptual frameworks dependent on the research
question. Three procedural tactics are commonly employed by researchers to specific network
boundaries (Marsden, 1990). Those based on the attributes of participants consider
membership criteria set by formal organizations such as schools (Coleman, 1988), or
occupancy of specific social positions deemed relevant by the researcher such as professional
communities (Wellman, 1983). The second tactic relates to using social relations to delimit
boundaries as in snowball sampling procedures (Erickson, 1979). The last approach involves
using participation in certain events as the basis for membership of a network, such as
publication in journals (Brieger, 1976).
102
For ego centric network data, the challenge is to determine the people who should be regarded
as part of a given individuals network. Usually, data is collected on direct contacts such as
ones friends or ones family. In principle one could also collect data on those who are linked
to the participant by one intermediary such as friend of a friend. However, pragmatic pressures
tend to restrict the attention of the researchers to direct contacts only (Marsden, 1990).
Boundaries for ego networks are typically set via one or more name generators that elicit
names of relevant people the participant shares a certain relationship with. These and
associated issues will be discussed in more detail in chapter 5, section 5.6.3.2.
A related issue is the type of tie to be examined. In practice, network analysts tend to focus on
one or more of four types of ties as described in section 4.2.2.
4.6.2 Respondent Burden

One of the challenges of conducting a network study is related to managing respondent burden.
In whole network studies, data about interrelationships between network members is either
collected from the network members through survey, observation or secondary data. In ego
network studies, structural data is collected from the respondents by asking them about the ties
between their network members. McCarty and colleagues (2007) argue that whole network
data collection is high on researcher burden, and low on respondent burden because the task of
collecting data on interrelationships of members is distributed across the participants who the
researcher must observe or interview individually. In contrast, ego network data collection is
low on researcher burden and high on respondent burden because the respondents are required
to provide the researcher with information related to the attributes of each network member
and the presence or absence of a relationship between every pair. This is a key difference.
103
Researchers address these issues by employing various methods to reduce respondent burden
such as asking for fewer alters or collecting detailed relational information on only a few alters
chosen from a larger list (McCarty et al., 2007). Each strategy has its own advantages and
disadvantages as will be discussed in more detail in chapter 5, section 5.6.3 and researchers are
often guided by their study aims in making the choice.
More recently a growing interest in Web-based computer-assisted self-interview (CASI)
applications has demonstrated that such methods can present substantial opportunities for
personal network data collection (Vehovar et al., 2008). Recent studies (Lackaff, 2010;
Lackaff, 2012) employ the web-based Propitious Aggregation of Social Networks (PASN,
http://pro.pitio.us), a survey instrument which reduces this burden by leveraging network data
already available in context of social network websites, and by providing an intuitive clickand-drag interface for survey responses. An experiment conducted on 85 participants using this
tool reports producing networks which were significantly larger and more diverse than those
produced using standard survey methods, yet required significantly lower time investments
from the participants. However, other studies (Matzat and Snijders, 2010) using web based
methods report that while such methods reduce costs and interviewer biases relative to face to
face data collection methods, they produce lower quality data as a consequence of the
respondents answering inattentively, almost mechanically to the questions.
4.6.3 Causality
Determining the exact nature of causal relationships between networks and their effect on an
individuals behaviour or perceptions is a challenging endeavor. Fowler and others (2011), in a
recent publication argue that four assumptions are critical in making causal inferences on
104
network data. First, it is necessary to assume that the network members elicited in a study are
an appropriate proxy for all the peer influences an individual will receive. Second,
identification of peer effects is only possible after assessing network selection based on the
principle of similarity or homophily. Third, it is necessary to assume that the respondent will
appropriately recall, and truthfully describe, the attributes and behaviours of network members
in relation to the area of interest. Fourth, it is necessary to elicit the respondents contextual
influences. A common criticism often raised in relation to observational studies is that
unobserved factors can influence the relationship between the respondent and the network
members.
Though well designed longitudinal studies can be a suitable way of addressing these concerns,
Doreian (2001) notes that there needs to be a very tight coupling of theory, mechanisms, and
credible empirical information before we can delineate the actual operation of causes in the
empirical world and before we can tell causal stories (p111)
4.7
Criticisms of SNA
Within social sciences, network research is subjected to a number of persistent criticisms.
4.7.1 Lacks Theory

The oldest and perhaps the most consistent of these criticisms, is that SNA lacks a theoretical
foundation and that it is merely descriptive or just methodology. Borgatti and colleagues
(2009) describe that this criticism has stemmed from several aspects which are central to
network theory such as the mathematical sophistication of the SNA methodological toolkit, the
105
apparent applicability and portability of the concepts to a wide variety of phenomena - just as
statistics is applicable to many problems and a tendency among some social scientists to
associate all things mathematical with methods. Network analysts (Wellman, 1983;
Wasserman and Faust, 2007; Borgatti et al., 2009) argue that it is indeed one of the strengths of
SNA that it provides excellent methods and concepts that have been shown to characterize
networks and the positions of nodes within them.
4.7.2 Lacks Agency

Another criticism often raised against network research is that it lacks agency in the sense that
it neglects subjectivity and human intentionality (Emirbayer and Goodwin, 1994). This
argument suggests that network science tends to conceptualize nodes as passive, wholly
determined by their positions and environments rather than active agents who are capable of
cognition. Borgatti and others (2009) argue that this is a misunderstanding and that a key
element of social networks is that the nodes are capable of cognition. People are reflective and
projective creatures and this affects how they react to their network positions, and how they
change their network positions in pursuit of their goals. For example, within the field of
management, researchers have shown that individuals with different psychological
characteristics actively construct different kinds of social networks in the workplace (Mehra et
al., 2001). This is in line with the work on social capital the notion that one can invest in
ones ties or position and obtain a return on that investment (Burt, 2005).
106
4.7.3 Ignores Dynamics of Relationships

A third criticism relates to the lack of attention that network researchers are thought to pay to
the evolution of networks; how ties form, are maintained, and decay over time. Borgatti et al
(2009) argue that while this is overstated and much is known about the principles of tie
formation such as the aforementioned management and social capital research, it is also true
that the bulk of work in social network analysis has focused on the consequences of networks
rather than the antecedents. The advocates of network science argue that this is natural in case
of a new field, which must gain legitimacy by showing that it provides new explanations for
existing problems before it explores the antecedents of such variables (Wasserman and Faust,
2007).
4.8
Conclusion
SNA is best understood as a perspective rather than as a theory or a methodology (Marin and
Wellman, 2010). It is based on the idea that social life is created by relations and the patterns
they form. Studying a problem from a network perspective is to study individuals as embedded
in a network of relations and seek explanations for social behaviour in the composition and the
structure of these networks rather than in the individuals. Unlike a theory, SNA provides us
with a way of looking at an issue but it does not predict what we will see.
As described in this chapter, SNA offers a set of powerful techniques to uncover the content
and patterning of ties in a given network. This can help addressing the important gap of norm
salience in SN literature, which has been repeatedly identified as being crucial to the
applicability and success of normative campaigns. The ability to measure large and small scale
107
attributes of peer relations and provide graphic knowledge of peer group structure is one of the
strengths of SNA and an important aspect in developing a comprehensive understanding of the
issue of salience. Further, information related to the content and structure of peer groups can
also help determining where the most salient referent groups are situated in a network and
examining their influence on individual behaviour and perceptions. This is crucial in
evaluating the suitability of a norm based marketing campaign in any given setting.
108
Methodology
Methodology
5.1
Introduction
A detailed methodological review of the procedures adopted in this study and a discussion of
how the approach fits the overall research design are presented in this chapter. Specifically, the
forthcoming discussion serves four key purposes not necessarily in the same order. First, it
formally states the objectives of this research that emerged from the preceding review of
literature and that in essence form the raison d'tre for this study. Second, it describes in detail
the research philosophy and design, the instruments and procedures used to collect data, their
pre testing, administration, rationale and issues of reliability and validity. Third, it draws
attention to some ethical concerns surrounding this study and the appropriate steps taken to
address them. Last, it provides a summary of how the data was analysed in this study and also
discusses some limitations of the methodological approach.
5.2
Research Objectives
As reviewed in chapter 3, section 3.6, the correction of misperceived peer drinking norms has
benefited several health interventions aimed at reducing drinking levels in the American
college system. The success of the approach in the US, backed by encouraging results in other
countries such as the UK, Denmark and Australia (McAlaney and McMahon, 2007; Hughes et
al., 2008; Balvig, 2009; EUDAP, 2010; Ringsted., 2010) has augmented the prominence of SN
theory. There is however, no published empirical work examining the application of SN theory
to alcohol consumption in Ireland. Given the heavy drinking culture of Ireland and the
109
Methodology
associated problems, this seems like a promising avenue for future prevention research and
policy making. However, there is an uncertainty about the success of the SN approach in a
culture which differs in important ways from the US where much of the normative theory was
developed and tested. These include differences in legal drinking age, consumption patterns,
and cultural influences. There is indeed evidence that prevalence of heavy drinking is
positively correlated with the legal purchase and drinking age (Keller et al., 2009). There are
also wider cultural differences in the attitude towards alcohol use and drunkenness in the two
countries. Drinking is a key element of Irelands social life since long as reviewed in chapter 2,
making the Irish much more tolerant and permissive in their attitudes towards drinking. This is
also evident from the latest comparison figures issued in the global status report on alcohol and
health published by the World Health Organization (WHO, 2011) which indicates that the per
capita alcohol consumption of adults in Ireland (aged 15 and above) has been consistently
higher than the US with the average Irish figures for (2003 2005) standing at 13.4 litres of
pure alcohol against 8.4 litres of pure alcohol in the US.
The primary prerequisite of any SN marketing intervention is to establish if the norm for
alcohol consumption among peers is overestimated by the target population and whether it
influences personal consumption (Haines et al., 2004), something which has not previously
been established to be the case in Ireland. Although alcohol related misperceptions have been
evidenced in similar heavy drinking cultures for example the UK (McAlaney and McMahon,
2007), however, local evidence is necessary and recommended in order to justify a possible SN
campaign in any population (Haines et al., 2004). This study serves as a first step to address
this broader question in an Irish context. In line with this, the first two objectives of this study
are:
110
Methodology
1. To investigate the extent of misperception among Irish students regarding peer

drinking norms
2. To examine the association of normative perceptions of prevalence with students own
drinking behaviour
As discussed in chapter 3, section 3.6, previous SN studies have consistently found college
students to overestimate the drinking of their peers. If a similar misperception does not occur in
the Irish population and predict personal consumption, then this will have serious implications
for the applicability of the increasingly popular normative marketing theory in Ireland. Thus
the first two objectives of this study aim to establish if students in Ireland do in fact
overestimate alcohol consumption on campus, as do their US counterparts. It is hypothesized
that this will indeed be the case, that in consistency with past research (Borsari and Carey,
2003; Korcuska and Thombs, 2003), the respondents will overestimate the drinking of other
students on campus, that they will perceive their proximal peers to drink more than them and
their distal peers to drink more than their proximal peers. It is also hypothesized that the
perceived descriptive norms of proximal social groups will be stronger predictors of personal
consumption compared to distal social groups.
When considering normative influence on behaviour, it is crucial to distinguish between the
is (descriptive) and the ought (injunctive) meaning of social norms because each refers to a
separate source of human motivation (Cialdini et al., 1990; Marcoux and Shope, 1997; Rimal
and Real, 2003; Rimal and Real, 2005). To date, majority of intervention efforts have been
directed at correcting misperceptions related to norms of prevalence (descriptive norms) as
described in chapter 3, section 3.9. While this has been an effective method in reducing self
111
Methodology
reported alcohol use, it is not clear from existing research if one of these norm types would be
more likely to change behaviour than the other and should therefore be preferred in SN
interventions. The third objective of this study is aimed at addressing this lack of research
directly comparing the prevalence and approval norms in terms of their influence on behaviour.
The third objective of this research is
3. To study the relative impact of injunctive and descriptive norms on drinking behaviour
In line with the idea that the two types of norms are conceptually and motivationally distinct
(Cialdini et al., 1991), it is hypothesized that both injunctive and descriptive norms will be
uniquely associated with drinking behaviour. If this does happen to be the case then this will
provide a new direction to the existing methods of providing normative feedback to target
populations by offering the possibility of incorporating injunctive norm education in practice.
This will add novelty to intervention based research and strengthen the SN theory by extending
our understanding of the potential roles of injunctive norms as behavioural guides.
Individuals interact with people from different social groups in varied contexts in their day to
day lives. These people may have similar or different norms and often share ties of varying
intensities with the focal individuals. The preceding review of literature (Cialdini et al., 1990;
Cialdini et al., 1991; Berkowitz, 2004; Berkowitz, 2005) points out that some or all of these
people in ones social circle may have different levels of influence on ones thoughts and
actions. While normative research has tended to provide friends or students at the same college
as a referent group, little is understood about how individuals visualize these groups and how
salient these groups are to them (McAlaney et al., 2011). As commented in chapter 3, section
3.9.3, one of the challenges that normative research faces today relates to identifying these
112
Methodology
influential others in the elaborate webs of social networks surrounding the population of
interest and evaluating the relative strength of these different norms in influencing individuals
normative perceptions and their consumption of alcohol. Given that many students at the onset
of college are in the midst of developing new social networks, college settings provide an
opportune atmosphere in which to further this understanding (McAlaney et al., 2011).
As discussed in chapter 4, section 4.2.4, SNA is a potentially suitable method to address this
issue as it is based on the study of relational data. It offers a way to examine the composition
and structure of networks surrounding individuals, which has been shown previously to
provide useful information about the peer context of substance use (Kirke, 2004; Ennett et al.,
2006; Reifman et al., 2006). The fourth and the fifth objectives of this research are aimed at
strengthening the SN theory by using network analysis as a potential methodology to study
salient norms. Specifically, the fourth and the fifth objectives of this research are
4. To identify and locate the salient peers in personal networks of students using social
network analysis
5. To examine the association of personal networks with normative perceptions and
drinking behaviour
It is hypothesized that the use of network techniques will allow extraction of the most salient
ties in personal networks of students; that it will be possible to examine the content and
evaluate the strength of these relationships and that network structure will provide valuable
insight into how drinking behaviours originate, develop and sustain in these networks. It is also
hypothesized that individuals normative perceptions will correspond closely to their personal
113
Methodology
networks and that their own drinking behaviours would be largely similar to their network
members and possibly influenced by them.
If these hypotheses are supported, then this will strengthen the SN theory by providing a new
direction to the study of norm salience. This will also have important implications for
incorporating network science in normative marketing research.
5.3
Research Philosophy
Research philosophy encompasses different research paradigms and clarifies a researchers

position on matters of ontology and epistemology. Ontology is the study of being and reflects
ones views on the nature of existence (Blaikie, 2000). Gray (2009) reflects that two opposing
ontological traditions dominate the western thought. The first tradition often referred to as the
ontology of becoming, places emphasis on a changing and emergent world. In contrast, the
second tradition, also known as the ontology of being stresses on a permanent and
unchanging reality which exists independent of human existence and comprises clearly formed
entities with identifiable properties. Between the ontology of becoming and being, it is the
latter that has held sway in Western philosophy (Gray, 2009). It is also the latter which
provides the foundation for this research.
Underling ontology is epistemology which describes our beliefs about possible ways of gaining
knowledge of social reality, whatever it is understood to be (Blaikie, 2000). Crotty describes
three epistemological positions in his seminal work (Crotty, 1998): objectivism, constructivism
and subjectivism. Only the first two positions are relevant to this research and are discussed
further.
114
Methodology
Objectivism refers to the idea that reality exists independently of consciousness which means
that there is an objective reality out there (Johnson and Onwuegbuzie, 2004; Gray, 2009).
Objectivism is closely related to the theoretical concept of positivism (Gray, 2009).
Researchers taking an objectivist/positivist stance believe that reality can be studied and
presented accurately by human knowledge. Further, for positivists, empirical knowledge alone
counts as true knowledge. Therefore, positivists commonly make use of quantitative methods
as research tools as these are objective and the findings can be generalized and replicated
(Teddlie and Tashakkori, 2008).
In contrast, constructivism refers to the belief that truth and meaning do not exist in some
external world but created instead by the subjects interaction with the world (Gray, 2009). It
means that individuals and groups contribute to the creation of their perceived social reality
(Tashakkori and Teddlie, 2003). A theoretical concept closely related to constructivism is
interpretivism (Crotty, 1998). Researchers taking a constructivist/interpretist stance typically
employ qualitative methods as research tools and unlike positivists they look to understand social
behaviour and focus on its meaning rather than explain it. While constructivism and objectivism
refer to different and contrasting epistemological positions, both are still based on a being
ontology (Gray, 2009).
Researchers have long argued in what has come to be known as the paradigms15 debate,
whether or not qualitative and quantitative methods can be combined in the same study
(Johnson and Onwuegbuzie, 2004). The debate has generated two opposing viewpoints. The
15
A set of beliefs, values, and assumptions that a community of researchers has in common regarding the nature
and conduct of research Johnson and Onwuegbuzie (2004, p 24)
115
Methodology
incompatibility thesis as the name suggests, argues that the integration of quantitative and
qualitative methods is inappropriate due to the incompatibility of the epistemological positions
that underlie these methods (Teddlie and Tashakkori, 2008). This idea has been largely
discredited because scholars have demonstrated that it is indeed possible to successfully
integrate mixed methods16 in research (Johnson and Onwuegbuzie, 2004). Instead, a
compatibility thesis based on pragmatism is suggested which offers a third choice that rejects
the either-or choices from the constructivism-positivism debate (Creswell and Clark, 2007).
Pragmatism embraces the consideration of perspectives from both sides of the paradigms
debate in interaction with the research question and real world circumstances (Teddlie and
Tashakkori, 2008). It is a set of ideas articulated by several people from historical figures such
as Dewey (1948), James (1907) and Peirce (1878) to contemporaries such as Cherryholmes
(1992), Murphy and Rorty (1990). A pragmatic approach employs what works, uses diverse
approaches and values both objective and subjective knowledge (Teddlie and Tashakkori,
2008). Tashakkori and Teddlie (2003) formally link pragmatism and mixed methods research
arguing that both quantitative and qualitative methods may be used in a single study and that
the forced-choice dichotomy between positivism and constructivism should be abandoned.
They further argue that the research question should be of primary importance, more important
than either the method or the philosophical world view that underlies the method. The general
characteristics of pragmatism as described by Johnson and Onwuegbuzie (2004) are
summarized in appendix 3.
16
The class of research where the researcher mixes or combines quantitative and qualitative research techniques,
methods, approaches, concepts or language into a single study. Johnson and Onwuegbuzie (2004, p17)
116
Methodology
The present study employs both quantitative and qualitative methods of data collection and
analysis (More on this in section 5.4). Philosophically, it is rooted in a pragmatic method and
system. It recognizes that knowledge is both socially constructed and based upon the reality of
the world we experience, interact and live in. In conjunction with the idea of pragmatism
(Johnson and Onwuegbuzie, 2004), the present study is oriented towards solving practical
problems in relation to the social norms approach particularly that of norm salience, rather than
on assumptions about the nature of knowledge. The focus therefore is on the consequences of
research as well as on the primary importance of the research questions and objectives which
have been described in the preceding section. It is believed that multiple methods of data
collection and analysis will inform the research questions under study.
5.4
Research Design
Study Layout and Population

As described in chapter 1, section 1.2.1, the population for this study comprises full time Irish
undergraduate students enrolled at DIT during the academic year 2010-11. A primary
difference between the population for this study and those studied in the bulk of American
research (Berkowitz and Perkins, 1986a; Perkins and Craig, 2002; Rimal and Real, 2005;
Rimal, 2008) relates to the lack of on campus accommodation in DIT. Most American studies
have focussed on colleges where students reside in on campus residential halls/dormitories. In
contrast, DIT students generally reside in self catering accommodations or their family homes.
Further, DIT is scattered across the city centre in 6 different campuses in contrast to most
American colleges which have a single campus. Another key difference between the
117
Methodology
population for this study and American college populations is the age at which alcohol
becomes legally available to students (18 years in Ireland compared to 21 years in the USA). In
this regards, the population studied by McAlaney and McMohan (2007) is closer to DITs
population as their study involved a commuter campus based in Scotland (University of
Paisley) where a majority of students reside in privately owned accommodations and the legal
drinking age is similar to Ireland. However, the University of Paisley differs from DIT in
relation to the age17 and gender18 composition of its student body.
The present study comprised two waves of data collection. The first wave comprised a campus
wide online survey for collecting data related to normative perceptions and personal
consumption of alcohol. The primary aim of the web survey was gathering base line statistics
on drinking behaviours of DIT students to assess if they overestimate alcohol consumption on
campus and whether proximal or distal peer norms are more predictive of personal
consumption and hence more salient. The second wave consisted of unstructured in depth
interviews with 26 students who had already participated in the web survey. These interviews
focussed on collecting ego network data and exploring the functioning of these networks and
the meaning behind the ties in order to understand norm salience.
Rationale and Alternative Strategies
The strategy of following a web survey by a SNA component was adopted because the former
meets important preconditions to exploring the phenomenon of norm salience. Establishing
17
18
Mean age in McAlaney and McMahon (2007) is 28 years compared to 21.13 years in the present study
Gender distribution 65% females and 35% males in McAlaney and McMahon (2007) compared to 43% females
and 57% males in the present study
118
Methodology
misperceptions of the campus norm and determining which social group (distal or proximal) is
more salient to DIT students were prerequisites to the second and the innovative element of
this research namely exploring the issue of norm salience via SNA. Therefore, while the web
survey and subsequent statistical analysis presented in chapter 6 tests SN theory in an Irish
context, SNA combined with a qualitative interpretation of the interviews presented in chapters
7 and 8 builds on this knowledge and explores the meaning and functioning of norm salience in
depth.
Some other strategies were also considered in finalizing the design of this study. Among these,
mail surveys and pen/paper surveys were considered as alternatives to the web survey. The
choice of a web survey as a data collection tool was guided by several aspects. First, several
previous studies in normative research have successfully employed online survey methods and
found them to be feasible for college based populations (Kypri et al., 2004a; Kypri et al.,
2004b; McAlaney and McMahon, 2007). McAlaney and McMahon (2007) which is
particularly relevant to this study investigates normative beliefs at a British university using a
web survey. Given the cultural similarities, it was expected that a similar methodological
approach will also be feasible in an Irish context. Second, studies employing web based
methods also report benefiting from lower non response rates and higher quality data at a lower
cost than traditional methods (Kypri et al., 2004a; Kypri et al., 2004b). Third, given the
dispersed campus structure at DIT (more on this in section 5.6.1.1) and the widespread and
regular use of college email system by the students; a web survey was the most appropriate
way of accessing the population largely and quickly. Finally, past research also suggests
several benefits of using web surveys such as greater perceived anonymity resulting in
respondents answering more honestly (Kiesler and Sproull, 1986; Davis et al., 2004) and the
119
Methodology
ability to collect data from a large number of participants within a short time frame and at a
low cost.
A single stage study was also considered as an alternative strategy to a two stage study. Such a
study would have involved conducting a web survey to collect data related to personal
consumption, perceived norms and ego networks simultaneously. There were two key
problems with this approach. First, while this strategy would have captured the basic features
of networks for the entire sample, it would not have been possible to generate as rich and deep
knowledge of ego networks as was achieved through in depth interviews. Second, such an
approach would have imposed unnecessary burden on the respondents. There is indeed
evidence that collecting network data via online tools lengthens the survey, increases the risk
of dropping out and affects data quality (Matzat and Snijders, 2010). Further, studies which
have collected network data via web surveys (Manfreda et al., 2004; Marin, 2004; Coromina
and Coenders, 2006; Vehovar et al., 2008) are not only limited in number but also tend to be
restricted in their coverage of topics in order to reduce respondent burden.
A second online survey was also considered as an alternative to in depth interviews. While on
one hand this approach would have offered greater coverage of the population, the problems
associated with online methods of collecting network data would still have been a concern.
Further, the idea of combining in depth interviews with formal network analysis techniques has
gradually gained traction in network science as commented in chapter 4, section 4.4.3.
Crossley (2010) examines the strengths this interaction offers. He emphasizes that the
combination of the two methods not only assists in generating and reproducing the ties that
constitute a network but also in reproducing various emergent properties such as shared
120
Methodology
meanings and norms/conventions which the quantitative tools of SNA are inclined to overlook
but which qualitative analysis is well placed to identify and analyze. Bellotti (2008), for
example uses standard network techniques and in-depth interviews to reconstruct ego networks
of friends and explores the meanings associated with these relationships, the kinds of resources
they offer and their development and evolution over time. This approach of combining network
analysis with qualitative methods was deemed appropriate because of its ability to tap into the
meaning and significance individuals associate with the ties in their networks in addition to
uncovering the content and the patterning of these ties. Further, as pointed out by Bellotti
(2008), this approach can provide knowledge about the evolution of relationships and their
transition from what they represented in the past to what they represent now for the respondent.
The data collection procedures and considerations for wave 1 and wave 2 are discussed in
detail in the following sections.
5.5
Wave 1: The Web Survey
5.5.1 Questionnaire Design

The formulation of any questionnaire to be used in a survey is an integral part of the research
design process. Several aspects have to be considered depending on the nature of investigation.
The research on survey methodology (Oppenheim, 2000; Cohen et al., 2007) draws attention to
the key elements that should be addressed in questionnaire design. At the most basic level,
important decisions have to be made about the purpose/objectives of the questionnaire, the
population and the sample to be chosen, generating the concepts and constructs that need to be
addressed and the data that is required therein. These choices are guided by the aims of the
121
Methodology
study and the theories to be investigated. The questionnaire used in this research was based on
the knowledge from a large body of SN research reviewed in chapter 3 and the specific
research objectives and proposed hypothesis presented in this chapter in section 5.2.
With regards to the measures used in the questionnaire, closed ended questions were preferred
instead of open ended questions. Cohen and colleagues (1983) describe that closed questions
are useful as they are quick to complete and straight forward to code compared to word based
data generated from open ended measures. These features render them amenable to statistical
techniques of analysis. Further, although several kinds of open, dichotomous or multiple
choice questions can be used in surveys, methodological research suggests some caveats about
the framing of questions (Cohen et al., 1983; Oppenheim, 2000). Specifically, it is
recommended to avoid leading, complex and ambiguous questions in the surveys. In keeping
with these principles, clear and simple language was used in the survey with no indications or
suggestions of an acceptable or a more favourable answer to any of the questions.
The initial questionnaire was subjected to cognitive interviewing to improve the quality of
questions and the design of survey (more on this in section 5.5.2) followed by a pilot study for
further fine tuning (more on this in section 5.5.3). The final questionnaire was designed and
launched using an online survey host website after reviewing several similar platforms as
described in section 5.5.5.
5.5.2 Cognitive Interviewing

In April 2010, three under graduate final year students from DIT Aungier Street filled out a
pen and paper survey followed by a debriefing session. This approach popularly known as
122
Methodology
cognitive interviewing served as a quick way of refining and improving the quality of the
initial questionnaire. Cognitive interviewing has emerged as one of the prominent methods for
identifying and correcting problems with survey questions since the 1980s (Beatty and Willis,
2007). What began through an interdisciplinary effort by survey methodologists and
psychologists has generated a body of methodological research (Jabine, 1984; Dippo, 1989;
Campanelli et al., 1991; Jobe and Mingay, 1991; Esposito and Hess, 1992; Willis, 1994;
DeMaio and Rothgeb, 1996; Campanelli, 1997; Sirken et al., 1999; Willis et al., 1999; Willis,
2004).
The aforementioned research identifies cognitive interviewing as involving administering the
draft survey questions and collecting additional verbal information about the survey responses.
This information is largely used to evaluate the quality of responses and to determine whether
the question is generating the information that the researcher intends (McColl, 2006). Verbal
information is either generated through explicit follow up questions (probes) from an
interviewer or based on general instructions to think out loud as much as possible with little
or no intervening by an interviewer (Willis, 1994). While the think out loud approach has the
benefits of an open ended format and being free from interviewer induced bias, it requires
substantial subject training and places the burden of intensive thinking on the respondents
which may result in divergence onto irrelevant tangents. On the other hand, advocates of a
more probing-centred paradigm argue that probing provides focus and that carefully selected
probes help to concentrate attention on relevant issues (Willis, 1994; Willis, 2004). Willis
(2004) further suggests that the use of non leading probes as is recommended for most
interviews can minimize the interviewer induced bias in such methods.
123
Methodology
In line with the discussed literature, the cognitive interviews conducted as part of this research
were aimed at capturing respondent elaborations regarding how they chose their answers, what
they interpreted the questions to mean and whether they found any difficulties in answering the
survey questions. This technique was useful in two ways. First, it incorporated the opinions of
students which helped in adapting the questions to suit an Irish context for better understanding
and second, it identified ambiguous or confusing questions and unrealistic response options
which were amended accordingly for the pilot study.
A point worth elaborating is the recruitment of just three participants for these interviews. This
is not unusual in cognitive interviewing research. Practitioners generally acknowledge that
participants for such interviews are chosen by convenience and that these interviews are not
intended to be representative of any larger population, but to reflect the detailed thoughts and
problems of the few respondents who do participate (DeMaio et al., 1993).
5.5.3 The Pilot Study

The term pilot study is used in two different ways in social science research. Sometimes it
refers to feasibility studies which are small scale versions, or trial runs, done in preparation for
a major study (Polit-O'Hara and Hungler, 1997). On other occasions, a pilot study entails the
pre-testing or trying out of a particular research instrument (Baker and Risley, 1994). In this
case, the latter was the main goal. The importance of pre-testing or piloting is paramount in
establishing that the design of the questionnaire works in practice and in identifying and
amending problematic questions and refining the instrument (Lancaster et al., 2004). Lancaster
and colleagues (2004) emphasize that problems relating to the content, wording, layout, length,
instructions and coding can be uncovered and dealt with by conducting pilot studies.
124
Methodology
The pilot study for this research comprised 102 students enrolled at DIT Aungier street campus
who completed a web survey through eSurveyspro.com between 4th April and 8th May, 2010.
The pilot survey was not extended to the wider DIT in efforts to minimize discussion of the
topic prior to the actual launch of the final web survey. The students were recruited instead
from pre selected modules through invites sent via their module instructors. They were briefed
online about the survey being part of a doctoral study related to drinking habits of young
people and were assured of confidentiality and protection of their identities. The pilot study
was useful in several ways. First, it helped to test the online format of the questionnaire and
identify inadequacies in design. Second, an optional feedback section asked participants if they
had any difficulty in answering the questions, if there were particular questions that confused
them and if they had any general comments or suggestions to offer. This assisted in further fine
tuning of the measures used. Third, it helped in testing the survey website itself for various
features such as ease of use, maximum responses allowed, reporting, data storage, retrieval and
options of importing the data into statistical packages. Last, the results of the pilot data were
found to be consistent with those from the final survey thus providing confidence in the
reliability of the measures.
One of the limitations of a pilot study as indicated by researchers (Baker and Risley, 1994;
Peat et al., 2002; Lancaster et al., 2004) concerns the problems arising out of data
contamination. Researchers agree that pilot data must not be included with that from the actual
study (Peat et al., 2002). The obvious concern is that if there are problems with the research
tool and modifications are made based on the findings of pilot study, then the data can be
flawed or inaccurate. In keeping with this literature, the pilot data was not included in the main
study.
125
Methodology
Another common problem is the inclusion of pilot study participants in the main study. Here
the concern is that such participants have already experienced the survey instrument and
therefore may respond differently from those who have not previously experienced it. This
issue has long been recognized by the literature and a run in period is generally
recommended when pilot participants are to be re-recruited for the main study (Lancaster et al.,
2004). The duration of these periods will differ depending upon the nature of the research
being conducted. In keeping with these guidelines, the pilot study and the main web survey
were spaced by 6 months which is roughly equivalent to two academic semesters. Less than
2% of the final sample had previously participated in the pilot study. As a further precaution,
an indicator variable was used in SPSS to keep track of the pilot respondents through their
student numbers. Their responses in the main survey and the pilot study were compared on the
variables of interest (specifically, control variables, normative belief items and self reported
personal consumption items) but no discrepancies were observed.
5.5.4 Sampling for the Web Survey

Since actual norms are typically based on the average drinking behaviour of students (Rimal
and Real, 2003; Rimal and Real, 2005; McAlaney and McMahon, 2007; Perkins, 2007), some
homogeneity in the student group is important for a reliable and consistent figure. The study
was therefore limited to one third level institute so that the normative influences would be
generally consistent across the sample. Specifically, DIT was selected because it simplified
several aspects of data collection from gaining access to students to the possibility of
conducting follow up interviews if required. A sample of DIT students was obtained via an
online survey between 2nd November and 25th November, 2010. Assistance was sought from
126
Methodology
the Public Affairs office at DIT which has access to a complete email directory of current
students. An email invitation was sent to all registered students in DIT on behalf of the
researcher explaining the purpose of study, the confidentiality arrangements and a link to the
online survey. In order to maximize the response rate, participants were offered a chance to
win an iPad by entering a draw. Past research provides evidence of higher response rates with
the use of an incentive (Ryu et al., 2006). 60% of the responses were received within the first
two days and by the end of the first week, 85% of the responses had been obtained. One week
later, a reminder was sent out with the assistance of the Campus Life Office, responsible for
providing several integrated services to students at DIT such as health, sports, counselling
services, career advice and student support. The web survey link was placed on the website of
Campus Life Office in the Whats Useful section which updates students on the most recent
happenings on campus. The website is accessed by DIT students on a regular basis for
checking emails and accessing several student services such as time tables and registration
updates. This generated another 15% of the responses in the following 2 weeks. 2115
completed responses were returned. The final sample size that was subjected to analysis,
comprised only full time undergraduate students (N=1700) (More on this in chapter 6, section
6.1). The total population of students at DIT is approximately 20,000 of which, 10,625
registered as full time under graduate students in the year 2010-11 (HEA, 2012), yielding a
response rate of 16%.
5.5.5 Evaluation of Online Survey Tools

Several online tools were evaluated for the preparation and launch of the web survey with the
search narrowing down to three options namely SurveyMonkey, eSurveyspro and
127
Methodology
SurveyGizmo. SurveyMonkey, though was found to be a popular online hosted survey tool,
worked well only for basic surveys. The free version offered only 10 questions and 100
responses per survey with little customization and no options for download of reports or data
import. Only the higher end options offered advanced features such as question/answer piping
(using previous answers in later questions) and integration with SPSS.
eSurveyspro was tested in the pilot study. The free version though offered reasonable features
such as unlimited surveys with unlimited questions and responses, basic piping and
customization; it did not allow direct import of the data to SPSS or Excel which left the
impractical and time consuming option of manual entry thus increasing the chances of data
entry errors. A lot of time was subsequently spent on cross checking the data entries in the pilot
study to ensure that these were error free.
The final web survey was prepared and launched via surveygizmo.com which was found to
have several advantages over the aforementioned options such as ease of use, direct imports to
SPSS and Excel with prior coding, basic piping and fully customizable survey look and feel.
Further, it offered a free student plan for researchers (upon provision and verification of a valid
college email address) which supported all features of higher end advanced packages.
Although the student plan allowed 1000 responses per month, the plan was flexible and
provided an option to get the overflow in responses unlocked and thus prevent any data loss. In
addition, the email support system was found to be very helpful and swift in resolving any
issues pertaining to the preparation and launch of the survey or handling of data.
128
Methodology
5.5.6 Measures
The measures that were used in the final questionnaire and the modifications and amendments
that were made along the way are described next. Please refer to appendix 5 for the final
questionnaire administered in the web survey. The details, sources, rationale and amendments
(if any) made in each measure are included in a spread sheet attached as appendix 2.
5.5.6.1 Control Variables
Several known predictors of personal consumption as informed by the literature were used in
the study. Past research has found age (Jones et al., 1992; McAlaney and McMahon, 2007),
gender (McAlaney and McMahon, 2007; Rimal, 2008), year in college (Perkins, 2002), age of
first drink (Real and Rimal, 2007), money available for drinking (Connolly et al., 1992;
Darmody et al., 2005; Bellis et al., 2007), drinking group size (Cutler and Storm, 1975;
Demers et al., 2002), communication about alcohol (Dorsey et al., 1999a; Jamison and Myers,
2008) and living arrangement (Jones et al., 1992) to be associated with personal consumption.
Age was coded as a categorical variable in the pilot study (see appendix 4 for pilot
questionnaire). 4 response options were provided namely, 18-21 years, 22-25 years, 26-29
years and 30 or above. However, the pilot analysis revealed that using a specific value of age
where possible is more useful, realistic and practical for the analysis. Hence in the final
questionnaire, it was included as an open ended numerical variable.
Gender was included as a dichotomous variable coded 1 for males and 0 for females.
Year in college was included as a 6 point variable coded 0 for less than 1 year to 5 for 5 or
more years.
129
Methodology
Age of first drink was included as a variable with 9 response options coded 0 for never had a
drink, 11 for 11 years or younger through to 18 for 18 and above. Based on feedback from
the pilot survey, an explanatory line other than just a sip was included in the final
questionnaire for clarification.
Money available for drinking in a typical week was an 8 point variable having response
options ranging from 20 or less to more than 140 with an increment of 20 in each step.
For simplification, the variable was coded by choosing the midpoint for each category. Based
on the feedback from the cognitive interviews, an explanatory line was included in brackets to
facilitate the choice of an appropriate response option. The final question was phrased as
How much money is available to you in a typical week for drinking? (Related costs such as
food and taxis included)?
Drinking group size was included as a variable with 5 response options in the pilot
questionnaire namely Alone, 1, 2-3, 4-9 and 10 or above based on Demers and others
(2002). Further, it was contextualized into two questions querying ones drinking group size on
a weekend night and when the respondent had college the following day. However, it was
also felt that contextualizing the measure caused confusion to the students and did not add any
usefulness to the analysis. Also, based on student feedback, narrower response options were
provided in the final questionnaire. Thus, the final measure comprised a single question
phrased as How many people do you usually drink with on a typical occasion?
The composition of drinking group was measured with a single matrix type question which
required participants to select all peer groups they usually drank with. The possible options to
choose from included ones boyfriend/girlfriend, class mates, flat mates, friends from college,
130
Methodology
friends from outside college, work mates, neighbours, siblings, parents and other relatives.
This question helped in formalizing the strategy for wave 2 of the study, as described in section
5.6.
Living arrangement was a categorical variable with 3 response options; I am living with
parents, I am living in a shared accommodation and I am living independently and alone.
A fourth option was added in the final questionnaire namely I am living with partner and/or
dependent others.
The pilot questionnaire examined Frequency of communication about alcohol by first asking a
general question about how often one communicated with the different people they drank with.
131
Methodology
How frequently do you communicate with the people you typically drink with? Please select the
appropriate answer for all categories that apply to you.
Less than
once a
month
Once a
month
Twice
a
month
Once
a
week
Twice
a
week
3-4
times a
week
5-6
times a
week
Everyday
Boy Friend /
Girl Friend
Class Mates
Flat Mates
Friends from
college
Friends from
outside college
Work Mates
Neighbours
Siblings
Parents
Other relatives
This was followed by a series of questions inquiring how often one had talked with the chosen
people about specific drink related topics in the past month. 6 topics were included based on
Dorsey and colleagues (1999a) namely, safer sex practices, effects of drinking too much
alcohol, unwanted sexual advances caused by drinking, binge drinking, physical violence,
injuries or fights and feeling sick as a result of drinking. This was a matrix type question which
is reproduced next.
132
Methodology
Think about the past month. Approximately how many times have you spoken with
the people you typically drink with about The Effects of Drinking too much Alcohol?
Please select appropriate answers for all categories that apply to you.
Never
1-2
times
3-4
times
5-6
times
7-8
times
9-10
times
11 or more
times
Boy Friend /
Girl Friend
Class Mates
Flat Mates
Friends from
College
Friends from
outside college
Work Mates
Neighbours
Siblings
Parents
Other relatives
The question was repeated for all 6 topics.

Respondent feedback indicated that the measure caused confusion, required students to do a lot
of thinking and often returned partially completed and doubtful responses as students perceived
their selves to be reflected negatively if they reported discussing certain topics.
Based on this feedback, the final questionnaire included two questions which are reproduced
below
1. How frequently do alcohol/drinking usually come up in your conversations with others?
133
Methodology
2. How often in a typical week do alcohol/drinking come up in your conversations with

others?
Such items on peer communication have been administered in prior studies such as Real and
Rimal (2007). These questions were accompanied by an explanatory note highlighting that the
question meant casual day to day conversations.
5.5.6.2 Personal Consumption
A variety of measures have been used in the past research to assess personal consumption over
varying time intervals. One approach is the daily diary or the recent recall methods that require
participants to record their alcohol intake on a daily basis or over a specified time period
respectively (Carney et al., 1998; Morrison et al., 2003). Although some researchers argue that
it is the most accurate approach to record drinking, it is one of the most challenging to apply
requiring substantial commitment and motivation from the respondents (McAlaney, 2007).
The QF (Quantity-Frequency) measures typically ask respondents to indicate their frequency
of alcohol consumption and the quantity of alcohol they drink. QF methods, of which there are
many, usually inquire about average or typical consumption patterns, usually over a
specific time period. A variation of QF measures inquires about each major alcoholic beverage
type (i.e., beer, wine, hard liquor) consumed over a certain time period and then summing
across beverage types (Rimal and Real, 2003; Rimal and Real, 2005). The QF approach is not
without reviews and critiques of associated problems (Polich and Kaelber 1985; Room 1990;
L.C. Sobell and Sobell 1992). While, the QF approach cannot capture un-patterned fluctuations
in drinking, researchers agree that it is most useful where information needs are limited to a
134
Methodology
rough estimate of typical drinking and where knowledge of atypical drinking is not required
(Allen and Columbus, 1997; Poikolainen et al., 2002).
The GF (Graduated Frequency) approach is based on the quantity-frequency method but it is
more complex and provides greater details about patterns of personal consumption. It measures
the quantity and frequency of drinking and, in contrast to the QF approach, also measures the
variability in drinking behaviour. For example
Source: McAlaney (2007)

The GF measures are the recommended method of choice for most purposes by WHO (Stockwell and Chikritzhs, 2002). This however, is more of a concern in measuring binge drinking
(McAlaney, 2007), as it has been noted that when asked to report average intake over a past
period respondents tend to ignore occasional episodes of heavy consumption (Gruenewald and
Nephew, 1994; McAlaney, 2007).
The test-retest reliability of both QF and GF measures has been found to be good (Poikolainen
et al., 2002). Although the GF measure escapes many of the limitations that befall other QF
methods, it is at the expense of a much longer administration time (Allen and Columbus,
135
Methodology
1997). By far, the most widely used method of measuring alcohol consumption in normative
research is the QF method (Wood et al., 1992; Baer and Carney, 1993; Baer, 1994; Perkins,
2002; Rimal and Real, 2003; McAlaney and McMahon, 2007; Perkins, 2007; Real and Rimal,
2007; Rimal, 2008). Additionally, Bloomfield and colleagues (2003) note that in countries
where drinking is typically frequent and regular, a simple questionnaire asking how often and
how much people drink (i.e., a quantity/frequency index) may correctly measure most of the
consumption.
In light of the above discussion it was deemed reasonable to use the QF measures in this study.
Respondents were asked to estimate how much and how often they typically drank. In addition,
the students were also required to estimate their frequency of drunkenness. This was done to
allow for heavy episodic drinking and to compensate for any under reporting of alcohol
consumption.
The personal consumption measures used in this study are a modified version of McAlaney
and McMahon (2007) and McAlaney (2007) who examine personal drinking behaviour,
normative beliefs and expectations in a sample of British college students. Given the
similarities between the British and the Irish cultures, it was thought reasonable to follow a
comparable research design and a similar measure of consumption to allow for a comparison
between the results as has been done in chapter 6, section 6.5. The questions in the original
study (McAlaney and McMahon, 2007) were phrased as:
136
Methodology
How many days in a typical month do you normally drink alcohol?

Never or
Less than
very
once a
rarely
month
Once
Once a
2-3 days
5-6
days a
days a
week
week
Twice
a
month
3-4
a month
Everyday
a week
week
How many alcoholic drinks would you normally drink during a night out in a pub or
club?
0
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
How many days in a month do you drink enough alcohol to become drunk?
Never or
Less than
Once
very
once a
Once a
2-3 days
rarely
month
month
a month
week
3-4
5-6
Twice
days a
days a
a week
week
week
Everyday
Based on respondent feedback from the cognitive interviews and the pilot study, three
modifications were made to the initial set of questions.
One, in the measure for frequency of drinking, the response option never or very rarely was
modified into never.
Two, in the measure for quantity of drinking, standard measures for beverage types as
prevalent in Ireland were provided for better estimates. For simplicity, an explanatory line
137
Methodology
stating Remember, 1 pint is the equivalent of 2 drinks; all other beverage is the equivalent of 1
drink each was added to each of the personal consumption and the normative belief items.
Three, the original study (McAlaney and McMahon, 2007) inquired about the number of
drinks consumed in a pub/club during a night out for the measure on the quantity of drinking.
The feedback from cognitive interviews suggested that this was confusing as the general
consensus among these students was that most drinking during a night out takes place at what
they call pre drinking sessions generally arranged at a friends house before heading out to
pubs and night clubs. Note that the commonality of this trend was later confirmed by the
qualitative accounts from the in depth interviews. The pilot questionnaire provided three
contexts in this measure and inquired about the number of drinks one had on a typical
drinking occasion, a typical weekend night and a typical college night (when you have
college the next day). The descriptive analysis did not show much variation between drinking
on a typical occasion and on a weekend and as such did not add any value to the measure.
Hence, the item pertaining to weekend drinking was dropped in the final questionnaire which
simply asked respondents to report the number of drinks they would have on a typical
occasion and on a college night.
5.5.6.3 Perceptions of Prevalence (The Descriptive Norms)
SN researchers argue that measures of descriptive norms on a questionnaire must have certain
characteristics to be valid and reliable (Perkins, 2003a). Specifically, two conditions must be
considered. Firstly, the items querying personal consumption must be paired with
corresponding perceived descriptive norms items and secondly, the perceived descriptive
norms items must be as closely worded to the personal consumption measures as possible. If
138
Methodology
these conditions are not met, any reported discrepancy between personal and perceived
behaviour may simply be an artefact of the questionnaire design. Further, Perkins (2003a)
emphasizes that items measuring perceptions of descriptive norms should be simple focusing
on key aspects of personal consumption so that the respondent may reply quickly and without
extensive cognitive consideration.
A related aspect is the social distance between the respondent and the referent groups
examined. As commented in section 5.2, this study examines the norms of both distal and
proximal social groups to evaluate the variability in perceptions and their subsequent impact on
drinking behaviour. It is also one of the unique aspects of this research as very few studies
have examined both distal and proximal social groups within the same study (McAlaney and
McMahon, 2007).
In keeping with the above discussion, the questions on personal consumption were repeated to
measure respondent perceptions of peer drinking for 5 peer groups namely most students at
DIT (distal peers), close friends (proximal peers), best friend (proximal peers), mother and
father. The choice of most students in DIT as a distal referent group, instead of the more
commonly used term typical student was guided by the need to establish as realistically as
possible, the extent to which students misperceived the norm on campus (DIT).
5.5.6.4 Perceptions of Approval (The Injunctive Norms)
Past studies on collegiate drinking have mostly conceptualized injunctive norms as the
perceived social acceptability of drinking. For example some studies have (Rimal and Real,
2003; Rimal and Real, 2005; Real and Rimal, 2007; Rimal, 2008) asked respondents if they
perceived that different social groups (such as most people in general, society in general and
139
Methodology
the university administration) consider having a drink or two four or more nights a week as
socially acceptable. In contrast, other more recent studies have used multiple items to assess
approval of specific drinking behaviours (LaBrie et al., 2010; Lewis et al., 2010). For example
Lewis and colleagues (2010) examined the perceived acceptance for a range of alcohol use and
negative consequences (15 items) from less severe (e.g., drinking to have fun) to more severe
(e.g., driving a car after drinking) drinking behaviour.
The former may not be equally useful in context of a culture like Ireland where drinking is
known to be socially acceptable. The latter may be accompanied by concerns of longer
administrating time and possible respondent fatigue. Therefore, the most appropriate measure
was deemed to be the one which is quick to consider and assesses not the social approval of
drinking but the social approval of heavy drinking conceptualized as drinking to get drunk, a
colloquial term frequently used by college students in Ireland. The injunctive norms were thus
assessed by querying whether individuals peers would approve of their drinking to get drunk.
The question was repeated for the 5 peer groups mentioned in the preceding section and two
contexts namely, approval in general and when one had college the next day.
5.5.7 Reliability and Validity

Reliability reflects the ability of a research instrument to yield the same results on repeated
trials (Oppenheim, 2000). One method of determining the reliability of empirical
measurements is the test-retest approach in which the same test is given to the same people at
two points of time (Cohen et al., 2007). Reliability is then estimated by examining the
consistency of responses between the two tests. Another way is to estimate the internal
consistency (Cronbach ) of items within a construct. It reflects how well various items
140
Methodology
complement each other in measuring different aspects of the same concept (Cohen et al.,
2007).
Validity defines the degree to which a research instrument measures what it set out to measure
(Oppenheim, 2000). There are several types. Construct validity (subdivided into convergent
and discriminant validity), examines the degree to which the test measures the construct it was
designed to measure (Neuman, 2005). Convergent validity occurs when measures of constructs
that are expected to correlate do so whereas discriminant validity occurs when constructs that
are expected not to correlate do not, such that it is possible to discriminate between them
(Neuman, 2005). Content validity occurs when the research instrument provides adequate
coverage of the subject being studied (Litwin, 1995). Its determination is principally
judgemental and intuitive which cannot be expressed numerically. It can be determined by a
panel of experts who can judge the extent of the measuring instruments standards (Litwin,
1995). Criterion validity examines the ability of a measure to predict a variable designated as a
criterion (Neuman, 2005). Sub divided into concurrent (other criteria assessed simultaneously)
and predictive validity (predicting future or past events), it is typically expressed as the
coefficient of correlation between test scores (Litwin, 1995). External validity occurs when
research results can be generalized to and across population, settings and times whereas
Internal validity occurs when conclusions related to cause and effect can be made (Proctor and
Capaldi, 2008).
The present study relied on self reports of students alcohol consumption. Much research about
alcohol related issues in the college environment relies on self-reports of students alcohol use
(Sobell and Sobell, 1995). The validity and reliability of self reported alcohol consumption has
141
Methodology
been studied extensively. Consequently, a substantial body of empirical research shows that self
reports are accurate proxies for objectively measured behavior both within (Midanik, 1988;
Harrison and Hughes, 1997; Johnston, 2001; Brener et al., 2003; Lintonen et al., 2004) and
out-with (Ramo et al., 2011) the alcohol consumption domain.
The present study utilized QF measures of self reported alcohol consumption the reliability
(test-retest) and validity (content, construct and criterion) of which is evidenced in prior
psychometric research (Romelsjo et al., 1995; Sobell and Sobell, 1995; Rehm et al., 1999;
Room, 2000). The ecological validity of these measures is also established in similar
(McAlaney and McMahon, 2007) and different settings (Perkins, 2007). The present study
estimated the internal consistency reliability of personal consumption measures by examining
the Cronbach . The Cronbach for the three items on personal consumption was found to be
0.735 which demonstrates good reliability.
The reliability and validity of SN measures (perceptions of descriptive and injunctive norms) is
an important and complex issue in the field created in part because these measures tend to be
tailored to the target population they are being used with. It also represents an under researched
weakness of the field. However, in general several studies show participants estimates of the
amount of alcohol consumed by their peers to be highly correlated with their own drinking
(Wood et al., 1992; Perkins, 2002; Kypri and Langley, 2003; Linkenbach and Perkins, 2003a;
Neighbors et al., 2006; McAlaney and McMahon, 2007; Perkins, 2007). In accordance with
theory and past research, the current study also found high correlations between items of
personal alcohol consumption and perceived norms (chapter 6, section 6.3) thus demonstrating
construct and criterion validity. In addition, the present study found modest correlation
142
Methodology
between descriptive and injunctive norms (r = 0.28 or less) indicating that these constructs are
conceptually distinct. It adds to the evidence supporting the discriminant and convergent
validity of descriptive and injunctive norms as has been noted in prior research (Grube et al.,
1986; White et al., 1994; Sheeran and Orbell, 1999; Rivis and Sheeran, 2003). Hierarchical
multiple regression procedure utilized in the present study revealed that perceived descriptive
norms contributed a substantial proportion to the variance19 in personal alcohol consumption
after several known confounding factors had been taken into account. This strengthens the
evidence of concurrent validity of the descriptive norm measures used in the study. The high
values of Cronbach for both descriptive and injunctive norm measures ( = 0.772 and 0.756
respectively) provide confidence in the internal consistency of these measures.
The sample for the web survey was found to be representative of full time undergraduate
student population at DIT in terms of gender and age. The information about student
enrolments and demographics was obtained from the Higher Education Authority (HEA,
2012). Chi square goodness of fit analysis revealed no significant differences between the
sample and population on these factors (p>0.01). Further, literature on issues related to nonresponse (Miller and Smith, 1983; Lindner et al., 2001) suggests that non-respondents tend to
be similar to late respondents in responding to surveys. In keeping with this literature, the
participants who responded to the first email invitation to the web survey within a week
(responses received till 8th Nov, 2010) were labelled as early. Similarly, those who responded
to the reminder invitation to the survey (responses received on or after 9th Nov, 2010) were
19
Perceived descriptive norms explained 20% of the variance in frequency of drinking in a typical month, 33.4%
of the variance in number of drinks consumed on a typical occasion and 20.9% of the variance in frequency of
drunkenness in a typical month. Detailed results can be found in chapter 6, section 6.4.
143
Methodology
labelled as late. Independent sample t tests were performed to compare early and late
respondent groups on key variables (attached as appendix 6). No significant differences were
found between early and late respondents indicating that the findings of the web survey can be
reasonably generalized to the population. The aforementioned tests and subsequent findings
strengthen the external validity of the present study.
The design of this study is cross sectional in nature which is why conclusions of cause and
effect cannot be made with certainty. This limits the internal validity of the study (More on this
in section 5.9.1.
5.6
Wave 2: In Depth Interviews
5.6.1 Sampling and Related Issues

5.6.1.1 Campus Structure at DIT
The Dublin Institute of Technology (DIT) was established officially in 1993. It was constituted
from six higher education colleges of the City of Dublin Vocational Education Committee
(CDVEC) (Duff et al., 2000). These colleges, located in the city centre both south and north of
the River Liffey, had provided applied and higher vocational education in the areas of
technology and business for over a century and in their new form came to be known as DIT.
While plans are underway for the consolidation of DIT to a new premises located at
Grangegorman, to date the institute comprises 6 separate campuses it originated from. The
following map shows the location of these campuses around the Dublin City encircled and
marked in red.
144
Methodology
Figure 4: Location of DIT Campuses around Dublin City
The following table describes the distribution of various academic disciplines across these
campuses
Aungier Street Campus
Kevin Street Campus
Bolton Street Campus
Cathal Brugha Campus
Mountjoy Square
Campus
Rathmines Campus
College of Business
College of Sciences and Health
College of Engineering and Built Environment
College of Arts and Tourism
College of Sciences and Health
School of Art and Design
School of Social Sciences and Legal Studies
College of Arts and Tourism
Table 1: Colleges at DIT
145
Methodology
5.6.1.2 Recruitment of Subjects

Participants for the second stage of this study were recruited from the Aungier Street campus,
which is one of the largest campuses of DIT. The choice of campus was aided by a preliminary
analysis of the data from stage 1. Three key results provided impetus for selecting Aungier
Street campus. One, 32% of the sample from stage 1 came from the Aungier Street campus
making the response rate from this campus the highest among all. Two, descriptive analysis
suggested that the gender distribution among the respondents from this campus (as extracted
from the main sample) was quite balanced (48% male Vs 52% female) as opposed to other
campuses. The Bolton Street and the Kevin Street for example are primarily male campuses
with the analysis reporting 87% and 70% males respectively whereas the Cathal Brugha Street
and the Mountjoy Square campuses are predominantly female with the survey reporting 71%
and 75% females respectively. These results were generally consistent with the records of the
registrations office at DIT. The Rathmines Road campus only comprised 1.6% of the sample
and was therefore not considered for stage 2. Last, the actual drinking norm at the Aungier
Street campus as deduced from the self reports of respective respondents was consistent with
the actual drinking norm of the main sample for all three dimensions of alcohol use. The
following table shows the mean values for all campuses.
146
Methodology
Frequency of drinking
(days in a typical month)
No. of drinks
Frequency of drunkenness
(typical occasion)
(days in a typical month)
Main Sample
4.91
8.68
3.26
Aungier St
5.06
8.65
3.44
Bolton St
5.12
10.21
3.47
Kevin St
4.51
8.49
2.76
Cathal Brugha
5.01
7.74
3.29
Mountjoy
4.66
7.24
3.24
Table 2: Mean alcohol consumption across DIT and individual campuses
Based on these findings, it was inferred that the attributes specifically gender and drinking
rates of the respondents from this campus best depicted the main sample.
Next, 545 potential people from Aungier Street campus were categorized into three cohorts
according to their drinking intensity. This categorization was based on Beck and Treiman
(1996) which provides a criteria of low, moderate and heavy drinking individuals. The
following table presents the categorization schema as laid out in the original study.
Low intensity drinkers

Moderate intensity
drinkers
Frequency of
No. of drinks
Frequency of
drinking
(per occasion)
Drunkenness
Once a month or less
2 drinks or less
Never get drunk
Exceeded low intensity drinkers on one or more of these

items but did not meet the criteria for high intensity drinkers
High intensity drinkers Once a week or more
5 or more drinks
Once a month or
more
Table 3: Drinking Intensity Criteria (Beck and Treiman, 1996)
147
Methodology
Two modifications were made to ensure consistency with the response options provided in the
web survey and appropriateness within an Irish context. The frequency of drunkenness for low
intensity drinkers was classified as once a month or less frequently as opposed to never and
that for high intensity drinkers as once a week or more frequently as opposed to once a
month or more in the original study. These amendments were considered necessary because
the original study was carried out in the US where the criterion for low, heavy and moderate
drinking might not necessarily be the same as in Ireland which is known to be quite tolerant of
alcohol use and drunkenness. These modifications were made in the light of Delaney and
Harmon et al (2007), a mixed methods study conducted by University College Dublin on the
perceptions of excessive drinking by Irish college students.

Moderate intensity
drinkers
High intensity drinkers
Frequency of
drinking
No. of drinks
(per occasion)
Frequency of
Drunkenness
Once a month or
less frequently
2 drinks or less
Once a month or
less frequently
Exceeded low intensity drinkers on one or more of these

items but did not meet the criteria for high intensity drinkers
Once a week or
more
5 or more drinks
Once a week or
more
Table 4: Drinking Intensity Criteria for the Current Study
Further, those who met the criteria on any one item for high intensity were also classified as
high intensity drinkers e.g. someone who drinks 2-3 days a month, consumes 9 drinks on an
occasion but believes that he/she never drinks to get drunk was also classified as a heavy
drinker.
148
Methodology
5.6.1.3 Sample Size

6 email lists of potential participants were prepared based on random sampling. The invitations
to take part in the interview were sent out to one list at a time followed by a reminder email a
week later. A total of 26 interviews were conducted between 28th January and 17th Feb, 2011.
The cessation of interviews occurred in response to theoretical saturation. Specifically, after
the 20th interview, there was saturation in the topics where additional interviews were
producing the same themes and material regarding drinking norms and their development and
sustainment in students' networks. A related point is that the goal of these interviews was not to
reach statistical representativeness but to understand the functional aspects of the ego networks
and examine how the composition and structure of networks contributed to the formation and
dissemination of drinking norms within them.
Small samples like this are typical of qualitative studies, where the goal is not to reach
statistical representativeness but to explore theoretical aspects of a social phenomenon (Curry
et al., 2009). For example, Bellotti (2008) is a qualitative ego network study of a non
representative sample of 23 single youth in Italy. Carpentier (2007) and Grosser (2010) are
other examples of qualitative network studies with non representative small samples of n = 49
and n = 30 respectively.
5.6.2 The Ego Network Approach

As discussed in chapter 4, section 4.3, most social network data collection can be classified as
either whole or egocentric networks. For the stage 2 of this research, the egocentric
approach was deemed more suitable for two key reasons. One, whole network studies are
149
Methodology
based on the assumption that actors be regarded for analytical purposes as bounded social
collectives (Marsden, 2005). This means that a network has boundaries which are known a
priori such as all members of an organization, all students in a college etc. The only reasonable
boundary that could have been used in this study was the college itself. Its not only difficult
but also impractical to study all possible ties in a sample as big as this let alone within a
college of more than 20,000 students. The egocentric approach focuses instead on a set of
selected individuals or egos where each person has his or her own network of relationships
that may traverse to different groups. These relationships contribute to an individuals
behaviour and attitudes (Christakis and Fowler, 2009). Two, students socialize and
communicate with a range of people who might not necessarily be from within the college
settings. In fact, one of the matrix type questions on the web survey required respondents to
select the people that they typically drink with given 10 peer group choices as explained in
section 5.5.6.1. In response to this measure, 86% people reported having friends from outside
college among the people they typically drink with. This finding supports the choice of ego
centric approach as being appropriate in the current context. The strength of egocentric
analysis lies in its ability to capture such diversity of a social environment which is not only
relevant but perhaps critical to develop an understanding of how and why some ties are more
relevant and salient in a persons network.
5.6.3 Interview Design

5.6.3.1 Settings and Protocol
Participants were contacted by email which they had provided during the online survey to set
up a convenient time for an interview. Each interview lasted anywhere between an hour and a
150
Methodology
half to two hours. The participants were advised to bring along their student cards for
identification. All interviews were conducted at DIT Aungier Street campus in pre booked
rooms and audio recorded in Audacity (Mazzoni and Dannenberg, 2011), a software
application that allows digital recording of sound. The basic format and sequence for the
interviews is presented below.
1. The overarching aims of the research and related ethical concerns (section 5.7) were
addressed in a 10 minute discussion preceding the actual interview.
2. The name generators were administered as will be explained in section 5.6.3.2.
3. Permission was sought to audio record the interview and an opening cue prompted the
participants to talk about the people that they named in response to the name generator
questions as will be explained in section 5.6.5.
4. Towards the end of the interview, the participants were required to fill up the adjacency
matrix (section 5.6.5.1) and an alter attribute chart (section 5.6.3.3). Finally, a short
activity comprising a concentric circles diagram as will be described in section 5.6.5.2
concluded each interview.
5. The participants were offered a cash incentive of 15 each for their participation in the
interview.
5.6.3.2 Name Generators
The technique for gathering data on personal networks generally consists of using a name
generator, which is a tool that uses a question(s) to produce names of people who share a
particular relation with the participant (Carpentier and Ducharme, 2007). Name generators can
151
Methodology
emphasize different aspects of a network (e.g. friendships, resource flows, information flows,
exchanges of social support), but the social-support perspective is the one most often used.
Researchers describe social support as an interactive process through which emotional,
instrumental, or financial assistance is obtained from ones social network (Wasserman and
Faust, 2007). Depending on the name generator and the selection criteria utilized (e.g., people
one takes advice from or people one plays sports with), the network will assume a different
size, form, and content. Since it is not possible for a researcher to study all contacts in an
individuals social surroundings, the name generator technique offers a way to extract a
fraction of respondents social contacts who are relevant to the research question (Marsden,
2005).
The key considerations then are to decide on the number of alters to be elicited, the use of one
versus multiple name generators and the choice of a criteria in generating alter names.
McCarty and colleagues (2007) suggest that free recall of 25 alters will capture the same
structural pattern as a network of 45 alters. For most ego centric studies where the focus in on
eliciting strong ties, free recall questions that place no definite upper limit on the overall
network size have been found to be generally useful (Ferligoj and Hlebec, 1999; Carrasco et
al., 2008).
While administrating a single name generator has the benefits of saving time and reducing
respondent burden, researchers examining the quality of network data argue that the use of
multiple name generators where possible is the preferred method of choice in most cases
(Marin and Hampton, 2007; Bidart and Charbonneau, 2011). This is because the use of a single
name generator induces people to remember the most obvious people whereas it may be
152
Methodology
relevant to prod memory into searching for significant persons not easily remembered (Bidart
and Charbonneau, 2011). For example, an individual may not be able to recall relevant ties
because they live far way or because the individual hasnt seen them for a long time.
The choice of criteria in a name generator is generally based on the nature of study. For
example two classic studies on ego networks use name generators identifying people with
whom the focal actor discussed important matters with in the past 6 months (Burt, 1984) or
with whom he or she frequently socialized with (Fischer, 1982). Many others follow suit.
On the whole, research indicates that people are generally better at recalling typical or routine
relationships and interactions than they are on transactions that occur within highly specific
time frames (Marsden, 1990).
In light of the above discussion, the following name generators were administered in this study.
1. From time to time people discuss important matters with other people. Who are the
people with whom you discuss matters important to you? Give me their first names and
last initial.
2. Who are the people you really enjoy partying/socializing with? Give me their first
names and last initial.
3. Please list anyone who is especially close to you, who you have not listed in one of the
previous questions. Give me their first names and last initial.
Naturally the outcome of such questions is dependent upon the habits of sociability in a
particular culture (Ruan, 1998; Bidart and Charbonneau, 2011). Given the magnitude and
153
Methodology
significance of drinking in the social life of Ireland, it was expected that these name generators
will elicit significant alters relevant to the phenomenon of interest i.e. drinking behaviour.
5.6.3.3 Name Interpreters
Name generators are generally followed by a series of name interpreter questions designed to
elicit information about the network members such as their attributes, properties of the
relationship they share with the ego and their ties with other named alters (Wasserman and
Faust, 2007). One of the key considerations in designing and administrating name interpreter
questions is to balance respondent burden and the information sought (Marsden, 2005). An
important decision that needed to be made in this regards relates to whether the number of
alters examined in the name interpreter questions be limited or not? The General Social Survey
(Burt, 1984) and McCallister and Fischer (1978) elicited alter data for only the first 5 and 8
people named respectively as a concession to time constraints. Burts reasoning for this is
quoted:
Somewhere between the time saving choice of asking too few names and the unacceptably
time consuming choice of asking too many names, is that number of alter names which reaches
the border of the respondents interpersonal environment to reveal social heterogeneity. We do
not know where this point is With no guide other than common sense, I expect the
corresponding upper limit to be a single digit number. There is evidence to suggest an upper
limit of five to seven (e.g. Miller (1956) and Simon (1974)). In the interest of increasing
measurement precision and decreasing measurement bias - under a severe time constraint the five alter limit proposed for seems judicious. (p 315)
154
Methodology
In context of the current research, it was deemed adequate to collect information on all alters
for two reasons. First, the studies mentioned above, were questionnaire based. Guided by
methods adopted in the past research (Bellotti, 2008), it was anticipated that the conversational
nature of the in depth interviews and the opportunity to ask for clarifications will make the
process less burdensome for the participants. Second, the Irish love of drinking and
conversation is, of course, proverbial and has earned them a reputation as one of the most
talkative and entertaining folk (Haining, 2003). The pilot interviews revealed that having an
hour or two of conversation related to ones social circle and drinking was not by any means
boring for the participants. Further, it was the beginning of the semester; students were
generally relaxed with no impending projects or class assignments that required immediate
attention.
In order to maintain the natural flow of conversation and avoid unnecessary disruption, the
name interpreter items related to basic attributes of the alters (age, gender, occupation, city of
residence, relation and time known) were presented to participants in the form of a chart at the
end of the interview. Most participants were able to complete the chart fairly quickly.
5.6.4 Interview Structure

In depth interviews are generally divided into three categories based on the degree of
structuring and the degree to which participants have control over the process and content of
the interview: structured, semi-structured, and unstructured interviews (Fontana & Frey, 2005,
Cassell, 1980; Morse, 2002).
155
Methodology
A structured interview is an interview that has a set of predefined questions, administered in

the same sequence and the same wordings to all the participants (Corbetta, 2003). This
standardization is intended to minimize the effects of the instrument and the interviewer on the
research results and is essentially similar to surveys except that structured interviews are
administered orally rather than in writing. There is an obvious danger of introducing rigidity in
the interview following this technique by adhering too closely to the predefined format and
subsequently being unable to incorporate appropriate probing (David and Sutton, 2004).
Consequently, themes unrelated to the interviewers focus may not emerge.
Semi-structured interviews are more flexible. Though channelled by an interview guide, the
interviewer has a certain amount of room to adjust the sequence of questions to be asked and to
add questions based on the context of participant responses. Corbetta (2003) describes that
within each topic, the interviewer is free to establish his own style of conversation by
conducting the dialogue as he thinks fit, asking the questions he deems appropriate and
requesting clarification if the answer is not clear.
5.6.4.1 Pilot Interviews
The flexibility offered by unstructured interviewing technique prompted its use in the two pilot
interviews conducted on 18th and 19th of November, 2010. The purpose was to test the
interview design, identify weaknesses in format and make appropriate changes for the final
version. Both interviews lasted about an hour and a half each. The participants were
volunteers, a male and a female doing a degree in marketing at DIT Aungier Street and aged 29
and 23 respectively.
156
Methodology
The pilot interviews differed from the final interviews in their approach towards administrating
the name interpreter questions. The semi structured technique offered a flexible way to collect
some data about all alters yet use probing where necessary to expand on a topic of interest.
Both interviews began with some basic questions about the participants to help them get
talking and feel less apprehensive about the whole process. The interview guide broadly
comprised three types of questions; those related to the attributes of the named alters
(demographics, shared activities and relationship with ego), those related to assessing the tie
strength (duration of relationship, frequency of communication and how close ego felt towards
each alter) and those related to the drinking behaviour of the named alters (frequency and
quantity of drinking and approval of drinking to get drunk). Information about each alter was
sought in the same sequence as the participant had named them.
Two shortcomings were noted in following a semi structured approach. One, the monotony of
responding to a similar line of questioning for all the people that one had named was tiresome
and boring for the respondents and reduced their focus and interest as the interview progressed.
Two, in efforts of collecting some information about all network members limited the
opportunities of following threads that emerged during the conversation. In order to overcome
these issues, the final interviews were unstructured thus allowing for a more flexible and
conversational format.
5.6.5 The Main Interviews

The final chosen technique for the in depth interviews was the unstructured interview
approach. Unstructured interviews are more flexible and casual than the aforementioned
techniques. Minichiello and Helms (1997) define them as interviews in which neither the
157
Methodology
question nor the answer categories are predetermined. Instead, such interviews rely on social
interaction between the researcher and the informant. Punch and Punch (2005) describe
unstructured interviews as a way to understand the complex behaviour of people without
imposing any a priori categorization, which might limit the field of inquiry. That being said,
such interviews are not completely non-directive and are generally guided by an aide memoire
which provides direction on issues of interest.
In each interview, an opening statement encouraged the participants to talk about the people
they named after the administration of the name generator questions. The opening statement
said,
Now that we have a list of names here, I want you to tell me about these people. You may start
from anywhere you like and take your time. Ill listen first and try not to interrupt you. I might
take some notes.
Note taking was aided by a tabulated chronicle that recorded names of alters, key themes,
situational contexts and specific phrases and expressions in the order mentioned by the
participants as they spoke. This allowed the researcher to generate further questions, cues and
probes in response to the interviewees initial narration following the same order in which the
participant spoke about the alters. After the initial narration, the interview took the form of a
conversation focussing on developing an understanding of the interviewees social
surroundings from their perspective.
Though the researchers control over the conversation was minimal, yet the conversation was
guided by a path or an aide memoire as is the case in most unstructured interviews
(Corbetta, 2003; Punch and Punch, 2005). The path served as a tool to organize the items to be
158
Methodology
investigated and keep track of interviewee narration (appendix 8). There were two main themes
of interest; the nature of relationships ego shared with network members and the drinking
behaviours of network members. Within these main themes the path aimed at generating
narratives related to several sub themes for example formation (how did they get to know each
other, common past experiences and common relationships) and strength of the relationship
(duration known, frequency of communication, intimacy, common shared activities, meaning of
important matters and exchange of support), drinking behaviours and attitudes, their
development and reinforcement and ways of socialization. The use of the aide memoire or the
path encouraged a certain degree of consistency across different interviews.
Probing was used to elaborate on topics of interest which is not unusual in unstructured
interviews. Researchers often probe, or ask for clarification during the course of unstructured
interviews (Fontana and Frey, 2005), the purpose of which is to deepen the response to a
question and to give cues to the interviewee about the level of response that is desired. Probing
often took the form of repeating interviewees exact words and asking to explain what they
meant by them or requesting concrete examples to aid understanding. For example, generally,
perceptions of peer drinking came up in the initial narrative of the participants in some form
and were further explored with the use of appropriate follow up cues and probes. In situations
when they didnt come up in the initial narrative, either direct probes were used such as lets
talk about the drinking behaviour of the people you named or you mentioned that X loves her
drink, can you tell me more about it?. Alternatively, gestures such as nodding or pausing a
little longer than usual were used to encourage participants to continue with the description.
Opinions and comparisons were sought (such as, what do you think about that? or In what
ways do you think that affected X?), clarifications and explanations of critical events and
159
Methodology
social jargon were requested (such as can you explain what you mean when you say
wasted?) and episodes and examples were encouraged (such as, can you think of an
example to help me understand? or can you describe an episode when you thought that
happened?).
Care was taken to remain unobtrusive by avoiding leading questions and suggesting outcomes
to minimize the interviewers influence on the participant. Methodological research
emphasizes the importance of using neutral inputs because directive questions may bias the
data by leading interviewees to respond in a way that they thought was expected or desired by
the researcher (Corbetta, 2003; Punch and Punch, 2005).
The resulting data was rich and progressively constructed in what proved to be an interactive
and empathetic process of dialogue.
5.6.5.1 Examining Interrelationships among Alters
Another important decision that needs to be made concerns whether ties between alters should
be examined at all? Without information on the inter-relationships among alters, no structural
analysis can be performed. Some studies related to social support choose to omit collecting
data on network structure (House et al., 1985). However, from a network perspective, although
burdensome, it is the structure of the network and how the structural properties affect
behaviour that is informative (Hawe et al., 2004). In fact, there is evidence that the process of
evaluating ties between network members is less burdensome than intuition might suggest
(McCarty et al., 2007). McCarty and colleagues (2007) explain that when alter pairs are
presented to respondents in a systematic way (such as one alter with all other alters, the next
160
Methodology
with the rest and so on), respondents can get a feel for the way their alter lists are organized
and anticipate their responses.
Based on the above discussion, the participants were requested to fill up an adjacency matrix; a
very simple square matrix with as many rows and columns as the people in a network. The
rows and columns in the network contain the names of the people (in the same sequence) as
shown in the hypothetical example in Figure 5.
Sam Adam Polly Mary Joe Sally

Sam
Adam
Polly
Mary
Joe
Sally
1
1
Figure 5: Example Adjacency Matrix
The scores in the cells of the matrix record information about the ties between each pair of
alters with 1 indicating the presence of a tie and 0 indicating an absence. For example, in
the matrix shown above, Sam knows Adam, Polly and Marry but does not know Joe and Sally.
This kind of a matrix is the starting point for almost all network analysis, and is called an
161
Methodology
"adjacency matrix" because it represents who knows whom in the "social space" mapped by
the relations that a researcher has measured.
Prior to filling out the matrix, the participants were informed that the objective of the exercise
was to examine who knows whom among the people that they named. Specifically, they were
told to pick one name (e.g. Sam in this case) at a time from the top of the matrix, go down the
corresponding row and identify the people Sam knows, meaning if Sam and Adam come
across each other on a street and stop to say hello, they know each other. The participants were
asked to repeat this process for all the people that they named. In consistency with past
research (McCarty et al., 2007), it did not take long for the participants to complete the matrix.
In fact most people were able to do it within 10 minutes.
5.6.5.2 Tie Strength
Another key consideration in network studies is the assessment of tie strength or the strength of
relationships between the participants and their network members. Multiple measures have
been used by researchers to obtain indices of tie strength (Marsden, 1990). Granovetter (1973)
for example approaches the subject by defining the concept of tie strength in terms of its
indicators which he describes are a combination of the amount of time, the emotional
intensity, the intimacy (mutual confiding) and the reciprocal services which characterize a tie
(p1361). Subsequent network research has utilized one or more of these indicators as the basis
to create indices of the strength of a relationship. The most common approach is to assess the
strength of a relationship based on indications of closeness, thus close friends are often
regarded as strong ties whereas acquaintances or friends of friends are said to be weak ties
(Erickson et al., 1978). A variation of this method is the presumption that the source of a
162
Methodology
relationship dictates its strength; therefore, kinship ties are assumed to be strong whereas
neighbours and co workers are treated as weaker ties (Murray et al., 1981). Other research
interprets the strong ties as those which feature greater frequency of contact (Lin and Dumin,
1986), greater duration of contact and the exchange of emotional support and assistance within
a relationship (Wellman, 1982).
Marsden and Campbell (1984) note in a study of best friend ties that measures of intensity of a
relationship are the best indicators of the unobserved tie strength concept because these are not
contaminated by other measures. The literature suggests that duration of the relationship tends
to overstate the strength of kinship ties and frequency of contact exaggerates the strength of
ties with co-workers and neighbours (Marsden, 1990). One innovation that has been adopted
by several researchers studying ego networks is to make use of a concentric circle diagram as a
measure to assess closeness (Curtis, 1979; Antonucci, 1986; Nadoh et al., 2004; Carrasco et al.,
2008).
The same approach was adopted in this study. The technique began by asking the participants
to look at a chart sized diagram of 6 concentric circles, with a smaller circle in the centre
containing the word 'You'. Each circle was viewed as representing a different level of intimacy
to the focal person such that the inner circles represented those who were closest to the ego and
so on and so forth. The names of all alters were written on removable Post-It notes. The
participants were then invited to position the tags on the diagram (reproduced below) based on
how close they felt towards each network member.
163
Methodology
Figure 6: Concentric Circles Diagram
Chua and colleagues (2011) suggest that this technique induces respondents to think about
their alters in relation to one another as well as to themselves and permits them to describe
their networks according to their own personal feelings of intimacy.
Further, the activity saved time and was less burdensome to the participants compared to other
measures such as rating the strength of each relationship on a predefined scale. It also provided
the participants with an opportunity to view their networks which nearly all interviewees
described as a fun thing to do.
164
Methodology
5.7
Ethics of Research
Addressing ethical dilemmas and concerns is an indispensable element of conducting research.

The present study was conducted in two stages using different methodologies thus resulting in
data that was unique, requiring different and appropriate measures to ensure adherence to
known ethical concerns. These included acquiring an informed consent from participants from
both survey and interview participants, making participation voluntary, ensuring privacy and
confidentiality of the collected data and dealing with issues relating to anonymity of the data
(Marsden, 1990; Buchanan, 2002; Eysenbach and Wyatt, 2002).
5.7.1 Informed Consent

In keeping with ethical guidelines provided by researchers (Eysenbach and Wyatt, 2002;
Mathy et al., 2003; Corden et al., 2005; Buchanan and Hvizdak, 2009), both the web survey
and in depth interviews were preceded with a full disclosure about the ramifications of the
study such as the purpose of research, the privacy/confidentiality assurance, the methods being
used and possible outcomes so that the participants may make an informed judgment about
their participation. Contact details of the researcher were also provided for any clarifications or
queries that respondents might have wanted to discuss. Further, participation in the survey and
interviews was made voluntary and the participants were informed that they were free to
withdraw from the study at any stage.
5.7.2 Privacy and Confidentiality

The principles of research ethics dictate that researchers must ensure adequate provision of
measures to protect the privacy of respondents and to maintain the confidentiality of data. A
165
Methodology
violation of privacy or breach of confidentiality can present the danger of exposing personal or
sensitive information (Buchanan and Hvizdak, 2009).
The protection of privacy and confidentiality was addressed through a combination of research
tactics and practices as is recommended by researchers (Mathy et al., 2003) for both stages of
this research. Only students 18 years of age and above were recruited in the study. The
participants were assured that their information will not be used or disclosed for purposes other
than academic research and that the data will be retained only for the duration of the current
project. Once collected, the survey data was imported in SPSS in a password protected
personal computer of the researcher. The data was also removed in entirety from the web
survey hosts employed in this research and the respective online accounts were deleted. Audio
recordings from the interviews were destroyed after being transcribed to a textual form for the
analysis. All textual and graphical data from the interviews such as the interviewers side
notes, adjacency matrices, concentric circle diagrams and alter attribute charts were converted
to a digital form and stored in the researchers password protected personal computer for the
duration of this study.
5.7.3 Anonymity
A related concern was the non anonymous nature of the data collected. As commented in
section 5.9.1, in order to cross check multiple submissions and prevent data contamination by
irrelevant subjects, the provision of a DIT student ID number was made mandatory for the
survey response to be submitted. Further, the sample for stage 2 was drawn from the web
survey respondents which necessitated a means to be able to contact them later on in the study.
This information was also needed to be able to contact the winner of the draw for the incentive.
166
Methodology
Students were thus asked to provide their email addresses in the web survey. The survey data
was subjected to statistical techniques of analysis where the focus is interpreting and reporting
the results at an aggregate level. Anonymity is therefore not easily compromised. The issue
however becomes more serious in network studies where asking for names of significant
others, requesting information on interrelationships of these people and asking to share
personal experiences related to those named is commonplace. Non participants are still
included in a sense that they are mentioned and discussed by others. Network studies therefore
differ from typical social science research because they reveal details of a persons social
surroundings (in the form of a network diagram and narratives) rather than reducing the data to
summary statistics. In order to protect the interests of the participants and preserve the viability
of their own academic field, network analysts propose a basic set of ethical guidelines
(Borgatti and Molina, 2003a) which recommend protecting identifiable information and
participant identity by means of processes designed to anonymize them. In keeping with this
research (Borgatti and Molina, 2003b), pseudo names and codes were used to refer to the
participants and their network members respectively to protect their identity. Further, any
information that could be traced back to them was anonymized prior to inclusion in this thesis.
5.7.4 DIT Ethics Committee Approval

The methods adopted in both stages of this research were reviewed and approved by the DIT
Research Ethics Committee. The unique ethical clearance reference number for this study as
issued by the committee was provided to the participants along with relevant contact details for
reporting any breaches to confidentiality and privacy of the information they provided.
167
Methodology
5.8
Data Transformation and Analysis
Survey data from wave 1 was subsequently imported in SPSS for a comprehensive descriptive
analysis and hierarchical multiple regression procedure to address the research objectives 1, 2
and 3. This analysis helped in determining the normative drinking rates prevalent at DIT.
These were then compared with perceived peer drinking norms to establish if misperceptions
existed as has been shown to be the case in other cultures (Perkins and Craig, 2002; McAlaney
and McMahon, 2007; Perkins, 2007). The regression procedure investigated the influence of
these perceived norms in predicting personal consumption. It also assisted in determining the
relative impact of descriptive and injunctive norms on drinking behaviour.
The data generated from the in depth interviews, comprised network data and qualitative data.
This data was analyzed by mixing formal measures of SNA with content analysis of the
interviews and interpreting the results combining the information from both methods. The
network data (such as the network members, their attributes and their ties with other members)
was input into UCINET (Borgatti et al., 2002) for the application of standard network
techniques. The qualitative data from the interviews was subjected to content analysis which is
a systematic coding and categorizing approach used for exploring large amounts of textual
information unobtrusively to determine trends and patterns of themes, their frequency and their
relationships (Vaismoradi et al., 2013). The interviews were transcribed, creating a verbatim
text of every interview by typing out each question and its response using the audio recordings.
A sense of the data was obtained by reading the transcripts several times. Open coding, which
refers to categorizing the data into segments and scrutinizing it for commonalities and
disparities (Pope et al., 2006) was then used to identify and isolate meaningful patterns and
168
Methodology
processes across the data. The data analysis process was not linear but recursive which
involved frequent reviews. The aforementioned procedure provided descriptions of the
composition and structure of the networks such as size, effective size, generation of specific
structural typologies, evaluations of tie strength and subjective meanings associated with these
feelings. This elaborate set of descriptions helped in the identification of the most salient and
possibly influential people in the ego networks of the participants addressing objective 4.
Data from the two data collection methods (web survey and interviews) was linked to evaluate
the possibility of normative perceptions and personal consumption been shaped and influenced
by ones personal network. This was aimed to address research objective 5. Further, this
analysis helped in examining the development, sustainment and dissemination of drinking
norms in the networks. Finally, the outcomes of the two data collection methods were linked to
demonstrate how the methods complemented each other and added value to the study.
5.9
Limitations of Methodological Approach
5.9.1 Specific Issues Related to the Web Survey

Sampling
The sample for the web survey was not randomly drawn as is generally the case with internet
based surveys. Indeed, most web surveys involving college students are conducted using non
probability sampling methods such as convenience sampling or purposive sampling (Couper,
2000). While strictly speaking statistical tests require probability sampling, in practice web
surveys are commonly analysed using such tests. McAlaney (2007) and McAlaney and
169
Methodology
McMahon (2007) are examples of SN studies conducted in the UK which have used web
surveys in college populations and analyzed the resulting data using statistical methods.
Therefore whilst it is acknowledged that the online data collection method utilized in this study
may have implications of representativeness of the sample to the student population because of non
probability sampling, the benefits in terms of quickly and largely accessing a population dispersed
across five campuses outweighed the costs.
Self Reports
Surveys related to drinking behaviours generally rely on self reports of personal consumption,
the validity of which is often questioned by substance use researchers (Midanik, 1988).
Although several studies suggest that memory aids can be used to enhance the recall of
drinking, (Midanik, 1988; Sobell and Sobell, 1992; Hammersley, 1994; Single and Wortley,
1994), it is important to remember that almost all drinking measures are retrospective and, as
such, they require people to provide the best estimate of their past drinking. Further, there is
a reasonable amount of evidence that describes self recall of drinking as a reasonably accurate
method to capture personal consumption without making huge compromises on the reliability
and validity of the results (Johnston, 2001; Brener et al., 2003; Lintonen et al., 2004). Studies
also support the reliability and validity of self reports in online surveys for health risk
behaviours (Ramo et al., 2011) and reflect that issues of reliability/validity are often specific to
measures being used rather than the online survey methodology.
Control over the Sample
The issue of self reports is even more important in online surveys because researchers cannot
be physically present at the time the participants take the survey and subsequently lose control
170
Methodology
over the makeup of sample. The ease of forwarding web addresses can make it difficult to
verify that the individuals who replied to the survey actually belong to the target population.
The anonymity of internet also makes it easier to submit multiple responses to online surveys
by the same person which increases the likelihood to respond mendaciously. It therefore cannot
be ruled out that self reporting can be a source of bias in this study. Several of these concerns
were addressed by the use of authentication measures during the web survey and screening the
data prior to analysis.
First, the survey invites were sent via the Public Affairs Office and the Campus Life Office at
DIT using the internal student email system which is used on daily basis by the students as has
been discussed in section 5.5.4. This measure whilst not completely eliminated the possibility
of non-students responding, was certainly a more secure method of sending invites than
posting the survey link on public forums like Facebook or Friendster. Second, supplying
student numbers and valid DIT email addresses was made compulsory for the survey to be
submitted and entered into the draw for an incentive. This prevented outsiders from taking the
survey and multiple submissions from the same individual.
Causality
The cross sectional nature of this study limits its internal validity. As such, causal inferences
cannot be made with certainty. This is an inherent limitation of cross sectional research and is
applicable to both phases of this study. However, the causal assumptions made in this study are
based on theory and past longitudinal research (Fearnow-Kenny et al., 2001). Further, the
combination of quantitative and qualitative research methods employed in this study provide
strength to the findings (More on this in chapter 8, section 8.2).
171
Methodology
Representativeness and Generalization

It has been noted in the US that internet users in the general public differ from non-users on
several demographic, social and psychological factors (Robinson et al., 2002). This means that
public samples obtained via the internet can be questioned in terms of how representative they
are of the larger population. This is a particularly important issue given the low response rate
associated with online surveys. While it is essential to acknowledge these concerns related to
data collection via online methods, the design of this study negates most of these concerns.
First, the sample in this study was drawn from the registered students at DIT and it can be
argued that these students will not differ greatly in terms of their demographic and social
psychological characteristics. Second, several researchers argue that response rates obtained
via online survey methods are not significantly different than would be obtained from
traditional postal surveys (Morphew and Williams, 1998). Third, previous studies note that the
difference between responders and non responders to online surveys tend to be minimal when a
population is clearly identified and comprises regular computer users (Kraut et al., 2004;
Hayslett and Wildemuth, 2005). The web survey was disseminated via the college email
system and the website of the Campus Life Office. Unlike public surveys, it can be reasoned
with confidence that DIT students are regular users of these channels thus minimizing errors
related to the coverage of population. Finally, as has been discussed in section 5.5.7, the
sample for the web survey was found to be representative of the population with regards to
gender and age and no significant differences were found between early and late respondents.
5.9.2 Specific Issues Related to Ego Network Analysis and In-depth Interviews
Generalization
172
Methodology
The sample for the in depth interviews was not statistically representative of the population as
is generally the case with qualitative methods of data collection. The purpose of utilizing SNA
in combination with in depth interviews in the present study was to discover the meaning and
understanding of norm salience rather than to verify or predict outcomes. Some qualitative
experts argue that qualitative findings can be generalized to other people, settings, times and
treatments to the degree to which they are similar to the people, settings, times and treatments
in the original study (Johnson and Christensen, 2007). Stake (1978) uses the term naturalistic
generalization to refer to this process of generalizing qualitative explanations on the basis of
similarity. Whilst it is accepted that the network results and related qualitative interpretations
uncovered in this study may not be generalized to the population in traditional sense of the
word, it is possible that the study may have uncovered a phenomenon which is particularly
related to commuter colleges (More on this in chapter 9, section 9.5). Therefore, the findings
may be tentatively and naturalistically generalized to the extent of being especially relevant to
commuter colleges.
Self Reported Network Data
One of the inherent features of ego network analysis relates to the process of determining
network structure. Unlike whole networks, which are based on a near-complete enumeration of
the population of interest and collect data from every member of the network, ego networks
view social environments from the eyes of the focal individual (ego). Egocentric network data is
therefore based purely upon the knowledge, reflection and recall of the ego (O'Malley et al., 2012).
It captures the diversity in the social environment of individuals and it is because of this
diversity that the boundaries of ego networks are dependent on the context being examined and
173
Methodology
generally not known at the onset of an inquiry. Since it is not possible to collect network data
from each and every member of an ego network, egos (participants) report on the presence or
absence of ties between each pair of nodes in their networks. The possibility that participants
might not have been aware of all possible ties in their networks or might not have recalled
them fully is thus inevitable. Network analysts believe that examining ego networks is still
relevant and useful considering that an egos perception of relationships may be more important
than whether or not the perceived relationship is validated via reciprocation by the alter (O'Malley
et al., 2012).
Missing Data
Social network studies are especially sensitive to missing data which often occurs due to non
response. It can either take the form of a participant leaving out a relevant node altogether
referred to as unit non-response or not completing specific items related to a node(s) referred
to as item non-response. Researchers caution that this can have negative effects on the
structural properties of networks (Burt, 1987; Ghani et al., 1998; Borgatti and Molina, 2003b;
Kossinets, 2006). The problems associated with sampling and missing data in whole network
studies often stem from the inability of researchers to interview or observe network members.
For ego networks, alters and ties are often missing because respondents either did not recall
them or were not asked about them in such a way so as to fully capture the network structure.
Several steps were taken to minimize the occurrence of missing data in this study. First, free
recall method was used to extract ego networks which is argued to be the most comprehensive
though time consuming method of eliciting network members (Ferligoj and Hlebec, 1999;
Carrasco et al., 2008). Second, multiple name generators were administered to ensure inclusion
174
Methodology
of all relevant nodes. In addition, a third name generator was administered probing if there was
anyone else that the participant might have missed in the first two questions and would want to
mention. Third, item non response was addressed by reviewing the alter attribute chart, the
concentric circles evaluation and the adjacency matrix prior to the termination of each
interview and complete any missing information. This step ensured that there was no missing
data with regards to alter attributes, tie strength scores and interrelationships of alters. Finally,
the understanding of the functionality of ego networks and the formation, dissemination and
reinforcement of drinking norms within them was based on a qualitative interpretation of the
interviews which strengthened the analysis and further reduced the chances of omitting socially
relevant and salient nodes.
Researcher Bias
As with any qualitative inquiry, researcher bias might have occurred and therefore it is
acknowledged. Researcher bias is an important concern in qualitative research because this
type of research tends to be exploratory, open ended and less structured than quantitative
methods (Johnson and Christensen, 2007). It may occur in the form of selective observation,
selective recording of information and also from allowing ones personal opinions and
perspectives to affect how qualitative data is interpreted and how the research is conducted.
While it is not possible to completely eliminate researcher bias from any qualitative study
(Johnson and Christensen, 2007), efforts were made to reduce its impact by improving the
rigor of data collection in keeping with research on qualitative methods (Johnson and
Christensen, 2007). This was achieved through the use of neutral probing, fewer assumptions
and avoidance of leading questions and/or premature interpretation. In addition, the use of low
175
Methodology
inference descriptors was maximized in deriving qualitative explanations and reporting them.
This was achieved by phrasing descriptions very close to the participants accounts and
researchers interview notes and using direct quotations where appropriate to expand on the
findings.
176
Analysis and Results of Survey Data
This chapter presents the results of analysis performed on the survey data and serves the
overarching aim of addressing objectives one, two and three of this research as outlined in
chapter 5, section 5.2. It begins with developing a feel about the data and presenting the key
descriptive findings which establish an understanding of alcohol consumption levels across
DIT. Self reported personal consumption is then compared with perceived peer norms to assess
if students misperceived these norms as has been found in prior studies described in chapter 3,
section 3.6.2. The results of hierarchical multiple regressions follow. These results address two
important aspects of this study. First, the effect of perceived peer norms of different referent
groups on personal consumption is investigated. Second, the relative impact of perceived
descriptive and injunctive norms on personal consumption is examined. The chapter concludes
with a short discussion of key findings.
6.1
Data Cleaning
A total of 2115 completed electronic surveys were retrieved from surveygizmo.com. This data
was first screened for missing values. Three cases of system missing values were removed.
Next, the ethnicity of respondents was examined. 14% of the sample comprised international
students. Since the study was designed for and aimed at Irish students only, these were
removed from the dataset. It was also felt necessary to limit the maximum age in the sample as
this study was aimed at studying drinking habits and normative mechanisms prevalent among
undergraduate students at DIT. Finally, only full time under graduate students were retained in
the sample. The effective sample size was N=1700.
177
6.2
Descriptive Statistics
57% of the sample was male while 43% was female. Mean age of the sample was 21.13 years.
68% of the students lived with parents, 24% in shared accommodations, 5% with their partners
or dependent others and only 3% lived alone. 50% of the students reported being in DIT 1 year
or less, 16% reported 2 years, 18% reported 3 years, 12% reported 4 years and 4% reported 5
or more years.
Variable
Mean
21.13
Std. Deviation
2.884
Year in college
1.70
1.596
Money available for drinking
46.94
34.516
Age of first drink
14.68
2.683
Drinking group size
2.08
20.776
Age
Table 5: Standard descriptive Statistics
The sample distribution across various campuses is presented below.

% Sample Distribution Across DIT Campuses
7
DIT Mountjoy Square
16
DIT Cathal Brugha Street
23
DIT Bolton Street
32
DIT Aungier Street
20
DIT Kevin Street
2
DIT Rathmines Road
100
Total
Table 6: Sample distribution across campuses in percentage
94% of those surveyed had been involved in underage drinking at some stage reporting that
they had their first proper drink before they turned 18. The mean age at which students
reported having consumed their first drink of alcohol (other than just a sip) was a little shy of
178
15 years. 60% of the sample reported having anywhere between 20 and 80 available in a
week for drinking. The average amount of money available for drinking in a week was around
47.
6.2.1 Personal Alcohol Consumption

Personal consumption of alcohol was measured along three dimensions (1) frequency of
drinking in a typical month, (2) number of drinks on a typical drinking occasion and (3)
frequency of drunkenness in a typical month as described in chapter 5, section 5.5.6.2. The
percentage distribution across these dimensions is presented below.
Frequency of drinking in a typical month
% Cumulative %
Never
Less than once a month
once a month
13
2-3 days a month
22
35
Once a week
32
67
Twice a week
26
93
3-4 days a week
99
5-6 days a week
100
No. of drinks on a typical occasion

Cumulative %
%
4
4
0
3
7
1-2
10
17
3-4
18
35
5-6
19
54
7-8
15
69
9-10
10
79
11-12
13-14
87
15 or more
13
100
Frequency of drunkenness in a typical month

%
Cumulative %
10
10
Never
11
21
12
33
once a month
25
58
2-3 days a month
28
86
Once a week
12
98
Twice a week
3-4 days a week
100
Table 7: Personal alcohol consumption

179
As can be seen from Table 7, 64% respondents reported drinking once a week or more
frequently in a typical month. Only 4% reported that they never drank alcohol. 83% people
reported drinking 5 drinks or more on a typical drinking occasion, 13% reported 3-4 drinks or
less and only 4% reported 0 drinks. 42% of those surveyed reported drinking enough alcohol to
get drunk once a week or more frequently in a typical month, 25% reported 2-3 days a month,
23% reported once a month or less and 10% reported never drinking enough alcohol to get
drunk. The mean number of drinks consumed on a typical occasion for females was 6.63
compared to 10.27 for males.
6.2.2 Descriptive Norms (DN) Vs. Personal Consumption

The means and standard deviations of personal consumption and perceived descriptive norms
in 5 referent groups are presented next.
Frequency of
drinking in a typical
month (Days)
Personal alcohol use

Most students at DIT
Close friends
Best friend
Mother
Father
Mean
4.91
8.51
7.23
6.55
5.17
7.05
St dev
3.67
3.87
3.98
4.52
6.46
7.6
No. of drinks on a
typical occasion
Mean
8.68
9.74
9.59
9.03
3.02
4.98
St dev
4.63
3.79
4.22
4.58
2.64
4.08
Frequency of
drunkenness in a typical
month (Days)
Mean
3.26
6.14
5.02
4.52
1.11
1.95
St dev
3.05
3.44
3.39
3.75
3.2
4.37
Table 8: Means and St Deviations of personal consumption vs. the perceived

descriptive norm
Four important findings can be deduced from Table 8. First, most students overestimated the
prevalence of drinking in DIT along all three dimensions of alcohol consumption used in this
study. Second, students also perceived their proximal peers (close friends and best friend) to
180
consume greater quantities of alcohol more often and to drink enough alcohol to get drunk
more frequently than they themselves. Third, most students perceived distal peers (most
students at DIT) to have more permissive drinking than that of proximal peers (close friends,
best friend) across all three dimensions of alcohol use. Fourth, most people perceived their
fathers to have a more permissive drinking behaviour as compared to their mothers along all
three dimensions of alcohol use. These trends were found to be consistent for both males and
females and across all campuses of DIT.
In addition, it was noticed that abstainers also overestimated the campus norm drinking and
that they also perceived their proximal peers to drink more than themselves. The following
table illustrates this.
Values by cohort:
Abstainers
in a typical month
(Days)
No. of drinks on
a typical occasion
Frequency of
drunkenness in a
typical month
(Days)

Close friends
Best friend
Mother
Father
Mean
8.88
6.96
5.06
3.76
5.2
Mean
9.00
6.77
5.23
0.99
3.11
Mean
4.84
3.08
2.22
0.26
0.79
Table 9: Means of perceived descriptive norm for abstainers
Further, as expected students reported consuming comparatively less alcohol when they had
college the next day (Mean: 6.3 drinks).
Figure 7, Figure 8 and Figure 9 provide a graphical illustration of personal consumption
behaviours and perceptions of these behaviours in distal and proximal referent groups.
181
The graphs illustrate that self reported personal consumption of the respondents was less than
the perceptions of the behaviour in other social groups who were used as reference. Also, most
students perceived their close friends and best friend to have less permissive attitudes towards
drinking than most students in DIT. For example, while 4% students reported that they never
drink alcohol or do so less than once a month; hardly any thought that other students in DIT
drink that infrequently. Similarly, while 22% students reported drinking 2-3 days a month, less
than 5% thought that other students did the same. However, 7% and 13% thought that their
close friends and best friend respectively drank 2-3 times a month. On the other hand, while
33% reported drinking twice a week or more, almost 80% thought that others in DIT did the
same and almost 60% and 50% respectively thought that their close friends and best friend did
the same.
Similar patterns were observed for the number of drinks on a typical occasion and the
frequency of drunkenness in a typical month. For example while 46% students reported
drinking 9-10 drinks on an occasion or more, 60%, 54% and 48% respectively thought that
most students in DIT, their close friends and best friend did the same. Likewise, while 42%
reported drinking enough alcohol to get drunk once a week or more frequently, twice (84%) as
many thought that most students in DIT did the same while the percentages for perceptions of
close friends and best friends behaviour were comparatively less (68% and 60%
respectively).
80% of those surveyed, overestimated the norm for frequency of drinking in a typical month,
60% overestimated the norm for the number of drinks on a typical drinking occasion and 85%
overestimated the norm for frequency of drunkenness in a typical month.
182
Personal Consumption Vs Perceived Descritpive Norm (DN):

Frequency of Drinking in a Typical Month
60
50
Response Perventage
Personal consumption
40
DN for Most students at DIT
30
DN for Close Friends
20
DN for Best Friend
10
0
Never
Less than
once a
month
once a
month
2-3 days a
month
Once a
week
Twice a
week
3-4 days a 5-6 days a

week
week
Everyday
Approximately 80% of
the respondents
overestimated the norm
Frequency of Drinking in a Typical Month
Figure 7: Frequency of drinking in a typical month Vs the perceived norm
183

No. of Drinks on a Typical Occasion
25
Response Percentage
20
15
DN for most students at DIT
10
DN for close friends

DN for best friend
the respondents
0
0
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Number of drinks on a typical occasion
Figure 8: No. of drinks on a typical occasion Vs the perceived norm
184

Frequency of Drunkenness in a Typical Month
45
40
Response Percentage
35
30
25
20
DN for most students at DIT

15
DN for close friends

DN for best friend
10
5
0
Never
Less than
once a
month
once a
month
2-3 days a
month
Once a
week
Twice a
week
3-4 days a 5-6 days a

week
week
Everyday
the respondents
Frequency of drunkeness in a typical month
Figure 9: Frequency of drunkenness in a typical month Vs the perceived norm
185
6.2.3 Injunctive Norms vs. Personal Consumption

Injunctive norms were conceptualized as the perceived approval for drinking to get drunk as
has been described in chapter 5, section 5.5.6.4. Most students perceived both distal and
proximal peers to have approving attitudes towards drinking to get drunk. Specifically, 76% of
the sample thought that for most students in DIT, drinking to get drunk was acceptable. Only
7% disagreed with the statement. 70% of the sample thought that their close friends would
approve if they drank to get drunk. Only 10% disagreed with the statement. 68% students
thought that their best friend would approve if they drank to get drunk. Only 15% disagreed
with the statement. Parents were an exception with most students perceiving their parents to be
less approving if they drank to get drunk. 45% and 38% of those surveyed respectively thought
that it was unacceptable to their mother and father if they drank to get drunk. Only 12% and
14% students respectively thought otherwise.
6.3
Correlations
In order to examine the association between personal consumption and perceived norms,
Pearsons correlations were carried out. The procedure was repeated for all three dimensions of
alcohol use and for all reference groups included in this study. These calculations were
performed twice, once with the abstainers included and once with them excluded. Given
similar results, the outcome from the analysis with abstainers included is presented here.
186
Referent Groups
Alcohol Consumption Measure
0.24**
0.66**
Frequency of drunkenness in a typical
0.39**
month
0.20**
0.64**
0.42**
month
0.16**
0.51**
0.40**
month
0.28**
0.57**
0.55**
month
0.21**
0.62**
0.40**
month
0.22**
0.62**
0.43**
month
Close
Friends
Best
Friend
Mother Father
DIT Aungier Street (n=545)

0.57** 0.56** 0.10**
0.84** 0.79** 0.29**
0.63** 0.52** 0.09**
DIT Bolton Street (n=393)
0.55** 0.45** no corr
0.80** 0.76** 0.22**
0.60** 0.46** 0.12*
DIT Kevin Street (n=347)
0.37** 0.45** 0.11**
0.72** 0.72** 0.27**
0.59** 0.57** no corr
DIT Mountjoy Square (n=121)
0.57** 0.53** no corr
0.70** 0.67** no corr
0.15**
no corr
0.28**
0.16**
no corr
0.32**
no corr
0.22*
0.21**
0.59** 0.64** 0.57** 0.49**

DIT Cathal Brugha Street (n=266)
0.5**
0.49** 0.17** 0.15*
0.76** 0.72** 0.25** 0.27**
0.44** 0.44** no corr
TOTAL SAMPLE (N=1700)
0.49** 0.47** 0.11**
0.80** 0.85** 0.25**
0.62**
0.69**
0.06**
Table 10: Pearson's correlation between personal consumption and perceived

descriptive norms (* p<0.05, **p<0.01)
187
0.15**
0.38**
no corr
0.13**
0.30**
0.10**
The correlation results indicated an overall positive relationship between individuals own
drinking behaviour and their perceptions of alcohol consumption in others. This is in
agreement with previous research regarding descriptive norms (McAlaney and McMahon,
2007). It was also noticeable that the order of strength of these correlations agreed with
previous normative research (McAlaney and McMahon, 2007) in that the perceptions of
alcohol usage of closer peers was more strongly associated with personal consumption
compared to distal peers (most students at DIT). Of all referent groups, perceived descriptive
norms for parents were the least associated with ones personal consumption. This suggests
that peers may be a more relevant source of normative information as opposed to parents as is
also noted by Borsari and Carey (2001, 2003). Further, it is interesting to note that the
association between personal consumption and perceived descriptive norms was stronger for
certain dimensions of alcohol consumption with the number of drinks consumed on a typical
occasion being the most strongly associated. This suggests that normative beliefs may have a
stronger influence on some dimensions of alcohol usage than others. McAlaney (2007) notices
a similar pattern. These results were consistent for both males and females and across all
campuses of DIT. This set of correlations was not performed for DIT Rathmines Road campus
because the number of respondents from this campus (n=28) were few.
The third research objective of this study, related to investigating the relative impact of
descriptive and injunctive norms on drinking behaviour. Pearsons correlation was performed
between personal consumption and perceived injunctive norms for each referent group in order
to investigate the association between the two. The results are presented in Table 11.
188
Referent Groups
Alcohol Consumption Measure
Most
students Close
Best
at DIT
Friends Friend Mother Father
DIT Aungier Street (n=545)
no corr 0.19** 0.21** 0.11** 0.11**
0.19**
0.40** 0.34** 0.12** 0.17**

month
0.16**
0.10*
0.14**
month
0.22**

no corr
0.13**
month
0.18**
0.32** 0.33** 0.13**

DIT Bolton Street (n=393)
0.30** 0.25** no corr
0.37** 0.38** 0.112*
0.11**
0.35** 0.34** 0.122*

DIT Kevin Street (n=347)
0.13** 0.17** no corr
0.31** 0.31** 0.13**
0.17**
0.10*
0.22**
no corr
no corr
0.29** 0.27** no corr no corr

DIT Mountjoy Square (n=121)
no corr 0.21*
0.18*
no corr 0.27**
no corr 0.30** 0.30** no corr no corr

month
no corr
0.21*
0.23** no corr no corr
DIT Cathal Brugha Street (n=266)
no corr 0.21** 0.22** 0.24** 0.24**
0.19**
0.22** 0.27** 0.27** 0.23**

month
no corr
0.08**
0.16**
month
0.20**
0.32** 0.32** no corr

TOTAL SAMPLE (N=1700)
0.21** 0.29** 0.10**
0.32** 0.39** 0.14**
no corr
0.37**
0.13**
0.40**
0.15**
0.11**
0.17**
Table 11: Pearson's correlation between personal consumption and perceived

injunctive norms (* p<0.05, **p<0.01)
189
The results reveal an overall weak but positive significant relationship between injunctive
norms and personal consumption. Similar to the previous set of correlations presented in Table
10, the order of strength of these correlations agree with existing research in that the
perceptions of injunctive norm for proximal peers were more strongly associated with personal
consumption compared to those of distal peers (most students at DIT). This pattern is
consistent across all campuses of DIT. For some peer groups (most students at DIT and
parents) no correlations were found during examination of campus wise data. This also
suggests that the relationship was not as strong when the dataset was examined in groups.
The results discussed above also seem to be in agreement with research which reports the
perceptions of descriptive norms to be more strongly related to personal consumption than
those of injunctive norms (McAlaney, 2007). These results were consistent for all three
dimensions of personal consumption and across all referent groups. The only exceptions were
parents. In this case, the Pearsons correlation coefficient values were comparable for both
descriptive and injunctive norms reflecting a weaker association between perceptions and
personal consumption. This may be because of the relatively less relevance of parents
compared to peers when it comes to alcohol consumption in a collegiate context as commented
earlier in this section.
6.4
Hierarchical Multiple Regression
Hierarchical multiple regression was used to investigate the predictive relationship between
perceived norms and personal consumption. The three measures of alcohol consumption were
treated as dependent variables. There were several predictor variables. The three dependent
190
variables were: (1) frequency of drinking in a typical month (2) the number of drinks
consumed on a typical drinking occasion and (3) frequency of drunkenness in a typical month.
Therefore, three sets of hierarchical multiple regressions were carried out, one for each
dependent variable in the study.
6.4.1 Data set considerations

Before conducting the hierarchical multiple regression procedure on each dependent variable,
the data set was subjected to the following considerations.
1) For the analysis of each dependent variable, the abstainers were excluded from the
dataset. This was felt necessary because the study focussed on examining the drinking
behaviour of college students which makes sense only if drinkers are considered. Also,
abstainers choose to not drink while being surrounded by students who do drink. This
suggests that the thoughts and motivations that drive the attitudes and behaviour of
abstainers might differ from those of drinkers, which is interesting to investigate but
beyond the scope of this study. Abstainers were conceptualized as those who responded
as never and 0 to the measures of frequency of drinking and drunkenness in a typical
month and number of drinks consumed on a typical occasion respectively.
2) In order to improve the skew and kurtosis of the underlying distributions and bring them
closer to the normal curve, the few cases with very high values were treated as outliers
and were not included in the regression analysis. However, before removing these cases,
the resulting implications on regression outcomes were considered by examining the R2
values and the predictive power of the independent variables. Given the same pattern of
191
results regardless of whether these cases were included or not, it was decided to exclude
them.
6.4.2 Assumption Testing

Once an appropriate model for hierarchical multiple regression was achieved for each
dependent variable, a detailed analysis was carried out to ensure that the basic assumptions of
regression were not being violated. Specifically the following assumptions were tested for each
set of regressions.
1. Assumption of collinearity
2. Assumption of linearity
3. Assumption of homoscedasticity
4. Assumption of non stochastic x
5. Assumption of zero mean of the error term
6. Assumption of normality
The details of this procedure along with relevant statistical inferences and plots are provided in
appendix 7.
6.4.3 First set of regressions

Dependent variable: Frequency of drinking in a typical month
There were 73 people who reported that they never drink in response to the question on frequency of drinking in a typical month. In addition, 10 people (0.6% of the sample) reported
drinking 5-6 days of the week and 1 person reported drinking every day. These were excluded
192
from the hierarchical multiple regression. The effective sample size for this set of regressions
was N=1616.
Single step regressions were first carried out for all predictor variables. The predictor variables
of interest in this study were gender, age, living arrangement (a categorical variable dummy
coded into three dichotomous variables namely, living with parents, living in a shared
accommodation and living alone and independently), number of years at DIT, money available
for drinking, age of first drink, drinking group size, frequency of weekly communication about
alcohol, frequency of communication about alcohol in general and the perceived descriptive
and injunctive norms for the 5 referent groups. Significant findings are summarized next.
Dependent Variable: Frequency of drinking in a typical month
Independent Variables
R2
B (unstandardized)
Gender
0.012
0.73
Living in a shared accommodation
0.008
0.709
Living with parents
0.003
-0.372
0.093
0.03
Age of first drink
0.024
-0.31
Drinking group size
0.005
0.1
Weekly communication about alcohol
0.099
0.842
General communication about Alcohol
0.054
0.808
Descriptive norm for most students at DIT
0.049
0.193
Descriptive norm for close friends
0.243
0.42
Descriptive norm for best friend
0.223
0.354
Descriptive norm for mother
0.01
0.052
Descriptive norm for father
0.017
0.057
Injunctive norm for most students at DIT
0.006
0.31
Injunctive norm for close friends
0.045
0.751
Injunctive norm for best friend
0.045
0.688
Injunctive norm for mother
0.011
0.382
Injunctive norm for father
0.013
0.399
Table 12: Single set regressions for frequency of drinking in a typical month,
significant at p<0.05
193
When entered on their own, the descriptive norm for close friends and best friend explained the
most variance in the dependent variable (24.3% and 22.3% respectively). This was followed by
money available for drinking (9.3%), weekly communication about alcohol (9.9%), general
communication about alcohol (5.4%), descriptive norm for most students at DIT (4.9%) and
the injunctive norms for close friends and best friend (4.5%). This step provided guidance
about which predictor variables to retain in the hierarchical regression model.
6.4.3.1 Model 1
After the independent testing for each predictor variable, the following hierarchical multiple
regression model was run. The goal was to try and increase the amount of variance explained
by the model in the dependent variable in each block.
Block 1: Gender, Living in a shared accommodation, living with parents
Block 2: Money available for drinking, age of first drink, drinking group size, weekly
communication about alcohol, general communication about alcohol
Block 3: Descriptive norm for most students at DIT
Block 4: Descriptive norm for close friends, Descriptive norm for best friend
Block 5: Descriptive norm for mother, Descriptive norm for father
The model explained 40.8% of the variance in the dependent variable. The change in R2 in
each block is presented below.
194
Block
1
2
3
4
5
R2 change
0.023
0.182
0.046
0.154
0.004
Table 13: Change in R2 for model 1 (first set of regressions)
Unstandardized regression coefficients for this model are presented next. DV denoted the dependent variable.
DV: Frequency of Drinking (typical month)
Gender
Living with Parents
Age of first drink
Drinking group size
Descriptive norm for most students at
DIT
Unstandardized Regression Coefficients (B)

Block 1 Block 2 Block 3 Block 4 Block 5
0.778
0.466
0.645
0.753
1.295
0.967
1.065
0.745
0.771
0.026
0.025
0.02
0.02
-0.157 -0.163
0.616
0.565
0.423
0.428
0.269
0.323
0.189
0.054
0.248
0.166
0.054
0.245
0.164
-
Significant at p<0.05, (-) indicates no significant relationship
Table 14: Unstandardized regression coefficients for model 1 (first set of regressions)
Given the large sample size, an increment of at least 1% in the explained variance was set as a
criterion to evaluate the significance of an effect on the dependent variable in following with
Rimal and Real (2005). As can be seen, adding the descriptive norm for most students at DIT
in block 3 improved R2 by 4.6% compared to adding the descriptive norm for proximal peers
195
(close friends and best friend) in block 4 which improved R2 by 15.4%. Addition of the
descriptive norm for parents in block 5 improved the explained variance in the dependent
variable by only 0.4% which is negligible.
Males and those who had their first drink at a younger age were found to be associated with
higher frequencies of drinking in a typical month. However, these effects were no longer
significant once the descriptive norms were introduced in the model. Also, living in a shared
accommodation, more money a student had available for drinking and greater occurrence of
alcohol/drinking in his/her weekly conversations were found to be associated with higher
frequencies of drinking in a typical month. The results also reveal that the descriptive norms
for proximal peers (close friends and best friend) were better predictors of frequency of
drinking in a typical month than distal peers (most students at DIT). This supports past
research (Baer et al., 1991; Thombs et al., 1997; Borsari and Carey, 2003; McAlaney, 2007;
McAlaney and McMahon, 2007). The descriptive norms for mother and father were not found
to be significant.
6.4.3.2 Model 2
Next, injunctive norms and descriptive norms were introduced into the model in successive
steps with the injunctive norms preceding descriptive norms. This order of entry allowed
estimation of the unique variance predicted by both types of norms after controlling for
confounding factors. Further, this also provided a way to evaluate whether adding injunctive
norms would improve the predictive ability of the overall model. Model 2 is presented below.
Block 1: Gender, Living in a shared accommodation, living with parents
196
Block 3: Injunctive norms for all peer groups
Block 4: Descriptive norms for all peer groups
Model 2 explained 41.4% of the variance in the dependent variable, 0.6% higher than the last
model. The change in R2 in each block is presented below
Block
1
2
3
4
R2 change
0.023
0.178
0.013
0.200
Table 15: Change in R2 for model 2 (first set of regressions)

The table illustrates that adding injunctive norms in block 3 of the model increased the R2
value by 1.3%.
Unstandardized regression coefficients for this model are presented next.
197
DV: Frequency of Drinking (typical month)
Gender
Living with Parents
Age of first drink
Drinking group size
Descriptive norm for most students at
DIT
Unstandardized Regression Coefficients

(B)
Block 1
Block 2
Block 3
Block 4
0.782
0.47
0.398
0.709
1.267
0.977
0.877
0.776
0.026
0.026
0.02
-0.15
-0.119
0.599
0.561
0.394
0.255
0.217
-0.223
0.333
0.304
0.067
0.231
0.171
-
Significant at p<0.05, (-) indicates no significant relationship
Table 16: Unstandardized regression coefficients for model 2 (first set of regressions)
Other than the findings of model 1, this model reveals three key results.
First, of all the peer groups, only the perceived injunctive norms for proximal peers were found
to positively predict ones frequency of drinking in a typical month in Block 3. A possible
explanation for this can be that injunctive norms are suggested in the past research to be most
influential within proximal social networks where social approval is particularly important and
sought for maintaining group membership and cohesion (Trafimow and Finlay, 1996; Larimer
et al., 2004a). Thus within tightly knit proximal groups, individuals might feel strongly
motivated to follow injunctive norms.
198
Second, the injunctive norms of most students at DIT were found to be negatively related to
ones frequency of drinking in a typical month in Block 3. This negative association of
injunctive norms with personal consumption is also noted in the past research (Chawla et al.,
2007; Neighbors et al., 2008). These studies found that the relationship between perceived
injunctive norms and alcohol consumption was dependent on the normative referent group
such that the perceived injunctive norms were negatively associated with alcohol consumption
when the normative referent was distal (such as typical students) and positively associated
when the normative referent is more proximal (such as close friends and family).
Third, injunctive norms lost their significance of predicting the dependent variable when
descriptive norms were entered into the regression equation. This suggests that the predictive
power of descriptive norms was stronger than that of injunctive norms. This finding also
supports past research (Armitage and Conner, 2001; McAlaney, 2007; Neighbors et al., 2008).
6.4.4 Second set of regressions

Dependent variable: Number of drinks on a typical drinking occasion
69 people who reported drinking 0 drinks on a typical occasion and 1 case of missing value
were excluded from the data set. The effective sample size for this set of regressions was
N=1630.
A procedure similar to the first set of regressions was followed. Single set regressions were
first carried out with predictor variables and the results are presented in Table 17.
199
When entered on their own in the regression equation, perceived descriptive norms for close
friends explained the greatest variance (63.6%) in the dependent variable. This was followed
by perceived descriptive norms of best friend (58.3%) and most students at DIT (39%), gender
(18.7%) and the perceived injunctive norms for close friends and best friend (10.5%
approximately).
Dependent Variable: number of drinks on a typical occasion
R2
B (unstandardized)
0.187
3.8
Gender
0.006
0.123
Age
0.003
0.157
Year in college
0.112
0.043
0.064
-0.66
Age of first drink
0.007
0.058
Drinking group size
0.077
0.978
0.041
0.93
0.39
0.722
0.636
0.828
0.583
0.733
0.059
0.396
0.096
0.33
0.026
0.823
0.103
1.494
0.108
1.393
0.021
0.683
0.03
0.781
Table 17: Single set regressions for number of drinks on a typical occasion, significant
at p<0.05
200
6.4.4.1 Model 1
Hierarchical multiple regression was then performed. The first model aimed at improving R2 in
each block and examining which peer group most strongly predicted the dependent variable.
The results are presented below
Block 1: Gender, age, year in college
Block 4: Descriptive norm for close friends, Descriptive norm for best friend
Block 5: Descriptive norm for mother, Descriptive norm for father
Over all, model 1 was able to explain 71.3% of the variance in the dependent variable. Block
wise change in R2 is presented below
Block
1
2
3
4
5
R2 change
0.186
0.166
0.206
0.150
0.005
Table 18: Change in R2 for model 1 (Second set of regressions)

The addition of perceived descriptive norm for most students at DIT in block 3 improved R2 by
20.6% and the addition of perceived descriptive norm for close friends and best friend in block
4 improved R2 by a further 15%. It indicates that both distal and proximal peers explain large
201
variance in the dependent variable. Adding perceived descriptive norm for mother and father in
block 5, improved R2 by 0.5% which is negligible.
At this stage, the residuals were examined for skew and kurtosis to make sure that the
distribution did not deviate substantially from a normal curve. The standardized residuals
within +/-4 standard deviations were retained to bring down the value for kurtosis. 4 such cases
were excluded. This improved R2 by 1.3%.
The unstandardized regression coefficients are presented next
Unstandardized Regression Coefficients (B)
Gender
Age
Year in college
Block 1
3.757
-
Age of first drink

Drinking group size
Block 2
Block 3
Block 4
Block 5
3.46
0.029
2.47
0.021
1.13
0.013
1.15
0.012
-0.412
0.029*
0.501
0.515
-0.296
0.404
0.449
0.553
-0.128
0.231
0.265
0.127
0.413
0.213
-0.111
0.227
0.266
0.114
0.408
0.223
0.067
p<0.01, *p<0.05, (-) indicates no significant relationship
Table 19: Unstandardized regression coefficients for model 1

(Second set of regressions)
More money available for drinking, having the first drink at a younger age and greater
frequencies of weekly and general conversations about alcohol were found to be associated
with greater number of drinks an individual would consume on a typical drinking occasion.
202
Also, a strong gender effect was noticed with being a male predicting an increase of 1.15
drinks on a typical drinking occasion after having controlled for confounding factors. Drinking
in bigger groups was also found to be associated with greater number of drinks an individual
had on a drinking occasion. However, this effect was no longer significant once the perceived
descriptive norms were entered into the equation.
Among descriptive normative beliefs, greater perceptions of prevalence of drinking in peers
was found to be associated with greater number of drinks an individual would consume on a
typical drinking occasion. The only exception was the perceived descriptive norm for mother
which was not found to be significant. Perceived descriptive norm for father produced a small
effect. Further, perceived descriptive norms for proximal peers (close friends and best friend)
were found to be better predictors of the dependent variable than those for distal peers (most
students at DIT).
6.4.4.2 Model 2
Next, perceived injunctive norms for all peer groups were included in Block 3 and perceived
descriptive norms for all peer groups were included in block 4. Those variables which were not
found to be significant in model 1 were also excluded (namely age and year in college from
block 1 and drinking group size in block 2). Model 2 looked took the following form.
Block 1: Gender, money available for drinking, age of first drink, weekly communication about alcohol, general communication about alcohol
Block 3: Descriptive norms for all peer groups
203
Model 2 was able to explain 73% variance in the dependent variable. Block wise change in R2
is presented below:
Block
1
2
3
R2 change
0.352
0.044
0.334
Table 20: Change in R2 for model 2 (Second set of regressions)

Adding perceptions of injunctive norms in block 2 improved R2 by 4.4% compared to the
addition of perceived descriptive norms in block 3 which improved R2 by 33.4%.
Unstandardized regression coefficients are presented next:
Gender
Age of first drink
Unstandardized
Coefficients (B)
Block 1 Block 2 Block 3
3.489
3.365
1.078
0.031
0.029
0.012
-0.413 -0.333 -0.088*
0.499
0.366
0.225
0.48
0.354
0.207
0.53
0.57
0.11
0.42
0.22
-
p<0.01, * p<0.05, (-) indicates no significant relationship
Table 21: Unstandardized regression coefficients for model 2 (Second set of

regressions)
204
In addition to the findings of model 1, this model reveals two key results.
First, perceived injunctive norms of close friends and best friend were found to be significant
in block 2 such that a unit increase in either caused an increase of about 0.5 drinks individuals
would have on a typical drinking occasion. This finding is similar to that observed in model 2
in the first set of regressions performed on the frequency of drinking in a typical month as a
dependent variable (section 6.4.3.2). Similar explanation therefore applies. However, similar to
the first set of regressions, perceived injunctive norms lost their predictive effect once
perceived descriptive norms were added into the regression equation in block 3. This suggests
greater predictive power of perceptions of descriptive norms over those of injunctive norms.
Two, in line with past research (Baer et al., 1991; Thombs et al., 1997; Borsari and Carey,
2003; McAlaney, 2007; McAlaney and McMahon, 2007), perceptions of descriptive norms of
proximal peers (close friends and best friend) predicted personal consumption more strongly
than those of distal peers (most students at DIT).
6.4.5 Third set of regressions

Dependent variable: Frequency of drunkenness in a typical month
99 people reported that they never drink enough alcohol to get drunk in a typical month.
Theoretically speaking, these could be treated as abstainers and removed from the regression
analysis. However, in reality, some of these people might not have meant that they dont drink
alcohol at all. They might have just meant that they do drink alcohol but never go to the extent
of getting drunk and hence chose never as their answer. Cross tabs were examined between
the frequency of drinking and drunkenness in a typical month. Those who reported never in
205
both questions were excluded. Further, those who reported drinking enough alcohol to get
drunk everyday of the month (n=1) or 5-6 days a week (n=4) were treated as outliers and were
also excluded from the analysis as they represented a negligible percentage of the sample and
skewed the underlying distribution. The final sample size subjected to regression analysis was
N=1623.
Following the same steps as before, single step regressions were first carried out with each
predictor variable. The results are as follows.
Dependent Variable: Frequency of drinking to get drunk in a typical month
B (unstandardized)
R2
0.006
0.442
Gender
0.015
-0.118
Age
0.001
-0.054
Year in college
0.003
-0.888
Living alone or independently
0.085
0.024
0.041
-0.334
Age of first drink
0.013
0.047
Drinking group size
0.166
0.911
0.101
0.94
0.014
0.087
0.138
0.265
0.13
0.223
0.008
0.037
0.013
0.043
0.043
0.663
0.139
1.11
0.142
1.016
0.023
0.468
0.018
0.391
Table 22: Single set regressions for frequency of drunkenness in a typical month,
significant at p<0.05
206
As can be seen from Table 22, greatest variance (16.6%) in the dependent variable was
explained by weekly conversations regarding alcohol followed by perceived descriptive and
injunctive norms for proximal peers (almost 14%). Money available for drinking was next
explaining 8.5% variance in the dependent variable.
6.4.5.1 Model 1:
Next, based on the knowledge from single set regressions, a full model was run, aimed at increasing R2 in each block and examining the predictive strength of perceived descriptive norms
for the 5 reference groups. Model 1 is presented below:
Block 1: Gender, age, year in college, living alone and independently
communication about alcohol, general communication about alcohol.
Block 4: Descriptive norm for close friends, descriptive norm for best friend
Block 5: descriptive norm for mother, descriptive norm for father
Model 1 explained 47.8% variance in the dependent variable. Block wise change in R2 is
presented next.
Block
1
2
3
4
5
R2 change
0.02
0.23
0.10
0.12
0.00
Table 23: Change in R2 for model 1 (Third set of regressions)
207
As can be seen from Table 23, block 5 did not improve R2 any further. Perceived descriptive
norms for both parents were therefore dropped from subsequent models.
The unstandardized regression coefficients for the significant relationships in model 1 are
presented next.
Gender
Age
Age of first drink
Unstandardized Regression Coefficients

(B)
Block Block Block Block Block
1
2
3
4
5
0.604 0.305* 0.505
-0.118 -0.074 -0.097 -0.057 -0.058
0.021 0.018 0.014 0.014
-0.162 -0.139
0.655
0.56 0.457 0.457
0.277
0.23 0.144*
0.28 0.113 0.113
0.267 0.267
0.112 0.112
Table 24: Unstandardized regression coefficients for model 1

(Third set of regressions)
More money available for drinking, older age and higher frequencies of general and weekly
conversations regarding alcohol were found to be associated with higher frequencies of
drinking enough alcohol to get drunk in a typical month. Being male and having started
drinking at a younger age were also associated with higher frequencies of drinking enough
alcohol to get drunk in a typical month. However, these variables lost their predictive effect
when perceived descriptive norms for proximal peers (close friends and best friend) were
entered into the model in block 4. Among normative mechanisms, higher perceptions of
208
prevalence of drinking enough alcohol to get drunk in distal and proximal peers was found to
be associated with higher frequencies of drunkenness. Perceived descriptive norms for close
friends were more strongly associated with personal drinking behaviour compared to those for
most students at DIT and best friend.
6.4.5.2 Model 2:
In order to examine the predictive power of perceived injunctive norms, another full model
was run with slight modifications. One, year in college and living alone and independently
were removed from block 1 as they were not found to be significant in model 1. Similarly,
drinking group size was removed from block 2. Two, block 1 and block 2 variables were
merged into one block, now block 1. Third, perceived injunctive norms for all peer groups
were included in block 2 and perceived descriptive norms for all peer groups (except for both
parents) were added in block 3. The model took the following form.
Block 1: Gender, age, money available for drinking, age of first drink, weekly
Block 3: Descriptive norm for most students at DIT, descriptive norm for close friends,
descriptive norm for best friend.
This model explained 49.8% variance in the dependent variable, 2% higher than model 1.
At this stage, standardized residuals were examined to ensure that the underlying distribution
did not deviate substantially from normality. The value for kurtosis exceeded the acceptable
range. To rectify this, standardized residuals outside +/-4 standard deviations were excluded.
There were 6 such cases. Model 2 was rerun and it was noticed that removal of these cases
209
improved R2 by 2.5%. The re-examination of standardized residuals revealed that kurtosis was
now within the acceptable range. A square root transformation of the dependent variable was
also attempted to bring down the value of kurtosis even further. The details of this procedure
and relevant comparison plots are provided in appendix 7. Given similar results and negligible
increase in R2, untransformed values for regression coefficients are reported here for ease of
interpretation.. Overall, the model was able to explain 52.6% of the variance in dependent
variable. Block wise change in R2 and unstandardized regression coefficients follow
Block
1
2
3
R2 change
0.259
0.058
0.209
Table 25: Change in R2 for model 2 (Third set of regressions)
Gender
Age
Age of first drink
Unstandardized
Coefficients (B)
Block 1 Block 2 Block 3
0.284*
-0.098 -0.093 -0.077
0.022
0.019
0.014
-0.188 -0.125 -0.032
0.605
0.506
0.358
0.314
0.216
0.384
0.161*
0.362
0.186*
0.05*
0.106
0.275
0.183
Table 26: Unstandardized regression coefficients for model 2 (Third set of

regressions)
210
Table 26 indicates four key results. First, more money available for drinking, older age, having
had the first alcoholic drink at a younger age and weekly and general conversations regarding
alcohol were found to be associated with higher frequencies of drunkenness in a typical month.
Being male was also associated with higher frequencies of drunkenness but the effect lost its
significance once the normative mechanisms were entered in the regression equation in block
2. Second, higher perceptions of injunctive norms for proximal peers were found to be
significantly associated with higher frequencies of drunkenness. Third, perceived injunctive
norms for close friends lost their predictive power when descriptive norms were entered into
the equation in case of other dependent variables. At the same time, higher perceptions of
injunctive norms for mother were found to be associated with higher frequencies of
drunkenness in block 3, an effect which wasnt noticed in earlier blocks. Fourth, higher
perceptions of descriptive norms for distal and proximal peers were found to be significantly
associated with higher frequencies of drunkenness with the predictive strength of the norm for
proximal peers being greater than that of distal peers.
6.5
Conclusion
Ireland is known for its infamous drinking culture which is a key element of social life for its
people. The results presented in this chapter demonstrate the prevalence of heavy drinking
among DIT students with the campus norm being drinking approximately 5 days a month,
consuming nearly 9 drinks on a typical occasion and getting drunk every week. As explained in
chapter 5, section 5.5.6.2, the measures for personal consumption and perceptions of the
descriptive norm used in this study are largely similar to McAlaney and McMahon (2007). The
comparison of the campus norm for DIT with that of the sample from McAlaney and
211
McMahon (2007) shows that DIT students consume twice as many drinks on a typical occasion
and get drunk twice as frequently in a typical month as their British counterparts. The results
also confirm the relevance of underage drinking in Irish policy making as it was found to be
commonplace in DIT with 94% of those surveyed reporting that they started drinking before
they turned 18.
The findings demonstrate that the respondents generally overestimated the prevalence of
drinking among other students perceiving them to drink almost twice as often and getting
drunk twice as frequently as they themselves in a typical month and consuming at least 1
additional drink on a typical drinking occasion. The extent of these misperceptions is
comparable to the results reported in McAlaney and McMahon (2007) and to those reported in
American college based studies such as Thombs et al (2005). These two studies are broadly
similar to the current research as both were conducted at campuses primarily attended by
students who commute to campus. Overall, the respondents in the current study and both
McAlaney and McMahon (2007) and Thombs et al (2005) perceived other students as
becoming drunk nearly twice as frequently as they themselves. Further, the results suggest that
the respondents also perceived their proximal peers to have more permissive drinking
behaviours than they themselves. However, the gap between participants own drinking
behaviours and their perceptions of the behaviour rose as social referents became more distal.
These trends were found to be consistent for both males and females, abstainers as well as
across all campuses of DIT.
The results presented in this chapter also provided evidence to support the effect of perceived
descriptive norms on personal consumption of alcohol. This effect was found to be stronger for
212
proximal peers than distal peers. Further, perceptions of parental norms were not found to have
a significant effect on participants own consumption. These results support the past research
which documents misperceptions of norms and investigates their association with personal
behaviour (Perkins, 1985; Berkowitz and Perkins, 1986a; Marks et al., 1992; Wood et al.,
1992; Beck and Treiman, 1996; Perkins and Wechsler, 1996; Thombs et al., 1997; Nagoshi,
1999; Page et al., 1999; Clapp and McDonnell, 2000; Botvin et al., 2001; D Amico et al., 2001;
Sher et al., 2001; Korcuska and Thombs, 2003; McAlaney and McMahon, 2007; Page et al.,
2008). The hierarchical multiple regressions also revealed some other effects. Specifically,
having more money available for drinking and higher frequency of engaging in drink related
conversations were associated with higher personal consumption. Being male and having had
the first drink at a younger age were found to be associated with greater number of drinks
consumed on a typical occasion. Living in a shared accommodation predicted higher frequency
of drinking in a typical month while being older predicted lower frequency of drunkenness in a
typical month.
Further, the results presented in this chapter provide evidence in support of perceived
descriptive norms having a greater effect on personal consumption than perceived injunctive
norms. Although, higher perceptions of approval for drinking to get drunk among proximal
peers (perceived injunctive norms) were found to be associated with higher personal
consumption, these effects disappeared when perceived descriptive norms were introduced into
the regression equations. The only exceptions were the perceived injunctive norm for proximal
peers and mother when they predicted individuals frequency of drunkenness in a typical
month. Further, perceived descriptive norms were able to explain a much higher proportion of
variance in personal consumption compared to perceived injunctive norms.
213
The implications of these findings and their contributions to the SN theory would be discussed
more fully in chapter 9.
214
Network Analysis and Results: Second Wave
7.1
Introduction
SN researchers consistently emphasize the importance of identifying significant social

referents and placing these associations in appropriate contexts to understand how and why
they are so salient (McAlaney et al., 2011). These are important questions but the complex and
multifaceted nature of human relations makes it difficult to examine these details of
individuals social lives following conventional survey methodology. SNA offers a way to
address this issue of norm salience as discussed in detail in chapter 4, section 4.4. The
powerful mathematical techniques embedded in graph theory and the ability to visualize
networks is the strength of network science which also makes it the best available method for
examining relational patterns.
The ego networks examined in this study were generated from 26 in depth interviews as
explained in chapter 5, section 5.6. The current chapter is aimed at addressing the issue of
norm salience by identifying and locating the most salient peers in the extracted ego networks.
The results presented in this chapter are based on the analysis of network composition and
structure combined with a qualitative interpretation of the interviews. It begins by providing an
overview of the sample and describing the composition of examined networks. It then assesses
the strength of network relationships by examining participant generated tie strength scores and
the subjective meanings associated with these ties. An analysis of structural configurations of
these networks is presented in parallel. In this regards, five distinct structural typologies are
215
identified and explained with the help of specific cases and network visualizations. The chapter
concludes with a brief discussion of the key findings.
Chapter 8 will examine the association of these salient relationships with participants
normative perceptions and their drinking behaviours. In doing so, it will also extend the
discussion of cases described here to explore how drink related behaviours and attitudes
developed and sustained in these networks. A cross case analysis will also be presented.
7.2
Sample Overview
Of the 26 students who participated in the interviews, there were 15 females and 11 males. The
average age of the sample was 20 years, the average age of females being 20 years and that of
males being 21 years. This difference is due to the fact that more females than males had spent
2 years or less in college. All participants were Irish and full time under graduate students at
DIT Aungier Street campus. 18 participants belonged to the high drinking cohort, 7 to the
moderate drinking cohort and only 1 participant belonged to the low drinking cohort. The
observation that only 1 participant belonged to the low drinking cohort is not surprising as only
3 respondents (from the web survey) based at Aungier Street campus fit the criteria for low
intensity drinkers only 1 of whom agreed to participate in stage 2.
The cut off points for these cohorts were determined on the basis of the criteria outlined in
chapter 5, section 5.6.1 and reproduced in Table 27 for convenience.
216
Moderate intensity drinkers
High intensity drinkers
No. of drinks (per

occasion)
Frequency of Drunkenness
Once a month or less frequently

Exceeded low intensity drinkers on one or more of these items but
did not meet the criteria for high intensity drinkers
5 or more
Once a week or more
Once a week or more
drinks
Once a month or less
2 drinks or less
Table 27: Drinking Intensity Criteria for the Current Study
7.3
Network Analysis
Before going into the structural and contextual details of ego networks, it will be useful to
develop a feel of the data by providing a general overview of network composition. Three
networks were examined for each participant based on the name generators used in this study.
The choice of these name generators is described in chapter 5, section 5.6.3.2. The first name
generator extracted networks of people with whom the participants shared important matters
(referred to as the important discussants network). The second name generator produced
networks of people with whom the participants liked socializing or partying (referred to as the
socialization network). The third name generator asked the participants if they wanted to name
anyone else who was especially close to them and who they did not mention in the first two
questions. Based on the data from these name generators, an overall ego network was
generated for each participant, which included everyone that they had named during the
interview (referred to as the Overall network).
UCINET software (Borgatti et al., 2002) was used to conduct network analysis in this study.
UCINET stores network data in matrix format. For illustration purposes, the following image
depicts the overall network of a participant named Ruth in matrix format.
217
Figure 10: Example of UCINET matrix20
The matrix indicates that Ruths network comprises 11 people (persons 71-711). The presence
of 1 in a cell indicates that a tie exists between the corresponding nodes whereas 0 depicts
that there is no relationship between the two nodes. Similar matrices were created to represent
the important discussants, socialization and overall networks of each participant. These
matrices were then used to calculate specific network measures in UCINET (Borgatti et al.,
2002) and to create network visualizations in NetDraw (Borgatti, 2002) which were then
assessed in combination with interview accounts to understand how these networks functioned.
20
Based on data from the present study

218
7.4
Overview of Network Composition
7.4.1 Size and Effective Size of Ego Networks

A basic indicator of interest in SNA is network size. The size of a network is determined by the
number of direct ties involving individual units (Marsden, 1990). In simple words, it is the
number of members in an individuals network referred to as degree in network terminology.
Effective size on the other hand, is a measure of network cohesion. It describes how cohesive
or connected a given network is. It is calculated as
=

(

)

A higher effective size means that a network is less connected. A lower effective size
conversely represents a network that is well connected with 1 being the highest value
indicating that everyone is connected to everyone in the network.
For clarity, consider the following two cases
219
Figure 11: Effective size in a 3 mode network
In the first case A has ties to three other actors. None of these three actors has ties to any of the
others. The effective size of As network is 3-(0/3) = 3. Alternatively, suppose that A has ties
to three others, and that all of the others are tied to one another. A's network size is three, but
the ties are "redundant" because everyone can reach all three neighbours by reaching any one
of them. The total number of ties between alters is 6 (each alter is tied to two other alters).
Hence, the effective size of this network is 3-(6/3) = 1.
The size and effective size of the 26 ego networks examined in this study are presented in
Table 28. Pseudo names have been used to indicate each participant in order to protect their
identities.
220
Respondent
1
Mary
2
Emma
3
Helen
4
Ruth
5
Linda
6
Edel
7
Pam
8
Sue
9 Debbie
10
Meg
11
Fay
12
Bella
13 Katie
14
Amy
15
Lisa
16
Tom
17 Peter
18
Sam
19
Alex
20
John
21 Adam
22 Brian
23 Cormac
24
Gary
25
Rob
26
Ken
Drinking
Gender
Cohort
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
M
M
M
M
M
M
M
M
M
M
M
H
H
H
L
H
H
H
H
H
M
H
H
H
M
H
H
M
H
M
H
M
H
M
H
H
M
Socialization Important Discussants

Network
Network
Effective
Effective
Size
Size
Size
Size
20
13.4
20
8.6
16
9.9
16
9.3
17
11.1
9
2.3
5
1.8
11
6.6
17
8.9
9
2.3
10
3.6
11
3.6
15
9
7
3.6
6
1.7
3
1
13
7.8
8
3
12
6.3
12
5.2
13
5
17
7.2
9
2.6
9
2.3
12
5
11
1.7
14
8
12
6
11
6.5
13
4.1
20
8.8
10
5
10
6
3
1.7
19
8.6
13
5.2
4
2.5
4
2
16
4.6
15
4.7
10
1
14
1
13
2.7
6
2.3
13
10.1
10
7.4
6
2.3
11
6.3
13
8.5
12
8.3
8
1
8
1
Overall Network
Size
32
20
20
11
20
13
17
7
17
19
20
10
18
20
24
23
13
26
5
23
18
14
21
12
17
17
Effective
Size
20.3
12.8
11
6.6
10
5.2
10.9
2.4
9.4
8.6
10.5
2.4
6.2
12.3
8.1
10.6
8.7
12.8
2.6
6.3
3.7
4
15.5
7
10.8
4.2
Table 28: Size and effective size of the examined ego networks. Gender: Female (F),
Male (M), Drinking Cohort: High Intensity (H), Moderate Intensity (M), Low Intensity (L)
221
The average size of overall networks was 17.6 nodes21, where the smallest network
comprised 5 nodes and the largest comprised 32 nodes. This reflects a lot of variation (SD: 6.0)
in networks with respect to size. Overall, the important discussants networks of females
(mean: 11.2) were slightly wider compared to those of males (mean: 9.6). These results
contrast with Moore (1990) who found men and women to have much smaller (mean: 3) but
similar sized networks of people with whom they discussed important matters. However,
Moores study was based on a national sample as opposed to the current study which was
restricted to DIT students only. It can be argued that network size is dependent on several
factors such as choice of name generators, population of interest and their age group. The
socialization networks and the overall networks of males and females, on the other hand
were found to be similar with regards to size (means: 12 and 17 respectively for both males
and females).
Table 28 indicates that in most cases, the important discussants networks were found to be
more connected (Mean 4.5) than the socialization networks (Mean 6). This is because the
important discussants networks generally comprised older and stronger ties such as family and
friends with whom the participants grew up or with whom they went to school. Generally,
these people knew the participants and each other for many years and in contexts outside of
college. The socialization networks on the other hand typically comprised subgroups of a
variety of people with whom the participants interact in various contexts in their day to day
lives.
21
Please refer to the Glossary of Common Network Terms in appendix 9 for definition.
222
7.4.2 Referent Group Composition of Ego Networks

A variety of peer groups emerged in the ego networks of participants in response to the name
generators. The following table (next page) presents the percentage distribution of different
referent groups in the overall network of each participant.
223
Respondent
Mary
Emma
3
Helen
4
Ruth
5
Linda
6
Edel
7
Pam
8
Sue
9
Debbie
10
Meg
11
Fay
12
Bella
13
Katie
14
Amy
15
Lisa
16
Tom
17
Peter
18
Sam
19
Alex
20
John
21
Adam
22
Brian
23 Cormac
24
Gary
25
Rob
26
Ken
1
2
School or
Area or
childhood
Friends
%
41
5
55
18
15
47
47
43
54
37
35
60
57
35
67
31
43
20
30
39
79
28
58
College
Friends
%
Work
Mates
%
15
50
5
18
10
23
14
29
26
30
30
22
35
8
4
15
38
20
22
5
17
29
Flat
Mates
%
Family
%
10
19
20
15
27
35
15
12
14
12
16
10
10
11
10
8
4
23
11
20
13
44
25
15
5
15
52
31
24
10
29
59
19
25
13
35
Other
Friends
Outside
Col22
lege
%
Activity
Friends
23
16
25
37
25
15
12
29
5
29
21
15
5
5
17
9
31
8
40
35
17
21
14
29
6
Important
Discussants
Outside
College
(%)
Overlapping
Peers or
Nodes
Outside
College
(%)
95
50
89
82
100
91
88
100
88
75
71
67
100
75
91
90
100
77
75
67
100
100
100
91
67
100
100
50
86
100
100
100
80
100
75
75
70
63
100
67
88
86
100
50
67
63
100
100
100
83
56
100
Table 29: Peer group composition (%) based on overall network size.
22
The category other friends from outside college included (1) boyfriends or girlfriends who didnt fit any other
peer group category (2) friends of friends and (3) people with whom the participants became friends in a
particular context for example Ruth has a friend who she sees every Saturday in a dance class and Adam is
friends with three girls with whom he and his friends like clubbing.
23
The category activity friends included friends who the participants made through membership in a specific
club or activity outside of college for example Pam is member of a juggling society and Mary is part of a
hockey club outside DIT.
224
Table 29 draws attention to three key findings.

First, in general, the participants named more peers from outside college than from within
college. The only exception to this was Emma who named as many people from within college
as from outside.
Second, in general, the important discussants networks comprised ties from outside college. In
13 cases of the important discussants networks, 90% or more peers were from outside
college whereas in 9 cases, between 70 and 90% peers were from outside college. In the
remaining 4 cases, between 50 and 70% peers were from outside college.
Third, in general, people who were named in both important discussants and socialization
networks (referred to as overlapping peers or nodes in Table 29) were related to the
participants in contexts outside of college. This was consistent across all networks with 11
cases comprising 90% or more, 8 cases comprising 70-90% and 7 cases comprising 50-80% of
overlapping peers from outside college.
The finding that college friends did not appear as frequently and as dominantly in the ego
networks as did peers from outside college suggests the possibility that perhaps the latter are
more relevant and salient to the participants - a point that will be explored more fully in section
7.6. These results also validate the web survey where 86% of the sample reported having
friends from outside college among the people with whom they usually engaged in drink
related socialization.
70% participants mentioned either or both their parents among people with whom they shared
important matters and 61% mentioned their siblings in either or both name generators.
225
7.4.3 Gender and Age Composition of the Ego Networks

Table 30 indicates the number of network members named by gender during the interviews.
Number of People Named by Gender
Male
Alters
Female
Alters
Total
Absolute Value
% value
117
62%
72
38%
189
100%
Female Egos
Absolute Value
% value
82
30%
186
70%
268
100%
Total
Absolute Value
% value
199
43%
258
57%
457
100%
Male Egos
Table 30: Number of people named by gender. Total of values in rows: total amount of
people named by ego gender, Total of values in columns: total amount of males and
females named.
Overall, the participants named 457 alters; 199 were male and 258 were female. Male
participants (11 cases) named 189 alters of whom 117 were male and 72 were female. Female
participants named 268 alters of whom 82 were male and 186 were female. These findings are
consistent with Bellotti (2008) and reflect that overall, the participants named more same
gender friends than opposite gender friends.
The average age of network members was 24 years compared to an average age 20 years for
the participants. The networks reflected that participants mostly socialized with people their
own age although it may seem from the average figures that they interacted more with those of
226
an older age group. This is because many people mentioned their parents and other family
(such as older siblings, uncles and aunts) in their important discussants networks which
subsequently raised the average. Kinship ties may not be categorized as friends who
participants would usually socialize with. Yet, being important sources of advice and support
as reflected in section 7.6, they appeared to play a significant role in the participants lives.
7.5
Evaluation of Tie Strength
The strength of relationships that the participants shared with their network members was
examined to identify the most salient peers in the given networks. The evaluation of tie
strength was based on two sources of information.
First, during the interviews, participants were asked to position their network members on a
diagram of 6 concentric circles based on how close they felt towards each member (refer to
chapter 5, section 5.6.5.2 for a detailed discussion on the procedure and its rationale). This
exercise generated participant aided numerical scores of tie strength ranging between 6
(strongest tie) and 1 (weakest tie). The tie strength scores for each of the 26 networks are
presented in appendix 10. The people with whom the participants felt extremely close
comprised approximately 43% (7 people) of the size of their networks. This is supported by
past research which found most American people to have about 6 close social contacts
(Christakis and Fowler, 2009).
Second, the qualitative accounts were examined to understand the meanings and sentiments
participants associated with these relationships. This qualitative interpretation provided context
to the aforementioned numerical scores by offering useful insights into why participants
227
perceived some ties to be stronger and more intimate than others. This will be explained further
with the help of examples and network visualizations in the following section, which presents a
structural and contextual analysis of networks in parallel with an evaluation of tie strength.
7.6
Variations in Network Structures
The analysis of network structures complemented by interview accounts draws attention to five
main typologies of networks. This categorization was partially based on how the participants
described their networks, how they interacted and socialized with their network members and
strength of these relationships. The five network typologies encountered in this study were
1. The Small clique
2. The Overlapping cliques
3. The Group
4. The Core/Periphery
5. The Contextualized network
Four of these typologies namely the small clique, the group, the core/periphery and the
contextualized network were derived from (Bellotti, 2008), which is an ego network study of
friendship networks among single people in Milan. Since the current study elicited ego
networks based on discussing important matters and socializing, of which friendship ties
are an imperative element among others, the structural types outlined in Bellotti (2008) were
deemed suitable for our purposes. That being said, some adaptations were made to the
conceptualizations of these typologies to suit the current context, a point that will be addressed
228
shortly. The fifth typology namely, the overlapping cliques, emerged during the analysis as
distinct from Bellotti (2008) and the one which appeared the most frequently.
Table 31 summarizes the structure type of each ego network examined.
Respondent
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Mary
Emma
Helen
Ruth
Linda
Edel
Pam
Sue
Debbie
Meg
Fay
Bella
Katie
Amy
Lisa
Tom
Peter
Sam
Alex
John
Adam
Brian
Cormac
Gary
Rob
Ken
Drinking
Cohort
H
H
H
L
H
H
H
H
H
M
H
H
H
M
H
H
M
H
M
H
M
H
M
H
H
M
Socialization Network
Overlapping cliques
Overlapping cliques
Overlapping cliques
Overlapping cliques
Core and periphery
Overlapping cliques
Overlapping cliques
Overlapping cliques
Overlapping cliques
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Overlapping cliques
Overlapping cliques
Small Clique
Overlapping cliques
Group
Group
Contextualized Network
Core and periphery
Contextualized Network
Group
Important Discussants
Network
Overlapping cliques
Overlapping cliques
Overlapping cliques
contextualized network
Overlapping cliques
Overlapping cliques
Overlapping cliques
Small Clique
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Overlapping cliques
Small Clique
Overlapping cliques
Group
Overlapping cliques
Group
Group
Overall Network
Overlapping cliques
Overlapping cliques
Overlapping cliques
Core and periphery
Overlapping cliques
Overlapping cliques
Overlapping cliques
Overlapping cliques
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Group
Overlapping cliques
Overlapping cliques
Overlapping cliques
Small Clique
Overlapping cliques
Group
Group
Overlapping cliques
Core and periphery
Overlapping cliques
Group
Table 31: Summary of the structures found. Pink: Females, Blue: Males. Drinking
Cohort: High (H), Moderate (M), Low (L)
Some cases that best illustrate the typologies are presented next.
229
7.6.1 The Small Cliques

In social psychology, the word "clique" is often used to describe a group of 2 to 12 (averaging
5 or 6) people who interact with each other more regularly and intensely than with others in the
same setting (Salkind, 2008). The concept of a clique in network science is more narrow and
precise. It is termed as a complete sub graph of at least three nodes without an upper size limit
(Wasserman and Faust, 2007). Simply put, it is a group of at least three people where everyone
knows everyone. While a clique has to have at least three members, there is no limit on the
maximum size.
Small cliques as the name suggests, have two features which distinguished these networks
from the other typologies. First, the size of such networks was small compared to other
networks in the data set. The minimum network size for a small clique was 3 nodes while the
maximum was noted to be 5 nodes. Second, the nodes in such networks were arranged in the
form of one or more cliques where each clique had 3 members at the most. This conception
varies from Bellotti (2008) who found the maximum size of such networks to be 3 nodes based
on her data compared to that of 5 nodes as noted in this study.
Of all the participants, only Alex was found to have all three networks (important discussants
network, socialization network and the overall network) shaped in this way. Another
example of this typology is the important discussants network of Sue, an 18 year old heavy
drinking female who had been going to DIT for less than a year at the time of the interview.
Sue lives in a shared accommodation in Dublin and named 3 females in her important
discussants network who all know each other forming a clique. Sue knows all three of them
in contexts outside of college.
230
Sues important discussants network is depicted in the following visualization.
Figure 12: Sue's important discussants' network: A small clique
Person 1101 who is 5 years older than Sue is her elder sister. Person 1102, whom Sue has
known for 6 years, is one of her best friends from school. Person 1103, whom Sue has known
for 12 years, is another friend from school. Persons 1102 and 1103 have the same wider circle
of friends as Sue. It appears from Sues interviews that they all socialize together in a big
group which comprises as many as 20 people.
The tie strength scores indicate that Sue feels extremely close to all three alters. She gave a 6 to
persons 1101 and 1102 and a 5 to person 1103. Her interview account compliments these
scores. Sue sources emotional support and companionship from these girls and shares inner
thoughts with them. She interacts with them on daily basis and enjoys spending time with
them. It appears from her description of these friends that her relationship with each of them is
unique in a sense that each girl serves a distinct role in her social life.
231
Sues sister suffered from anorexia when she was younger as a result of being bullied on her
weight in primary school. Her teenage years were marked by excessive drinking after falling in
the company of a wrong crowd subsequently resulting in alcohol induced epileptic seizures.
She is a recovering alcoholic since the past 2 years. Sue has supported her sister since younger
years and encouraged her to overcome her eating issues and become a happier and a healthier
person. She feels she understands her sister like she understands herself.
Even though there is 5 years between us, its more like we are twins...I know her inside out...I know
what she hates...I know what she likes...I know what makes her uncomfortable...I know how to deal with
her when she is in a bad mood...she is basically me in another person.
Sue communicates with her sister everyday and sees her in person at least once a week when
they spend the day together doing varied activities such as going to the cinema, exercising and
talking about their lives. Sue feels that her relationship with her sister is based on mutual trust
and understanding where she provides emotional support most of the time such as by helping
her mother pick out her sisters food and monitoring her eating. However, she believes that her
sister will do the same when and if needed. For example Sue described a time when she was
very worried about her fathers illness and her sister had provided comfort by explaining things
and putting her mind at ease.
Sues interview account highlights that her sisters health is the most important matter that the
two of them often talk about.
Well obviously eating is one of the important matters and I think that has in a way helped our
relationship because I know that I can ask her anything. She knows she can tell me if she is struggling
with it and then Id jump on board and be like ok lets fix it now and I know that I can tell her about
232
anything like if I am having a bad day at college or if its just stressing about exams, she is always
there, I think it works the same way around because I am always there for her and I know that if I need
her she is in the background ready to jump
Person 1102 is a complete opposite of Sue. During the interview Sue described that person
1102 is more outgoing and sporty whereas Sue herself is not. Yet, person 1102 is her support
system and the first to know about any good or bad things that happen in her life.
She is my best friend, she is my other half. I can tell her anything under the sun. If I have a problem,
she knows it even before Ive told her nearly. I couldnt live without her. If anything happened to her
Id be devastated. I think wed be friends forever because I cant see me coping in life without her
The intimacy Sue feels for person 1102 is evident from her narration of several instances when
she had shared personal issues with her such as her fathers illness. Sue feels that person 1102
provided emotional support and comfort to her in return.
I just remember going into my exam and I left early because I just couldnt focus on it. I went down to
the toilet just sat there because I was afraid to go home and because I did not know what awaited me. I
could have gone home and my dad could have been passed away. It was just an awful moment. Person
1102 actually left her exam early and followed me down to the toilets and sat beside me. I just cried and
cried and didnt even say anything and she didnt even ask why I was crying. She just sat there beside
me and eventually I blurted everything out. She gave me a hug and that was all I needed
Sue derives gratification and fulfilment from her friendship with person 1102 and feels that she
has become more open about her feelings after having become friends with her. Person 1102
attends college in Galway and the two make an extra effort to keep in regular contact. They
233
see each other on weekends and talk every night on phone, their conversations typically lasting
an hour.
Sues relationship with person 1103 is multifaceted. They are childhood friends and have
grown up together. Since the start of college, they are also roommates and see each other
throughout the week which has strengthened their relationship in many ways. Sue describes her
relationship with person 1103 as being based on mutual exchange of social and emotional
support. For example, Sue accompanied person 1103 several times on her class nights out at
the start of college. She felt that her presence will help her friend socialize with others who
otherwise is a very quiet person. They often discuss relationships and family matters with each
other. For example, Sue talked to person 1103 about her sisters drinking issues. The fact that
they are both insomniacs offers them lots of opportunities to interact frequently and engage in
personal conversations.
I think its hilarious that we share a room because at 5 o clock in the morning Id turn
around and shed be awake and wed just go out and watch TV and its just nice to have
someone else up at that mad hour. If she cant sleep, Id get up and if I cant sleep, shed get
up. I think you are so vulnerable when you are up at that hour that you just talk about
whatever is bothering you
Sue also named persons 1102 and 1103 in her socializing network.
From the above description of the functioning of this small group, it is clear that this clique is
salient and comprises long standing intimate ties based on reciprocal exchange of social and
emotional support. At the same time, the ego acknowledges and identifies with the unique
contribution of each tie in her life.
234
Chapter 8, section 8.3.1.3 will explore how drinking behaviours developed in Sues network
and how are they reinforced by the network members.
7.6.2 The Overlapping Cliques

This typology represents a network structure that comprises multiple cliques connected to each
other. In some networks, these multiple cliques were connected by the overlap of common
nodes (people) (e.g. Toms network) whereas in others, people in different cliques happened to
know each other (e.g. Sams network). An important feature of this typology is that, generally,
the strongest ties were not confined to a single dominant group as will be noticed in the next
typology, the group. In fact, they were found to be scattered in different cliques and each
clique was unique in its significance to the participant.
This is the most common structure found in the examined networks. 16 socialization networks,
13 important discussants networks and 16 overall networks were shaped in this manner. There
were 11 participants who had all three networks (socialization, important discussants and
overall) shaped in this manner. 8 of these participants were females and 3 were males.
An example of this typology is Sams network, a 23 year old male student who had been in
DIT for a little over 3 years at the time of the interview. He lives with his parents and belongs
to the high drinking cohort. Sam named 13 people as those he shares important matters with
and 19 as those he likes socializing or partying with. His overall network comprised 26 people
(11 females and 15 males), 6 of whom appeared in both his important discussants and
socialization networks. These have been enclosed in parenthesis in Table 32. The tie strength
scores for his network are presented next.
235
Tie strength score

6
5
4
3
2
1
Alters
(61, 62), 63, 64, 65, 66, (611), 618
(67, 68), 617
(69), 613, 620, 625, 626
614, 615, 616, 619, 621, 624
610, 622, 623
612
Table 32: Tie strength scores for Sams network
As has been commented earlier, Sams network is an example of a case where the overlap
between multiple cliques occurs because some people located in different cliques know each
other. The following visualization shows his overall network
Figure 13: Sams overall network: An overlapping clique
The orange nodes encircled in orange, form a clique comprising his college friends who know
one another and with whom he interacts on a daily basis. Sam is a final year student and
236
therefore has known these people for over 3 years. His course in DIT has as many as 30
females and only 6 males which is why Sam feels that the lads tends to stick together.
Among these friends, Sam feels particularly close to three lads; persons 67, 68 and 69 as
depicted by Table 32. Sams relationship with them is largely based on companionship and
exchange of informational support such as talking about college related matters and seeking
related advice and opinion. He socializes with them regularly both in and outside college of
which nights out and drinking form a central part.
Theyd be the three Id be most close to and if we ever have a night out or I was having trouble with
college, theyd be the first three Id cal. A lot of things we do would be college based. While we are
editing and stuff in college, we find ourselves with a lot of free time so we would hang out and maybe
go shopping around town... just sort of hang out and be in each others company and then at the same
time theres nights out when we get together. Thats the kind of stuff we do
Sam also feels quite close to two girls, persons 617 and 618 who are also from college. Person
618 is enrolled in a different course at DIT yet Sam saw her a lot in the first two years of
college because she lived in the same area as him. More recently, she has moved to a new area
and the interaction between them has decreased. Person 617 is Sams classmate. She has gone
though some hard times as an adolescent such as dealing with bullying in younger years and
being homosexual. Sam admires her for being a strong character and having gone through
some common problems at the start of college, he feels he can relate to her. The two of them
have often talked about their deep personal matters and offered encouragement to one another.
However, their interaction is limited to college.
I dont see her as often as Id like but we have had quite in deep conversations where we have
confided an awful lot into each other and thats why I feel quite close to her. I was having trouble with
237
the first year in college and social anxiety as well and she had the same sort of thing. Thats something
we openly talked about, how we are dealing with it and how it affected us. She confided in me that she
was actually on medication for it. And that was very early on in our friendship. We also talked quite
openly about even her bullying and my experience with it and about her sexuality. Its not an easy thing
to talk about with someone but I always appreciated that she felt that she can do that with me and she
wouldnt do that with that many people , thats why I sort of get on well with her
The nodes encircled in blue form a second clique comprising people Sam has known for 15 or
more years. These are his school friends (blue) and his family (red). They all live in the same
general area in close proximity of each other. People with whom Sam shares the strongest ties
are embedded in this clique as is indicated by the tie strength scores in Table 32. Of this clique,
persons 61, 62, 63, and 613 form what Sam describes as his tight circle of friends whom he
has known since early teens and with whom he interacts on daily basis. These friends are also a
source of emotional and instrumental support for him. Sam recalled several occasions during
the interview when he had turned to them for opinion and advice on issues important to him
such as relationships, academics and career related matters.
For example, he described person 61 whom he has known for 15 years as a confidant and the
one person hed feel comfortable confiding in on anything.
If I was to say to him that I am really struggling with college or something of the sort, hed be the first
to help me and actually that reminds me, when I was doing my dissertation I was stressed time wise to
get stuff typed. He stayed up till 4 in the morning and he did it for me. So...thats the kind of character
he is
Similarly, when Sam failed the first year in college, person 62 who also goes to DIT (different
campus) was the first to know and offer advice. Lately, Sam has started accompanying person
238
62 to the gym where Sam helps him in his new fitness routine. Sam feels that his company
makes the gym atmosphere less intimidating for his friend.
Person 63 is a girl who Sam has been friends with for 15 years and with whom he also had a
relationship for two years. They are still good friends and Sam feels he misses her when he is
not able to talk to her. Over the years, he has discussed personal issues with her and feels
grateful for her support.
We were in a relationship for 2 and a half years so basically anything that I went through, I disclosed
to her, it was in the middle of this relationship that I went through this social anxiety thing and I was in
pretty deep to be honest. I was having trouble with my folks about it because I was kind of ashamed of
the fact that I was almost afraid of my own shadow and she was the first person that I talked to about
that
Person 613 is also a girl with whom Sam has been friends for 15 years. Sam described that he
seeks her advice and opinion whenever he has to shop for presents. More recently, he has also
begun to give her driving lessons. It appears from his interview that Sams relationship with
this tight group of friends is based on years of companionship, mutual trust, shared activities
and being able to confide in one another.
The other people in this clique include persons 611, 619, 620, 621 and 624 whom Sam
described as the jock group that he grew up with. These lads were always heavily involved in
sports during school years and are still on Sams foot ball team. Sams friendship with them is
based on having grown up together and sharing childhood memories. Much of the socialization
with these friends is drink related as will be described in more detail in chapter 8, section
8.3.1.1. Sam also plays foot ball with them 3-4 times a week. Among them Sam feels closest to
239
person 611 to whom he assigned a 6 on tie strength scores. Sam seeks his advice on sports
related matters and values his opinion.
I was having doubts about playing for a certain Gaelic football team and person 611 was the
first person that I confided in. Its an honour to play for this team and I was feeling two ways
about it. He just said ,Sam, if I could play for that team and I didnt, Id look back in 20 years
and think what else was I doing that was so important. It has stuck with me and its something
that I am going to pursue because of his advice
Sam is part of a youth club where he works as a volunteer. The youth club holds a special place
in his social life and he referred to it several times during the interview. The nodes which are
larger in size represent the people who are or have been part of the youth club at some stage.
This includes most of Sams school friends as well as his brother. It is apparent from the
visualization that the youth club is central to his interaction with most people from school. He
describes it as a cult which most of his school friends have been part of at some stage.
Its a difficult thing to explain to people because I am sure everybody feels very special about a youth
club they are involved in but our particular one was has kind of been a family. And its where I forged
as you can see my longest standing relationships and friendships
This older clique also includes Sams family; his parents and a younger brother. He described
his parents as very welcoming and considers their support in his life as constant. The
encouragement, emotional support and backing that they provide, means a lot to him. Sam for
example described how his parents supported him when he wasnt managing very well in
academics at the start of college and how his father encouraged his love of sports from a very
young age.
240
When I was younger, I wanted to play foot ball but I was too young to play for a team, so dad went
down to a club and started a team for people my age and he managed to find 10 kids from 6 to 7 and
hed coach them. Every week hed organize parents and get them to matches, lined the pitches, put up
the flags, put up the nets and stuff and he coached me for 10-15 years and then he coached my brother
Sam also feels quite close to his brother with whom he shares several activities such as playing
football on the same team and being involved in voluntary work in the same youth club. Sams
brother seems to know a lot of Sams friends through the youth club.
In addition, there are two more people in Sams network. The green node is the person who
founded the youth club and whom Sam holds in high esteem as he initiated him into youth
work. The yellow node is a counsellor at college who helped Sam deal with some social
anxiety issues at the start of college. Sam still goes to her for counselling from time to time.
The nodes boxed in red show the overlap between the two cliques from college and school.
Several people from the two cliques know each other. For example, person 69 from college
lives in the same area as Sam and has known him for 10 years. He knows many of his school
friends who also live in the vicinity. Likewise, two of his school friends namely persons 61 and
62 are ex and current DIT students respectively (though at a different campus than Sam) and
know his friends from college. Because of these friendships across the two cliques, the two
groups (college friends and school friends) often get to interact with each other in night clubs
and social events such as birthdays. However, it appears from the interview that Sam also
socializes with these two groups separately and sources different types of support from them.
Though he feels very close to some college friends as described earlier, Sam only discusses
college related matters with most of them and much of the socialization is limited to college
241
based activities and class nights out. The older clique of school friends and family on the other
hand forms his support system where he feels comfortable confiding about personal matters.
He socializes with these friends in several ways such as playing football, going to the gym,
playing computer games in each others houses and activities based around the youth club.
Thus while he engages in drink related activities with both his college friends and school
friends who serve unique purposes in Sams social circle, it is clear that the most salient ties
are embedded in the cohesive group of older friends outside college.
The discussion of Sams network will be extended in chapter 8, section 8.3.1.1 to assess how
drinking norms developed and sustained in his network.
Another example of this typology is the network of Meg, a 19 year old moderate drinking
female who has been a student at DIT for 2 years and lives with her parents. Meg named 19
people in all of whom 4 were males and 15 were females. Her important discussants and
socialization networks comprised 12 people each. 8 people overlapped between her important
discussants and socialization networks (enclosed in parenthesis in Table 33). The tie strength
scores for Megs network are presented below.
Tie strength score
6
5
4
3
2
Alters
(1304), 1308
1301, (1302, 1303, 1305), 1312
(1306), 1307, (1309, 1310, 1311)
1313, 1314, 1315, 1316, 1317, 1318
1319
Table 33: Tie strength scores for Meg's network
Megs overall network is shown in the following visualization
242
Figure 14: Megs overall network: An overlapping clique
The visualization shows two distinct cliques. The nodes encircled in pink represent 5 class
mates from college, all girls, whom Meg has known for 2 years and with whom she interacts
daily. They spend a lot of time in each others company in college and sometimes socialize
outside college settings such as on class nights out and in the cinema. Among these girls, Meg
feels closest to persons 1306, 1307 and 1311 whom Meg named in both her important
discussants and socialization networks. During the interview, it emerged that like Meg, these
three girls have boyfriends and Meg feels she can talk to them openly about relationships and
boy friend troubles. She described a time when her boyfriend (person 1304) had broken up
with her for a while and she had discussed it with these friends who were very supportive and
243
offered encouragement in return. Megs relationship with her college friends is largely based
on social companionship.
The nodes encircled in blue represent the people who Meg knows from outside college. This
clique includes her school friends (green nodes), some friends she has made through her
involvement in a voluntary organization (blue nodes and large nodes) and some family
members (orange nodes). These people live or have lived in the same general area as Meg as a
result of which they all know each other and interact frequently.
There are 6 friends from school in this clique, all females with whom Meg communicates and
socializes regularly. She takes a holiday with these friends and person 1315 (who lies outside
the clique) every year when they all go down to a holiday home that Megs family owns in the
country side and spend time together doing varied activities. Among these girls, Meg feels
closest to person 1302, who she has known for 15 years and persons 1303, 1305 and 1312
whom she has known for 7 years. She interacts with them frequently as a group as well as
individually. For example person 1302 though lives in another town, works in the same
company as Meg as a result of which they get to see each other a lot in relation to work. Person
1305 works for the same voluntary organization as Meg and they interact a lot during trainings
and related events. Person 1312, who Meg described as the first friend she made in secondary
school currently resides outside Ireland in relation to her studies, yet Meg keeps in touch with
her regularly and they write letters to each other. Meg described an occasion when she felt
overwhelmed because person 1312 came home especially for Megs surprise birthday party.
During the interview, Meg described that she feels she can discuss most matters with these
friends and that she relies on them for advice on relationships and family issues. Persons
244
1309 and 1310 are also Megs school friends who she has known for 9 and 4 years
respectively. They both have boyfriends and Meg feels they are good to talk to about
boyfriend related issues. She also feels she can discuss family related issues with person
1310 who has some family problems herself which she often shares with Meg.
All male friends in her network come from the voluntary organization that she is involved in.
Of these lads, she feels closest to person 1304 who has been her boyfriend for a little over 2
years. Meg interacts with him frequently doing varied activities such as going to the gym,
sitting in and watching TV, going to the cinema and socializing on nights out and on civil
defence related activities. Meg feels she can discuss anything with him and values his advice.
She described how he had organized a surprise birthday party for her and made sure all her
friends were there. Person 1316 is a friend of Megs brother whom she has known since 10
years. He is also in a relationship with Megs school friend (person 1309). During the
interview Meg described him as an older brother. Besides the voluntary work, they often
play football together. The other two lads (persons 1318 and 1319) are friends she interacts
with at the voluntary organization and with whom she also socializes in group settings.
Though, she does not feel as close to them as some of the other friends in the clique, Meg
described them as good friends, whom she values and trusts. She also described how their
lives were complicated by family related issues and how they deal with it suggesting that these
friends share personal matters with her.
Meg belongs to a big family comprising her parents and 8 siblings. Of her family, she
mentioned only her mother (person 1302) and two sisters (persons 1308 and 1317). Meg
described during the interview that she feels very close to her mother with whom she can talk
245
about anything. Person 1308 is her sister who is 18 years old and whom Meg described as a
loner and someone who hates drinking and crowds. She resides out of town in relation to
her studies but Meg talks to her several times during the week. Meg socializes with her sister at
trainings related to her voluntary organization as she is also a member. Person 1317 is Megs
older sister who works out of country. Meg does not feel very close to her as she has lived
away from home for a very long time. They communicate once in a while via email.
Unlike Sams network described earlier, Megs friends from college do not know most of her
friends from outside college except person 1309 who knows two of her college friends
(persons 1306 and 1307). As a result, there is very little interaction between the two cliques.
Meg socializes with them separately and it appears from the interview that her strongest and
most intimate ties are those she has known since many years generally embedded in the bigger
clique outside college.
It is clear from the functioning of this network that of all network members, Meg feels the
closest with her group of female friends from school and her boyfriend. Her relationship with
these people is based on frequent interaction, reciprocal exchange of support, mutual confiding
and shared activities.
Chapter 8, section 8.3.1.2 will describe how drinking behaviours developed in Megs network
and how these norms are reinforced by the network members.
7.6.3 The Group

This typology represents a network which comprises a predominant group of people in a
participants social life. The group was found to provide social companionship and support
246
and came across as particularly salient in the focal individuals life. In networks of this type,
the group generally featured a whole entity where single ties were less important than the
group itself. However, within the group, sub groups sometimes arose from shared interests as
will be discussed later with examples. In a network of this type, the strongest ties were usually
found to be embedded in the group. The group generally existed outside college settings
and comprised older friends, family or a combination of both. In some cases, the group also
included people who were comparatively recent friends yet, had become very close to the
participants and were well integrated in the wider network.
An important distinction between the group and overlapping cliques is that while the latter
comprises multiple cliques (where everyone knows everyone), the group is dense where most
people knew most others but it does not necessarily have to be a clique - a point that will be
clarified with examples shortly.
This is the second most common typology noted in the data set. 6 important discussants,
socialization and overall networks follow this typology. There were 5 participants (3 females
and 2 males), who had all three networks (important discussants, socialization and overall)
shaped in this manner.
An example of this typology is the network of Lisa, who is a 20 year old female doing a
business degree in DIT since three years. She lives with her parents and belongs to the high
drinking cohort. Lisa named 24 (14 females and 10 males) people in all. The tie strength scores
for her network are illustrated in the following table. The people who appeared in both her
important discussants and socialization networks are enclosed in parenthesis for identification
.
247
Tie Strength scores

6
5
4
3
2
1
Alters
(2601, 2602, 2603, 2604, 2605), 2606, 2607, 2608, 2618
2611, 2617, 2619
2610, (2612, 2613), 2620
2614, 2616, 2621, 2623
2615, 2622, 2624
2609
Table 34: Tie strength scores for Lisas network
The following visualization represents her total network
Figure 15: Lisas overall network: A group
The group in this network is comprised by 16 friends who Lisa has known from school and a
friend who lives close to her. These people are represented by the red nodes which have been
encircled in black for emphasis in the above visualization. The group is predominantly
female. There are only 3 male friends. Also, as commented earlier, the group does not
248
necessarily have to be a clique (where everyone knows everyone). In this example the group
is very dense but not everyone is connected to everyone. For example, persons 2620 and 2624
do not know person 2607. Similarly, person 2621 does not know persons 2607 and 2606. Yet,
these friends are part of the group because they know most of the people in it.
Lisa has known some of these friends longer than the others. Generally, the duration of
friendship in this group ranges between 7 and 14 years. There is variation in the extent of
closeness that Lisa feels towards her friends in the group. There is a sub group of 5 girls within
the main group that she described as her tight group of girls and with whom she feels the
closest.
I have a group of girlfriends and that would include persons 2601, 2602 and 2606, 2607 and 2608.
We all live in the same area so we have known each other since primary school and I would be really
close to them. Person 2602 is my best friend and persons 2601 and 2608 are best friends with each
other while persons 2606 and 2607 are also best friends with each other. We meet up like all the time,
the six of us
Their relationship is strengthened by years of companionship, shared activities and reciprocal

exchange of emotional support. Their families also know each other. 4 of these girls go to
Trinity which is located close to DIT. Lisa interacts with them several times a week during her
college breaks and at home. They all hang out together, go on holidays together and socialize
by engaging in both drink and non drink related activities. Lisa recalled several of these
holidays during the interview and described how much she and the girls enjoyed them. At the
time of the interview, the girls were planning to go to Australia for a year. Lisa feels that these
friends know everything about her and that she wouldnt have any problem talking to them
249
about anything. Person 2601 is Lisas neighbour and the two have known each other since
they were 8 years old. She goes to Trinity and knows her school friends as most of them also
go to Trinity.
During the interview, Lisa spoke about her 6th year holiday across Europe in celebration of
clearing her leaving certificate. Many friends from the group such as persons 2602, 2603,
2610, 2617, 2618, 2619, 2621 and 2624 accompanied her on the trip which lasted a month.
Lisa described that during this time she and her friends travelled together, stayed in hostels,
met new people, enjoyed nights out and did a lot of sightseeing. Lisa feels that she has gotten
closer to the group after the holiday because she interacted a lot with these friends on the trip
and feels that she knows them better now. For example she described how her relationships
with some of the friends such as persons 2603 and 2617 strengthened after the trip. Lisa gave a
5 to these friends on tie-strength scores.
I knew person 2603 before but not as closely as I do now because there were only 6 of us like the
whole time (on the holiday). Person 2602 who is my best friend left after two and a half weeks to go to
the (holy) walk in Spain so I got really close to person 2603 whose best friend was in another country
at the time. We got really close because our best friends werent there. Similarly, I got really close to
person 2617 because she was also in the group that I went on holiday with and I didnt know her that
well before like I obviously knew her I was friends with her but afterwards, when I came back, I was
really close with her
The three males in the group (persons 2620, 2621 and 2624) are also Lisas school friends.
Of them, she feels closest to person 2620, whom she described as her oldest guy friend and
who she has known since she was 3 years old. He does not live in the same area any more
250
however, Lisa still stays in contact with him and they often meet up in their free time and at
group get-togethers.
Lisa described that several people in the group were best friends with one another. This also
explains why the group is very closely knit. The following visualization shows an enlarged
image of the group where people who are best friends with each other are represented in
similar colours.
Figure 16: Enlarged Image of the 'group' in Lisa's network
251
Outside of the group, there are two recent college friends who Lisa has known for less than a
year. She only socializes with them in college or at class nights out, and discusses college
related matters with them. She does not however share personal matters with them. These
college friends also do not know anyone from the group.
Lisa feels very close to her two brothers namely persons 2604 and 2605 and thats the only
family that she named.
I dont know as far as brothers and sisters go Id say I am quite close to them. Recently, I would go
out with person 2604 and his friends or like I would see them out you know and I would hang around
with them or Id go to lunch with him and his girlfriend
In addition, there are some friends outside the group that she knows through people in the
group. They sometimes come out with the group and spend time together. Lisa only socializes
with them in the group rather than on her own.
Lisas interview account reflects that the group is predominantly salient to her in several
ways. Her strongest friendships are embedded in the group. She confides in these friends and
values their opinion. Not only does the group provide her companionship and support, it also
provides her with fulfillment as she does not feel the need to form new associations or
introduce her recent contacts to the group. For example, she has been in college for three years
and yet she didnt mention anyone from college that she has known for longer than a year.
Also, her college friends do not know anyone in her wider network nor does Lisa share deep
personal issues with them.
Chapter 8, section 8.3.2.1 will examine how drinking behaviours developed within the group
and how do members reinforce these norms.
252
7.6.4 The Core/Periphery

This typology represents a network with a highly dense core (density = 1) and less dense
peripheries that interact with the core to a greater or lesser extent (Borgatti and Everett, 2000).
In simple words, a core characterizes a group of people where everyone knows everyone
(hence density = 1), while the peripheries are connected to the core but not to each other.
Following the classification of (Bellotti, 2008), the core is conceptualized as a very cohesive
subgroup that provides multiplex24 support and whose ties were older than the others. The
peripheries on the other hand are generally less dense and represent ties that are more recent
and weaker than those in the core. Also these ties are largely dependent on the context in which
they were formed and sustain. Two participants, Linda and Gary, had their socialization and
overall networks based on this model.
Gary is a heavy drinking, 23 year old male, studying Art in DIT since 4 years. He lives in a
shared accommodation in Dublin. Gary named 11 people in all (3 females and 9 males). His
socialization network comprised 6 people all of whom also appeared in his important
discussants network which comprised 11 people.
The tie strength scores for Garys network are tabulated below with the nodes enclosed in
parenthesis representing people who appear in both the important discussants and
socialization networks.
24
Multiplexity in network science means an overlap of roles, exchanges or affiliations Wasserman, S. and K.
Faust (2007). Social Network Analysis: Methods and Applications, Cambridge Univ Pr..
253
Tie strength scores

6
5
4
3
2
1
Alters
(2301, 2304), 2307
(2308, 2309)
(2310)
(2302), 2312
2306, 2311
2303, 2305
Table 35: Tie strength scores for Garys network
The following visualization (Figure 17) shows his overall network
Figure 17: Garys overall network: A core/periphery
Here we find a cohesive core of friends encircled in blue, which Gary described as his close
group of friends and less dense peripheries with whom he interacts in a certain context. The
254
core comprises 5 lads who he knows in contexts outside of college and who overlap between
the important discussants and the socialization networks. Gary interacts with them several
times a week. Among them he describes persons 2301 and 2304 as his best friends. These are
Garys longest standing friends who he has known for over a decade and with whom he shares
inner thoughts and personal matters. Gary shares his love for American foot ball with person
2304 who is a friend who he knows from school. Gary does not get to see him as often as the
others because person 2304 works full time. He also has a steady girl friend (person 2312)
since 5 years with whom he spends most of his time. Yet, Gary confides in him and person
2301 and relies on them for an honest and candid advice on matters that are important to him
and in turn values their opinion.
I am planning to do masters next year and its coming down to the time when I have to start applying
where I want to go and I met person 2304 last week and we were able to discuss. He gave me his
outline, what he thinks Id be good at. If Id go that I want to do law or be a doctor, hed go I have
known you, you dont like those kind of things, when it gets down to it you are very good with people
and you are very good with business but you wouldnt like to sit and do 4 years and be a doctor. I can
only imagine himself and person 2301 to be able to say it to me...to tell me the truth rather than other
people whod just fill my head like my mum or the guys that I dont know as well like persons 2302 and
2303. These two lads (persons 2301 and 2304) will tell me what I need to hear rather than what I want
to hear
Person 2301 is Garys childhood friend with whom he has grown up. Although they went to
different schools, the fact that they were going to the same college brought them closer and in
the past five years they have been inseparable. Their relationship is strengthened by several
shared activities, for example, person 2301 lives close to Gary, plays on the same rugby team,
255
goes to the same college (DIT) and the same gym and their moms work together. They interact
daily and go on holidays together every summer without an exception. Gary described during
the interview that he feels he can confide in person 2301 on any matter from college to
relationships and family. Their relationship of being able to confide in one another and look
out for each other is mutual and reciprocal. They also socialize in the same extended group of
friends most of whom are person 2301s school friends including the remaining 3 members of
the core namely persons 2308, 2309 and 2310.
Gary always hangs out with persons 2301, 2308, 2309 and 2310 together. The relationship that
he shares with these lads as a group is based on shared activities as much of the socialization
revolves around football, gym and going out. He also shares an interest in music, politics and
foreign affairs with person 2310. They often play music together or discuss current affairs at
length. Gary sees the lads at least once a week if not more.
The peripheries comprise two friends from college (persons 2302 and 2303), two neighbours
(2305 and 2312) and 3 family members (persons 2306, 2307 and 2311). The relationship with
the peripheral friends is largely dependent on the context. The college friends for example
provide informational support in college related work. Gary only sees them in college or on
college nights out, and discusses course related matters with them. He met these friends on
Erasmus in Spain and shared an apartment with them. Person 2301 came over to visit him and
thats how he also knows Garys college friends. Among college friends, Gary feels closer to
person 2302 because he got to spend more time with him compared to person 2303, who has a
girlfriend with whom he spends most of his time. Person 2302 is the only person from outside
the core who appears in both the important discussants and socialization networks.
256
Of neighbours, person 2312 is person 2304s girlfriend and Garys closest female friend. She
has a large group of friends and Gary often comes across her on nights out where they usually
chat briefly. Person 2305 is a girl he has known since childhood. As children they always hung
out together and Gary also had a relationship with her for a short while. Now, they are not as
close and mostly interact through social networking websites. While their relationship has
changed in terms of support and frequency of contact, Gary values her and mentioned her as
one of the people he likes socializing with. This agrees with Bellotti (2008) in that friendship
cannot always be reduced to the kind of support it provides. The relationship history must be
taken into consideration.
She is just another neighbour Ive had for a long time and someone I can talk to if I have a problem.
We wouldnt be that close. She is just one of those people that you dont see that often but whatever you
do its like nothing has change. As in with some people if I havent seen them in a while, you grow
apart. But with her, its like youve seen her yesterday even if its been two months. Nothing changed
Among his family members, Gary feels closest to his younger brother who is 19 years old.
They spend a lot of time together at home doing various things. Gary describes his family as
very sporty and he and his brother often support each others teams in foot ball and rugby. His
relationship with his family members largely revolves around discussing issues as a family and
talking about fitness and sports.
It is apparent from the tie strength scores and Garys description of his network during the
interview that the people who are most salient to him lie outside college settings and embedded
in a very cohesive core. All of them also appear in both his important discussants and
257
socialization networks.
Chapter 8, section 8.3.3.1 will explore how drink related norms
developed in Garys network and how are they socially reinforced by the members.
Another example of this typology is the network of Linda, a heavy drinking 19 year old female
who lives with her parents and has been studying business and management in DIT since 3
years. Linda named 9 people as those she shares important matters with and 17 as those she
likes partying or socializing with. Over all, her network comprised 20 people 6 of whom
appeared in both the important discussants and socialization networks. From gender
perspective, her network comprised 8 females and 12 males.
The tie strength scores for Lindas network are tabulated below with the overlapping nodes in
parenthesis.
Tie strength scores
6
5
4
3
2
Alters
81, 82, (83, 85)
(84, 86, 87), 89
(88), 812, 813, 816, 819
810, 811, 814, 815, 818, 820
817
Table 36: Tie strength scores for Linda's network
The following visualization (Figure 18) shows her overall network.
258
Figure 18: Lindas overall network: A core/periphery
A very cohesive core comprising Lindas family (red: parents and two cousins), school friends
(pink), boy friend (blue) who she met on a pub crawl and a work mate (orange) has been
encircled in thick black.
Linda feels extremely close to her mother (person 81) and step father (person 82) whom she
regards as her real dad. She described that her parents are really supportive and that she
shares inner thoughts and personal issues with them and values their advice and opinion. For
example, during the interview she described a time in secondary school when she had felt that
she had no friends. Linda feels that her step-dad let her pour her heart out at the time and
offered emotional support and encouragement to cheer her up. She also explained how her
259
mother takes care of her despite ill health. She described a time when she was passing through
a depressive phase in college and nearly decided on dropping out. Linda feels that her mothers
support at the time helped her make the right decision which she appreciates.
I just wasnt happy at that time in college and I was regretting my decision of going to college and
mom just helped me get through it. Its very hard to describe... Id just come home and Id be upset and
Id be crying all the time and she would just come up to my room and talk for hours and even though I
didnt want anybody around me, shed just be there comforting me. She was very supportive but at the
same time she was like just think about it, its only for this time like it will pass. I only had a tunnel
vision at the time and all I could see was the worst but she showed me the light at the end of the tunnel
Lindas two cousins (persons 87 and 816) are almost like sisters to her as she grew up with
them doing various activities. Of these, she feels closer to person 87 with whom she socializes
frequently in both drink and non drink related ways. She also helped person 816 settle in
Dublin by introducing her to many new people that helped her get along. Of school friends,
Linda has known person 84 the longest. Her relationship with person 84 is nurtured by several
years of companionship and mutual trust. They grew up together, went to the same school and
shared many activities as teenagers. Although their schedules have become busier because of
college work, they still make a conscious effort to spend time with each other.
I have known her since I was 4 years old. We met on the first day of school and have done everything
together, every sport, and every dance class. Because we are so busy now, we try to have a girls night
at least once a month. Wed share it so one month wed go to her house and the next month itd be my
house. We spend around 4 hours getting ready (chuckles) and drinking before we go out and then we
just stay in each others houses and talk the whole night
260
Person 83 is Lindas new boy friend who she met on a pub crawl a few months ago. He is 26
years old and Linda feels extremely close to him. She socializes with him several times a week
in both drink and non drink related ways and confides in him on personal matters such as those
related to family and career. She also interacts with person 83s wider group of friends.
Person 85 is a female friend Linda made at work. She describes person 85 as her party buddy.
Though person 85 does not drink, Linda goes dancing with her regularly. They have gone on
two holidays together in the three years of knowing each other and discuss almost everything
with one another. Person 85 also introduced Linda to some of her own close friends.
Lindas description of the core relationships reflects the strong bond she shares with these
people. However, sometimes, conflicts arise in these relationships, yet the longstanding nature
of these ties keeps the core intact. For example Linda described a time when she believed that
her cousin (person 87) was deliberately creating a wedge between Linda and her step sister
(person 88) which she and her sister realized and sorted out. She also recalled an instance when
she caught her cousin (person 816) kissing her ex-boyfriend (person 813). Linda described the
incident as being stabbed in the back, but explained that she didnt confront her cousin for
the sake of their mothers who are sisters.
The peripheral friends are also encircled in the visualization. These are the people Linda came
to befriend through people in the core. Her relationship with these friends is based on specific
contexts. For example, the orange nodes are friends of person 85 with whom she works. She
met these friends on one of the holidays she took with person 85. She now socializes with them
on nights out. Whenever she goes out with person 85, theyd probably be there too. She
interacts with them once a week when their schedules match.
261
The blue nodes are friends of her boy friend (person 83). She met them through nights out and
much of her socialization with them is drink related. She interacts with them in group settings
and her boy friend is usually around as well. She wouldnt see these friends unless there is a
night out or a house party which would be about once a week.
The green node (person 817) is her cousins (person 816) roommate. Linda has only socialized
with him a few times at his apartment and described that she gets on well with him.
In addition to the peripheral friends, we see a clique outside the core. These are the people she
came to know through work (yellow). It includes her ex boy friend (person 813), whom she
described as a goth. He is also Lindas work mate and though they are not in a relationship
any more, he remains a good friend. Linda socializes with him at work and on nights out.
Person 812 is her exs (person 813) best friend and also works with Linda. Linda describes him
as a good friend who she sees on nights out as well as individually. She described that they
sometimes go for drives and often talk about their relationships as he also has a girlfriend.
They dont see each other a lot but whenever they do (once a month), Linda loves to party with
him. Person 819 is a work mate who also goes to DIT. He is a year ahead of Linda in college
and sometimes helps her with college work. She socializes with him at work as well as on
nights out. Linda spends half an hour with him every day after work while they wait for their
respective buses.
Some family is also visible outside the core. These are her step sister (person 88), a younger
cousin (person 818) and her uncle (person 89). Linda does not see them a lot. Her step sister is
married and has a child. When they were younger they were not very close but Linda feels that
her relationship with her sister has improved and they make an effort to get along with each
262
other for their parents sake. Her uncle is doing up scaling in DIT because of which she feels
she can discuss college related issues with him. He is married and has his own family. Linda
only socializes with him on family occasions. She described that when she was younger there
were a lot of family occasions to go to such as weekly dinners and Christmas at the great aunts
home however, now it is not as frequent. Then there is a younger cousin (person 818) who she
was much closer to in younger years. Now they have drifted apart mainly because of separate
social circles. They go out together very rarely.
The core in general is salient to Linda in several ways. First, Linda feels very close to most of
these people which is reflected by the tie strength scores and her interview account. Second,
these relationships are based on shared activities, regular interaction (though not as a group)
and mutual exchange of emotional and social support. Third, 5 of the 6 people that were named
in both name generators happen to be part of the core (persons 83-88). And last, people in the
core share long standing relationships with Linda. The exceptions are her boy friend (person
83) whom she has known only a few months and a work mate (person 85) who she has known
for two and a half years. However, both are extremely close friends of Linda and appeared in
both name generators.
Chapter 8, section 8.3.3.2, will explore what kind of drinking behaviours and attitudes are
prevalent in the core and how are these norms transmitted.
7.6.5 The Contextualized Network

The last typology represents a network structure where a participant is connected to
independent sub groups or nodes specialized in a specific context. An important feature of this
263
typology which distinguishes it from those described earlier is that in a contextualized network,
various groups and individuals that a participant is connected to are not connected to each
other except through that participant.
There was only one person whose overall network followed this structure. Ruth is a 22 year old
female who has been studying business and management in DIT for 4 years. She lives with her
parents in Dublin and happens to be the only person in the sample who belonged to the low
drinking cohort. She named 11 people in all, 8 females and 3 males. She discusses important
matters with all 11 and socializes with 5 all of whom are females.
The tie strength scores for Ruths network are illustrated next with the nodes enclosed in
parenthesis representing overlapping nodes between the important discussants and socializing
networks.
Tie strength score
6
Alters
71, (72), 73, (74, 75), 76, (710, 711)
77, 78, 79
Table 37: Tie strength scores for Ruths network
The following visualization (Figure 19) represents her overall network.
264
Figure 19: Ruths overall network: A contextualized network
As is apparent from the visualization there are three main components of this network which
are not connected to one another except through Ruth. These are the orange nodes, the green
node and the cluster of nodes visible in the right hand side of the above visualization.
The orange nodes, a male and a female, are Ruths class mates in college (Persons 78 and 79).
She met them in the first year through group projects in class. What is interesting is that
although Ruth has known these friends for 4 years, yet their friendship is limited to having
lunch in college and engaging in casual chats about college related matters such as assignments
and dissertation. Ruth never interacts with these friends outside college. During the interview
she described that that both live far from her and that she has other friends with whom she
prefers to spend time after college. In 4 years, Ruth has gone out with these friends only once.
She described that it was a class night out in the first of college. These friends do not know
anybody else in Ruths broader network.
265
Ruth has been doing tap dancing since she was a child. Person 77 who she has known for 4
years is one of the friends she made in her dance class. They meet every Saturday morning for
the class and talk about college, assignments and how their week went. Their friendship is
restricted to the dance class and Ruth never meets up with person 77 outside the class. Person
77 also does not know anybody else in Ruths network.
The nodes enclosed in the pink box form a clique which comprises Ruths family (red), two
friends from school (pink) and a friend she made on Erasmus while she was in Canada (blue:
Person 710). Ruth feels extremely close to everybody in this clique as is apparent from the tie
strength scores. She described that she often discusses matters related to money, going away to
another country and college with her mother. She also feels very close to her father and
describe that they often discuss issues as a family. Ruths sister is about her age. They see each
other at home over dinner and sometimes go to the cinema together. Ruth mentioned during the
interview that they have always gotten along well and that she is able to discuss anything with
her sister such as matters related to college and friends.
Persons 74 and 75 are Ruths two best friends from school who she has known for 6 years and
with who she interacts several times a week. Her relationship with these girls is based on
regular interaction, social companionship and mutual exchange of social and emotional
support. They often hang out in each others homes, go shopping and watch movies in the
cinema. Ruth feels she can talk to these friends about anything. She described several instances
when she confided in these friends. For example she described how she sought their advice and
opinion when she was assigned to go for Erasmus or when she decided to join a music society.
266
She also described a time when person 74 faced some family issues and discussed the matter
with the girls.
Her dad got remarried and her mom unfortunately died a few years ago. Her dad and stepmom decided they were moving to a new area and she didnt have any choice but to go with
them. She wasnt too happy about it at the. We discussed it a lot. But she is fine with it now
Persons 710 and 711 are two girls Ruth made friends with in Canada while she was there on
Erasmus. Both lived on the same floor as her. Person 710 was also her class mate during that
time. The girls spent a lot of time with each other in Canada doing varied activities largely
such as going to the cinema, doing college work together and going shopping together. Both
girls reside in Canada and Ruth communicates with them via Facebook. Person 710 also
visited Ireland to celebrate Christmas with Ruth and her family. During this time Ruth also
introduced person 710 to her school friends, showed her around the city and spent a lot of time
with her at home.
Person 76 (yellow node) is a friend Ruth made randomly on Facebook. She described him as
her best male friend who she has known for 3 years. Ruth does not see him much because he
lives in another town. However, she communicates with him daily through Facebook and
texting. Ruth described during the interview that she feels as close to him as she does to her
two best friends although she has known them since much longer. Ruth discusses college
related matters with him and seeks his opinion if a disagreement occurs between her friends.
Person 76 knows Ruths two Erasmus friends and her two best friends but does not know her
family.
267
Ruths network is diverse in a sense that she prefers to maintain friendships in specific
contexts. It appears from her interview that she is somehow aware of this contextualized nature
of her friendships and prefers it that way. For example, she described that person 76 who she
became friends with on Facebook is good to take advice from because he does not know her
other friends.
If I was having some sort of problem with my friends, I would talk to him because he doesnt
know them very well. You know people have friends that are kind of friends with everyone they
know. They maybe take sides or something but he wouldnt know most of my friends very very
well. I think its the same for other people I mentioned... I dont see persons 78 and 79 much
outside college or person 77 other than the dance class. Yeah its good to have different
friends like that I think
The tie strength scores and Ruths interview account suggests that her strongest relationships
are those she nurtures outside college settings. Of these, she feels closest to the clique
described above and the two friends who are more or less connected to the clique. Chapter 8,
section 8.3.4.1 will discuss what kind of drinking behaviours are prevalent in this network and
how are they reinforced.
7.7
Conclusion
The results presented in this chapter contribute to the SN theory by addressing an important
gap in the literature that of identifying and locating peers who are salient and relevant to an
individual. In doing so, the chapter draws attention to several key features of the examined ego
networks which are described next.
268
First, the analysis of network composition revealed that in general, the participants named a
higher percentage (more than 70%) of peers whom they knew in contexts outside of college
than those from within college. Further, in most cases, more than 70% of important
discussants and the overlapping nodes were peers from outside college. The evaluation of tie
strength scores complemented by a qualitative interpretation of the meanings participants
attached with the ties in their networks, confirmed that the peers who were most salient to the
participants indeed existed outside college settings. These ties generally featured a high tie
strength score (6 or 5), greater interaction, more shared activities, reciprocal exchange of social
and emotional support and mutual confiding. Also, the people with whom the participants felt
the closest comprised about 43% of the size of their networks which is equivalent to about 7
people. These figures are comparable with past research on an American sample (Christakis
and Fowler, 2009) which suggests that most people have about 6 close social contacts. This
provides confidence that this study was productive in eliciting the most important and salient
people in the social environments of participants.
Second, the analysis of network structure combined with a qualitative interpretation of the
interviews, provided evidence that the strongest ties were usually embedded in cohesive and
tightly knit sub groups. For example, in networks shaped like a group and core/periphery,
the most salient ties were always embedded in the group and the core respectively.
Similarly, in networks shaped like the overlapping cliques, the strongest ties were found to
exist in different cliques which often overlapped through cross-clique friendships. In
consistency with the first finding, these subgroups in most cases evolved and sustained in
contexts outside of college. While the overlapping cliques differed from the other typologies
269
in that the most salient ties were found to be scattered in different cliques, the majority of the
salient ties were almost always embedded in the cliques outside of college.
Third, there is a general lack of network studies which have explored various structural forms
ego networks can take. The present study contributes to this literature. The identification of
overlapping cliques, extends the previous mixed methods network research examining the
structural variations in ego networks such as Bellotti (2008). The overlapping cliques was
found to be the most common structural typology in this study and as described previously,
featured a network where different cliques overlapped through common nodes or cross clique
friendships.
The implications of these findings will be described more fully in the final discussion in
chapter 9. The next chapter will examine the association between these salient relationships
and individuals normative perceptions and drinking behaviours.
270
Personal Networks, Normative Perceptions and Drinking Behaviours
Personal Networks, Normative Perceptions and

Drinking Behaviours
8.1
Introduction
The preceding chapter focused on identifying the most salient ties in the social surroundings of
the participants and locating these peers in their personal networks. This chapter serves the
overarching aim of investigating if there is a relationship between participants personal
networks and their drinking behaviours and related normative perceptions. A qualitative
approach is adopted to investigate the correspondence between the data from web survey and
that from in depth interviews. The key considerations of this procedure and related results are
presented in section 8.2. In order to establish a better understanding of these results, the chapter
then explores how drinking behaviours developed in participants networks, how are they
reinforced in social situations and whether drinking behaviours of the participants appear to be
shaped by their networks? In doing so, the chapter extends the discussion of networks
described in chapter 7 to provide a well-rounded picture of these cases. Some other interesting
examples are also examined. This discussion is presented in section 8.3. A tabulated cross case
analysis follows providing an overall picture of each participants survey responses, network
characteristics and themes discovered across the interviews. The chapter concludes with a
discussion linking the outcomes of the two data collection methods employed in this study.
The implications of these findings and their contribution to SN theory are discussed more fully
in chapter 9.
271
8.2
Linking the web survey and interview data
One of the objectives of this study was to examine the association of personal networks with
individuals perceptions of peer drinking and their own drinking behaviours. This was
addressed by comparing participants responses to normative belief and personal consumption
items on the web survey with their interview descriptions of the drinking behaviours of their
network members. This comparison was carried out for each of the 26 cases. Specifically, the
analysis involved examining how the participants perceived and described the drinking
behaviours prevalent in their networks and whether these perceptions corresponded with those
they reported in the web survey. The results from this analysis are summarized in Table 38. It
was also investigated if the drinking behaviour of the participants was similar to that of their
network members and if similar, whether it was influenced by their networks. These results are
summarized in Table 39.
An important consideration in integrating the web survey and the interviews was the difference
in nature of data. As described in chapter 5, section 5.5.6, the web survey assessed personal
consumption and perceived descriptive norms by asking students to respond to closed ended
questions inquiring about the frequency and quantity of their drinking as well as that of
different referent groups. Similarly, injunctive norms were assessed by inquiring whether these
different referent groups would approve or disapprove if the participants drank to get drunk.
This generated data which was numerical or quantitative. However, inquiring about the
frequency and quantity of peer drinking was neither appropriate for the unstructured format of
in depth interviews nor was it always mentioned by the participants, who mostly described the
drinking behaviours and attitudes of their friends through recollection of shared experiences.
272
In general, the participants used a variety of colloquial terms and jargon to describe the level of
intoxication in their experiences of drinking with their network members. For example, the
people who described the drinking of their peers as being tipsy, mostly associated this phrase
with moderate drinking after which one feels giddy and buzzing from alcohol yet able to
hold a conversation and react normally to situations. On the other hand, those who
described the drinking of their peers as being hammered, messy wasted or plastered
generally associated these phrases with heavy drinking which subsequently results in
unpleasant experiences such as getting sick or falling over, slurring ones speech, getting
into fights, having blackouts and not remembering much of the night.
Similarly, the participants used different expressions to describe their network members
attitudes towards drinking to get drunk. For example, those who perceived their peers to
believe that its a waste of money to go out and not get drunk, that drinking to get drunk is
the thing to do, and that its something that everyone does, associated these phrases with
high approvals towards drinking with an intention to get drunk and encouraging and endorsing
it regularly in social situations. Those who perceived that their peers would not judge
someone if they drank to get drunk, used this phrase to describe that their peers neither
approved nor disapproved of drinking to get drunk. Similarly, those, who thought that their
peers do not like drinking to get drunk or that they think that it is stupid to do so or that they
think that it is a waste of money, used these expressions to describe that their friends
disapprove the idea whatsoever.
As is evident from the above description, the in depth interviews produced qualitative data as is
generally the case with this mode of data collection. This data described (1) the relationships
273
participants shared with their network members and the significance they attached to these ties
as has been discussed in detail in chapter 7, section 7.6 and (2) the drinking behaviours and
attitudes prevalent in these networks and how they developed and sustained over time.
The results presented in Table 38 and Table 39, are thus based .on a qualitative interpretation
of the interviews. These results are guided by the sense and the meaning participants
attached with descriptions of their network members drinking behaviours and attitudes.
274
Respondents
Mary
2
Emma
3
Helen
4
Ruth
5
Linda
6
Edel
7
Pam
8
Sue
9
Debbie
10 Meg
11 Fay
12 Bella
13 Katie
14 Amy
15 Lisa
16 Tom
17 Peter
18 Sam
19 Alex
20 John
21 Adam
22 Brian
23 Cormac
24 Gary
25 Rob
26 Ken
1
Correspondence
between web survey and interviews
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Network members whose drinking behaviours correspond

closely to perceived norms of proximal peers on the web
survey
Close Friends
Best Friend
35, 38, 36, 312, 319

43, 45, 48, 412, 415, 416
52, 57, 59, 519
74, 75, 710, 711
83, 84, 86, 87, 820
91, 94- 99
101, 102, 103, 105, 1014
1102, 1105-1107
1201, 1203, 1206, 1207, 1209, 1210
1302, 1303, 1305, 1309, 1310, 1312
1501-1503, 1509-1511, 1513
1701-1703, 1705, 1707, 1708
1802-1804, 1806, 1809, 1810
1904, 1911, 1912, 1915, 1918
2601-2603, 2606-2608, 2610-2613
11-15, 111, 112, 115, 118, 120
22, 23
61- 63, 67-69, 611, 613
1402, 1405
1604-1607, 1612, 1619, 1620
2003-2005, 2008, 2009, 2013-2017
2101-2103, 2105, 2106, 2108-2110
2203, 2213
2301, 2308-2310
2406- 2408, 2410-2412, 2415, 2417
2502, 2506
31
41
52
74, 75
83,84,86,87,820
91, 93- 97
101, 107
1102
1203, 1209
1303, 1305
1501-1503
1701-1703
1802, 1803
1903
2602
11-13, 15
22, 23
61-63, 611, 613
1402, 1405
1604, 1605
2003-2005
2106
2203, 2213
2301
2403
2501
Table 38: Association between perceived norms and personal networks. Gender: Pink
(Female), Blue (Males), Drinking Cohort: Red (High), Orange (Moderate), Green
(Low)
Table 38 demonstrates a close correspondence between perceived drinking of proximal peers

as reported in the web survey and the perceived drinking of some network members as it
275
appears from the interviews. These findings provide validation to the web survey. Further, the
peers identified in the preceding table (Table 38) were found to be very salient in the social
lives of participants and contributed to their emotional well being. The saliency of a
relationship was established on the basis of tie strength scores and the meanings participants
associated with these ties as described in detail in chapter 7. In general, the peers identified in
Table 38 shared strong and long standing friendships with the participants based on mutual
trust and confiding, regular interaction, multiple shared activities and social companionship.
Further, these peers were found to be embedded in tightly knit and intimate sub groups.
Participant generated numerical scores indicated that these relationships also scored high on
tie-strength. Given the significance and social proximity of these peers, the above findings
suggest the possibility that the participants might have been referring to these salient network
members as their proximal peers in the web survey. It means that the participants might have
been thinking of these individuals when they responded to questions on perceived norms of
proximal peers in the web survey.
The association between participants own drinking behaviours and that of their network
members is presented in Table 39. The table illustrates the correspondence between
participants self reported personal consumption in the web survey and perceived drinking
behaviour of their network members as it appears from the interviews. It also identifies the
salient networks members whose drinking behaviours were found to be very similar to that of
the participants and appeared to influence them. As commented earlier, these results are based
on a qualitative interpretation of the interviews. In addition, the table also presents the gender,
drinking cohort and network typology of each participant for easy reference.
276
Respondents
Mary
2
Emma
3
Helen
4
Ruth
5
Linda
6
Edel
7
Pam
8
Sue
9
Debbie
10 Meg
11 Fay
12 Bella
13 Katie
14 Amy
15 Lisa
16 Tom
17 Peter
18 Sam
19 Alex
20 John
21 Adam
22 Brian
23 Cormac
24 Gary
25 Rob
26 Ken
1
Network Typology
Overlapping Cliques
Overlapping Cliques
Overlapping Cliques
Contextualized
Core/periphery
Overlapping Cliques
Overlapping Cliques
Overlapping Cliques
Overlapping Cliques
Overlapping Cliques
Group
Overlapping Cliques
Group
Overlapping Cliques
Group
Overlapping Cliques
Overlapping Cliques
Overlapping Cliques
Small Clique
Overlapping Cliques
Group
Group
Overlapping Cliques
Core/periphery
Overlapping Cliques
Group
Extent of agreement
between drinking
behaviours
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Correspond Closely
Salient network members whose

drinking behaviours appear to influence that of the participants
31, 35, 38
41*, 45
52
74, 75
83*, 87
94-99
101, 102*
1102, 1106, 1107
1201, 1203, 1206, 1207*, 1209, 1210
1303-1305, 1309, 1310, 1312
1509-1511, 1513
1701-1703*, 1705*, 1707, 1708
1802-1804*, 1809, 1810
1901*
2601, 2602, 2606, 2608
13, 14, 111, 112
22*
61-63, 613
1402*, 1405*
1605, 1619, 1620
2003-2005
2101, 2103, 2105, 2106
2203, 2213
2301
2403, 2412
2501*
Table 39: Association between individual drinking behaviour and personal networks.
Gender: Pink (Female), Blue (Male). Drinking cohort: Red (High), Orange (Moderate),
Green (Low).
277
Table 39 depicts a close correspondence between self reported drinking behaviours of the
participants and the drinking behaviours of some network members. These peers were
identified to be very salient to the participants and were found to be embedded in dense
subgroups within the participants networks. In general, this similarity in drinking behaviours
appears to be an effect of peer influence. It is important to acknowledge that the cross sectional
design of this study makes it difficult to draw causal inferences about peer influence with
certainty. The alternative reciprocal explanation that similarity in drinking behaviours of the
participants and their salient peers could be an effect of peer selection rather than influence
may be true. It means that the participants may have chosen to become friends with people
whose drinking behaviours were similar to their own. However, the combination of survey data
and qualitative interviews in the present study offered a way to determine if the age at which
the participants started drinking preceded the friendships in their networks or followed them.
As described in chapter 5, section 5.5.6.1, one of the items on the web survey inquired about
the age at which respondents began drinking alcohol. In addition, the alter attribute chart filled
out as part of each interview, captured among other details, the duration for which participants
had known each of their network members. Based on this information, it was possible to
distinguish the ties where either peer influence or selection might have resulted in similarity of
drinking behaviours from those, where peer selection could not have been a likely
explanation. This is because in case of the latter, friendships were formed before the
participants started drinking. Most of the peers identified in Table 39 fall in this latter category.
It appears from the interviews that most participants started experimenting with alcohol in the
company of these salient friends developing very specific attitudes and behaviours towards
drinking (More on this in section 8.3 and 8.4). In addition, the qualitative nature of in depth
278
interviews allowed examining the contexts in which friendships and drinking behaviours
developed, sustained and sometimes changed in a network. This knowledge about the
functioning of networks strengthens the aforementioned findings. The asterisks (*) in Table 39
indicate the relationships where age of first drink preceded a friendship. In these cases, it can
be argued that either peer selection or influence might have resulted in similarity of drinking
behaviours between network members and participants. However, as will be demonstrated in
sections 8.3 and 8.4, the latter offers a more likely explanation.
In order to understand, clarify and elaborate the aforementioned results the following section
presents a discussion of some networks and explore how drinking behaviours developed and
sustain within them.
8.3
Development and Reinforcement of Drinking Behaviours in

Networks
Chapter 7 presented the networks of some participants, discussed their composition and
structural variations and identified the strongest ties within these networks. This section
explores how drinking behaviours and attitudes developed within these networks and how are
they reinforced and upheld by the members in social situations.
8.3.1 The Overlapping Cliques

8.3.1.1 Sams network
As has been described in chapter 7, Sam is a 23 year old participant who drinks heavily. Sams
network is reproduced in the following visualization
279
Figure 20: Sam's overall network
It appears from Sams interview that most people in his network are heavy drinkers like
himself. These are represented by circles in Figure 20. Sams network is shaped like
overlapping cliques where the people encircled in blue represent the clique of school friends
and family who he has known for 15 or more years and those encircled in orange represent the
clique of college friends whom he has known for about 3 years. As described in detail in
chapter 7, section 7.6.2, majority of the people most salient to Sam are embedded in the clique
of older friends with whom he interacts regularly in both drink and non drink related ways and
with whom he shares inner thoughts and personal matters.
Sam started drinking when he was 15 years old. His earliest experiences involving alcohol
occurred in the company of his older friends from school who are embedded in the blue clique
and who live in the same area as him. Of his school friends, Sam drinks regularly and the most
280
frequently with his tight group of friends which includes person 61, 62, 63 and 613. Persons
611 and 625 also accompany them on most occasions though the latter was not drinking at the
time of the interview because of a health condition.
Sam described that him and his friends drank heavily throughout school years. He still drinks
with them every week. He described that a typical drinking night with these friends would
start with a pre drinking session at someones house where each of the lads would bring a 6pack of beer and the girls would bring a bottle of wine each. They would all drink in the house
till mid night after which theyd head out to night clubs or pubs where more drinks would
usually follow. Sam described that sometimes theyd visit two or three pubs during the same
night, enjoying a few drinks in each and would usually end up getting fairly hammered.
Sam usually drinks with his school friends in group settings. The only exception is person 61
who is one of his best friends and with whom he sometimes drinks outside of the group. He
described these occasions as their random nights and explained that every once in a while,
they meet up in their homes and enjoy an informal night of TV and conversation over drinks.
Sam described that these nights are not aimed at getting drunk however he feels that usually
as the night progresses and drinks flow more freely; they do get quite hammered.
During the interview, Sam talked about the drinking behaviours of his network members. It
appears from his interview that heavy drinking is considered as normal in his circle of school
friends where it is very rare for someone to drink just 4-5 bottles of beer and head home on
a night out. Sam feels that drinking to get drunk is regarded as a positive thing to do, and
frequently reinforced at house parties and get-togethers with these friends. He described
several occasions where drinking had resulted in one of his friends getting sick, engaging in
281
unwanted sexual acts and even getting injured. For example, he described an occasion when
person 62 got drunk and slept with a girl mistaking her to be his girlfriend, later getting in
trouble with both girls. He also spoke about how person 611 injured his hand from a broken
bottle of beer during a game they were playing in a pub and lost sensation in one of his fingers.
We used to find the game hilarious when we were young. But that time, he sliced his finger and it was
a really deep gash. That night he got in an ambulance. They wouldnt take us in the ambulance because
we were fairly hammered...me and another guy. We had to walk to the hospital without knowing where
exactly it was and we spent about two hours trying to find him.
Sams description of his school friends drinking behaviours also suggests that these
youngsters respond to group norms by trying to keep up with each other when they are
drinking. For example, Sam described how in the younger years, person 62 used to get very
drunk because he tried to keep up with those who were physically bigger than him and thus
had a higher tolerance towards alcohol. Similarly he feels that two of his female friends,
persons 63 and 626 drink too much almost every time they are together.
They drink a bit too much for what they can handle. They always end up quite giddy I would
say in trying to keep up with others I suppose. There have been a couple of times when I have
held person 63s hair while she was throwing up somewhere
Sam feels that he has deliberately distanced himself from some of his school friends,
particularly, the jocks (persons 619, 620, 621 and 624). He explained that his friendship with
these lads was primarily based on drinking. A night out with these lads would involve heavy
drinking of which drinking games were a regular feature. Sometimes Sam and the lads would
play 3-4 drinking games at the same time. They would also intentionally upset other people,
282
verbally abuse them and encourage each other to do the same. Sam felt that the reputation of
this particular group was suffering because of their notorious attitude and he felt that he had
to pull away.
I have found myself distancing myself from them because the nights out were the friendship if you get
me. As a group I found it a bit artificial and forced. It was like the lads were living in past glory and
remembering the good old days and not really moving on with their lives. I am still mates with them
and I meet up with them because we still train for football and I feel bad because they are a
part of my childhood. I have grown up with them
He also described how different people in the group felt pressurized to conform to their ways.
Person 619 is a lovely lad almost childlike in a way. He would never openly hurt someones
feelings but he is swayed by the group. He will get himself involved and even if he is not
verbally slagging anybody, hed laugh along. Person 620 is similar, he just does whatever the
group does, he doesnt like making his own decisions. Person 624 is more territorial. He likes
the group of friends to remain a group of friends and he doesnt like outside interferences and
he doesnt like when people leave from a group of friends. Person 621 and a couple of others,
when they have a few drinks in them, they are assholes
Sams description of the drinking norms of this particular group indicates that he rejects their
norms after having conformed to them for several years because he felt that his reputation was
at stake. Reflecting on his decision to distance himself from the jocks, he recalled that it took
a while before he was able to stand up and disagree with the drinking habits of the jocks. On
the other hand, person 611 was able to do it more easily.
283
I was sort of saying it in a polite way that I am not (stresses) going to be doing this, whereas
person 611 was like no, I just told you I am not doing it, its bullshit and he tore people out of
it. Whereas I was in a passive way saying that I dont want to play this game.
The majority of people Sam knows from college are also heavy drinkers. Of these, persons 67,
68 and 69 who are in his class often socialize with his bigger group of friends outside college.
Sam feels their drinking behaviour is similar to that of his school friends.
They are the same like us. They wouldnt just have like 4 or 5 cans. They would have the pre
drinking and then night out drinking shots and a lot more. Theyd keep pace with the best
drinkers that are out there
There are some people in the orange clique with who Sam became friends through college
nights out. These are persons 614, 615, 616 and 623. He describes them as a good crack and
feels that they are very outgoing people who are fun to hang out with. It appears that Sams
friendship with them is based around nights out and drinks related socialization though he does
not go out with them as often as he does with his old friends.
Of family, his parents are occasional drinkers and do not approve drinking to get drunk
whereas his younger brother who is 22, drinks heavily. Sam described him as a stereotypical
party boy. Sam does not drink with his family except on special family occasions. He and his
brother socialize with their own groups of friends. Sam described that sometimes they end up
in the same night club because they all live in the same general area. On these occasions, he
and his brother chat briefly before going back to their own circles. It is clear from his
description that his family does not have any impact on his drinking habits.
284
Sams survey response suggests that he perceives his proximal peers (close friends and best
friend) to drink to drunkenness once a week and consume 15 or more drinks on an occasion.
These perceptions correspond closely with those of his strongest ties. Sams own drinking
behaviour is largely similar to that of his school friends and appears to be influenced by his
tight circle of friends comprising persons 61, 62, 63 and 613. These friends as has been shown
are embedded in a very cohesive clique. He feels attached to these friends, confides in them
and shares multiple activities with them as has been described in chapter 7. He has been
drinking with these friends for several years and responds to their norms. For example, he
mentioned that he finds it very difficult to not drink when in the company of these friends
because it is their main way of socializing. He also described that he sometimes finds it
difficult to voice his opinion clearly if it disagrees with that of his friends and tries to please
people.
I am kind of more of a people pleaser. I try and smooth things over rather than become
confrontational and say it loud.
8.3.1.2 Megs Network
Meg is a 19 year old moderate drinking participant who has a network shaped like
overlapping cliques. Megs network comprises a clique of college friends (pink encircled)
and a clique of friends and family from outside college (blue encircled). As described in
chapter 7, the people Meg feels closest to lie in the bigger clique of older friends that she
knows from school or through a voluntary organization she works in.
Her network is reproduced in the visualization that follows.
285
Figure 21: Meg's overall network
Meg feels that nobody in her network drinks to excess. As is illustrated in the above
visualization, most people in her network are either light drinkers or do not drink at all. Her
mother drinks occasionally and one of her sisters is a complete abstainer. Only 5 people (out of
19 named) in her network drink heavily. She explained during the interview that those who do
drink alcohol know how to handle their drink.
Meg started drinking when she was 18 years old. She is 19 now and had been drinking for
about a year and a half at the time of the interview. When she started drinking alcohol, it was
286
in the company of her school friends (green nodes in the above visualization) with who she
interacts regularly. They drink together once every two weeks. Among these girls, persons
1303, 1305, 1309 and 1310 do not drink much. Meg recalled that person 1303 had one bad
experience when she started drinking alcohol after which Meg has never seen her drink to
excess. It happened at their Debs25 when person 1303 had to be carried home because she had
gotten sick as a result of drinking. Meg described that her friend had not like the feeling and
has never put herself in a similar situation.
In general, Megs school friends drink in moderation and do not believe in drinking to get
drunk. Meg feels that they do drink for fun but never for the sake of drinking. During the
interview she described that most of her friends do not give in to peer pressure.
When we were in school person 1303 never felt like having to agree with someone just
because she thought they would like it. Thered be groups in school and she never had to be
friendly with one particular group if she didnt want. Shed always be friendly because she
wanted to be not because she had to. I think most of my friends are the same. We have our
own opinions and we are always quite confident about them.
It appears from her interview that the norms of this group of school friends are strongly
reinforced in social situations and that the girls neither appreciate deviant behaviour nor
encourage it by mutual participation. She described several instances when opportunities for
25
A Debs is a formal ball for students in their final year of secondary school in Ireland,
analogous to the Prom in American schools.
287
heavy drinking were turned down by the girls. For example she described a time when person
1305 had a breakup with her boyfriend who had cheated on her. Meg feels that her friend
wanted to drink to get over her problem and ease her mind. However, the girls did not
encourage her behaviour.
She broke up with a boyfriend recently and she wanted to go out and just get drunk. I think that was
mostly to do with her being upset and not wanting to think about it. We wouldnt go out with her
because we knew what she wanted to do, so we just gave her a bit of time to think. Eventually she
realized that thats not going to work and I think she calmed down a little bit after that. She is her
normal self now
Similarly, Meg recalled how person 1302 who is one the heavier drinkers among her friends
changed her drinking habits when she became friends with Meg and the other girls.
She was the type that would meet up with the wrong kind of people. Just the type that would not go to
school quite as much and be very anguish all the time, that yell at people and cause trouble just for fun.
She drank in fields and stuff but she stopped that when she ended up making friends with me, persons
1303, 1305 and 1312. We wont be late partying and wouldnt go out as much. She had to adjust to
things that we did, like going down for shopping (laughs), watching movies things like that
As described in chapter 7, section 7.6.2, Meg takes a holiday with these friends every year. She
described one of these holiday with the girls at her familys holiday home. They had just
turned 18 at the time when person 1315 insisted on heavy drinking during the holiday. Meg
recalled having felt very annoyed at that.
The first year was grand because we were all 17 and couldnt drink, the next year she wanted to drink
an awful lot during the holiday. I didnt want to drink because I thought we didnt want it to be a
288
drinking holiday and then she said this is a drinking holiday and I was like No this is not. We were
away from families and away from authority and she just wanted to buy drinks and go out every night.
At that stage I and the other girls still didnt drink that much at which she was annoyed that you need
to get drunk and I replied no, we really dont, But it didnt really work out that way anyway. Nobody
got really bad
Although Meg feels that her friends do not feel compelled to drink heavily because other
people do so, it appears from her interview that some girls drink differently when they are with
their other friends. For example she described that person 1302 lives in another town because
of studies and often goes out with her college friends.
Just heard some stories from where she lives. I think thats what happens down there. They just drink
an awful lot. I wont hear that much because I dont drink that much, she doesnt tell me too much
about it either. But I know that she knows her limits now and can handle herself
Similarly she described that person 1315 whom she does not feel very close to often socializes
with her college friends who drink a lot as a result of which she often gets drunk.
It appears from her interview that she tries and conforms to the norms of this particular group
of friends even when she is drinking away from them such as with her college friends and her
boy friend.
Meg described that of college friends, persons 1307 and 1313 do not drink for religious
reasons. The former is a Hindu and the latter, a pioneer. The remaining 3 girls (persons 1307,
1311 and 1314) drink heavily. Meg described that whenever they go out, persons 1307 and
1314 typically spend most of the night at the bar and that she has never seen them come home
sober. Meg feels that these friends drink with a deliberate intention to get drunk and wouldnt
289
stop until they couldnt drink any more. Although person 1311 also drinks heavily, Meg thinks
she does so because she feels compelled to drink as much as person 1307 who is her very close
friend.
Shed try to keep up with person 1307. She kind of wants to keep up with her. The two of them
are quite close and she just wants to be the same. They just drink the same amount. If person
1307 went out for the night and didnt drink at all, Id say person 1311 would only have 1 or 2
because then she wouldnt see the point
Meg goes out with these girls once or twice a month depending on college work. It appears
from her interview that she does not feel pressurized to drink like the girls when she is with
them. For example, she explained that on a typical night out with these girls, she is usually the
only one who remains sober whereas the other girls drink a lot.
Person 1304 who is Megs boyfriend works full time and Meg feels that he likes to drink, to
relax himself in social situations but never gets drunk. She described that he is usually reluctant
to go on the dance floor unless he has had a few drinks. Meg recalled that at the start of their
relationship he was a little uncomfortable about the fact that she didnt drink as much as him.
Although Meg described that he has accepted it now, she feels that she has to drink every now
and then with him because he likes it.
He asked me in the beginning of our relationship if I would drink as well and I had said No.
So just at the beginning, he would have liked if I drank and sometimes hed even ask. I do drink
every now and then when I am with him but not to get drunk
290
They drink in social situations as well as enjoy a quiet drink at home when they are by
themselves. Meg works in a voluntary organization as explained in chapter 7. Most of her
friends in the voluntary organization (represented by the larger nodes in the above
visualization) are either abstainers or moderate drinkers. Two of her school friends and her
sister are also part of the voluntary organization. They often socialize together as a big group at
the local pub or at the events related to voluntary work though not everyone drinks.
Megs response to the normative belief items on the web survey indicates that she perceives
her proximal peers to drink 2-3 days a month, have 3-4 drinks on an occasion and never drink
to drunkenness. There appears a close correspondence between these perceptions and those of
her school friends which she described during the interview. Her own behaviour matches
closely and appears to be influenced by the norms of her group of school friends who she feels
closest to of all the network members. These friends are embedded in the clique of older
friends she knows in contexts outside of college. It also appears that she feels compelled to
drink more often than she prefers when she is with her boyfriend. However, her boyfriend
understands and accepts that she will not drink with an intention to get drunk.
8.3.1.3 Sues Network
Sue is an 18 year old participant who drinks heavily. In chapter 7, her important discussants
network was discussed which is shaped like a small clique. In order to understand the
drinking behaviour of Sues peers, it is the wider network that needs to be examined rather
than only those with whom she discusses important matters. Therefore, her overall network is
discussed here which resembles overlapping cliques as shown in the following visualization.
There are 7 members in this network.
291
Figure 22: Sue's overall network
Clique 1 which is encircled in red, comprises Sues school friends. The pink nodes are the girls
and the blue nodes are the lads. Persons 1102 and 1103 who were also part of her important
discussants network, have already been introduced in chapter 7, section 7.6.1. They are both
very salient to Sue. Persons 1105, 1106 and 1107 are three male friends she made in the
transition year26 of school. Sue feels that she can discuss anything with the lads. Her
relationship with them is based on mutual trust and exchange of social and emotional support.
They socialize regularly in both drink and non drink related ways, have the same wider circle
26 Transition Year (TY) is an optional one-year school programme that may be taken in the year after the Junior
Certificate in Ireland. It is intended to make the senior cycle a three year programme encompassing both the
Transition Year and the Leaving Certificate.
292
of friends and understand each other very well. Of the lads, she feels closest to person 1107
with who, she was also in a relationship. They are no longer together but Sue described that
they have worked very hard on their friendship and that she feels they will be friends
forever
It was hard at first to remain friends because we had to see each other in social situations. We have
the same group of friends. A lot of people would cut the strings and never see each other again but I
preferred to have him as a friend than to have him as nothing at all. He went out with me for a very
difficult period of my life because that was when my dad was recovering and my granny just passed
away and my mum was not coping so well with all the stress. The fact that he managed to stick around
with my crazy ass when I was so moody says a lot about him
Persons 1107 and 1106 share a flat with each other. Sue feels that person 1106 is like an elder
brother to her. She described that he is very protective of her which she appreciates.
If someone is not getting the message or is giving you the creeps on a night out, person 1107 would
come over and hed just be like no, she is alright, leave her alone. Hes like a bodyguard at that stage,
hell stay with you the whole night and walk you home and you know nothings going to happen when
he is there. You just feel safe. Its almost like he is the bodyguard of the group and he minds all the
girls
Person 1105 goes to a college in Galway which is why Sue does not see him as often as the
others. She however talks to him every day and meets him every month. Sue described that
person 1105 is one of her best guy friends with whom she can discuss personal issues and
whose advice she values.
293
It appears from Sues interview that clique 1 is part of an extended group of friends
comprising at least 20 people. These friends have known each other from school and interact
on regular basis. Sue described that they meet regularly for DVD nights in each others homes,
go to rugby matches and socialize on nights out two to three times a week. Of this group, she
mentioned only the 5 friends embedded in clique 1 because she feels they are her closest
friends.
Sue began drinking when she was 13 years old. She described that everyone in the group began
drinking around the same time. It appears from her interview that the group was heavily
involved in underage drinking for several years. Sue recalled that within the group,
drinking to drunkenness was approved and regularly reinforced in social situations.
When we were young, we thought drinking as much as you could was cool. If you drank so much then
everyone would think you are class. So we did that a lot
Sue still drinks regularly with friends in the group. A typical night involves pre drinking at
someones house before heading out to the night clubs. Sue described that pre drinking is
seen as a time to socialize with friends and drink so that one would not have to buy drinks in
the night club. She further described that there is a perception that one needs to drink a bit
more so that the effect will last through the night club.
During the interview she talked about the drinking behaviour of her friends. It appears that
most people in her network are heavy drinkers (indicated by the red stars) who engage in heavy
episodic drinking from time to time. Even those who drink moderately or abstain from
drinking alcohol did drink heavily at some stage in the past.
294
She described that person 1102 who is her best friend, is heavily involved in sports because of
which she has to be wary of her drinking. She prefers not to drink when she has trainings and
matches. It is clear from the interview that though person 1102 does not drink frequently, she
engages in sessions of heavy drinking when she gets the chance and considers drinking to get
drunk as a positive and an acceptable thing to do. Sue feels that it is alright because person
1102 does not drink every week. She described an occasion when they were drinking as a
group and person 1102 had to be taken care of because she had gotten too drunk.
She drank too much and she was quite bad. She was all over the shop, stumbling and walking and
slurring her speech. But its never a problem. When someone is bad in our group which does happen
sometimes, you know that you are with people you can trust to mind you and to bring you home, no
questions asked and they are always there for you if you need it. I brought her home that night and the
next day she apologized but there was no need for an apology
Sue feels that person 1106 has a similar drinking behaviour because he too is involved in
sports. She thinks that he has a very good attitude towards drinking because he enjoys the
night but does not drink to an extent that he wont be able to remember the night.
Persons 1105 and 1107 drink to excess regularly. Sue described that person 1105 gets drunk
and causes trouble by picking fights with people almost every time he goes out. There have
been instances when he has broken his hand or has required stitches as a consequence. Sue
feels that it has reached a point where the lads in the group are somewhat reluctant to go out
with him because they always find themselves looking out for him and joining in on fights
because they are friends.
295
He becomes fearless. He will stand up to 3 guys 10 times his size and just run his mouth saying stupid
(stresses) remarks and get himself into trouble. He thinks he can do anything that he is untouchable. I
fucking hate it. Its almost as if he takes on a different persona when he drinks. He is such a nice guy
and then to become this rowdy person looking for fights and smashing bottles. I know for a fact that
everyone in the group doesnt like when he drinks
Sue described that person 1107 relies on the drink to deal with his problems and becomes a
mess when there is something upsetting him. She recalled several instances when he has been
drunk because there was a problem in the background. For example, she described how he had
become unconscious as a result of drinking continuously for several hours at his brothers
wedding. Sue feels that he was upset because his brother who is very close to him was
moving away to another country after the wedding.
Hell drink until he cant walk, cant talk, cant do anything and its almost like an outlet for him. We
know that it is his way of dealing with a problem, just to escape from it. Everyone deals with stuff
differently. And maybe its not a positive thing but it works for him
Of all the friends in the group, person 1103 is the only one who began drinking at the legal age
of 18 years. Sue feels that it was a very hard thing to do. She described that person 1103 is
the most mature drinker in the group who does not approve drinking to get drunk and knows
her limits. Sue described an occasion when person 1103 drank to excess because she was
upset over her breakup. This was the only time in their friendship when Sue had to mind her.
That night when we went out as a group she drank to forget how upset she was. We have the same
group of friends and her ex is in that group. Although they have broken up she still has to see him in
social situations. That made it a bit harder for her. She just drank too much to get rid of the
awkwardness
296
Clique 2 comprises persons 1101, 1102, 1103 and 1107. There is only person in this clique
Sues sister, who was not a member of clique 1. The other three members of this clique overlap
between cliques 1 and 2. As has been described in chapter 7, Sues sister (person 1101) is a
recovering alcoholic who suffers from alcohol induced epileptic seizures. Sue is extremely
close to her sister and spoke at length about how her sister was bullied in primary school as a
consequence of which she began drinking excessively.
When she started secondary school she was still raw from the bullying in the primary school. She just
got in with a bad (emphasizes) crowd of people who were smoking and drinking all the time. She didnt
enjoy it, she was doing it to fit in. When she was 16, it really started going downhill. She went mental
basically, drinking till odd hours and not coming home, not telling my parents where she was.
Eventually her body just gave in from dehydration and she started having epileptic seizures
Sues sister cannot drink alcohol anymore because of her health. While Sue spends a lot of
time with her, she never drinks with her sister.
Clique 3 is formed by persons 1103, 1104, 1106 and 1107. The only person in this clique who
was not a part of clique 1 is person 1104 who is also the only friend she named from college.
The remaining three members of this clique overlap between cliques 1 and 3. Sue has known
person 1104 for only a few months but feels that she is very similar to her closest friends persons 1102 and 1103. Sue spends a lot of time with her in college and socializes with her on
nights out. They go out drinking several times during the week and person 1104 usually stays
at Sues flat after the night. Sue described that she is a good drinker who knows her limits.
However, it appears from the interview that person 1104 does drink to excess from time to
297
time. Sue described several occasions when she got sick as a result of drinking and had to be
taken care of.
Of clique 1, persons 1102 and 1105 live outside Dublin. Sue does not drink with them as often
as with the others. However, they meet once a month when they come home or when Sue visits
them. On these occasions they go on nights out and drink. Persons 1103, 1104, 1106 and 1107
live in Dublin. They socialize a lot and Sue drinks with them several times a week. Sues
response to the web survey suggests that she perceives her proximal peers to drink twice a
week and consume 9-10 drinks on an occasion. It appears from her interview that these
perceptions correspond closely to several of her friends particularly those embedded in clique 1
who are very salient to her. Sues drinking behaviour is similar to theirs. However, it appears
that Sue does not consider herself or her friends to be heavy drinkers. While she feels that
some of her friends such as persons 1102, 1103 and 1106 are capable of going on nights out
and not drink, she tends to overlook the instances when they do engage in heavy drinking. For
example, Sue feels that since person 1102 does not drink often because of her sports
commitments, she is entitled to drinking as much as she likes when she gets the chance.
Similarly she thinks that though person 1107 drinks to drunkenness to deal with the slightest
of problems, his drinking habits are positive.
I definitely wouldnt say that he has a bad (stresses) attitude to drinking but Id just that his attitude
isnt as good as some of my other friends such as person 1106
It appears from her interview that Sues drinking behaviour matches closely with that of her
extended group of school friends of which she named only her closest friends who are
embedded in a clique. These people knew her before she began drinking. They started drinking
298
together, drank heavily for several years and as a group, reinforced the belief that it was a
favourable thing to do because it makes one look classy. Sue still drinks with them several
times a week and shares inner thoughts with these friends. It strengthens the possibility that
Sues drinking behaviour is influenced by friends in this particular group as it appears to be the
most salient in her social life. Although she did not name her parents in her network, it
appeared from her interview that they do not influence her drinking behaviour. Unlike her, they
are occasional drinkers who do not approve drinking to get drunk.
8.3.1.4 Peters network
Peter is a 20 year old male participant who drinks moderately. Peter named 13 people in all, of
which there were 4 males and 9 females. The tie strength scores for his network are as follows.
The nodes enclosed in parenthesis represent the people he named in both the important
discussants network as well as the socialization network
Tie Strength Scores
6
5
3
2
1
Alters
21, (22, 23)
210, 213
24, 211
27, 29
25, 26, 28, 212
Table 40: Tie strength scores for Peter's network
Peters network resembles overlapping cliques. There are three distinct cliques in this
network as has been highlighted in the following visualization.
299
Figure 23: Peter's overall network
Peter started drinking when he was 16 years old. He did not name any friends who would have
known him before he began drinking (except his family) which is why it is not possible to say
with certainty that anybody in his network influences his drinking. However, it appears from
his interview that he used to drink a lot before he met his boyfriend (person 22). He not only
changed his drinking behaviour after having met person 22 but also rejected the norms of some
of his heavy drinking friends and distanced himself from them.
The first clique in this network which has been encircled in pink is formed by Peters work
mates (pink nodes) and his boyfriend (person 22). Peter works in a pub and has known his
work mates for a year. Of these friends, he feels closest to person 23 who is also his
housemate. Peter described her as his best friend. He spends a lot of time with her at work as
300
well as at home. They socialize in both drink and non drink related ways. Peter recalled several
occasions when he discussed personal issues with person 23. He feels he can discuss his
relationships with her openly which is a very private matter for him. He described her as a very
sensitive person who has been through several harsh experiences such as rape, unwanted
pregnancies, abortion and miscarriages. Peter feels that these experiences have left her deeply
traumatized and unable to trust people. His friendship with her is based on social
companionship, mutual confiding and exchange of emotional support. Peter described her as a
moderate drinker. However, it appears from his interview that although she does not drink
often (twice a month); she drinks to drunkenness when she does go out.
She wouldnt be too keen on having just a drink. She thinks its cool and that you need it to relax. She
drinks to get drunk and wouldnt drink otherwise really
Peter described that his other work mates namely persons 26, 27 and 28 are heavy drinkers
who drink most nights of the week. They are all females who work full time in the pub and
Peter sees them every day in the pub. Peter described that when he started working with them
he felt that he needed to fit in with his work mates. He started going out with them most
nights and drank to drunkenness because that was the routine of this group.
We started going out together to socialize and to know each other better. Every night after work wed
go out to a pub across the road and get in for free because we were staff. Wed always drink there until
4 in the morning and do the same the next day
Similar to Sam who rejected the norms of some of his school friends as has been described
earlier, Peter chose to reject the norms of these workmates. He described that since he met his
boyfriend, he has distanced himself from them because he believes that they indulge in the
drink excessively and have very unhealthy life styles. He recalled that he was mocked and
301
slagged when he stopped drinking with them but he did not give in. Now they dont meet up
unless it is to go out which is not very often. Peter described that person 23 feels the same way
towards them and does not go out with the girls as much as she used to.
I did that for a few months. I had my phase of drinking all the time because I was going through a
break up but now I am past it. I still get on great with them but I wouldnt be as close to them because
the only way to get close to them is to go drinking with them and I dont want to do that anymore
The fact that Sam dissociated himself from his childhood friends and Peter found himself
distancing from his work mates because they no longer preferred these friends drinking
norms, reflects how dynamic and fluid network associations can be. It is well known in alcohol
use literature that sometimes people become friends with each other because of similar
drinking behaviours (Ennett and Bauman, 1994). The above finding draws attention to a
different process that of people rejecting old friendships or consciously reducing them
because they have dissimilar drinking behaviours. Although this is a relatively uncommon
phenomenon, it happens in youngsters lives and warrants further attention.
The second clique in this network is formed by Peters college friends (persons 24 and 25), his
boy friend (person 22), his work mate/flat mate (persons 23) and a friend he made randomly on
a night out (person 29). Of these friends, persons 22 and 23 overlap between cliques 1 and 2.
Peter has known his college friends for two years. He used to go out with them a lot but since
he has met his boyfriend, he chooses to go out less often. Peter feels that these friends drink a
lot. He described that he used to go out with them a lot when he was recovering from a bad
relationship. They used to drink twice a week and get wasted. However, since he has met his
boyfriend, he only sees these friends in college and does not socialize with them as much as he
used to.
302
The third clique is formed by Peters family and his boyfriend. Among family, Peter mentioned
his parents and an aunt who is her mothers sister. He described that his mother and his aunt do
not drink alcohol and disapprove of it completely. Peter feels that this is because their father
was an alcoholic and their husbands (Peters uncle and his dad) also have issues with alcohol.
Peter described that his father (person 212) used to be an alcoholic and that even now his
addiction often relapses. He still drinks every day which often leads to unpleasant arguments
between his parents. Peter described that he does not feel like discussing anything important
with his father nor does he feel very close to him. In fact he finds himself unable to forgive his
father for some of the experiences which he had to go through as a child because of his fathers
drunkenness. Among his family, Peter feels closest to his mother and his aunt and discusses
personal issues with them. However, since Peter does not live at home, he interacts with his
mother and his aunt less frequently than he does with other members in his network.
Peters boy friend (person 22) overlaps between all three cliques in his network. It means that
he knows most of his contacts. He is 11 years older than Peter and the two of them met at a pub
less than a year ago. Peter feels very close to him. He described that person 22 has moved to
another country for a year in relation to his work. However, they talk every night and visit each
other every month by travelling back and forth. Peter feels he can discuss anything with person
22 and often seeks his advice on work related matters. He described that person 22 might have
drank more when he was younger but now he is a moderate drinker who drinks twice a month
and very rarely drinks to excess.. Peter feels person 22 has been a very good influence on
him because he has tamed down since he met him. He described that he socializes with his
boyfriend in non drink related ways and that they prefer to go to restaurants most of the times
rather than pubs and clubs.
303
I think we are really good for each other because, I wouldnt really know when to stop before I met
him. Now he is always there to tell me I think youve had enough in a nice way so that I dont get all
fluttered cause when you drink you always get Ah I didnt drink enough. I used to drink much more
before I met him but I think I have changed. He doesnt agree with drinking to get drunk whatsoever.
He just thinks its stupid and ridiculous and thats why he doesnt let me do it and I agree with him
when I am sober
Peters survey response suggests that he perceives his proximal peers to drink to drunkenness
2-3 times a month and consume 3-4 drinks on an occasion. These perceptions correspond
closely with the drinking behaviours of persons 22 and 23 who are very salient in this network.
They appear in both his important discussants and socializing networks. Peter interacts with
them regularly, shares inner thoughts with them and seeks their advice and opinion on matters
of importance. Peters own drinking behaviour is very similar to that of his boyfriend and
appears to be influenced by him. It is clear from his interview that though he has known him
for less than a year, he chose to change his drinking behaviour after having met him and feels
that his boyfriend has been a good influence on him in this regards. It is also clear from the
visualization, that the most salient nodes in this network are embedded in cohesive sub groups
outside of college.
8.3.2 The Group

8.3.2.1 Lisas Network
Lisa is a 20 year old heavy drinking participant. Her network is shaped like a group.
304
Figure 24: Lisa's overall network
The heavy drinkers in this network are identified by a red star in the visualization. As is
evident, everyone in Lisas network drinks heavily.
As described in chapter 7, section 7.6.3; the most salient ties in her network are embedded in
the group of 16 friends encircled in black in the above visualization. They all live in the same
general area. The group has been central in Lisas life since early teens. It was in the
company of these friends that Lisa began drinking. She was 13 years old at the time. In their
teenage years the group would usually drink at parties and get-togethers because they were
underage to go to pubs/clubs. Lisa recalled that drinking has been a central element of
socialization within the group since their teenage years and continues to be so. It appears
from her interview that over the years, the group has developed very specific drinking
behaviours and attitudes to which everyone conforms. For example, they all drink similar
305
beverages in similar amounts and in the same manner which Lisa described as the standard
way. She described that heavy drinking is not only common but also almost taken for
granted within the group.
If we are going out and someones not drinking, eyebrows would be raised, wed be like why not. If
someone wasnt drinking they wouldnt come with us and itd be for a reason, it wouldnt be just
because I dont want to. Itd be because I have work tomorrow or I have no money. It would never be
like O I just dont want to. So drinking is a big factor in our going out. We usually have a naggin27 of
spirits before we go out and 1 or2 drinks in the club. Thats the standard. Drinking a naggin a week is
quite a lot. We dont think that we are massive alcoholics but we know that we drink more than we
should
She described that they always drink pre-purchased alcoholic beverages in a pre-drinking
session in their area before heading out to the licensed venues in town. Lisa and her friends
think that it saves them money. It appears from her interview that a typical night out is aimed
at getting drunk. Lisas description also indicates that social pressures exist within the group
to conform to the drink related norms and expectations.
Usually when we go out, we would pre drink in our area. If we are in rush for the train wed
just down our drink, get on the train and then wed want to be drunk for the rest of the night. If
you are going out and you are going to be drinking then youd want to be drunk rather than
feel that you wasted your money on drinking and you are not even drunk. So yeah most times
when we go out wed drink to get drunk
27
A naggin of spirit is equivalent of a 200ml serving (6 standard drinks) - (drinkaware.ie)

306
She also described that drinking is encouraged in the group in the form of drinking games
which form a regular feature of pre drinking sessions in the group.
When there are a few of us we play kings. You have a cup in the middle surrounded by cards. You
have to flip over a card and do what the card says. If you are the last to do whatever that card says then
you have to drink and if you get a king card then you have to pour some drink into the cup and the last
person to get the 4th king has to down the king cup. Its like a mixture of vodka and beer. We play that a
lot, you just have to (laughs)
Within the group, Lisa has a sub group of friends (comprising persons 2601, 2602, 2606,
2607 and 2608) with whom she spends most of her free time. She drinks with these girls twice
a week, relies on them for advice and emotional support and never goes out without them.
Outside the group, Lisa has two college friends who she has known for only a few months. It
appears from her interview that Lisa does not go out with them often and that most of their
activities are college based. She described that she does not have much idea about their
drinking behaviours because they usually drink within their own circles of friends outside
college. Lisa however feels that their drinking habits are similar to those of her own friends
because they often appear hung over and tired in college after a night out with their friends
and tell her how drunk they were the night before.
When we chat I know they go out and that they would drink to get drunk as well. They would imply it
when they tell me that they went out last night or what happened. I dont know how much or whatever
they drink but I think they are like me and my friends
Of family, Lisa only mentioned her brothers. Lisas two brothers also drink heavily. Of them
person 2605 is still under age. Yet, it appears from Lisas interview that he drinks every week
307
and consumes up to 10 standard drinks or more on a single occasion. He is a rugby player and
knows many people who are older than him. He is not old enough to go to clubs and pubs but
he drinks with his older friends at house parties and sports get-togethers. Lisa never drinks with
him.
Lisa did not mention her parents in either of the name generators however it appears from her
interview that unlike her and her siblings, they drink only on special occasions.
Lisas survey response indicates that she perceives her proximal peers to drink to drunkenness
twice a week and have 9-10 drinks on an occasion. These perceptions correspond closely with
those of the cohesive group of school friends. Her own drinking behaviour matches closely
with that of the group. It appears from her interview that she prefers to enjoy nights out with
her closest friends in the group and derives fulfilment from her relationship with these girls.
These girls namely persons 2601, 2602, 2606, 2607 and 2608 are the most salient in her
network and also appear to influence her behaviour by regularly reinforcing heavy drinking in
social situations, associating it with the group and expecting each other to engage in it.
8.3.2.2 Kens network
Ken is a 20 year old participant who drinks moderately. He named 17 people in all of which 11
were males and 6 were females. Kens network is also shaped like a group.
The tie strength scores for this network are presented next. The nodes enclosed in the
parenthesis overlap between Kens important discussants network and his socializing
network.
308
Tie Strength Scores

6
5
4
3
2
1
Alters
(2501)
2504, 2512, 2513, 2514
2502, (2505, 2506)
2503, (2508), 2510, 2515
2507, 2509, 2511, 2517
2516
Table 41: Tie strength scores for Ken's network
Kens network is shown in Figure 25. The group in this network is encircled in blue. In addition to the group, Kens network includes a clique (encircled in green) of family members
comprising his parents, siblings and two cousins. Its a predominantly male network.
Figure 25: Ken's overall network
309
The group in this case comprises Kens girl friend (person 2501) and 10 friends who he
knows from school. They all live in the same general area. Ken has known most friends in the
group for over 10 years. The only exception is his girl friend who he has known for a year.
However, she is well integrated into Kens wider network and knows all his school friends as
well as his immediate family. She is part of the group because she socializes with Ken and
his school friends regularly in both drink and non drink related ways.
Ken and his school friends socialize through several shared activities such as playing football,
music and drinking. As a group, they often hang out in each others houses to watch movies,
play video games or just to talk and drink. Ken is also involved in a music band with some of
his school friends which include persons 2503, 2505, 2506, 2508, 2509 and 2510. He feels that
music connects them. Of his school friends, he feels closest to persons 2502, 2505, 2506 and
2508 who he has known for 16 years. Ken feels that he can confide in these friends about
private matters and seek their advice.
The group means a lot to Ken because he grew up with these friends. He interacts with these
friends every week and relies on them for social companionship as well as emotional support.
Ken began drinking in the company of the group. He was 15 years old at the time. During the
interview he recalled that he and his friends did a lot of heavy drinking in the early years. Ken
described several instances from his teenage years when his friends ended up doing things they
later regretted as a result of being drunk. He recalled that in the earlier years, it was really
difficult to avoid drinking in group situations because his friends relied on drinking to
socialize.
310
Its almost isolating yourself if you are not getting involved. Its very hard for all my friends to get
together at one place without alcohol being involved. Theres a pressure and a huge emphasis on
drinking apart from family really. One of the first questions youd be asked when you are out is are
you drinking.
Most of the people in the group still drink heavily (encircled in red). It appears from the
interview that those who drink moderately also engage or used to engage in episodes of heavy
drinking from time to time. Ken described that drinking to get drunk is considered as a
normal thing to do among his friends and no one really disapproves it.
Of all the network members, Ken feels closest to his girl friend (person 1501). While he sees
the group at least once a week, he interacts with his girl friend on daily basis. They spend
their free time together in each others houses watching movies, cooking food and talking. Ken
described that they have gone on several trips together. At the time of the interview, Ken and
his girlfriend were planning to go on a holiday together. He feels that they dont keep anything
from each other and can openly talk about their troubles and personal issues. Whenever Ken
socializes with his school friends, his girl friend is usually there as well.
Kens parents abstain from drinking alcohol and his siblings only drink occasionally, yet Ken
began drinking when he was still underage and continued to drink heavily through secondary
school. This reflects that Kens family does not influence his drinking. Ken never drinks with
his family. His siblings are several years older than him. He recalled that his brother used to
drink heavily in teenage years now he only drinks on occasions. His cousins are also heavy
drinkers but Ken never drinks with them either. They live in different towns and Ken only
communicates with them through social networking sites.
311
It appears from the interview that Kens drinking behaviour changed after he met his girl
friend. For example he feels that his drinking has reduced because of her. Ken described that
he used to drink heavily at least once a week with his school friends before he met her.
Although he did not like to drink heavily, he relied on his friends to socialize and spend his
free time with. Ken feels that there was always an unspoken pressure to drink heavily.
When you hang out with big groups of lads like me, you drink a lot because not having a drink in
ones hand is almost looked down upon. It was an understood thing
Ken described that when he was single, it was harder to say no to drink because one needs to
go out to meet potential partners.
When you dont have a girl friend you rely more on your friends circle because you have to go out
with your friends to meet someone in the first place so you do what your friends do and you need to go
out to the pub and you do the usual pub and the club and drink and then go home at 3 o clock in the
morning. That was just the general routine in my case
After meeting his girl friend, Ken felt he didnt need to drink as much. He still sees his school
friends once a week and enjoys watching movies with them and playing sports or music.
However, he only drinks with them once a month because he prefers to spend most of his time
with his girl friend doing non alcoholic evening activities because she does not like to drink
as much as his other friends.
I probably drank more before I met her. When we are together we dont drink unless we want
to go to a pub to have cocktails or something nice taste wise which would be twice a month at
the most. We dont ever feel the need to go out drinking together. We are just happy just like
staying in and doing something non alcoholic
312
Kens girl friend does not approve drinking to get drunk and thinks that it is stupid to do so.
He described an incident when he had gotten drunk by accident and his girl friend had been
extremely angry at him. Ken feels she has been a positive influence on him.
Kens survey response suggests that he perceives his close friends to drink to drunkenness once
a week and consume 7-8 drinks on an occasion. These perceptions correspond closely to the
drinking behaviour of most of his school friends who drink heavily. His survey response also
suggests that he perceives his best friend to drink 2-3 days a month have 3-4 drinks on an
occasion and never get drunk. These perceptions correspond closely with the drinking
behaviour of his girl friend. There was a time when Kens drinking behaviour was similar to
the group and he felt pressurized to drink like them. However, it is clear from the interview
that his current drinking behaviour is very similar to that of his girl friend and appears to be
influenced by her. She is also the most salient person in his network and embedded in the
group which is still very important to him and holds a central place in his social life. It is also
interesting to note that though Ken has been in college for 2 years, he did not name any friends
from college in his important discussants or socialization networks. This reflects the social
fulfilment he derives from the group.
8.3.3 Core/Periphery
8.3.3.1 Garys network
Gary is a 23 year old participant who drinks heavily. Garys network which resembles a
core/periphery type structure is reproduced in Figure 26.
313
Figure 26: Gary's overall network
Gary belongs to a family (yellow nodes) where most members abstain from drinking alcohol or
do so only occasionally whereas he himself is a heavy drinker. During the interview he
described that both his grand fathers were alcoholics. His parents have seen some very hard
times as young adults and decided to completely abstain from alcohol when they had their first
child Garys sister (person 2306). They havent drunk alcohol in 30 years and strongly
oppose heavy drinking. Both his siblings drink only occasionally. Gary described his elder
sister as very ambitious and career focused. His younger brother does not like the idea of
drinking and Gary feels that his grandfathers alcoholism has affected his brother deeply.
I think that had a direct effect on my brother. He has only met my mums father who spent all
his money in the pub. My brother saw him drunk several times and I can only imagine that put
314
him off. It didnt have much of an effect on me, I dont really know why but it did on him. My
sister doesnt drink because she is always doing something, she always has to be up early and
drinking is not an important part of her social situation
Gary thinks that his brother does not make any friends because he does not like to drink. He
believes that the opportunities to make friends and socialize in the Irish culture are
dramatically reduced if one does not drink because there is no point of coming out on a
night out if it is not for drinking. Gary is the only person in his family who drinks. He started
drinking when he was 13 years old.
Garys network is shaped like a core/periphery where the core is formed by 5 male friends
encircled in black in the visualization (persons 2301, 2304, 2308, 2309 and 2310). As has been
described in chapter 7, section 7.6.4, these are the most salient ties in Garys network. Further,
of all network members, Gary drinks with these lads the most frequently. Person 2304 is one of
Garys best friends from school whom he has known for 10 years. He has a steady girlfriend
(person 2312) since he started drinking. Gary described that person 2304 does not like
drinking to get drunk because he usually prefers to remain in control on nights out and return
home with his girlfriend while he is still sober. Gary does not drink with him as often as he
would with the other friends in the core because person 2304 has a busy job and a different
circle of friends with whom he likes to drink. However, they usually meet up at least twice a
month and drinking is usually involved on these occasions.
Person 2304 is a completely different kind of friend. Hes been going out with person 2312 for many
years. He knows my group of friends but he doesnt go out with them. He lives 15 minutes from us,
works full time and crazy hours. When he does have time off he likes to have a few beers with his own
friends. But Id never go to his house and not have a few beers. Its just the culture that we live in. No
315
matter what, everybody seems to have alcohol with whatever they do. I plan seeing him on Sunday for a
match because we are both big American football fans, I am sure well have a few beers while we do it
Gary usually drinks with persons 2301, 2308, 2309 and 2310. They always go out together as a
group. Gary described that while these friends drink different amounts on a night out; it is not
unusual for them to engage in episodes of heavy drinking from time to time. For example Gary
described a time when persons 2301 and 2310 went through phases of heavy drinking after
breaking up with their girl friends. It appears from his interview that drinking to get drunk is
considered as normal in his group of friends.
I dont think anybody would really have a problem with it. Its normal. Weve all had experiences
about it where you get out of control and wasted, you dont know the one thats one too many
Among these friends, person 2308 drinks the most. Gary described that he drinks to get drunk
every night.
He is always drunk when we go out. Every night he goes out and hell drink crazy amounts
which he cant handle and wont remember a lot of the stuff the next day. If there is a sign of a
party, no matter who gives it, hell go. He rarely takes a night off
It also appears from his interview that while as friends, they look out for each other on nights
out and never leave each other alone when they are drunk yet they do not intervene when
someone drinks heavily. Gary feels that it is unnecessary.
If somebody wants to get wasted you let them do it, its their life and who are you to stop
them. Its quite common. Weve all done that. You may advise them differently if they are a
really (stresses) good friend or a family member but again its not your place to step in
316
As has been described earlier in chapter 7, these friends are part of a wider group of lads who
Gary has become friends with through person 2301. During the interview, Gary described that
he has a group of 12 friends of which he mentioned these 4 because they are his closest friends.
They go out every week and sometimes more frequently if their schedules allow
We have a page on Facebook just for our group of 12 friends. Its a lot of person 2301s school
friends and the friends that we have from rugby. We all have the same interests, we all play
foot ball and go out together. Its very male orientated. Every now and then somebody would go
does anyone fancy a pint and some people will reply and wed go down to the local pub because its
close enough. We all live in close proximity to each other
Gary started going out with this group some years ago. He recalled that in the beginning he
was very shy and reserved and found himself spending more time drinking rather than
engaging in conversations with them. Then he and person 2301 went on a holiday together
with some of the lads after which their friendship strengthened.
Of core friends, Gary described person 2301 as his best friend who he has known for 15 years.
They started drinking together. Gary described that person 2301 was introduced to beer by his
father who works as a bar man. Both Gary and Person 2301 began drinking together and drank
heavily through school years. Both were involved in rugby which Gary describes was closely
related to drinking.
We played a lot of rugby in school and its just one of those sports which go hand in hand with
drinking. If youre involved in it youll know how it is. So we did a lot of drinking in school
Gary recalled that there were several instances of not remembering the night and getting
sick.
317
When we were younger, I, person 2301 and a lot of other people didnt know when to stop. A
lot of the times, wed keep drinking and then thered be like a switch thatd go and then wed
change and wouldnt be ourselves and wouldnt probably remember anything after that point
Gary and person 2301 drink in the company of other friends as well as when they are on their
own. Gary recalled that he never goes out without him (person 2301) whether it is to watch a
match or a night out. He further described person 2301 as his drinking buddy on a night out
when they keep up with each other and drink the same amount of alcohol.
Of the peripheral friends, person 2305 is an abstainer. She plays a lot of sports and does not
engage in drink related activities. Person 2312 drinks within her own circle of female friends.
Gary does not drink with her specifically.
The two people from college are friends he made while he was in Spain for Erasmus. Gary has
only known them a few months and only meets them in college. They are heavy drinkers but
Gary feels that they know their limits. His perceptions of their drinking are largely based on
the three months that they spent together in Spain. Gary described that the lads had built up a
lot of tolerance for alcohol because they lived in an area which was very focused around
drinking. On the other hand, Gary lived with people from different nationalities in the first few
months. Their cultures were not as permissive about drinking as is the Irish culture. Gary
recalled that he had to reduce his drinking out of necessacity rather than desire during this
time. During the last month of his stay in Spain, person 2301 visited Gary. He recalled that
they spent most of their time doing drink related activities and visiting various pubs and club.
Garys survey response suggests that he perceives his proximal peers to drink to drunkenness
at least twice a week and drink 9-10 drinks on an occasion. These perceptions correspond
318
closely with the drinking behaviour of the majority of his core friends who drink regularly and
heavily. Garys own drinking behaviour is very similar to theirs. It is clear from his interview
that Gary mostly drinks with his core friends, that they often engage in heavy drinking and that
they reinforce it in social situations. It appears that of these friends, person 2301 is the most
salient and influences Garys drinking the most. They began drinking together, drank heavily
all through school years and continue to do so both in group situations as well as when they are
together by themselves. Gary never drinks without him and they drink similar amounts
whenever they go out. It is also clear that his peripheral contacts do not influence his drinking
as many of them are abstainers and those who are heavy drinkers are recent friends with whom
Gary drinks very rarely. It is apparent that the most salient and influential people in this
network are embedded in the core which exists outside of college. It is also interesting to note
that though Gary has been in college for 4 years, he only named two people from college who
he has known for only a few months.
8.3.3.2 Lindas Network
Linda is a 19 year old participant who drinks heavily. Lindas network is shaped like a
core/periphery, where the most salient people in her life are embedded in the core as
described in detail in chapter 7, section 7.6.4. Unlike Garys network where the core comprised
5 male friends, Lindas core contacts are diverse and include her parents, cousins, school
friends, boyfriend and a work mate. Most of the peripheral friends are people she came to
know through her core contacts. Linda did not name any college friends in her network though
she has been in college for a year.
319
Lindas network is reproduced in Figure 27 in which the core is encircled in thick black and
the peripheral friends are encircled in thin black.
Figure 27: Linda's overall network
Lindas interview suggests that most people in her network are heavy drinkers (encircled in
red) like herself. As will be discussed, there are several people in her network with whom she
become friends through drinking such as persons 810, 811, 812, 814, 815, 819 and 820. Most
of these people are her peripheral friends.
Linda started drinking when she was 17 years old. She recalled being introduced to different
types of alcoholic beverages by her parents (persons 81 and 82) in a very controlled
320
environment. Her parents would encourage her to try different drinks at home and to learn to
pace herself while she drank. She described that this helped her know her cut off point at
which she usually stops drinking. She described that both her parents drink in moderation,
however, recently there have been some episodes when her father returned home drunk during
the week. This has led to some arguments at home which Linda finds unpleasant and out of
the ordinary. During the interview, she described that every Friday she and her parents enjoy
a quiet evening at home where they have drinks with dinner followed by dancing. Her parents
would put on their wedding song and dance to it. Linda described these evenings as very
enjoyable and relaxing. This is the time when they socialize together as a family, discuss
their problems and engage in light hearted conversation.
Person 87 is Lindas cousin who is three years older than her. Lindas earliest recollections of
drinking outside her home involve this particular cousin. She introduced Linda to drinking in
clubs/pubs and taught her to dress up and put on makeup. Person 87 drinks heavily and Linda
feels that her cousin influences her to drink more than she should on a night out. She recalled
an occasion when she had gotten terribly drunk after a night out with her cousin. Her father
had been very angry and upset at her cousin for getting Linda drunk and influencing her.
When I was 17 we started talking more because I was starting to go out then. She
would take me out and show me what to drink and how to do it. My step dad thinks she
is a bad influence on me when drink is concerned because I wouldnt be technically a
heavy drinker. When we go out together shes always like Lets do shots and wed be
mixing drinks. She would drink most guys under the table and Id try to keep up with
her
321
Linda drinks with person 87 at least twice or thrice a week. It appears from her interview that
these occasions frequently involve drinking excessively where Linda feels pressurized to drink.
She described that person 87 often pokes fun at her if she does not keep up with her and
encourages her to continue drinking even when she feels she should stop.
She is always encouraging me to keep up with her. If I dont she is like O, you scary
cat, why arent you drinking, go on. Two years ago, we were drinking in the house and
playing a drinking game. I said O I need to slow down and she was like But I am
over, you must keep drinking and I felt like O yeah I should be ok, Ill feel fine if I just
keep drinking and then I was just completely drunk. I should have probably stopped a
long time ago but I didnt, I was like O, She is still standing, Ill be ok too
Person 84 is Lindas school friend with whom she grew up. Linda recalled that as kids they
used to have weekly sleepovers in each others houses. Now they have different friends.
However, they have a girls night every month when they drink and catch up with each
other. Linda feels that person 84 did not use to drink as much as she does now. She thinks that
her friend has gotten into the company of a group that drinks a lot all the time. On a typical
night Linda and person 84 usually have up to two full bottles of wine28 before they head out to
the night clubs where they would drink more.
Linda works with person 85. She described that persons 85 and 810 are best friends with each
other and used to be heavily involved in underage drinking as teenagers.
28
A full bottle of wine is equivalent of a 750 ml serving (7.4 standard drinks) Hope, A. (2009). A Standard Drink
in Ireland: What strength?, Health Service Executive - Alcohol Implementation Group.
322
She and person 810 used to bring the drink into school and actually drink in school. To me
thats seems extraordinary. They used to go out and intentionally get drunk
Linda described that person 85 would often end up losing her clothes and do crazy things.
Over the years, she has realized that she does not know her limit which is why she now
abstains from drinking alcohol completely. Linda still goes out with her every week to dance.
Person 86 who is Lindas school friend as well as her work mate also usually accompanies
them. Linda described that she too is a heavy drinker who drinks to get drunk and can be a
burden. More recently, Linda and person 85 have stopped going out with her because they
felt they had to baby sit her, each time she drank. Linda described several occasions when
they had to hold her hair while she threw up somewhere or take her home and put her to bed.
Persons 810 and 811 who are Lindas peripheral contacts are the people with whom she
became friends through person 85. They are person 85s best friend and boy friend respectively
and Linda met them on a holiday on which they all had gone together. Person 811 is an
occasional drinker and Linda does not drink with him. Person 810 on the other hand is a heavy
drinker whom Linda described as her drinking buddy on the holiday.
When we were on the holidays, I and person 810 were drinking buddies and dancing buddies
and we were like lets party this place and that place. He is very extrovert which is a good
attitude to have whereas Id be a bit shyer. We drank a lot together and it was so much fun
Linda, person 85 and her friends often socialize together once a week on nights out.
Of all the network members, Linda drinks the most frequently with person 83 who is her
boyfriend. He is 7 years older than her. As described in chapter 7, they met on a pub crawl
where they became friendly. Linda recalled that he had bought her drinks all night. It appears
323
from her interview that person 83 loves partying, drinks heavily throughout a night and
encourages others to do the same by buying them drinks. He is admired among his friends for
his always party attitude towards drinking. Linda described that when she first started
socializing with his circle of friends, theyd tell her that she hasnt partied with anybody till
she parties with person 83
He is lethal. When he gets going there is no stopping him. We went out a few weeks ago with
8 people and he bought everyone around five rounds of Jagerbombs within the space of three
hours (sounds impressed). He is such a different person who keeps saying we are having a
good time, lets just keep on drinking and thats his attitude. If everyone is having a good time
he buys everyone a drink. He has bought a round of drinks for people he doesnt even know.
He gets drunk obviously but he also gets others drunk (laughs)
Linda drinks with him at least twice a week. They drink both in the company of other friends
as well as when they are by themselves. It appears from Lindas interview that her boyfriends
attitude towards drinking and partying prompts her to drink excessively. For example she
described that she always drinks beyond her cut off point and gets hammered when she is out
with him. She recalled several occasions when she felt she should have stopped drinking but
she didnt because her boyfriend kept buying her drinks in the swing of having a good time.
She also explained that she has stopped buying drinks for herself because he does it for her.
Some of her peripheral contacts are person 83s friends who she became friends with through
nights out. These are persons 814 and 815. They too are heavy drinkers and Linda feels that
their drinking is influenced by person 83.
324
I think they are more influenced by person 83. When they are together are they just want to have a good
time. They have a if we get drunk well then we get drunk, its great sort of attitude. .
Linda drinks with them whenever she socializes with her boyfriends circle of friends. She
does not see them otherwise.
Person 820 is a friend Linda made 4 years ago. He was going out with one of her school
friends at the time and Linda was introduced to him on a night out. Linda is still friends with
him and interacts with him frequently. Linda described that when she was underage, he used to
help her get into night clubs.
I met him at a party when I was 17 and still underage. He was heading off to a local night
club and he said come with me and I said I wont get in, I am only 17, he was like O I
know the bouncer, I know everybody, cling to me, you wont even be asked for an ID and he
got me in
He and Linda live close to each other. Linda described that whenever she feels like sitting in
the local for a while and have a few drinks, she sends him a text and he joins her. They
socialize in this way regularly.
Of the yellow nodes, person 813 is her ex-boyfriend. Linda has known him for 3 years. When
she started going out with him, he didnt drink much. Linda described that his parents are
pioneers and he wasnt really taught how to drink. Linda introduced him to drinking and
taught him about different types of drinks. She however feels that he cannot pace himself very
well because of which he get drunk almost every time that he is out.
325
The first time I brought him out, we were just buying pitchers. I had stopped drinking because
I knew my limit whereas he was this guy who is 6 ft 4 and he was on the floor while I was still
there because I was pacing myself. He was just downing it like coke or 7up. Any time that wed
go out drinking he would just get plastered
Although they are not in a relationship any more, they go out with work friends. Linda used to
accompany him to his goth club every week when they were together. They sometimes still
go there. Linda described that the club is a great place to meet different people and that she and
person 813 often end up going to someone elses house at the end of the night where the party
would usually continue well into the morning hours.
Person 812 is her exs best friend. Linda drinks with him frequently and describes him as a
party animal. He drinks several times a week and Linda feels that he is fun to hang around
with. Person 819 is another work mate who drinks moderately. He and Linda get to spend half
an hour together after work while they wait for their buses. During this time, they try out the
local pubs where they have a few drinks. Linda described that they do this regularly and take
turns in buying drinks each week. They also try to organize nights out at work.
Person 818 is Lindas younger cousin. Linda introduced her to drinking, make up and boys just
like person 87 did to her. Linda recalled that she would take her cousin to parties and sneak her
in the night clubs because she would look older than she was. Linda does not drink with her
very frequently but feels that person 818 drinks to get drunk from time to time.
Linda does not drink much with person 816 who is her cousin nor does she have much idea
about her drinking habits.
326
It is clear from the functioning of Lindas network that drinking is a key element of her
socialization with friends. She also drinks regularly with her parents though not as much as she
does with her friends. It is interesting to note, that she tends to be friends with people who are
older than her. Her survey response indicates that she perceives her proximal peers to drink to
drunkenness twice a week and drink 7-8 drinks on an occasion. These perceptions correspond
closely with the drinking behaviour of most friends in her network whom she described as
heavy drinkers. Although she drinks heavily, Linda does not consider herself to be a heavy
drinker. For example, she described that she is not technically a heavy drinker and talked
about how she always paces herself and knows her cut off point. However, it appears from
her interview that she does drink heavily and engage in heavy episodic drinking from time to
time. It is also clear that she responds to normative pressures to drink by trying to keep up
with other friends and drinking more alcohol in the company of friends who favour and
encourage heavy drinking (such as her boy friend-person 83 and her cousinperson 87).
Although Lindas boyfriend did not know her before she began drinking, it is evident from her
interview and the above discussion of her network that she feels compelled to drink to
drunkenness when she is in his company. While Linda seems surrounded by people who drink
heavily, it appears from her interview that she feels especially compelled and encouraged to
drink heavily when in the company of her boy friend and her cousin (person 87). Both persons
83 and 87 are embedded in the cohesive core, are very salient to her and she drinks with them
on regular basis.
327
8.3.4 Contextualized Network

8.3.4.1 Ruths network
Ruth is a 22 year old participant who belongs to the low drinking cohort. It appears from her
interview that most people in her network are light drinkers, especially those who are the most
salient to her and with whom she usually drinks. These people have been encircled in green in
the following visualization.
Figure 28: Ruth's overall network
Ruth belongs to a family where no one drinks to excess. Her father (person 73) completely
abstains from alcohol because he does not like drinking. Her mother (person 71) only drinks
occasionally. Ruth described that she might have a glass of wine with her colleagues once in a
few months. Her sister (person 72) goes out on the weekends with her own friends. Ruth
described that she only drinks socially with her friends and never with an intention to get
drunk.
328
Ruth started drinking when she was 17 years old. Persons 74 and 75 who are Ruths best
friends from school are the only people in the network (besides her family) who knew her
before she began drinking. It appears from her interview that they started drinking together and
formed similar attitudes and behaviour towards drinking. Ruth recalled that she and her friends
did not like drinking as much as some of the other students in her school right from those early
days. Her interview draws on several similarities between the drinking behaviours of these
friends and Ruths own behaviour. She described that she and her friends have never liked
going to pubs and clubs unless there was a special occasion such as a friends birthday. They
prefer to socialize in each others homes or go for shopping or to the cinema.
They are pretty much the same as me. We dont really drink that much because we dont like it. Wed
have an odd night once every few weeks if even and wed be happy with that. We just dont see the point
of going out and getting drunk and not remembering anything or not knowing what were doing. We see
that as really (stresses) pointless
Ruth recalled during the interview that the last time they went out was on New Years eve.
They had enjoyed a couple of drinks with dinner to celebrate the New Year and had returned
home by mid night. This was at least 6 weeks before the interview.
Of the other friends who are very salient to her, persons 710 and 711, have a similar drinking
behaviour. Ruth recalled that both girls have the same thoughts on drinking as she. Ruth
spent a year with them in relation to Erasmus. She did not mention any occasions or episodes
which suggest that they might have socialized in drink related ways during this time. It
appears that most of their activities did not involve alcohol.
329
We did a lot of things together. I was new to the place so they showed me around. We did homework
together, wed have classes together, and wed have dinner. During the week wed go to the cinema
and on the weekends if we had time wed go shopping
Person 76 who is a friend Ruth made on Facebook drinks more frequently than her. She
described during the interview that he goes out with his mates every weekend and sees
drinking to get drunk as more acceptable than she or her school friends would. Although Ruth
has never drunk with him, she feels that he drinks in moderation.
Persons 78 and 79 are her classmates in college. Ruth does not have much idea about their
drinking behaviours because she never drinks with them. As commented in chapter 7, Ruth has
only been on a night out with them once. It was a class night out in the first year. She described
that they have their own friends from home with whom they go drinking on the weekends.
Ruth does not see them outside college. Similarly, Person 79, who Ruth knows from the dance
class drinks at the weekends with her own friends. Ruth has never gone out with her but she
feels that person 79 does not like the idea of drinking with an intention to get drunk.
Ruths survey response suggests that she perceives her proximal peers to drink less than once a
month, have 1-2 drinks on an occasion and never drink to drunkenness. It appears from her
interview that these perceptions correspond closely with those of her school friends and the
two friends she made on Erasmus. It is clear from her interview that her own behaviour is very
similar to theirs. It is also apparent that whatever little Ruth drinks, it is with these people
especially persons 74 and 75 who are her best friends. These friends have known her from the
time when she didnt drink at all. They developed similar drinking attitudes together and
started reinforcing them in their day to day interaction. It is also possible that Ruths drinking
330
behaviour is also influenced by her parents however it appears from her interview that she and
her parents do not talk much about drinking. Ruth is in the final year of college and yet her
friendship with her class mates is restricted to the class room. The most salient peers in her
network are embedded in the clique (boxed in pink), which exists outside college.
8.4
Cross Case Analysis
A detailed and tabulated cross case analysis follows. It presents participants responses to key
survey items against their network features and themes identified across the cases as well as
the number of occasions these were evidenced. A check () in the following tables indicates
occurrence of a theme whereas a shaded cross (X) indicates absence of a theme.
331
332
333
334
The first spread sheet clearly depicts that the most salient peers were always found to exist
outside college settings and that they were embedded in dense sub groupings often comprising
long standing friendships dating back to school years. In most cases, these salient friends lived
in the same general area as the participants and/or shared an accommodation with them.
Further, these friendships featured regular interaction, reciprocal exchange of social
companionship and emotional support and socialization in varied ways. The only exception
was Cormac who described that his closest friends with whom he went to school now live in
other towns. Cormac works long hours most days of the week as well as attends college. He
does so to support himself financially. It is because of these reasons that he gets very little time
to socialize or keep in touch with his friends. The themes related to regular interaction,
socialization in varied ways and provision of social companionship were therefore not reflected
in Cormacs interview. Although Cormac does not interact with his friends regularly, he does
rely on them for emotional support. It is also interesting to note that 21 of 26 participants had
spent 3 or more years in DIT, yet their salient subgroups mostly comprised friends from
outside DIT and interaction with college friends remained limited to college settings, college
nights out or exchange of instrumental support. In only a few cases, a college friend was found
to be salient in addition to older friends from school/home. These included the networks of
Emma, Debbie, Pam and Bella each of whom had one or two friend(s) from college who was
(were) salient to them.
The second spread sheet depicts participants responses to personal consumption items on the
web survey against interview themes related to the development and dissemination of drinking
norms in their networks. It provides a summary of evidence supporting the influence of the
335
most salient peers on participants personal consumption. As is evident from the tabulated
findings, drinking behaviours of the most salient peers and the participants were found to be
homophilous (similar). This finding was consistent across all networks examined in this study.
Further, in majority of cases, salient friendships preceded age of first drink and participants
earliest experiences involving alcohol occurred in the company of their school friends.
Drinking norms in these cases developed over several years during which participants shared
alcohol related experiences with their salient friends. Most participants described that they
mainly drink with their salient friends; that group norms related to alcohol are mutually
reinforced through behaviour in social situations and that conformance to alcohol related group
norms is expected and encouraged among their salient friends. These findings reflect that tacit
pressures to follow the norms of the group do exist within most networks. Further, parents
were not found to influence drinking behaviours of the participants. This was consistent across
all ego networks.
The cases where the theme salient friendship preceded age of first drink did not occur
included the networks of Amy, Peter and Alex. For Amy and Peter, the most salient peer was a
relatively recent girlfriend and boyfriend respectively whereas for Alex the most salient peers
were friends he had recently made during his study period outside of Ireland. In cases where a
salient friendship followed age of first drink, it was noted that the participants either changed
their drinking behaviours after befriending such peer(s) and/or described feeling compelled and
encouraged to drink similarly when in the company of such friend(s). In these instances, peer
influence offers a more likely explanation of similarity in drinking behaviours than peer
selection. Some of these cases such as the networks of Linda, Peter and Ken are discussed in
detail in section 8.3. Therefore, while the design of this study does not preclude the possibility
336
of a reverse causal explanation, it adds weight to the presented results. In some cases,
participants described rejecting the norms of some friends after having conformed to them for a
certain period of time. These cases (Sam, Peter and Ken) are discussed in section 8.3.
Table 38, presented in section 8.2 identified the salient network members whose drinking
behaviours corresponded closely to participants survey reports of perceived norms (proximal
peers). The third spreadsheet of cross case analysis demonstrates how this correspondence was
established for both descriptive and injunctive norms. Specifically, it provides survey measures
of perceived descriptive and injunctive norms, categorizes them in accordance with the
drinking intensity criteria presented in chapter 5 (Table 4) and presents this information against
comparable and related themes found in the interviews. As described in section 8.2 the
participants used various colloquial terms and drinking jargon in describing their friends
drinking levels and in recalling specific examples and episodes. The themes used in this
analysis were therefore largely derived from this knowledge. The tabulated findings reflect
consistency in perceived norms of proximal peers reported in the web survey and those of the
most salient peers as it appears from the interviews.
8.5
Linking the Outcomes of Web Survey and SNA
The use of SNA in combination with the web survey uncovered important findings in this
study. On one hand, the analysis of networks and interviews complemented the findings of the
web survey by demonstrating agreement with them and offering probable explanations. On the
other hand, this combination extended and enriched the quantitative findings by providing a
novel perspective on norm salience and bringing to light important aspects of participants
337
networks which could not have been explored and uncovered following a survey based
methodology alone. Further, this approach also uncovered some unexpected findings which
open up new fields of inquiry within normative research. These aspects are discussed next.
First, SNA and the qualitative analysis validated the web survey in that a very close
correspondence was found between the perceived norms of proximal peers in the survey and
those of the most salient peers in the ego networks. In general, these peers shared strong and
long standing friendships with the participants based on mutual trust, companionship and
reciprocal exchange of emotional support. These peers also scored high on numerical scores of
tie strength. The literature on saliency emphasizes that it is crucial to understand how students
decide which of their friendship groups to draw from when answering items querying
normative perceptions in surveys (McAlaney et al., 2011). This information is directly related
to the effectiveness of SN interventions. Given that the ties identified as being the most
salient were very significant in the social lives of the participants and contributed to their
social and emotional well being, it is highly likely that the participants were thinking of these
individuals when conceptualizing proximal peers in the web survey. These findings suggest
that individuals are likely to think of specific rather than generic peer groups when they
respond to normative items on SN surveys.
Second, participants self reported drinking behaviours as reported in the web survey were
found to be very similar to the drinking behaviours of their most salient peers. In general, this
similarity in drinking behaviour appeared to be due to peer influence rather than selection. This
is in agreement with the web survey which found perceived drinking of proximal peers to be
closest to and predictive of participants own drinking. This finding that our networks influence
338
and shape our behaviour is also backed by recent network research (Christakis and Fowler,
2009; Rosenquist et al., 2010) which provides evidence of influence up to three degrees. This
research was based on the Framingham Heart Study data base which contains detailed
information on over 12,000 individuals monitored over more than three decades since 1971.
The researchers found that a person was 50 percent more likely to drink heavily if a person
they are directly connected with also drinks heavily and 36 percent more likely to drink heavily
if a friend of a friend drinks heavily. The impact extended up to three degrees of separation:
our friend (one degree), our friends friends (two degrees), our friends friends friends (3
degrees).
Third, one of the key findings from the web survey as discussed in chapter 6 was that proximal
peers (close friends and best friends) were found to be better predictors of personal drinking
behaviour than distal peers and that parents were not found to be a significant source of
influence. Qualitative analysis of the interviews complements these findings by providing
evidence that participants drinking behaviours were primarily influenced by friends who were
socially salient to them in several ways and that parents were not the primary influence.
Further, SNA and qualitative interviews extended this finding by identifying who these salient
peers were, where they were situated in the networks and how they influenced participants
drinking behaviours. SNA combined with qualitative interviews allowed examining how
drinking behaviours and attitudes developed and sustained over the years in participants
networks, and how the members currently socialize. The results provided useful and consistent
evidence that the structure of individuals networks was related to the drinking behaviours and
attitudes of their inhabitants and provided useful information about the development,
reinforcement and transmission of drinking norms. In general, it was found that the participants
339
developed their perceptions of what is normative drinking behaviour within tightly knit sub
groups and that it was within these sub groups that they regularly reinforced these norms in
social situations. It was demonstrated with the aid of examples and network visualizations that
the most salient peers were embedded in cohesive sub groups which existed outside college
settings. For example, the participants whose networks were shaped like a group described
the group as the central and the most dominant element of their social lives. The most salient
and influential nodes were found to be always embedded in the group. Similarly, in networks
shaped like core/periphery, the most salient and influential peers were always rooted in the
core. The overlapping cliques type networks were different in that, the most salient peers
were scattered in different cliques, each clique being distinct in its relevance to the participant.
However, it was noticed that the most salient and influential peers were almost always
embedded in a clique which existed outside college settings and generally comprised older
friends.
Fourth, the web survey found that DIT students misperceive and more specifically
overestimate the drinking of other students on campus and that perceived norms predict
personal consumption. On its own, this finding justifies the use of SN marketing campaigns in
DIT to reduce alcohol consumption rates on campus. However, the analysis of networks and
interviews suggests otherwise and emphasizes that establishing misperceptions should not be
the only basis for a SN marketing campaign and that norm salience is a key factor in this
equation. Given that the most salient and influential peers were found to exist outside of DIT,
this study does not support SN marketing campaigns in DIT as these mostly address the norms
of a typical student rather than those of a more intimate group of friends.
340
Last, the qualitative interpretation of interviews provided useful information about network
dynamics. It was found that sometimes individuals rejected the drinking norms of friends
because they no longer found them preferable or appropriate. As a consequence of this
mismatch between drinking behaviours, friendships in a network sometime decay and even
diminish altogether. Although relatively uncommon, this phenomenon does occur and requires
further understanding.
The
implications
of
these
findings
are
341
discussed
in
the
next
chapter.
Discussion of Results
9.1
Introduction
This chapter addresses the research objectives set out in chapter 5, section 5.2, in light of the
key findings. It also highlights the implications of these findings and their contributions to SN
theory. In doing so, the research objectives are revisited and their origin and rationale restated.
The evidence regarding each research objective is presented. The implications and theoretical
contributions of the findings are discussed next. The limitations of this study are then
acknowledged and some useful recommendations for future research follow.
9.2
Summary of Evidence Against Research Objectives
9.2.1 Research Objectives 1 and 2

The first two objectives of this research stemmed from the apparent inadequacy of existing
policies in Ireland to reduce alcohol consumption and related harm and the need to consider
alternative strategies such as the SN marketing campaigns to deal with this problem. However,
there is an apparent lack of published empirical work examining the potential of SN theory in
Ireland. In order to justify the use of SN marketing campaigns to reduce alcohol consumption
in any population of interest, an important first step is to establish that misperceptions of
alcohol related norms exist in the population. In line with this rationale, the first objective of
this study was
342
1. To investigate the extent of misperception among Irish students regarding peer drinking
norms
The campus norm was found to be clearly inclined towards heavy drinking with the average
consumption figures showing that DIT students drink approximately 5 days a month, consume
nearly 9 drinks on a typical occasion and get drunk every week. As hypothesized, the findings
indicated an overall misperception of the campus norm. Specifically, students overestimated
the drinking of most students at DIT perceiving them to drink more often (80% of the sample
overestimated the norm), consume higher quantities of alcohol (60% of the sample
overestimated the norm) and drink to drunkenness more frequently (85% of the sample
overestimated the norm) than they themselves. The results demonstrated that DIT students on
average perceive other DIT students to drink almost twice as often as them in a typical month,
consume at least 1 additional drink on a typical occasion and get drunk twice as frequently as
them in a typical month. Further, students also estimated their proximal peers to drink more
alcohol more frequently than their own selves. These perceptions rose as the social distance
from the individuals increased, with the perceived norms of proximal peers being closest to
students own drinking behaviours. These findings were consistent across all campuses of DIT
and across gender. These were also consistent for abstainers who comprised 4% of the sample.
Related to the first objective, the second objective of this research was
2. To examine the association of normative perceptions of prevalence with students own
drinking behaviour
As hypothesized, an overall positive and statistically significant correlation was found between
individuals own drinking behaviours and their perceptions of the behaviour in peers. The
343
strength of this association was found to be stronger for proximal peers than distal peers and
among proximal peers, stronger for best friend than close friends. Hierarchical multiple
regression analysis revealed that higher perceptions of prevalence of drinking (permissive
descriptive norms) predicted higher personal consumption after several known confounding
factors were accounted for. This effect was stronger for proximal pees than distal peers.
Parental alcohol use however, correlated positively but weakly with personal consumption and
did not significantly predict the latter. Overall, regression models were able to explain 41.4%
of the variance in frequency of drinking (typical month), 73% of the variance in number of
drinks (typical occasion) and 52.6% of the variance in frequency of drunkenness (typical
month).
9.2.2 Research Objective 3
The literature reviewed in chapter 3, section 3.9, drew attention to an overall lack of research
examining the potential of injunctive norms in inducing behaviour change and identified it as
of the key weaknesses of SN theory. It is unclear from the existing research whether
descriptive or injunctive norms are more likely to change behaviour and should therefore be
preferred in SN marketing campaigns. In line with this, the third objective of this research was
3. To study the relative impact of injunctive and descriptive norms on drinking behaviour
The findings demonstrated that descriptive norms have a stronger effect on personal
consumption than injunctive norms. Specifically, it was found that the perceptions of
descriptive norms were more strongly correlated with personal consumption than those of
344
injunctive norms and that this was consistent across all reference groups examined in the
study29. Similarly, hierarchal multiple regression analysis revealed that while perceived
injunctive norms of proximal peers, predicted personal consumption of alcohol, these effects
were not significant once the perceived descriptive norms were introduced into the regression
model30. Further, the addition of perceived injunctive norms to regression models did not
improve the explained variance in dependent variables as much as the addition of perceived
descriptive norms did (chapter 6, section 6.4).
9.2.3 Research Objectives 4 and 5
One of the primary considerations for the success of any SN marketing campaign is to target
those reference groups which are the most salient and relevant to the population of interest. As
discussed in chapter 3, section 3.9.3, a key challenge faced by SN research relates to
identifying the most salient reference groups for a population and understanding how
individuals visualize these groups. SNA is a trans-disciplinary field which uses sophisticated
mathematical techniques embedded in graph theory to study relational data. The fourth and the
fifth objectives of this study were aimed at addressing this gap of norm salience via SNA and
were stated as
4. To identify and locate the salient peers in personal networks of students using social network analysis
29
The only exceptions were the perceived injunctive and descriptive norms of parents which correlated positively
but weakly with personal consumption.
30
The only exception was the perceived injunctive norms of proximal peers and mother when they predicted
ones frequency of drunkenness. These effects were significant even after controlling for perceived descriptive
norms
345
5. To examine the association of personal networks with normative perceptions and drinking
behaviour
Research objective 4 was addressed by examining the composition and structure of ego
networks in parallel with a numerical and qualitative assessment of tie strength. Research
objective 5 was addressed by integrating survey and interview data to examine the
correspondence between self reported personal consumption/perceived norms as reported in
the survey and the drinking behaviours of network members as transpired from the in depth
interviews.
Compositional analysis revealed that in general, the networks comprised a higher percentage
(more than 70%) of peers from outside college than from within college. Five distinct
typologies were identified based on structural variations in the networks which were explained
with the help of network visualizations in chapter 7. A key finding which emerged from the
analysis of network structures combined with an assessment of tie strength and a qualitative
interpretation of the interviews was that the strongest and the most salient ties existed outside
college settings generally embedded in cohesive and intimate sub groups. Synonymous with
Granovetters (1973) indicators of tie strength, these ties generally comprised long standing
friendships based on regular interaction, intimacy, mutual trust and exchange of social
companionship and emotional support.
As hypothesized, the integration of qualitative and quantitative data revealed that the drinking
behaviours of salient network members (based on interviews) corresponded closely with both
self reported drinking behaviours of the participants and their perceptions of the behaviour in
proximal peers (based on survey). This reinforces the results of the web survey which found
346
the perceived norms of proximal peers to be closely related to students own drinking
behaviours. Further, it appeared that the similarity in drinking behaviours of the participants
and their salient peers was attributable to the influence of these peers rather than the reverse
causal explanation which implies self selection into networks with similar drinking behaviours.
Finally, the structure of individuals networks was found to provide useful information about
how norms developed, sustained and transmitted in networks. Specifically, the study provided
strong and consistent evidence that participants perceptions of what is normative drinking
behaviour were largely developed within cohesive sub groups of strong and intimate ties
mostly comprising school friends and that it was within these sub groups that they regularly
reinforced these norms in social situations. This was exemplified with a discussion of specific
cases and cross case analysis in chapter 8, section 8.3 and 8.4.
Further, these friends were found to be more salient and influential than family. This supports
classic research on college drinking which argues that there is a pronounced shift in influence
from parents to peers at the onset of college (White et al., 1991). Subsequently, peers become
increasingly important as youngsters become relatively independent from parental oversight
and control (Brown et al., 1997). This is also in agreement with the web survey which found
the perceived norms of proximal peers to be more influential than parental norms.
9.3
Implications and Theoretical Contributions
9.3.1 First Evidence of Alcohol Related Misperceptions in Ireland

One of the contributions of this study is that it provides the first evidence which documents
misperceptions of alcohol related norms in Ireland and establishes their impact on personal
347
consumption. In doing so, it extends and supports the research which has already validated SN
marketing theory in other heavy drinking cultures which are similar to Ireland (McAlaney and
McMahon, 2007). SN marketing campaigns induce behaviour change by reducing
misperceptions related to the problem behaviour (Berkowitz, 2004). Therefore, a primary
prerequisite of any SN marketing campaign aimed at reducing drinking rates is to first establish
if misperceptions of peer drinking norms exist in the population and then reduce these
misperceptions via intervention. As commented in chapter 3, section 3.8.4, SN theory is often
criticized in relation to its applicability in situations where a majority of the population exhibits
heavy drinking behaviour thus making it normative. The proponents of SN approach argue that
even in such situations, individuals will overestimate peer drinking and that consequently
misperceptions will still hold. Although misperceptions of alcohol related norms have been
documented in other heavy drinking environments like the UK (McAlaney and McMahon,
2007) which is culturally similar to Ireland, local evidence was still required in order to justify
a SN campaign in Ireland and establish base line statistics for campaign evaluation. Further,
this step was also a pre requisite to exploring the issue of norm salience which was the key
focus of this study.
9.3.2 Integration of Norm Salience and SNA

A unique aspect of this study was the integration of norm salience and SNA. It is the first study
to have utilized ego network analysis to examine the SN theory in context of college drinking.
In doing so it provides a unique perspective on norm salience based on structural
configurations and functioning of networks.
348
To date, most SN marketing interventions have utilized the typical or average college student
as a normative referent (Perkins, 2003b; Lewis and Neighbors, 2006b). The present study
provides evidence that the typical or average student may not always be an ideal normative
referent. The finding that the most salient and relevant ties of participants existed outside DIT
embedded in cohesive subgroups often comprising older friends from school implies that a SN
marketing campaign targeting the norms of a typical or average student may not be
successful in DIT. This finding calls in question the applicability of SN marketing campaigns
to reduce alcohol consumption in Ireland particularly in DIT. It also has important implications
for current practice and policy on SN approach.
Misperceptions alone are not conclusive
The study contributes to theory by providing evidence that establishing misperceptions of
alcohol related norms is not conclusive evidence on which to base a SN marketing campaign to
reduce drinking rates in a college population. Even when students misperceive the norms of
other students on campus, a SN based campaign may not be justified as was found to be the
case in DIT. This is because norm salience is an equally important consideration which must
be addressed in parallel for improved institutional and policy measures. This implies that
practitioners of SN theory should preferably reassess their procedures to ascertain that socially
salient referent groups are being targeted in campaigns. As commented in chapter 3, section
3.9.3, while SN theory acknowledges the importance of salient norms in influencing
behaviours (Berkowitz, 2004), in practice, interventions are impeded by challenges in
addressing this key issue appropriately (McAlaney et al., 2011). This in part reflects why most
studies have primarily focussed on documenting misperceptions alone. The present study
349
demonstrates that the integration of SNA and norm salience is a useful way to address this gap
in literature. Incorporating SNA in the planning stages of a campaign to determine salient
referent groups can be an effective strategy towards improved policy making.
Accessing groups outside of college
The present study also contributes to SN theory by raising several practical questions in
relation to accessing salient referent groups when they inhabit networks outside of college
settings as was found to be the case in DIT. For example, if these groups were to be targeted
with a norm change strategy such as personalized normative feedback, what would be the best
way to access these peers, ascertain their consumption levels and establish that misperceptions
exist within these groups? Similarly, what kind of normative messages would be most
appropriate for such tightly knit groups? Finally, why would the target audience believe in the
content of normative feedback over their own experiences with their friends? These issues pose
serious difficulties for the implementation of a SN campaign to target these groups. There is
another potential barrier which policy makers must be aware of. Normative belief interventions
operate on normative misperceptions. The extent of overestimations decrease as does the social
distance between an individual and referent group as has been shown to be the case in DIT.
Researchers warn that there may come a point in which a referent group is so close to an
individual that normative belief interventions based on that referent group would have a
negligible effect (McAlaney et al., 2011). This implies that SN campaigns may not be an
appropriate strategy in such situations.
350
Phenomenon of commuter colleges: Focussing campaigns on pre college years

The present studys findings related to norm salience stand in contrast to the majority of
normative intervention research conducted in the US, which has primarily focused on the
typical student and found it to be a successful strategy in reducing problem behaviours. It is
highly likely that this deviation is related to the differences in residential arrangements at the
American and the Irish colleges and that the study may have uncovered a phenomenon which
is especially applicable to commuter colleges. This is a key contribution which can have
important implications for policy making. Most of the colleges in the US academic system
offer on-campus accommodation for students generally referred to as dorms or residential
halls. In addition, most of these colleges are based away from major cities constituting what
can be described as self contained cities boasting a vibrant campus life. Perkins (2003a),
explains that the socialization in residential colleges is extremely peer intensive. Reflecting
on the above, it is only natural for these students to develop close and intimate ties with each
other because they live together, interact frequently in a variety of ways and lack frequent
contact with other reference groups such as previous circles of friends and family. On the other
hand, the majority of Irish colleges have very limited on-campus accommodation with some
like DIT offering no such option. As a result, most college students in Ireland and specifically
those studying at DIT reside in self catering accommodation or live with their families and
commute to college every day from their homes. While the American colleges lie on one end
of the spectrum secluded from the major cities and constituting a thriving campus life, DIT lies
on the farthest extreme with no on-campus accommodation and therefore relatively fewer
opportunities to develop intimate ties with in-college peers and several campuses scattered
across the city with very little interaction between them. It is a natural rather than a scientific
351
observation that the social circles of students in such situations will generally comprise friends
from their area/locality with whom they spend most of their free time. Further, the major
colleges of the country are based in Dublin which is the centre of social activity in Ireland.
Since most of the Irish students go to colleges in Dublin, they do not have to reduce
communication with older friends who probably also go to a college in Dublin. The
opportunities to interact and develop deep friendships with college peers are thus markedly
reduced.
The above observations are supported by SN literature. Berkowitz (2004) reflects that when
most students in a college live off campus, they may differ from the on-campus students in
terms of the saliency of campus norms. Given that this study may have reflected a trend
especially prominent in commuter campuses, it may be time to re-examine how policy is being
framed and to attempt alternative approaches to target such populations with socially salient
normative content. One possible solution can be to shift focus of interventions on pre college
years. The present study demonstrated that the development, reinforcement and transmission of
drinking norms primarily occurred in cohesive sub groups comprising friends with whom the
participants had gone to school and with whom they continued to drink regularly. Schools may
therefore be the most suitable and practical environments to maximize access to these sub
groups via SN marketing campaigns. This approach may also be beneficial in addressing some
of the practical concerns discussed under the preceding heading. Institutional and policy
measures may potentially benefit from this approach as it offers the opportunity to attempt to
curtail unhealthy drinking behaviours in adolescent groups before they transform into
irrevocable and binding group norms. Recent SN research in Europe (Balvig and Holmberg,
2011) and Australia (Hughes et al., 2008) also proposes similar suggestions. Balvig and
352
Holmberg (2011) argue that targeting school children with SN interventions to reduce smoking
can prevent the onset of smoking before it turns into problem behaviour. In contrast, when
interventions are focussed on college students, the risk behaviours which are sought to be
modified are already well established among those participating. Reflecting on the above, this
may be an even greater concern in situations where alcohol becomes legally available at a
younger age (18 years in most of the Europe compared to 21 in USA) and most youngsters
begin drinking even earlier than the legal drinking age as is the case in Ireland.
9.3.3 Descriptive norms vs. Injunctive norms

The present study is one of the few studies to directly compare the two types of norms in terms
of their influence on drinking behaviour. It strengthens SN theory by providing evidence that
perceptions of descriptive norms have a markedly stronger effect31 on personal consumption
than those of injunctive norms and should therefore be the preferred target for SN marketing
interventions. This means that while targeting misperceptions of descriptive norms is an
effective strategy to induce behaviour change, targeting only the perceptions of injunctive
norms may not be as successful. Further, the finding that injunctive norms of only the proximal
peers were found to have any significant effect on personal consumption implies that
individuals are largely unaffected by injunctive norms of distal peers and that SN campaigns
based on injunctive norms of distal referents are not likely to succeed. Past research has
pointed out that one of the primary motivations to conform to injunctive norms is the fear of
negative evaluation and that social approval is particularly sought for maintaining group
31
Based on the assumption that perceptions preceded drinking behaviour

353
memberships (Borsari and Carey, 2001). It is only natural then that within proximal groups,
individuals might feel more motivated and even compelled to conform to injunctive norms
because of the fear of social exclusion or negative evaluation. The problems associated with
accessing these proximal groups and the possibility that SN campaigns will have a negligible
impact when the deviance between personal behaviours and perceived norms of such peers is
little further complicate addressing injunctive norms of proximal groups in practice.
9.4
Other Findings and Their Contributions
The study demonstrated that some participants distanced themselves from certain friends
because they no longer deemed their drinking behaviours acceptable and/or appropriate. As
described with the help of specific examples in chapter 8, section 8.3.1, these individuals
described feeling that their reputation or health would have been at stake had they maintained
close ties with such friends. This finding implies that friendships do deteriorate and decay as a
result of people rejecting the drinking norms and practices of their peers if they dont find such
norms preferable or apt. It contributes to SN theory by drawing attention towards a relatively
uncommon but important phenomenon which suggests that just as people select friends based
on similar drinking behaviours (Mundt et al., 2012), they also sometimes reject friends based
on dissimilar drinking behaviours. This merits further consideration and opens up a new field
of inquiry within normative research.
Another finding that emerged from the analysis of network structures was the identification of
overlapping cliques, which was found to be the most common structural configuration in this
study. As described in chapter 7, such networks featured multiple cliques which overlapped
354
through common nodes or cross clique friendships. The identification of this typology
contributes to network literature by extending and adding to the previous research by Bellotti
(2008) which examines structural variations in ego networks and reports the occurrence of four
network types namely, small cliques, groups, core/periphery and contextualized.
9.5
Limitations
The following factors must be considered when evaluating the results of this study.
1. Causality: This study was cross sectional in nature making it difficult to derive causal
inferences. Although, it was assumed on the basis of theory and past longitudinal
research (Fearnow-Kenny et al., 2001) that normative perceptions cause personal
consumption, the research design does not preclude an alternative reciprocal
relationship. As described previously, the design of this study allowed assessing if
relationships in the networks preceded or followed participants age of first drink.
Further, the qualitative nature of in depth interviews allowed examining the contexts in
which relationships formed and flourished. These aspects of the study strengthen the
analysis and suggest peer influence as a more plausible explanation for similarity in
drinking behaviours of the participants and their networks compared to selection.
2. Self Reports: As is often the case with alcohol related research, the results of this study
are based entirely on self reports, the validity of which is a known issue of long standing debate in substance use research (Midanik, 1988). However, there is evidence that
self reports in college samples can be reasonably reliable and valid (Johnston, 2001).
This issue was discussed in detail in chapter 5, section 5.5.1 in relation to web surveys.
355
3. Generalizing the results: This study was based on a student sample which naturally limits the applicability of findings to the general population. However, such a sample
makes it easier and practical to compare the results with existing normative belief research, which originates almost exclusively from the American college system. Findings from the web survey may be generalized to undergraduate student population at
DIT. However, they might not accurately reflect consumption behaviours and normative beliefs prevalent at other colleges in Ireland.
It must also be acknowledged that the network results are not statistically representative
of the population. They must therefore be interpreted with caution and the design of
this research replicated on a wider scale to validate the results. However, this is the first
study to examine, SN theory in relation to college drinking using ego network approach
combined with in depth interviews. The over arching aim of utilizing network science
in this study was to understand the theoretical aspects of the phenomenon of norm
salience rather than achieving statistical representativeness. Further, as described in
section 9.2.3, DIT has a commuter campus without any live-in accommodation, at the
extreme end of the spectrum from many American colleges which feature on-campus
residence halls and a thriving campus life. It is possible that network results and related
qualitative explanations which emerged in this study reflect a phenomenon which is
especially relevant to commuter colleges. To the extent that this is a phenomenon of
commuter colleges, the results may be tentatively and naturalistically generalized.
4. The structural typologies derived in this study are not exhaustive of all possible
typologies. Ideally, a statistically representative survey can identify other types of
networks and assess their distribution and impact in the population. However,
356
generating such detailed ego network data from a statistically representative sample
remains prohibitively costly in terms of time and money.
5. It must also be acknowledged that ego network studies view networks from the eyes of
participants. It is generally not possible to collect network data from all alters which is
why the connections examined in this study are based entirely on how the participants
perceived their ego networks rather than individual reports from each network member.
6. From a network perspective, a whole network study would have allowed mapping a
complete network of students at DIT based on self reports. This would have provided
maximum coverage of the population. However, an inherent limitation of such an
approach would have been its inability to capture relationships extending beyond the
college settings which in this study were found to be salient. The incorporation of
whole network approach in intervention work may be more beneficial in situations
where salient referent groups lie within college boundaries as has been explained in sec
9.6 (Point 4).
7. Including multiple institutes in this study would have allowed examining if
misperceptions of alcohol related norms and explanations related to norm salience are
also reflected in other Irish third level institutes.
8. This study was based on undergraduate students at DIT. It would have been interesting
to examine if the networks of post graduate students have a similar or different
functioning.
9. Finally, social networks are dynamic and fluid. They change over time. A longitudinal
study collecting network data at multiple time points would have allowed examining
357
how the composition and structure of networks, their functioning and the norms they
foster change and evolve over time.
9.6
Recommendations for Future Research
This section draws attention to some important theoretical and practical aspects which future
research may wish to explore.
1. The one unexpected finding which emerged from the analysis of networks in this study is
that the peers who were the most salient in determining participants drinking behaviours
were found to exist outside college settings. This finding implies that a SN marketing
campaign targeting a typical student may not be an effective strategy within DIT to
reduce drinking levels. As explained in section 9.2.3, this may be a phenomenon which is
especially relevant to commuter campuses like DIT and many other colleges in Ireland.
Future research should begin by examining the extent to which this phenomenon is
consistent in bigger samples. Ideally, a representative sample of full time undergraduate
students at DIT is recommended to validate these results. The present study might be
regarded as a valuable exploratory investigation which clearly showed the benefit of
combining network analysis with in depth interviews. However, given the constraints of
time and money, it can be impractical to conduct a large scale representative study which
collects such detailed ego network data via in depth interviews. Future research may
address this problem by utilizing web based methods and tools for collecting ego network
data and exploring attributes of relationships and drinking behaviours. The use of web
based methods in SNA is growing (Coromina and Coenders, 2006). There is mixed
evidence about the potential of these methods. While some researchers report that the
358
quality of ego network data collected via web methods seems to compare well with that of
traditional modes (Coromina and Coenders, 2006), others argue that the quality of data
may suffer due to respondents answering in a time saving and almost mechanical manner
(Matzat and Snijders, 2010). Studies that have used such methods include Marin (2004),
Manfreda et al (2004), Coromina and Coenders (2006), Vehovar et al (2008), Matzat and
Snidjers (2010) and Tubaro et al (2012). These studies reflect that careful pretesting of the
survey instrument, visual design, use of graphic stimuli to prevent invalid entries, format of
the name generators and possibilities of dynamic interaction between the respondent and
the questionnaire are some of the challenges associated with web based network studies.
Further, this body of research suggests that the number of alters (network members to be
named) should be limited in some substantial manner in web based methods as it lengthens
the survey and increases the risk of respondents dropping out. In response to these
concerns, a forthcoming publication (Tubaro et al., 2012) suggests the use of participant
generated web based sociograms in internet surveys. The study reports on the potential of
an interactive graphical interface which enables the participants to draw their own personal
networks with simple and intuitive tools. The authors report that the method reduces
respondent burden and the risk of drop out.
2. It will be beneficial if other Irish colleges with commuter campuses replicate the current
study to validate and cross examine the findings among their students. It will help in
establishing figures that most closely represent the drinking patterns of Irish students. It
will also help to confirm the DIT finding, that misperceptions of alcohol related norms
exist among students. In addition, if the other institutions also find that the most salient
relationships exist beyond the college frontiers, then this will reinforce the findings of this
359
study and endorse the conclusion that SN marketing campaigns focusing on a typical
student are not an effective strategy within Irish colleges to reduce alcohol consumption.
3. Another potentially interesting area to be explored by future SN research in Ireland relates
to assessing the applicability of SN marketing interventions to second level (school)
students in Ireland. The results of this study indicate that the most salient peers existed in
cohesive structures outside college. These structures often comprised people who went to
the same schools and lived in the same general area. It is possible that socially salient
referent groups for school students will comprise on-campus peers. Thus, SN marketing
interventions targeting typical student as a normative referent may be more useful in
changing perceptions of drink related norms in Irish school students. This is an important
hypothesis based on the main findings of this study and should be investigated further
because it would have important implications for SN research in Ireland. It is crucial that a
research focusing on populations of school children address related ethical concerns. For
example, children under 18 years of age are not legally competent to provide consent to
participate in a research. This necessitates obtaining appropriate consent from first the
school authorities and then the parents or guardians. Clearly laying out the purpose of the
investigation and the involvement of children in the outcomes of research in a manner
which is appropriate to the participants is also paramount. Last, compliance with legal
requirements of child protection and safeguarding children during research are also central
to the ethics of conducting such a study.
4. The approach taken by the present research shows the real benefit to be gained from
incorporating network science in SN marketing interventions to reduce alcohol
360
consumption and future researchers should follow this lead. This recommendation is
especially relevant to those institutions where salient reference groups reside within college
settings and campus wide normative marketing campaigns are hence deemed suitable. As
described earlier, this is not the case with DIT based on the results of this study. The
incorporation of network science in the existing line of standard and more customary
intervention work can be achieved via the whole network approach which is well suited for
campus wide programmes. Indeed, the idea of utilizing the whole network approach in
interventions targeting norm change has generated increased interest in academic research
circles. A recent study (Paluck and Shepherd, 2012) examining the descriptive norm for
peer harassment in a high school setting reports on the first randomized experiment to
assign individuals to treatment based on their position in the schools complete social
network. As elaborated in chapter 4, section 4.5.4, they used a survey and SNA to map the
entire social network of a school in Connecticut. Widely known people and clique leaders
were identified as social referents. A widely known individual was conceptualized as a
person who had many ties to individuals across the network whereas clique leaders were
defined as those who had a large number of ties within a clique. A subset of these social
referents was assigned to a randomized intervention which involved these people
demonstrating their opposition to peer harassment in the school assembly. The results
showed that these social referents were able to significantly change the perception that peer
harassment was typical. This shows that SNA is a cutting edge and novel approach in SN
research. However, the SN literature on college drinking reflects a lack of similar studies
and this is a gap where future research should contribute. If network ties can be discerned
on a campus wide scale, the influential individuals or those most at risk for being affected
361
by normative pressures can be targeted by virtue of their network positions. It can also
enable researchers to design interventions which target groups of inter connected people.
This is because network science allows one to identify centrally located hubs that have
access to most of the larger network (Valente, 1996; Valente, 2003; Valente et al., 2004).
For example, the usual approach to reduce drinking on a campus is to either broadcast
feedback to everyone or work with a small group that is felt to be at risk because survey
data identified them to be heavy or problem drinkers. An alternative strategy can be to
identify hubs in the social network comprising people who might or might not be drinkers
and target them with corrective normative feedback. This can have important implications
because research confirms that people can be influential in the spread of norms even if they
do not exhibit the behaviour themselves. They can still be the carriers of the idea and thus
promote it unintentionally (Valente, 1996; Valente, 2003; Valente et al., 2004).
Encouraging results have been documented with the use of this strategy in promoting better
diets and safer sex previously (Buller et al., 1999; Sikkema et al., 2000). A further benefit
of incorporating network science in normative research at campus level is that it will allow
us to study the change in student networks over time and the implications this has on peer
drinking norms. The elaborate knowledge of connections at campus level will allow for
comparing and monitoring the network dynamics, patterns of drinking and the relation
between them at later points in time. This will improve programme monitoring and
implementation by offering improved and specifically tailored follow up normative
feedback embedded in longitudinal research. The above discussion highlights some of the
avenues that the future SN research can explore and benefit from.
362
5. A relevant and interesting area of research is that of network dynamics. Networks are
vibrant and change perhaps is the only constant in them. They grow, shrink and even die
with time. Norms are no different. They too change as we age, move up in college, befriend
new people or let go of older ones. The present study demonstrates that sometimes
individuals rejected the drinking norms of their friends because they no longer found them
acceptable and in line with their own behaviours and attitudes. Though these friends still
formed part of the participants networks, the participants made a conscious decision to
distance themselves from them. The occurrence of this phenomenon may be relatively
uncommon but the fact that it exists and that it is practiced opens up a new field of inquiry.
6. Finally, the majority of past SN studies have been cross sectional in nature. It is
recommended that the increased use of longitudinal research will provide increased
methodological rigor to SN marketing interventions. This will provide the benefits of
observing and measuring changes over time, adapting the interventions accordingly and
making stronger causal interpretations.
363
9.7
Conclusions
In conclusion, norm salience is an important consideration in the applicability, planning and

success of SN marketing interventions. However, the complexities associated with examining
socially salient environments of target audiences render it an under researched area.
Consequently, most studies tend to focus on documenting misperceptions of alcohol related
norms as the basis of implementing a norm change strategy to reduce drinking rates among
college students. The integration of norm salience and SNA in this study improves our
understanding of how norms function and impact individuals behaviours. The results suggest
that evidencing misperceptions may not always be a conclusive rationale justifying a SN
campaign and that establishing the salience of normative referents to be targeted is equally
important. It is possible that the results of this study are especially relevant to commuter
colleges like DIT where salient social groups exist outside college, embedded in cohesive,
intimate and long standing sub groups of school friends. In such colleges, a typical or average
student may not be an ideal referent and policy measures may benefit by shifting the focus of
SN campaigns to pre college years and school settings. This may be a better approach to target
socially proximal groups who may be harder to access in college populations. The standard
adage is in this regards very much the case further research is needed and recommended.
In terms of the aetiology of personal drinking behaviours, the results of this study support
previous research on the importance of perceived peer norms. However, SN interventions have
mostly focussed on changing perceptions related to the norms of prevalence. The present study
supports this to be an effective strategy and suggests that perceptions of descriptive norms have
a stronger impact on personal consumption compared to those of injunctive norms.
364
References
References
AAI. (2011). "How much do we drink?" Alcohol Action Ireland Retrieved 16th November,
2011, from http://alcoholireland.ie/alcohol-facts/how-much-do-we-drink/.
Abel, G., L. Plumridge and P. Graham (2002). "Peers, networks or relationships: strategies for
understanding social dynamics as determinants of smoking behaviour." Drugs:
education, prevention and policy 9(4): 325-338.
Agostinelli, G., J. M. Brown and W. R. Miller (1995). "Effects of normative feedback on
consumption among heavy drinking college students." Journal of Drug Education
25(1): 31-40.
Ajzen, I. (1991). "The theory of planned behavior." Organizational behavior and human
decision processes 50(2): 179-211.
Ajzen, I. and M. Fishbein (1980). Understanding attitudes and predicting social behavior,
Prentice-Hall.
Alexander, C., M. Piazza, D. Mekos and T. Valente (2001). "Peers, schools, and adolescent
cigarette smoking." Journal of Adolescent Health 29(1): 22-30.
Allen, J. P. and M. Columbus (1997). Assessing alcohol problems: A guide for clinicians and
researchers, DIANE Publishing.
Aloise-Young, P. A., J. W. Graham and W. B. Hansen (1994). "Peer influence on smoking
initiation during early adolescence: A comparison of group members and group
outsiders." Journal of Applied Psychology 79(2): 281-287.
Anderson, P., A. De Bruijn, K. Angus, R. Gordon and G. Hastings (2009). "Impact of alcohol
advertising and media exposure on adolescent alcohol use: a systematic review of
longitudinal studies." Alcohol and Alcoholism 44(3): 229-243.
Andreasen, A. R. (1995). Marketing social change: Changing behavior to promote health,
social development, and the environment, Jossey-Bass San Francisco, CA.
Antonucci, T. C. (1986). "Measuring social support networks: Hierarchical mapping
technique." Generations 3: 10-12.
Armitage, C. J. and M. Conner (2001). "Efficacy of the theory of planned behaviour: A meta
analytic review." British Journal of Social Psychology 40(4): 471-499.
Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgment
In H. Guetzkow (Ed), Oxford: Carnegie Press.
Baer, J. S. (1994). "Effects of college residence on perceived norms for alcohol consumption:
An examination of the first year in college." Psychology of Addictive Behaviors 8(1):
43-50.
365
References
Baer, J. S. and M. M. Carney (1993). "Biases in the perceptions of the consequences of alcohol
use among college students." Journal of Studies on Alcohol 54(1)(1): 54-60.
Baer, J. S., A. Stacy and M. Larimer (1991). "Biases in the perception of drinking norms
among college students." Journal of Studies on Alcohol 52(6)(6): 580-586.
Baker, T. L. and A. J. Risley (1994). Doing social research, McGraw-Hill New York, NY.
Balvig, F. and L. Holmberg (2011). "The Ripple Effect: A randomized trial of a social norms
intervention in a Danish middle school setting." Journal of Scandinavian Studies in
Criminology and Crime Prevention 12(1): 3-19.
Balvig, J. (2009). Reducing Risk Behavior among Young People in Denmark: The Ringsted
Project. Social Norms Forum, Brussels, Belgium.
Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ, Prentice
Hall.
Barnes, J. A. (1979). "Network analysis: Orienting notion, rigorous technique, or substantive
field of study." Perspectives on Social Network Research: 403-423.
Barnett, L. A., J. M. Far, A. L. Mauss and J. A. Miller (1996). "Changing perceptions of peer
norms as a drinking reduction program for college students." Journal of Alcohol and
Drug Education.
Bauerle, J. (2003). The University of Virginias Social Norms Marketing Campaign: The
Report on Social Norms: Working Paper #11. Little Falls, NJ: PaperClip
Communications.
Bauman, K. E. and S. T. Ennett (1994). "Peer influence on adolescent drug use." American
Psychologist 49(9): 820-822.
Bavelas, A. (1950). "Communication patterns in task-oriented groups." Journal of the
Acoustical Society of America 22: 725-730.
Beatty, P. C. and G. B. Willis (2007). "Research synthesis: The practice of cognitive
interviewing." Public Opinion Quarterly 71(2): 287-311.
Beck, K. H., A. M. Arria, K. M. Caldeira, K. B. Vincent, K. E. O' Grady and E. D. Wish
(2008). "Social context of drinking and alcohol problems among college students."
American journal of health behavior 32(4): 420.
Beck, K. H. and K. A. Treiman (1996). "The relationship of social context of drinking,
perceived social norms, and parental influence to various drinking patterns of
adolescents." Addictive Behaviors 21(5): 633-644.
Bellis, M., K. Hughes, M. Morleo, K. Tocque, S. Hughes, T. Allen, D. Harrison and E. FeRodriguez (2007). "Predictors of risky alcohol consumption in schoolchildren and their
implications for preventing alcohol-related harm." Substance Abuse Treatment,
Prevention, and Policy 2(1): 15.
366
References
Bellotti, E. (2008). "What are friends for? Elective communities of single people." Social
Networks 30(4): 318-329.
Bendor, J. and P. Swistak (2001). "The Evolution of Norms." American Journal of Sociology
106(6): 1493-1545.
Berkowitz, A. (2004). "The social norms approach: Theory, research, and annotated
bibliography August 2004." Newton, MA: Higher Education Center for Alcohol and
Other Drug Prevention.
Berkowitz, A. (2005). An overview of the social norms approach. Changing the Culture of
College Drinking. L. C. Lederman and L. Stewart, Cresskill, NJ: Hampton Press: 193214.
Berkowitz, A. D. (1997). "From reactive to proactive prevention: Promoting an ecology of
health on campus." Substance abuse on campus: A handbook for college and university
personnel: 119139.
Berkowitz, A. D. and H. Perkins (1986a). "Resident Advisers as Role Models: A Comparison
of Drinking Patterns of Resident Advisers and Their Peers." Journal of College Student
Personnel 27(2): 146-53.
Berkowitz, A. D. and H. W. Perkins (1986b). "Problem drinking among college students: A
review of recent research." Journal of American College Health 35(1): 21-28.
Bertholet, N., J. Gaume, M. Faouzi, J. B. Daeppen and G. Gmel (2011). "Perception of the
amount of drinking by others in a sample of 20-year-old men: The more I think you
drink, the more I drink." Alcohol and Alcoholism 46(1): 83-87.
Bewick, B. M., K. Trusler, B. Mulhern, M. Barkham and A. J. Hill (2008). "The feasibility and
effectiveness of a web-based personalised feedback and social norms alcohol
intervention in UK university students: a randomised control trial." Addictive
Behaviors 33(9): 1192-1198.
Bidart, C. and J. Charbonneau (2011). "How to Generate Personal Networks: Issues and Tools
for a Sociological Perspective." Field Methods 23(3): 266.
Bigsby, M. J. (2002). "Seeing Eye to Eye? Comparing Students' and Parents' Perceptions of
Bullying Behaviors." School Social Work Journal 27(1): 37-57.
Blaikie, N. (2000). Designing social research, Polity Press.
Blau, P. M. (1994). Structural contexts of opportunities, University of Chicago Press.
Bloomfield, K., T. Stockwell, G. Gmel and N. Rehn (2003). "International comparisons of
alcohol consumption." Alcohol Research and Health 27(1): 95-109.
Boorman, S. A. and H. C. White (1976). "Social structure from multiple networks. II. Role
structures." American Journal of Sociology 81(6): 1384-1446.
Borgatti, S. P. (2002). Netdraw Network Visualization Analytic Technologies: Harvard, MA.
367
References
Borgatti, S. P. and M. G. Everett (2000). "Models of core/periphery structures." Social

networks 21(4): 375-395.
Borgatti, S. P., M. G. Everett and L. C. Freeman (2002). Ucinet for Windows: Software for
social network analysis. Harvard, MA: Analytic Technologies. 2006.
Borgatti, S. P., A. Mehra, D. J. Brass and G. Labianca (2009). "Network analysis in the social
sciences." Science 323(5916): 892.
Borgatti, S. P. and J. L. Molina (2003a). "Ethical and strategic issues in organizational social
network analysis." The Journal of Applied Behavioral Science 39(3): 337.
Borgatti, S. P. and J. L. Molina (2003b). "Ethical and strategic issues in organizational social
network analysis." The Journal of Applied Behavioral Science 39(3): 337-349.
Borsari, B. and K. B. Carey (2001). "Peer influences on college drinking: A review of the
research." Journal of substance abuse 13(4)(4): 391-424.
Borsari, B. and K. B. Carey (2003). "Descriptive and injunctive norms in college drinking: A
meta-analytic integration." Journal of Studies on Alcohol 64(3)(3): 331-341.
Borsari, B. and K. B. Carey (2005). "Two brief alcohol interventions for mandated college
students." Psychology of Addictive behaviors 19(3): 296.
Botvin, G. J., E. Baker, E. M. Botvin, L. Dusenbury, J. Cardwell and T. Diaz (1993). "Factors
promoting cigarette smoking among black youth: A causal modeling approach."
Addictive Behaviors 18(4): 397-405.
Botvin, G. J., K. W. Griffin, T. Diaz and M. Ifill-Williams (2001). "Preventing binge drinking
during early adolescence: One-and two-year follow-up of a school-based preventive
intervention." Psychology of Addictive Behaviors 15(4): 360-365.
Brener, N. D., J. O. G. Billy and W. R. Grady (2003). "Assessment of factors affecting the
validity of self-reported health-risk behavior among adolescents: evidence from the
scientific literature." Journal of Adolescent Health 33(6): 436-457.
Brieger, R. L. (1976). "Career attributes and network structure: A blockmodel study of a
biomedical research specialty." American Sociological Review: 117-135.
Bronfenbrenner, U. (1977). "Toward an experimental ecology of human development."
American psychologist 32(7): 513-531.
Brown, B. B., M. M. Dolcini and A. Leventhal (1997). "Transformations in peer relationships
at adolescence: Implications for health-related behavior."
Bruce, S. (2002). "The A Man campaign: Marketing social norms to men to prevent sexual
assault." The report on social norms: Working paper 5.
Buchanan, E. A. and E. E. Hvizdak (2009). "Online survey tools: Ethical and methodological
concerns of human research ethics committees." Journal of Empirical Research on
Human Research Ethics 4(2): 37-48.
368
References
Buchanan, M. (2002). Nexus: Small worlds and the groundbreaking science of networks. New
York, WW Norton & Company.
Buller, D. B., C. Morrill, D. Taren, M. Aickin, L. Sennott-Miller, M. K. Buller, L. Larkey, C.
Alatorre and T. M. Wentzel (1999). "Randomized trial testing the effect of peer
education at increasing fruit and vegetable intake." Journal of the National Cancer
Institute 91(17): 1491-1500.
Burt, R. S. (1984). "Network items and the general social survey." Social Networks 6(4): 293339.
Burt, R. S. (1987). "A note on missing network data in the general social survey." Social
Networks 9(1): 63-73.
Burt, R. S. (1995). Structural holes: The social structure of competition, Harvard Univ Pr.
Burt, R. S. (2004). "Structural Holes and Good Ideas." American Journal of Sociology 110(2):
349-99.
Burt, R. S. (2005). Brokerage and closure, Oxford University Press.
Byrne, S. (2010). Costs to Society of Problem Alcohol Use in Ireland. Health Service
Executive. Dublin.
Campanelli, P. (1997). "Testing survey questions: New directions in cognitive interviewing."
Bulletin de Methodologie Sociologique 55(1): 5-17.
Campanelli, P. C., E. A. Martin and J. M. Rothgeb (1991). "The use of respondent and
interviewer debriefing studies as a way to study response error in survey data." The
Statistician: 253-264.
Carew, A. M., D. Bellerose and S. Lyons (2011). Treated problem alcohol use in Ireland 20052010. Health Research Board Trends Series. Dublin.
Carney, M. A., H. Tennen, G. Affleck, F. K. Del Boca and H. R. Kranzler (1998). "Levels and
patterns of alcohol consumption using timeline follow-back, daily diaries and realtime" electronic interviews"." Journal of studies on alcohol 59: 447-454.
Carpentier, N. and F. Ducharme (2007). "Social network data validity: The example of the
social network of caregivers of older persons with Alzheimer-type dementia." Canadian
Journal on Aging 26(1): 103-116.
Carrasco, J. A., B. Hogan, B. Weilman and E. J. Miller (2008). "Collecting social network data
to study social activity-travel behavior: an egocentric approach." Environment and
Planning B: Planning and Design 35(6): 961-980.
Carter, C. A. and W. M. Kahnweiler (2000). "The efficacy of the social norms approach to
substance abuse prevention applied to fraternity men." Journal of American College
Health 49(2): 66-71.
369
References
Chawla, N., C. Neighbors, M. A. Lewis, C. M. Lee and M. E. Larimer (2007). "Attitudes and
perceived approval of drinking as mediators of the relationship between the importance
of religion and alcohol use." Journal of Studies on Alcohol and Drugs.
Cherryholmes, C. H. (1992). "Notes on pragmatism and scientific realism." Educational
researcher 21(6): 13-17.
Chipperfield, B. and M. Vogel-Sprott (1988). "Family history of problem drinking among
young male social drinkers: Modeling effects on alcohol consumption." Journal of
Abnormal Psychology 97(4): 423-428.
Christakis, N. A. and J. H. Fowler (2009). Connected: The surprising power of our social
networks and how they shape our lives, Little, Brown and Company.
Chua, V., J. Madej and B. Wellman (2011). Personal communities: the world according to me.
Cialdini, R. B. (1993). Influence: Science and practice. New York, HarperCollins.
Cialdini, R. B., L. J. Demaine, B. J. Sagarin, D. W. Barrett, K. Rhoads and P. L. Winter
(2006). "Managing social norms for persuasive impact." Social Influence 1(1): 3-15.
Cialdini, R. B., C. A. Kallgren and R. R. Reno (1991). A focus theory of normative conduct: A
theoretical refinement and reevaluation of the role of norms in human behavior.
Advances in experimental social psychology. M. P. Zanna. London, Academic Press
Inc. 24: 202-232.
Cialdini, R. B., R. R. Reno and C. A. Kallgren (1990). "A focus theory of normative conduct:
Recycling the concept of norms to reduce littering in public places." Journal of
Personality and Social Psychology 58(6): 1015-1026.
Cialdini, R. B. and M. R. Trost (1998). Social Influence, Social Norms, Conformity and
Compliance. The handbook of social psychology. D. T. Gilbert, S. T. Fiske and G.
Lindzey. New York, McGraw-Hill, Oxford University Press. 2.
Clapp, J. D., J. E. Lange, C. Russell, A. Shillington and R. B. Voas (2003). "A failed norms
social marketing campaign." Journal of Studies on Alcohol 64(3): 409-414.
Clapp, J. D. and A. L. McDonnell (2000). "The relationship of perceptions of alcohol
promotion and peer drinking norms to alcohol problems reported by college students."
Journal of College Student Development 41(1)(1): 19-26.
Cohen, J., P. Cohen, S. G. West and L. S. Aiken (1983). "Applied multiple
regression/correlation analysis for the behavioral sciences. Hillsdale." NJ Eribaum.
Cohen, L., L. Manion, K. Morrison and K. R. B. Morrison (2007). Research methods in
education, Psychology Press.
Cohen, S., W. J. Doyle, D. P. Skoner, B. S. Rabin and J. M. Gwaltney (1997). "Social ties and
susceptibility to the common cold." Journal of the American Medical Association
277(24): 1940-1944.
370
References
Coleman, J. S. (1988). "Social capital in the creation of human capital." American journal of
sociology: 95-120.
Coleman, J. S. (1990). Foundations of social theory. Cambridge MA, Harvard University
Press.
Connolly, G. M., S. Casswell, J. Stewart and P. A. Silva (1992). "Drinking context and other
influences on the drinking of 15 year old New Zealanders." British Journal of
Addiction 87(7): 1029-1036.
Cooper, A. M., G. J. Waterhouse and M. B. Sobell (1979). "Influence of gender on drinking in
a modeling situation." Journal of studies on alcohol 40(7): 562-570.
Corbetta, P. (2003). Social research: Theory, methods and techniques, Sage Publications Ltd.
Corden, A., R. Sainsbury, P. Sloper and B. Ward (2005). "Using a model of group
psychotherapy to support social research on sensitive topics." International Journal of
Social Research Methodology 8(2): 151-160.
Coromina, L. and G. Coenders (2006). "Reliability and validity of egocentered network data
collected via web: A meta-analysis of multilevel multitrait multimethod studies." Social
networks 28(3): 209-231.
Couper, M. P. (2000). "Review: Web surveys: A review of issues and approaches." The Public
Opinion Quarterly 64(4): 464-494.
Creswell, J. W. and V. L. P. Clark (2007). Designing and conducting mixed methods research,
Wiley Online Library.
Crossley, N. (2010). "The social world of the network: Combining qualitative and quantitative
elements in social network analysis " Sociologica 1.
Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research
process, SAGE Publications Limited.
Curry, L. A., I. M. Nembhard and E. H. Bradley (2009). "Qualitative and mixed methods
provide unique contributions to outcomes research." Circulation 119(10): 1442-1452.
Curtis, R., S. R. Friedman, A. Neaigus, B. Jose, M. Goldstein and G. Ildefonso (1995). "Streetlevel drug markets: network structure and HIV risk." Social Networks 17(3-4): 229249.
Curtis, W. R. (1979). Social networking mapping. The future use of social networks in mental
health. D. Todd. Boston.
Cutler, R. E. and T. Storm (1975). "Observational study of alcohol consumption in natural
settings: The Vancouver beer parlor." Journal of studies on alcohol 36(9): 1173-1183.
D Amico, E. J., J. Metrik, D. M. McCarthy, K. C. Frissell, M. Appelbaum and S. A. Brown
(2001). "Progression into and out of binge drinking among high school students."
Psychology of Addictive Behaviors 15(4): 341-349.
371
References
Darmody, M., E. Smyth, P. O Connell, J. Williams and B. Ryan (2005). "Eurostudent survey
II: Irish report on the social and living conditions of higher education students
2003/2004." HEA: Dublin.
David, M. and C. D. Sutton (2004). Social research: The basics, Sage Publications Ltd.
Davis, M., G. Bolding, G. Hart, L. Sherr and J. Elford (2004). "Reflecting on the experience of
interviewing online: perspectives from the Internet and HIV study in London." Aids
Care 16(8): 944-952.
DeJong, W. (2000). "Note to the field: the case of the missing misperception." Higher
Education for Alcohol or Drug Prevention News Digest 155.
Delaney, L., C. Harmon, C. Milner, L. Sweeney and P. G. Wall (2007). Perception of
excessive drinking among Irish college students: a mixed methods analysis. Discussion
Paper Series, UCD Geary Institute.
Delaney, L., A. Kapteyn and J. P. Smith (2008). Why do some Irish drink so much? Discussion
Paper Series, UCD Geary Institute.
DeMaio, T., N. Mathiowetz, J. Rothgeb, M. E. Beach and S. Durant (1993). "Protocol for
pretesting demographic surveys at the Census Bureau." Report prepared for the Bureau
of the Census, Center for Survey Methods Research.
DeMaio, T. J. and J. M. Rothgeb (1996). "Cognitive interviewing techniques: In the lab and in
the field."
Demers, A., S. Kairouz, E. Adlaf, L. Gliksman, B. Newton-Taylor and A. Marchand (2002).
"Multilevel analysis of situational drinking among Canadian undergraduates." Social
Science & Medicine 55(3): 415-424.
Dericco, D. A. (1978). "Effects of peer majority on drinking rate." Addictive Behaviors 3(1):
29-34.
Dericco, D. A. and W. K. Garlington (1977). "The effect of modeling and disclosure of
experimenter's intent on drinking rate of college students." Addictive Behaviors 2(2-3):
135-139.
DeRicco, D. A. and J. E. Niemann (1980). "In vivo effects of peer modeling on drinking rate."
Journal of Applied Behavior Analysis 13(1)(1): 149-152.
Dewey, J. (1948). Reconstruction in philosophy, Courier Dover Publications.
Dimeff, L. A. (1999). Brief alcohol screening and intervention for college students (BASICS):
A harm reduction approach, The Guilford Press.
Dippo, C. S. (1989). The use of cognitive laboratory techniques for investigating memory
retrieval errors in retrospective surveys.
DoH (2012). Steering Group Report on a National Substance Misuse Strategy.
372
References
DoHC (2004). Strategic Task Froce on Alcohol, 2nd Report. Department of Health and
Childern. Health Promotion Unit, Dublin.
Donato, F., S. Monarca, R. Chiesa, D. Feretti and G. Nardi (1994). "Smoking among high
school students in 10 Italian towns: patterns and covariates." Substance Use & Misuse
29(12): 1537-1557.
Donohew, L., R. R. Clayton, W. F. Skinner and S. Colon (1999). "Peer networks and sensation
seeking: Some implications for primary socialization theory." Substance use & misuse
34(7): 1013-1023.
Doreian, P. (2001). "Causality in social network analysis." Sociological Methods & Research
30(1): 81-114.
Dorsey, A. M., C. W. Scherer and K. Real (1999a). "The college tradition of" drink'til you
drop": The relation between students' social networks and engaging in risky behaviors."
Health Communication 11(4): 313-334.
Dorsey, A. M., C. W. Scherer and K. Real (1999b). "The College Tradition of" Drink Til You
Drop": The Relation Between Students Social Networks and Engaging in Risky
Behaviors." Health Communication 11(4): 313-334.
Dubuque, E., C. Ciano-Boyce and L. Shelley-Sireci (2002). "Measuring Misperceptions of
Homophobia on Campus." The Report on Social Norms: Working Paper 4.
Duff, T., J. Hegarty and M. Hussey (2000). The Story of the Dublin Institute of Technology,
DIT Press.
Durkheim, E. (1951). Suicide: A study in sociology. New York, Free Press.
Elster, J. (1989). The cement of society: A study of social order. New York, Cambridge
University Press.
Emirbayer, M. (1997). "Manifesto for a relational sociology." American Journal of Sociology
103(2): 281-317.
Emirbayer, M. and J. Goodwin (1994). "Network analysis, culture, and the problem of
agency." American journal of sociology: 1411-1454.
Engels, R., R. A. Knibbe, M. J. Drop and Y. T. de Haan (1997). "Homogeneity of cigarette
smoking within peer groups: Influence of selection?" Health Education & Behavior
24(6): 801-811.
Ennett, S. T. and K. E. Bauman (1993). "Peer group structure and adolescent cigarette
smoking: a social network analysis." Journal of Health and Social Behavior 34(3): 226236.
Ennett, S. T. and K. E. Bauman (1994). "The contribution of influence and selection to
adolescent peer group homogeneity: The case of adolescent cigarette smoking." Journal
of Personality and Social Psychology 67(4): 653-663.
373
References
Ennett, S. T., K. E. Bauman, A. Hussong, R. Faris, V. A. Foshee, L. Cai and R. H. DuRant

(2006). "The peer context of adolescent substance use: Findings from social network
analysis." Journal of Research on Adolescence 16(2): 159-186.
Erickson, B. H. (1979). "Some problems of inference from chain data." Sociological
Methodology 10(1): 276-302.
Erickson, B. H. (1996). "Culture, class, and connections." American Journal of Sociology
102(1): 217-251.
Erickson, B. H., T. A. Nosanchuk, L. Mostacci and C. F. Dalrymple (1978). "The flow of crisis
information as a probe of work relations." Canadian Journal of Sociology/Cahiers
canadiens de sociologie: 71-87.
Esposito, J. and J. Hess (1992). The use of interviewer debriefings to identify problematic
questions on alternative questionnaires.
EUDAP. (2010). "The European Drug Addiction Prevention."
from http://www.eudap.net/.
Retrieved 1st August, 2010,
Eysenbach, G. and J. Wyatt (2002). "Using the Internet for surveys and health research."
Journal of Medical Internet Research 4(2).
Fabiano, P. M. (2003). "Applying the social norms model to universal and indicated alcohol
interventions at Western Washington University." The social norms approach to
preventing school and college age substance abuse: A handbook for educators,
counselors, and clinicians: 83-99.
Fang, X., X. Li, B. Stanton and Q. Dong (2003). "Social network positions and smoking
experimentation among Chinese adolescents." American Journal of Health Behavior
27(3): 257-267.
Far, J. and J. Miller (2003). "The small group norms challenging model: Social norms
interventions with targeted high risk groups." The social norms approach to preventing
school and college age substance abuse: A handbook for educators, counselors, and
clinicians: 111132.
Farrow, S. M. (2009). A closer look at the impact of peer influences on college drinking|
Testing moderation of normative drinking perceptions by social network
characteristics, State University of New York at Buffalo.
Fearnow-Kenny, M. D., D. L. Wyrick, W. B. Hansen, D. Dyreg and D. B. Beau (2001).
"Normative Beliefs, Expectancies, and Alcohol-Related Problems among College
Students; Implications for Theory and Practice." Journal of Alcohol and Drug
Education 47(1): 31-44.
Feld, S. L. (1981). "The focused organization of social ties." American journal of sociology
86(5): 1015-1035.
374
References
Ferligoj, A. and V. Hlebec (1999). "Evaluation of social network measurement instruments."

Social networks 21(2): 111-130.
Festinger, L. (1954). "A theory of social comparison processes." Human relations 7(2): 117140.
Finlay, K. A., D. Trafimow and E. Moroi (1999). "The importance of subjective norms on
intentions to perform health behaviors." Journal of Applied Social Psychology 29(11):
2381-2393.
Fischer, C. S. (1982). To dwell among friends: Personal networks in town and city, University
of Chicago Press.
Fisher, L. A. and K. E. Bauman (1988). "Influence and Selection In the Friend-adolescent
Relationship: Findings from Studies of Adolescent Smoking and Drinking1." Journal of
Applied Social Psychology 18(4): 289-314.
Flay, B. R., F. B. Hu, O. Siddiqui, L. E. Day, D. Hedeker, J. Petraitis, J. Richardson and S.
Sussman (1994). "Differential influence of parental smoking and friends' smoking on
adolescent initiation and escalation and smoking." Journal of Health and Social
Behavior: 248-265.
Foley, A. (2012). Estimates of Average Adult Alcohol Consumption 2001-2011 and
International Comparison. Dublin, Commissioned for The Alcohol Beverage
Federation of Ireland.
Fontana, A. and J. H. Frey (2005). "The interview: From neutral stance to political
involvement." The Sage handbook of qualitative research 3: 695-727.
Foss, R., S. Deikkman, A. Goodman and C. Bartley (2004). Social norms program reduces
measured and self reported drinking at UNC-CH. The report on social norms: Working
Paper # 14, Little Falls, NJ, PaperClip Communications.
Foss, R., S. Diekman, A. Goodwin and C. Bartley (2003). Enhancing a norms program to
reduce high-risk drinking among first year students, Chapel Hill: University of North
Carolina, US Department of Education.
Fowler, J. H., M. T. Heaney, D. W. Nickerson, J. F. Padgett and B. Sinclair (2011). "Causality
in political networks." American Politics Research 39(2): 437-480.
Franca, L. R., B. Dautzenberg and M. Reynaud (2010). "Heavy episodic drinking and alcohol
consumption in French colleges: the role of perceived social norms." Alcoholism:
Clinical and Experimental Research 34(1): 164-174.
Francessca, C. M. (2009). What are norms? A study of beliefs and action in a Maya
community. Cambridge, Cambridge University Press.
Freeman, L. C., D. R. White and A. K. Romney (1992). Research methods in social network
analysis, Transaction Pub.
375
References
Friedman, S. R., A. Neaigus, B. Jose, R. Curtis, M. Goldstein, G. Ildefonso, R. B. Rothenberg

and D. C. Des Jarlais (1997). "Sociometric risk networks and risk for HIV infection."
American Journal of Public Health 87(8): 1289-1296.
Fugas, C. S., J. L. Meli and S. A. Silva (2011). "The 'is' and the 'ought': How do perceived
social norms influence safety behaviors at work?" Journal of Occupational Health
Psychology 16(1): 67-69.
GAAG (2008). "Report of the Government Alcohol Advisory Group."
Ghani, A. C., C. A. Donnelly and G. P. Garnett (1998). "Sampling biases and missing data in
explorations of sexual partner networks for the spread of sexually transmitted diseases."
Statistics in medicine 17(18): 2079-2097.
Giles, S. M., D. Helme and M. Krcmar (2007). "Predicting disordered eating intentions among
incoming college freshman: An analysis of social norms and body esteem."
Communication Studies 58(4): 395-410.
Gill, J. S. (2002). "Reported levels of alcohol consumption and binge drinking within the UK
undergraduate student population over the last 25 years." Alcohol and Alcoholism
37(2): 109-120.
Glider, P., S. J. Midyett and K. Johannessen (2001). "Challenging the collegiate rite of
passage: A campus-wide social marketing media campaign to reduce binge drinking."
Journal of Drug Education 31(2): 207-220.
Gorman, D. M., P. J. Gruenewald, P. J. Hanlon, I. Mezic, L. A. Waller, C. Castillo-Chavez, E.
Bradley and J. Mezic (2004). "Implications of systems dynamic models and control
theory for environmental approaches to the prevention of alcohol-and other drug userelated problems." Substance use & misuse 39(10-12): 1713-1750.
Graham, J. W., G. Marks and W. B. Hansen (1991). "Social influence processes affecting
adolescent substance use." Journal of Applied Psychology 76(2): 291-8.
Granfield, R. (2002). "Can you believe it? Assessing the credibility of a social norms
campaign." Report on Social Norms 2: 1-8.
Granovetter, M. S. (1973). "The strength of weak ties." The American journal sociology 78(6):
1360-1380.
Gray, D. E. (2009). Doing research in the real world, SAGE Publications Limited.
Grosser, T. J., V. Lopez-Kidwell and G. Labianca (2010). "A social network analysis of
positive and negative gossip in organizational life." Group & Organization
Management 35(2): 177-212.
Grube, J. W., M. Morgan and S. T. McGree (1986). "Attitudes and normative beliefs as
predictors of smoking intentions and behaviours: A test of three models." British
Journal of Social Psychology 25(2): 81-93.
376
References
Grube, J. W., M. Morgan and M. Seff (1989). "Drinking beliefs and behaviors among Irish
adolescents." Substance Use & Misuse 24(2): 101-112.
Gruenewald, P. J. and T. Nephew (1994). "Drinking in California: theoretical and empirical
analyses of alcohol consumption patterns." Addiction 89(6): 707-723.
Gryczynski, J. and B. W. Ward (2011). "Social Norms and the Relationship Between Cigarette
Use and Religiosity Among Adolescents in the United States." Health Education &
Behavior 38(1): 39-48.
Haines, M. and G. Barker (2003). "The NIU experiment: A case study of the social norms
approach." The Social Norms Approach to Preventing School and College Age
Substance Abuse: A handbook for Educators, Counselors, and Clinicians. San
Francisco, Jossey-Bass.
Haines, M. and S. F. Spear (1996). "Changing the perception of the norm: A strategy to
decrease binge drinking among college students." Journal of American College Health
45(3): 134-140.
Haines, M. P. (1996). "A social norms approach to preventing binge drinking at colleges and
universities." Newton, MA: Higher Education Center for Alcohol and Other Drug
Prevention, Department of Education.
Haines, M. P., G. P. Barker and R. Rice (2003). "Using social norms to reduce alcohol and
tobacco use in two Midwestern high schools." The social norms approach to preventing
school and college age substance abuse: A handbook for educators, counselors, and
clinicians: 235244.
Haines, M. P., H. W. Perkins, R. M. Rice and G. Barker (2004). A guide to marketing social
norms for health promotion in schools and communities. IIllinois, National Social
Norms Resource Center: 2-92.
Haining, P. (2003). Great Irish Drinking Stories, Barnes & Noble Publishing.
Halgin, D. and R. DeJordy. (2008). "Introduction to Ego Network Analysis, Academy of
Management, PDW. Boston College and the Winston Center for Leadership and
Ethics."
Retrieved 21st August, 2010, from http://www.analytictech.com/enet/PDWHandout.pdf.
Hammersley, R. (1994). "A digest of memory phenomena for addiction research." Addiction
89(3): 283-293.
Hancock, L. and N. Henry (2003). "Perceptions, norms and tobacco use in college residence
hall freshmen: evaluation of a social norms marketing intervention." The Social Norms
Approach to Preventing School and College Age Substance Abuse: A Handbook for
Educators, Counselors, and Clinicians. San Francisco, CA: Jossey-Bass.
Hansen, W. B. (1997). "A social ecology theory of alcohol and drug use prevention among
college and university students." Designing alcohol and other drug prevention programs
in higher education: Bringing theory into practice: 155-176.
377
References
Hansen, W. B. and J. W. Graham (1991). "Preventing alcohol, marijuana, and cigarette use
among adolescents: Peer pressure resistance training versus establishing conservative
norms* 1." Preventive Medicine 20(3): 414-430.
Harrison, L. and A. Hughes (1997). "Introduction--the validity of self-reported drug use:
improving the accuracy of survey estimates." NIDA research monograph 167: 1.
Haug, S., S. Ulbricht, M. Hanke, C. Meyer and U. John (2011). "Overestimation of Drinking
Norms and its Association with Alcohol Consumption in Apprentices." Alcohol and
Alcoholism 46(2): 204-209.
Hawe, P., C. Webster and A. Shiell (2004). "A glossary of terms for navigating the field of
social network analysis." Journal of epidemiology and community health 58(12): 971975.
Hayslett, M. M. and B. M. Wildemuth (2005). "Pixels or pencils? The relative effectiveness of
Web-based versus paper surveys." Library & Information Science Research 26(1): 7393.
HEA. (2012). "2010/11 Higher Education Statistics."
http://www.hea.ie/en/node/1424.
Retrieved 12th Nov, 2012, from
Hechter, M. and S. Kanazawa (1993). "Group solidarity and social order in Japan." Journal of
Theoretical Politics 5: 455-455.
Hechter, M. and K. D. Opp (2001). Social norms, Russell Sage Foundation Publications.
Hibell, B., U. Guttormsson, S. Ahlstrom, O. Balakireva, T. Bjarnason, A. Kokkevi and L.
Kraus (2012). "The 2011 ESPAD report: substance use among students in 36 European
countries."
Hollingshead, A. B. (2007). Elmtown's Youth-The Impact of Social Classes on Adolescents,
Case Pr.
Homish, G. G. and K. E. Leonard (2008). "The social network and alcohol use." Journal of
studies on alcohol and drugs 69(6): 906-916.
Hope, A. (2007). Alcohol consumption in Ireland 1986-2006. Report for the Health Service
Executive-Alcohol Implementation Group, Dublin: Health Service Executive.
Hope, A. (2009). A Standard Drink in Ireland: What strength?, Health Service Executive Alcohol Implementation Group.
Hope, A. (2011). Hidden realities: Childrens' Exposure to Risks from Parental Drinking in
Ireland.
Hope, A. and N. McCrea (2009). "Get'em Young: Mapping young people's exposure to alcohol
marketing in Ireland." Dublin: National Youth Council of Ireland.
378
References
House, J. S., R. L. Kahn, J. D. McLeod and D. Williams (1985). Measures and concepts of
social support. Social Support and Health. S. Cohen and S. L. Syme. New York,
Academic: 83-108.
Hughes, C., R. Julian, M. Richman, R. Mason and G. Long (2008). "Harnessing the Power of
Perception: Reducing Alcohol-Related Harm among Rural Teenagers." Youth Studies
Australia 27(2): 26-355.
Hussong, A. M., R. E. Hicks, S. A. Levy and P. J. Curran (2001). "Specifying the relations
between affect and heavy alcohol use among young adults." Journal of Abnormal
Psychology 110(3): 449-461.
Iannotti, R. J. and P. J. Bush (1992). "Perceived vs. actual friends' use of alcohol, cigarettes,
marijuana, and cocaine: Which has the most influence?" Journal of Youth and
Adolescence 21(3): 375-389.
Jabine, T. B. (1984). Cognitive aspects of survey methodology: building a bridge between
disciplines: report of the Advanced Research Seminar on Cognitive Aspects of Survey
Methodology, National Academies.
James, W. (1907). Pragmatism. New York, Dover.
Jamison, J. and L. B. Myers (2008). "Peer-Group and Price Influence Students Drinking along
with Planned Behaviour." Alcohol & Alcoholism 43(4): 492-497.
Jeffrey, L. R., P. Negro, D. Miller and J. D. Frisone (2003). "The Rowan University social
norms project." The social norms approach to preventing school and college age
substance abuse: 100-110.
Jenkins, J. E. and S. T. Zunguze (1998). "The Relationship of Family Structure to Adolescent
Drug Use, Peer Affiliation and Perception of Peer Acceptance of Drug Use."
Adolescence 33(132): 811-813.
Jobe, J. B. and D. J. Mingay (1991). "Cognition and survey measurement: History and
overview." Applied Cognitive Psychology 5(3): 175-192.
Johannessen, K. and P. Glider (2003). "The University of Arizona's campus health social
norms media campaign." The Social Norms Approach to Preventing School and
College Age Substance Abuse: A Handbook For Educators, Counselors, and Clinicians.
San Francisco, CA: Jossey-Bass: 65-82.
Johnson, B. and L. Christensen (2007). Educational research: Quantitative, qualitative, and
mixed approaches, SAGE Publications, Incorporated.
Johnson, R. B. and A. J. Onwuegbuzie (2004). "Mixed methods research: A research paradigm
whose time has come." Educational researcher 33(7): 14-26.
Johnston, L. D. (2001). Monitoring the Future: National Survey Results on Drug Use, 19752008: Volume II: College Students and Adults Ages 19-50, DIANE Publishing.
379
References
Jones, D. H., Y. Harel and R. M. Levinson (1992). "Living arrangements, knowledge of health
risks, and stress as determinants of health-risk behavior among college students."
Journal of American College Health 41 (Sep): 43-43.
Juarez, P., S. T. Walters, M. Daugherty and C. Radi (2006). "A randomized trial of
motivational interviewing and feedback with heavy drinking college students." Journal
of Drug Education 36(3): 233-246.
Kahler, C. W., J. P. Read, M. D. Wood and T. P. Palfai (2003). "Social environmental
selection as a mediator of gender, ethnic, and personality effects on college student
drinking." Psychology of Addictive Behaviors 17(3): 226-234.
Kandel, D. B. (1985). "On processes of peer influences in adolescent drug use." Advances in
Alcohol & Substance Abuse 4(3): 139-162.
Keller, A., L. Frye, J. Bauerle and J. C. Turner (2009). "Legal ages for purchase and
consumption of alcohol and heavy drinking among college students in Canada, Europe,
and the United States." Substance Abuse 30: 248-252.
Kelly, J. A., Y. A. Amirkhanian, E. Kabakchieva, S. Vassileva, B. Vassilev, T. L. McAuliffe,
W. J. DiFranceisco, R. Antonova, E. Petrova and R. A. Khoursine (2006). "Prevention
of HIV and sexually transmitted diseases in high risk social networks of young Roma
(Gypsy) men in Bulgaria: randomised controlled trial." Bmj 333(7578): 1098.
Kiesler, S. and L. S. Sproull (1986). "Response effects in the electronic survey." Public
Opinion Quarterly 50(3): 402-413.
Killworth, P. D., E. C. Johnsen, H. R. Bernard, G. Ann Shelley and C. McCarty (1990).
"Estimating the size of personal networks." Social Networks 12(4): 289-312.
Kirke, D. M. (2004). "Chain reactions in adolescents' cigarette, alcohol and drug use: similarity
through peer influence or the patterning of ties in peer networks?" Social networks
26(1): 3-28.
Kirke, D. M. (2006). "Gender and chain reactions in teenagers social networks." Connections
27(1): 15-23.
Kirke, D. M. (2009). "Gender clustering in friendship networks: some sociological
implications." Methodological Innovations Online 4: 23-36.
Klein, H. (1992). "College students attitudes toward the use of alcoholic beverages." Journal
of Alcohol and Drug Education 37(3)(3): 35-52.
Klovdahl, A. S. (1981). "A note on images of networks." Social Networks 3(3): 197-214.
Korcuska, J. S. and D. L. Thombs (2003). "Gender Role Conflict and Sex-Specific Drinking
Norms." Journal of College Student Development 44: 204-216.
Kossinets, G. (2006). "Effects of missing data in social networks." Social networks 28(3): 247268.
380
References
Krackhardt, D. (1999). "The ties that torture: Simmelian tie analysis in organizations."
Research in the Sociology of Organizations 16(1999): 183-210.
Kraut, R., J. Olson, M. Banaji, A. Bruckman, J. Cohen and M. Couper (2004). "Psychological
research online: Report of Board of Scientific Affairs' Advisory Group on the Conduct
of Research on the Internet." American Psychologist 59(2): 105-117.
Kusch, J. M. (2002). "Test of a social norms approach to understanding disordered eating
practices in college women." Unpublished masters thesis, Washington State
University, Pullman, Washington.
Kypri, K., S. J. Gallagher and M. L. Cashell-Smith (2004a). "An Internet-based survey method
for college student drinking research." Drug and Alcohol Dependence 76(1): 45-53.
Kypri, K., J. Hallett, P. Howat, A. McManus, B. Maycock, S. Bowe and N. J. Horton (2009).
"Randomized controlled trial of proactive web-based alcohol screening and brief
intervention for university students." Archives of Internal Medicine 169(16): 15081514.
Kypri, K. and J. D. Langley (2003). "Perceived Social Norms and Their Relation to University
Student Drinking*." Journal of Studies on Alcohol 64(6): 829-835.
Kypri, K., S. Stephenson and J. Langley (2004b). "Assessment of nonresponse bias in an
internet survey of alcohol use." Alcoholism: Clinical and Experimental Research 28(4):
630-634.
LaBrie, J. W., J. F. Hummer, C. Neighbors and M. E. Larimer (2010). "Whose opinion
matters? The relationship between injunctive norms and alcohol consequences in
college students." Addictive behaviors 35(4): 343-349.
Lackaff, D. (2010). Propitious aggregation: reducing participant burden in ego-centric network
data collection, ACM.
Lackaff, D. (2012). New Opportunities in Personal Network Data Collection. HumanComputer Interaction: The Agency Perspective: 389-407.
Lancaster, G. A., S. Dodd and P. R. Williamson (2004). "Design and analysis of pilot studies:
recommendations for good practice." Journal of Evaluation in Clinical Practice 10(2):
307-312.
Lapinski, M. K. and R. N. Rimal (2005). "An explication of social norms." Communication
Theory 15(2): 127-147.
Larimer, M. E., D. L. Irvine, J. R. Kilmer and G. A. Marlatt (1997). "College drinking and the
Greek system: Examining the role of perceived norms for high-risk behavior." Journal
of College Student Development 38(6): 587-598.
Larimer, M. E. and C. Neighbors (2003). "Normative misperception and the impact of
descriptive and injunctive norms on college student gambling." Psychology of
Addictive Behaviors 17(3): 235-243.
381
References
Larimer, M. E., A. P. Turner, K. A. Mallett and I. M. Geisner (2004a). "Predicting drinking

behavior and alcohol-related problems among fraternity and sorority members:
examining the role of descriptive and injunctive norms." Psychology of Addictive
Behaviors 18(3): 203-212.
Larimer, M. E., A. P. Turner, K. A. Mallett and I. M. Geisner (2004b). "Predicting drinking
behavior and alcohol-related problems among fraternity and sorority members:
examining the role of descriptive and injunctive norms." Psychology of Addictive
Behaviors 18(3): 203.
Latane, B. (1981). "The psychology of social impact." American Psychologist 36(4): 343-356.
Latkin, C., W. Mandell, M. Oziemkowska, D. Celentano, D. Vlahov, M. Ensminger and A.
Knowlton (1995a). "Using social network analysis to study patterns of drug use among
urban drug users at high risk for HIV/AIDS." Drug and Alcohol Dependence 38(1): 19.
Latkin, C., W. Mandell, D. Vlahov, A. Knowlton, M. Oziemkowska and D. Celentano (1995b).
"Personal network characteristics as antecedents to needle-sharing and shooting gallery
attendance." Social Networks 17(3-4): 219-228.
Latkin, C. A. (1998). "Outreach in natural settings: the use of peer leaders for HIV prevention
among injecting drug users' networks." Public Health Reports 113(Suppl 1): 151-159.
Latkin, C. A., V. Forman, A. Knowlton and S. Sherman (2003). "Norms, social networks, and
HIV-related risk behaviors among urban disadvantaged drug users." Social Science &
Medicine 56(3): 465-476.
Laumann, E. O., P. V. Marsden and D. Prensky (1989). "The boundary specification problem
in network analysis." Research Methods in Social Network Analysis 61.
Leavitt, H. J. (1951). "Some effects of certain communication patterns on group performance."
Journal of abnormal and social psychology 46(1): 38-50.
Leonard, K. E., J. Kearns and P. Mudar (2000). "Peer networks among heavy, regular and
infrequent drinkers prior to marriage." Journal of studies on alcohol 61(5): 669-673.
Lewis, M. A. and C. Neighbors (2004). "Gender-specific misperceptions of college student
drinking norms." Psychol Addict Behav 18(4): 334-9.
Lewis, M. A. and C. Neighbors (2006a). "Social norms approaches using descriptive drinking
norms education: A review of the research on personalized normative feedback."
Journal of American College Health 54(4)(4): 213-218.
Lewis, M. A. and C. Neighbors (2006b). "Who is the typical college student? Implications for
personalized normative feedback interventions." Addictive Behaviors 31(11): 21202126.
Lewis, M. A. and C. Neighbors (2007). "Optimizing personalized normative feedback: The use
of gender-specific referents." Journal of Studies on Alcohol and Drugs 68(2): 228-237.
382
References
Lewis, M. A., C. Neighbors, I. M. Geisner, C. M. Lee, J. R. Kilmer and D. C. Atkins (2010).

"Examining the associations among severity of injunctive drinking norms, alcohol
consumption, and alcohol-related negative consequences: The moderating roles of
alcohol consumption and identity." Psychology of Addictive Behaviors 24(2): 177-189.
Lewis, M. A., C. Neighbors, L. Oster-Aaland, B. S. Kirkeby and M. E. Larimer (2007).
"Indicated prevention for incoming freshmen: Personalized normative feedback and
high-risk drinking." Addictive behaviors 32(11): 2495-2508.
Lied, E. R. and G. A. Marlatt (1979). "Modeling as a determinant of alcohol consumption:
Effect of subject sex and prior drinking history." Addictive Behaviors 4(1)(1): 47-54.
Lin, N., K. S. Cook and R. S. Burt (2001). Social capital: theory and research, Aldine de
Gruyter.
Lin, N. and M. Dumin (1986). "Access to occupations through social ties." Social networks
8(4): 365-385.
Lindner, J. R., T. H. Murphy and G. E. Briers (2001). "Handling nonresponse in social science
research." Journal of Agricultural Education 42(4): 43-53.
Linkenbach, J. W. and H. W. Perkins (2003a). "Misperceptions of peer alcohol norms in a
statewide survey of young adults." The Social Norms Approach to Preventing School
and College Age Substance Abuse: A Handbook for Educators, Counselors, and
Clinicians. San Francisco, CA: Jossey-Bass.
Linkenbach, J. W. and H. W. Perkins (2003b). "Most Of Us Are Tobacco-Free: An EightMonth Social Norms Campaign Reducing Youth Incitation Of Smoking In Montana."
The social norms approach to preventing school and college age substance abuse: a
handbook for educators, counselors, and clinicians. San Francisco, CA: Jossey-Bass.
Linkenbach, J. W., H. W. Perkins and W. DeJong (2003). "Parent's perceptions of parenting
norms: Using the social norms approach to reinforce effective parenting." The Social
Norms Approach to Preventing School and College Age Substance Abuse: A
Handbook for Educators, Counselors, and Clinicians. San Francisco, CA: Jossey-Bass.
Lintonen, T., S. AhlstrM and L. Metso (2004). "The reliability of self-reported drinking in
adolescence." Alcohol and Alcoholism 39(4): 362-368.
Lintonen, T. P. and A. I. Konu (2004). "The misperceived social norm of drunkenness among
early adolescents in Finland." Health education research 19(1): 64-70.
Litwin, M. S. (1995). How to measure survey reliability and validity, SAGE Publications,
Incorporated.
MacCormick, N. (1998). "Norms, institutions, and institutional facts." Law and Philosophy 17
(3): 301-345.
383
References
Maisto, S. A., K. B. Carey and C. M. Bradizza (1999). Social Learning Theory. Psychological
theories of drinking and alcoholism. a. H. T. B. K. E. Leonard. New York, Guilford
Press: 106.
Mallett, K. A., R. L. Bachrach and R. Turrisi (2009). "Examining the unique influence of
interpersonal and intrapersonal drinking perceptions on alcohol consumption among
college students." Journal of Studies on Alcohol and Drugs 70(2): 178-185.
Manfreda, K. L., V. Vehovar and V. Hlebec (2004). "Collecting ego-centred network data via
the Web." Metodoloki zvezki, 1(2): 295-321.
Marcoux, B. C. and J. T. Shope (1997). "Application of the theory of planned behavior to
adolescent use and misuse of alcohol." Health Education Research 12(3): 323-331.
Marin, A. (2004). "Are respondents more likely to list alters with certain characteristics?:
Implications for name generator data." Social Networks 26(4): 289-307.
Marin, A. and K. N. Hampton (2007). "Simplifying the Personal Network Name Generator."
Field Methods 19(2): 163-193.
Marin, A. and B. Wellman (2010). Social Network Analysis: An Introduction. Handbook of
Social Network Analysis. P. Carrington and J. Scott. London, Sage.
Marks, G., J. W. Graham and W. B. Hansen (1992). "Social projection and social conformity
in adolescent alcohol use: A longitudinal analysis." Personality and Social Psychology
Bulletin 18(1): 96-101.
Marsden, P. V. (1987). "Core discussion networks of Americans." American Sociological
Review 52(1): 122-131.
Marsden, P. V. (1990). "Network data and measurement." Annual review of sociology 16(1):
435-463.
Marsden, P. V. (2005). Recent developments in network measurement. Models and methods in
social network analysis. P. Carrington, J. Scott and S. Wesserman. New York,
Cambridge University Press. 8.
Marsden, P. V. and K. E. Campbell (1984). "Measuring tie strength." Social Forces 63: 482501.
Martinez, A., Y. Dimitriadis, B. Rubia, E. Gomez and P. De La Fuente (2003). "Combining
qualitative evaluation and social network analysis for the study of classroom social
interactions." Computers & Education 41(4): 353-368.
Mason, M., I. Cheung and L. Walker (2004). "Substance use, social networks, and the
geography of urban adolescents." Substance Use & Misuse 39(10-12): 1751-1777.
Mathy, R. M., D. L. Kerr and B. M. Haydin (2003). "Methodological rigor and ethical
considerations in Internet-mediated research." Psychotherapy: Theory, Research,
Practice, Training 40(1-2): 77-85.
384
References
Matzat, U. and C. Snijders (2010). "Does the online collection of ego-centered network data
reduce data quality? An experimental comparison." Social Networks 32(2): 105-111.
Maxwell, K. A. (2002). "Friends: The role of peer influence across adolescent risk behaviors."
Journal of Youth and Adolescence 31(4): 267-277.
Mazzoni, D. and R. Dannenberg (2011). Audacity (Version 2.0.2)[Computer software].
McAlaney, J. (2007). The relationship between normative beliefs, expectancies and heavy
episodic alcohol consumption, University of the West of Scotland. PhD.
McAlaney, J., B. Bewick and C. Hughes (2011). "The international development of the'Social
Norms' approach to drug education and prevention." Drugs: Education, Prevention and
Policy 18(2): 81-89.
McAlaney, J. and J. McMahon (2007). "Normative beliefs, misperceptions, and heavy episodic
drinking in a british student sample." Journal of Studies on Alcohol and Drugs 68(3):
385-392.
McCallister, L. and C. S. Fischer (1978). "A procedure for surveying personal networks."
Sociological Methods & Research 7(2): 131-148.
McCarty, C., P. D. Killworth and J. Rennell (2007). "Impact of methods for reducing
respondent burden on personal network structural measures." Social Networks 29(2):
300-315.
McColl, E. (2006). "Cognitive interviewing. A tool for improving questionnaire design."
Quality of Life Research 15(3): 571-573.
McDonnell, K. A., J. G. Burke, A. C. Gielen, P. O'Campo and M. Weidl (2011). "Women's
perceptions of their community's social norms towards assisting women who have
experienced intimate partner violence." Journal of Urban Health 88(2): 240-253.
McGee, H., R. Garavan, M. de Barra, J. Byrne and R. Conroy (2002). "The SAVI report:
sexual abuse and violence in Ireland."
McPherson, M., L. Smith-Lovin and M. E. Brashears (2006). "Social isolation in America:
Changes in core discussion networks over two decades." American Sociological
Review 71(3): 353-375.
Mehra, A., M. Kilduff and D. J. Brass (2001). "The social networks of high and low selfmonitors: Implications for workplace performance." Administrative Science Quarterly
46(1): 121-146.
Meijer, R. R., A. M. M. Muijtjens and C. P. M. van der Vlueten (1996). "Nonparametric
Person-Fit Research: Some Theoretical issues an Empirical Example." Applied
Measurement in Education 9(1): 77-89.
Midanik, L. T. (1988). "Validity of Self-reported Alcohol Use: a literature review and
assessment*." British Journal of Addiction 83(9): 1019-1029.
385
References
Miller, D. T. and C. McFarland (1991). "When social comparison goes awry: The case of
pluralistic ignorance." Social comparison: Contemporary theory and research: 287313.
Miller, D. T. and D. A. Prentice (1996). The construction of social norms and standards. Social
psychology: Handbook of basic principles. E. T. H. a. A. W. Kruglanski. New York,
Guilford Press: 799829.
Miller, L. E. and K. L. Smith (1983). "Handling nonresponse issues." Journal of extension
21(5): 45-50.
Minichiello, V. and N. Helms (1997). In-depth interviewing: Principles, techniques, analysis,
Royal Victorian Institute for the Blind. Special Request Service.
Mongan, D. (2007). Health-related consequences of problem alcohol use. Overview 6. Dublin,
Health Research Board.
Mongan, D., A. Hope and M. Nelson (2009). Social consequences of harmful use of alcohol in
Ireland. HRB Overview Series 9. Dublin: Health Research Board.
Moore, G. (1990). "Structural determinants of men's and women's personal networks."
American Sociological Review: 726-735.
Moreira, M. T., L. A. Smith and D. Foxcroft (2009). "Social norms interventions to reduce
alcohol misuse in university or college students." Cochrane Database Syst Rev 3.
Moreno, J. L. (1953). "Who shall survive?" Journal of Consulting Psychology 17(6): 466.
Morphew, C. C. and A. N. Williams (1998). "Using electronic mail to assess undergraduates'
experiences: Assessing frame and mode effects." Journal of Computing in Higher
Education 10(1): 38-55.
Morris, M., J. Zavisca and L. Dean (1995). "Social and sexual networks: Their role in the
spread of HIV/AIDS among young gay men." AIDS Education and Prevention 7: 2435.
Morrison, D. M., M. R. Gillmore, M. J. Hoppe, J. Gaylord, B. C. Leigh and D. Rainey (2003).
"Adolescent drinking and sex: findings from a daily diary study." Perspectives on
Sexual and Reproductive Health: 162-168.
Mundt, M. P., L. Mercken and L. I. Zakletskaia (2012). "Peer selection and influence effects
on adolescent alcohol use: a stochastic actor-based model." BMC pediatrics 12(1): 110.
Murphy, J. P. and R. Rorty (1990). Pragmatism: From Peirce to Davidson, Westview Press
Boulder, CO.
Murray, S. O., J. H. Rankin and D. W. Magill (1981). "Strong ties and job information." Work
and Occupations 8(1): 119-136.
386
References
Nadoh, J., P. Podreberek and V. Hlebec (2004). "Cognitive evaluation of the hierarchical
approach for measuring ego-centered social networks." Metodoloki zvezki 1(2): 379393.
Nagoshi, C. T. (1999). "Perceived control of drinking and other predictors of alcohol use and
problems in a college student sample." Addiction Research & Theory 7(4)(4): 291-306.
Nahapiet, J. and S. Ghoshal (1998). "Social capital, intellectual capital, and the organizational
advantage." Academy of Management Review: 242-266.
Neighbors, C., A. J. Dillard, M. A. Lewis, R. L. Bergstrom and T. A. Neil (2006). "Normative
misperceptions and temporal precedence of perceived norms and drinking." Journal of
Studies on Alcohol 67(2): 290-299.
Neighbors, C., R. M. O'Connor, M. A. Lewis, N. Chawla, C. M. Lee and N. Fossos (2008).
"The relative impact of injunctive norms on college student drinking: The role of
reference group." Psychology of Addictive Behaviors 22(4): 576-581.
Neuman, W. L. (2005). Social research methods: Quantitative and qualitative approaches,
Allyn and Bacon.
Newcomb, T. M. (1943). Personality & social change: attitude formation in a student
community. New York, Holt, Rinehart, and Winston.
Newcomb, T. M., E. K. Wilson and L. Baird (1966). College peer groups: Problems and
prospects for research. Chicago, Aldine Pub. Co.
O'Malley, A. J., S. Arbesman, D. M. Steiger, J. H. Fowler and N. A. Christakis (2012).
"Egocentric Social Network Structure, Health, and Pro-Social Behaviors in a National
Panel Study of Americans." PloS one 7(5).
Odahowski, M. and C. Miller, Eds. (2000). Results of the 2000 Health Promotion Survey on
the Class of 2003. Office of Health Promotion. Charlottesville: University of Virginia.
Oppenheim, A. N. (2000). Questionnaire design, interviewing and attitude measurement,
Continuum Intl Pub Group.
OTC (2006). Children, Youth and Tobacco: Behaviour, Perceptions and Public Attitudes. A
report by Office of Tobacco Control, Ireland.
Page, R. M., F. Ihasz, I. Hantiu, J. Simonek and R. Klarvora (2008). "Social normative
perceptions of alcohol use and episodic heavy drinking among Central and Eastern
European adolescents." Substance Use & Misuse 43(3): 361-373.
Page, R. M., A. Scanlan and L. Gilbert (1999). "Relationship of the estimation of binge
drinking among college students and personal participation in binge drinking:
Implications for health education and promotion." Journal of Health Education 30: 99104.
387
References
Paluck, E. L. and H. Shepherd (2012). "The Salience of Social Referents: A Field Experiment
on Collective Norms and Harassment Behavior in a School Social Network." Journal of
Personality and Social Psychology (17 September, 2012).
Park, H. S. and S. W. Smith (2007). "Distinctiveness and influence of subjective norms,
personal descriptive and injunctive norms, and societal descriptive and injunctive
norms on behavioral intent: A case of two behaviors critical to organ donation." Human
Communication Research 33(2): 194-218.
Pearson, M. and L. Michell (2000). "Smoke rings: social network analysis of friendship
groups, smoking and drug-taking." Drugs: education, prevention and policy 7(1): 2137.
Pearson, M. and P. West (2003). "Drifting Smoke Rings." Connections 25(2): 59-76.
Peat, J. K., C. Mellis and K. Williams (2002). Health science research: a handbook of
quantitative methods, Sage Publications Ltd.
Peeler, C. M., J. Far, J. Miller and T. A. Brigham (2000). "An analysis of the effects of a
program to reduce heavy drinking among college students." Journal of Alcohol and
Drug Education 45(2): 39-54.
Peirce, C. S. (1878). "How to make our ideas clear." Popular Science Monthly 12: 286-302.
Perkins, H. (1992). "Gender patterns in consequences of collegiate alcohol abuse: A 10-year
study of trends in an undergraduate population." Journal of Studies on Alcohol and
Drugs 53(5): 458.
Perkins, H. and D. W. Craig (2002). "A Multifaceted Social Norms Approach To Reduce
High-Risk Drinking: Lessons from Hobart and Williams Smith Colleges."
Perkins, H., M. P. Haines and R. Rice (2005). "Misperceiving the college drinking norm and
related problems: a nationwide study of exposure to prevention information, perceived
norms and student alcohol misuse." Journal of Studies on Alcohol 66 (4): 470-478.
Perkins, H. and J. Linkenbach (2003). "Harvard Study of Social Norms Deserves An F Grade
for Flawed Research Design." Alcohol Problems and Solutions.
Perkins, H. W. (1985). "Religious Traditions, Parents, and Peers as Determinants of Alcohol
and Drug Use among College Students1." Review of Religious Research 27(1): 15-31.
Perkins, H. W. (1997). College student misperceptions of alcohol and other drug norms among
peers: Exploring causes, consequences, and implications for prevention programs.
Designing alcohol and other drug prevention programs in higher education, Newton,
MA: The Higher Education Center for Alcohol and Other Drug Prevention: 177-206.
Perkins, H. W. (2002). "Social norms and the prevention of alcohol misuse in collegiate
contexts." Journal of Studies on Alcohol 14 (Suppl): 164-172.
388
References
Perkins, H. W. (2003a). The emergence and evolution of the social norms approach to
substance abuse prevention. The Social Norms Approach to Preventing School and
College Age Substance Abuse: A Handbook for Educators, Counselors, and Clinicians.
San Francisco, CA: Jossey-Bass. H. W. Perkins.
Perkins, H. W. (2003b). The promise and challenge of future work using the social norms
model. The social norms approach to preventing school and college age substance
abuse: A handbook for educators, counselors, and clinicians. H. W. Perkins, CA:
Jossey-Bass: 280296.
Perkins, H. W. (2007). "Misperceptions of peer drinking norms in Canada: Another look at the
"reign of error" and its consequences among college students." Addictive Behaviors
32(11): 2645-2656.
Perkins, H. W. and A. D. Berkowitz (1986). "Perceiving the Community Norms of Alcohol
Use among Students: Some Research Implications for Campus Alcohol Education
Programming*." Substance Use & Misuse 21(9-10): 961-976.
Perkins, H. W. and D. W. Craig (2003a). The Hobart and William Smith Colleges experiment:
A synergistic social norms approach using print, electronic media, and curriculum
infusion to reduce collegiate problem drinking. The Social Norms Approach to
Preventing School and College Age Substance Abuse: A Handbook of Educators,
Counselors and Clinicians. San Fransisco, JosseyBass.
Perkins, H. W. and D. W. Craig (2003b). "The imaginary lives of peers: Patterns of substance
use and misperceptions of norms among secondary school students." The Social Norms
Approach to Preventing School and College Age Substance Abuse: A Handbook for
Educators, Counselors, and Clinicians. San Francisco, CA: Jossey-Bass.
Perkins, H. W., P. W. Meilman, J. S. Leichliter, J. R. Cashin and C. A. Presley (1999).
"Misperceptions of the norms for the frequency of alcohol and other drug use on
college campuses." Journal of American College Health 47(6): 253-258.
Perkins, H. W. and H. Wechsler (1996). "Variation in perceived college drinking norms and its
impact on alcohol abuse: A nationwide study." Journal of Drug Issues 26: 961-974.
Perry, C. L., C. L. Williams, K. A. Komro, S. Veblen-Mortenson, M. H. Stigler, K. A.
Munson, K. Farbakhsh, R. M. Jones and J. L. Forster (2002). "Project Northland: longterm outcomes of community action to reduce adolescent alcohol use." Health
Education Research 17(1): 117-132.
Pischke, C. R., H. Zeeb, G. Van Hal, B. Vriesacker, J. McAlaney, B. M. Bewick, Y. Akvardar,
F. Guillen-Grima, O. Orosova and F. Salonna (2012). "A feasibility trial to examine the
social norms approach for the prevention and reduction of licit and illicit drug use in
European University and college students." BMC public health 12(1): 882.
Podolny, J. M. and J. N. Baron (1997). "Resources and relationships: Social networks and
mobility in the workplace." American Sociological Review 62(5): 673-693.
389
References
Poikolainen, K., I. Podkletnova and H. Alho (2002). "Accuracy of quantity-frequency and

graduated frequency questionnaires in measuring alcohol intake: comparison with daily
diary and commonly used laboratory markers." Alcohol and Alcoholism 37(6): 573576.
Polit-O'Hara, D. and B. P. Hungler (1997). Essentials of Nursing Research: Methods,
Appraisal, and Utilization, Lippincott.
Pollard, J. W., J. E. Freeman, D. A. Ziegler, M. N. Hersman and C. W. Goss (2000).
"Predictions of Normative Drug Use by College Students." Journal of College Student
Psychotherapy 14(3): 5-12.
Pope, C., S. Ziebland and N. Mays (2006). Analysing qualitative data. Qualitative research in
health care. Pope C and M. N, Oxford: Blackwell Publishing, 3rd Edition: 6381.
Portes, A. (2000). "Social capital: Its origins and applications in modern sociology." LESSER,
Eric L. Knowledge and Social Capital. Boston: Butterworth-Heinemann: 43-67.
Prentice, D. A. and D. T. Miller (1993). "Pluralistic ignorance and alcohol use on campus:
Some consequences of misperceiving the social norm." Journal of Personality and
Social Psychology 64(2): 243-256.
Presley, C. A., P. W. Meilman and J. R. Cashin (1996). Alcohol and Drugs on American
College Campuses: Use, Consequences, and Perceptions of the Campus Environment:
Volume III: 1991-93, Southern Illinois University.
Primmer, E. and H. Karppinen (2010). "Professional judgment in non-industrial private
forestry: Forester attitudes and social norms influencing biodiversity conservation."
Forest Policy and Economics 12(2): 136-146.
Proctor, R. W. and E. J. Capaldi (2008). Internal and External Validity. Why Science Matters:
Understanding the Methods of Psychological Research. Oxford, UK. , Blackwell
Publishing Ltd: 180-194.
Punch, K. F. and K. Punch (2005). Introduction to social research: Quantitative and qualitative
approaches, Sage Publications Ltd.
Purshouse, R., Y. Meng, R. Rafia and A. Breenan (2009). "Model-based appraisal of alcohol
minimum pricing and off-licensed trade discount bans in Scotland: A Scottish
adaptation of the Sheffield Alcohol Policy Model version 2." ScHARR, University of
Sheffield.
Putnam, R. D. (1993). "The prosperous community." The American Prospect 4(13): 35-42.
Quan-Haase, A. and B. Wellman (2006). Hyperconnected net work. The Firm as a
Collaborative Community: Reconstructing Trust in the Knowledge Economy. C.
Heckscher and P. Adler, Oxford University Press: 281333.
Quigley, B. M. and R. L. Collins (1999). "The modeling of alcohol consumption: A metaanalytic review." Journal of Studies on Alcohol 60(1): 90-91.
390
References
Rabinovich, L., P. B. Brutscher, H. de Vries, J. Tiessen, J. Clift, A. Reding, R. Europe, C.

Rand and C. European (2009). The affordability of alcoholic beverages in the European
Union: Understanding the link between alcohol affordability, consumption and harms,
RAND.
Rabow, J. and M. Duncan-Schill (1995). "Drinking among college students." Journal of
Alcohol & Drug Education 40(3): 52.
Rai, A. A., B. Stanton, Y. Wu, X. Li, J. Galbraith, L. Cottrell, R. Pack, C. Harris, D.
D'Alessandri and J. Burns (2003). "Relative influences of perceived parental
monitoring and perceived peer involvement on adolescent risk behaviors: An analysis
of six cross-sectional data sets." Journal of Adolescent Health 33(2): 108-118.
Ramo, D. E., S. M. Hall and J. J. Prochaska (2011). "Reliability and validity of self-reported
smoking in an anonymous online survey with young adults." Health Psychology 30(6):
693-701.
Read, J. P., M. D. Wood and C. Capone (2005). "A prospective investigation of relations
between social influences and alcohol involvement during the transition into college."
Journal of Studies on Alcohol 66(1): 23-34.
Real, K. and R. N. Rimal (2007). "Friends Talk to Friends About Drinking: Exploring the Role
of Peer Communication in the Theory of Normative Social Behavior." Health
Communication 22(2): 169-180.
Rehm, J. r., T. K. Greenfield, G. Walsh, X. Xie, L. Robson and E. Single (1999). "Assessment
methods for alcohol consumption, prevalence of high risk drinking and harm: a
sensitivity analysis." International Journal of Epidemiology 28(2): 219-224.
Reifman, A., W. K. Watson and A. McCourt (2006). "Social networks and college drinking:
Probing processes of social influence and selection." Personality and Social Psychology
Bulletin 32(6): 820-832.
Reno, R. R., R. B. Cialdini and C. A. Kallgren (1993). "The transsituational influence of social
norms." Journal of Personality and Social Psychology 64(1): 104.
Rhodes, T., G. V. Stimson and A. Quirk (1996). "Sex, drugs, intervention, and research: from
the individual to the social." Substance Use and Misuse 31(3): 375-407.
Rice, R. E. (1993). Using network concepts to clarify sources and mechanisms of social
influence. Progress in communication sciences Vol 12: Advances in communication
network analysis
W. D. Richards and G. A. Barnett. Norwood. NJ: Ablex. 12: 43-62.
Rice, R. E., L. Donohew and R. Clayton (2003). "Peer network, sensation seeking, and drug
use among junior and senior high school students." Connections 25(2): 32-58.
391
References
Rimal, R. N. (2008). "Modeling the Relationship Between Descriptive Norms and Behaviors:
A Test and Extension of the Theory of Normative Social Behavior (TNSB)." Health
Communication 23(2): 103-116.
Rimal, R. N. and K. Real (2003). "Understanding the influence of perceived norms on
behaviors." Communication Theory 13(2): 184-203.
Rimal, R. N. and K. Real (2005). "How behaviors are influenced by perceived norms: A test of
the theory of normative social behavior." Communication Research 32(3): 389-414.
Ringsted. (2010). "Ringsted Project."
http://www.ringstedprojektet.dk/.
Retrieved
1st
August,
2010,
from
Rivis, A. and P. Sheeran (2003). "Social influences and the theory of planned behaviour:
Evidence for a direct relationship between prototypes and young people's exercise
behaviour." Psychology and Health 18(5): 567-583.
Robinson, J. P., A. Neustadtl and M. Kestenbaum (2002). Why public opinion polls are
inherently biased: Public opinion differences among Internet users and non-users.
annual meeting of the American Association for Public Opinion Research, St.
Petersburg Beach, FL.
Roethlisberger, F. J. and W. J. Dickson (2003). Management and the Worker, Routledge.
Rogers, E. M. (1995). Diffusion of Innovations, Free Pr.
Romelsjo, A., H. Leifman and S. Nystrom (1995). "A comparative study of two methods for
the measurement of alcohol consumption in the general population." International
journal of epidemiology 24(5): 929-936.
Room, R. (2000). "Measuring drinking patterns: the experience of the last half century."
Journal of substance abuse 12(1): 23-31.
Rosenquist, J. N., J. Murabito, J. H. Fowler and N. A. Christakis (2010). "The spread of
alcohol consumption behavior in a large social network." Annals of Internal Medicine
152(7): 426-W141.
Ruan, D. (1998). "The content of the General Social Survey discussion networks: an
exploration of General Social Survey discussion name generator in a Chinese context."
Social Networks 20(3): 247-264.
Ryu, E., M. P. Couper and R. W. Marans (2006). "Survey incentives: Cash vs. in-kind; face-toface vs. mail; response rate vs. nonresponse error." International Journal of Public
Opinion Research 18(1): 89-106.
Salkind, N. J. (2008). Encyclopedia of educational psychology, Sage Publications, Inc.
Sampson, R. J., S. W. Raudenbush and F. Earls (1997). "Neighborhoods and violent crime: A
multilevel study of collective efficacy." Science 277(5328): 918-924.
392
References
Schall, M., A. Kemeny and I. Maltzman (1992). "Factors associated with alcohol use in
university students." Journal of Studies on Alcohol 53(2): 122-136.
Schmidt, E., S. Mide Kiss and W. Lokanc-Diluzio (2009). "Changing social norms: a mass
media campaign for youth ages 12-18." Canadian Journal of Public Health 100(1): 4145.
Schroeder, C. M. and D. A. Prentice (1998). "Exposing Pluralistic Ignorance to Reduce
Alcohol Use Among College Students1." Journal of Applied Social Psychology 28(23):
2150-2180.
Sheeran, P. and S. Orbell (1999). "Augmenting the Theory of Planned Behavior: Roles for
Anticipated Regret and Descriptive Norms1." Journal of Applied Social Psychology
29(10): 2107-2142.
Sher, K. J., B. D. Bartholow and S. Nanda (2001). "Short-and long-term effects of fraternity
and sorority membership on heavy drinking: A social norms perspective." Psychology
of Addictive Behaviors 15(1): 42-51.
Sherif, M. (1972). Experiments on norm formation. Classic contributions to social psychology.
E. P. H. R. G. Hunt. New York, Oxford University Press: 320329.
Shore, E. R., P. C. Rivers and J. J. Berman (1983). "Resistance by college students to peer
pressure to drink." Journal of Studies on Alcohol 44(2): 352-361.
Sieving, R. E., C. L. Perry and C. L. Williams (2000). "Do friendships change behaviors, or do
behaviors change friendships? Examining paths of influence in young adolescents'
alcohol use." Journal of Adolescent Health 26(1): 27-35.
Sikkema, K. J., J. A. Kelly, R. A. Winett, L. J. Solomon, V. A. Cargill, R. A. Roffman, T. L.
McAuliffe, T. G. Heckman, E. A. Anderson and D. A. Wagstaff (2000). "Outcomes of
a randomized community-level HIV prevention intervention for women living in 18
low-income housing developments." American Journal of Public Health 90(1): 57-63.
Simmel, G. (1964). The web of group affiliations. Conflict and the web of group affiliations.
K. Wolff. Glencoe, IL, Free Press: 125-195.
Simmel, G. and D. N. Levine (1971). On individuality and social forms: Selected writings,
University of Chicago Press Chicago, IL.
Simmel, G. and K. H. Wolff (1950). The Sociology of Georg Simmel, Free Pr.
Single, E. and S. Wortley (1994). "A comparison of alternative measures of alcohol
consumption in the Canadian National Survey of alcohol and drug use." Addiction
89(4): 395-399.
Sirken, M. G., D. J. Herrmann, R. Tourangeau, J. M. Tanur, N. Schwarz and S. Schechter
(1999). Cognition and Survey Research, Wiley New York.
393
References
Sobell, L. C. and M. B. Sobell (1992). Timeline follow-back: A technique for assessing selfreported alcohol consumption. Measuring Alcohol Consumption: Psychosocial and
Biological Methods. R. Z. Litten and J. P. Allen. Totowa, NJ, Humana Press: 41-72.
Sobell, L. C. and M. B. Sobell (1995). "Alcohol consumption measures." Assessing alcohol
problems: A guide for clinicians and researchers 4: 55-76.
Stake, R. E. (1978). "The case study method in social inquiry." Educational researcher 7(2): 58.
Steffian, G. (1999). "Correction of normative misperceptions: An alcohol abuse prevention
program." Journal of Drug Education 29(2): 115-138.
Stockwell, T. and T. Chikritzhs (2002). "International guide for monitoring alcohol
consumption and related harm." Geneva: World Health Organization.
Streeter, C. L. and D. F. Gillespie (1993). "Social network analysis." Journal of Social Service
Research 16(1-2): 201-222.
Suls, J. and C. K. Wan (1987). "In search of the false-uniqueness phenomenon: Fear and
estimates of social consensus." Journal of Personality and Social Psychology 52(1):
211-217.
Summer, W. G. (1906). Folkways: A study of the sociological importance of usage, manners,
customs, mores, and morals. Boston, Ginn and Compnay.
Sutherland, E. H. and D. Cressey (1974). Criminology (9th edn), Philadelphia: Lippincott.
Tajfel, H. and J. C. Turner (1986). The social identity theory of inter - group behaviour.
Psychology of Intergroup Relations. S. W. a. W. G. Austin. Chicago, Nelson - Hall.
Tashakkori, A. and C. Teddlie (2003). Handbook of mixed methods in social & behavioral
research, SAGE Publications, Incorporated.
Teddlie, C. and A. Tashakkori (2008). Foundations of mixed methods research: Integrating
quantitative and qualitative approaches in the social and behavioral sciences, SAGE
Publications, Incorporated.
Thomas, W. I. and D. S. Thomas (1928). The child in America: Behavior problems and
programs. A. A. Knopf. New York, Inc.
Thombs, D. L. (2000). "A test of the perceived norms model to explain drinking patterns
among university student athletes." Journal of American College Health 49(2): 75-83.
Thombs, D. L., S. Dotterer, R. S. Olds, K. E. Sharp and C. G. Raub (2004). "A close look at
why one social norm campaign did not reduce student drinking." Journal of American
College Health 53: 61-68.
Thombs, D. L., J. Ray-Tomasek, C. J. Osborn and R. S. Olds (2005). "The role of sex-specific
normative beliefs in undergraduate alcohol use." American Journal of Health Behavior
29(4): 342-351.
394
References
Thombs, D. L., B. J. Wolcott and L. G. E. Farkash (1997). "Social context, perceived norms
and drinking behavior in young people." Journal of Substance Abuse 9: 257-267.
Trafimow, D. and K. A. Finlay (1996). "The importance of subjective norms for a minority of
people: Between subjects and within-subjects analyses." Personality and Social
Psychology Bulletin 22(8): 820-828.
Traxler, C. (2010). "Social norms and conditional cooperative taxpayers." European Journal of
Political Economy 26(1): 89-103.
Trockel, M., S. S. Williams and J. Reis (2003). "Considerations for more effective social
norms based alcohol education on campus: An analysis of different theoretical
conceptualizations in predicting drinking among fraternity men." Journal of Studies on
Alcohol 64(1): 50-59.
Tubaro, P., A. A. Casilli and L. Mounier (2012). "Eliciting personal network data in web
surveys through participant-generated sociograms (in press)." Field methods 26(2): 116.
Turner, J. C., M. A. Hogg, P. J. Oakes, S. D. Reicher and M. S. Wetherell (1987).
Rediscovering the social group: A self-categorization theory. New York, Blackwell.
Turrisi, R. (1999). "Cognitive and attitudinal factors in the analysis of alternatives to binge
drinking." Journal of Applied Social Psychology 29(7): 1512-1535.
Turrisi, R., K. A. Mallett, N. R. Mastroleo and M. E. Larimer (2006). "Heavy drinking in
college students: who is at risk and what is being done about it?" The Journal of general
psychology 133(4): 401-420.
Unger, J. B. and X. Chen (1999). "The role of social networks and media receptivity in
predicting age of smoking initiation:: A proportional hazards model of risk and
protective factors." Addictive Behaviors 24(3): 371-381.
Urberg, K. A., S. M. Degirmencioglu and C. Pilgrim (1997). "Close friend and group influence
on adolescent cigarette smoking and alcohol use." Developmental Psychology 33(5):
834-844.
Vaismoradi, M., H. Turunen and T. Bondas (2013). "Content analysis and thematic analysis:
Implications for conducting a qualitative descriptive study. ." Nursing & Health
Sciences.
early online review (doi: 10.1111/nhs.12048).
Valente, T. W. (1996). "Network models of the diffusion of innovations." Computational &
Mathematical Organization Theory 2(2): 163-164.
Valente, T. W. (2003). "Social network influences on adolescent substance use: An
introduction." Connections 25(2): 11-16.
395
References
Valente, T. W. and R. L. Davis (1999). "Accelerating the diffusion of innovations using

opinion leaders." The Annals of the American Academy of Political and Social Science
566(1): 55-67.
Valente, T. W., P. Gallaher and M. Mouttapa (2004). "Using social networks to understand and
prevent substance use: A transdisciplinary perspective." Substance Use & Misuse
39(10-12): 1685-1712.
Valente, T. W., B. R. Hoffman, A. Ritt-Olson, K. Lichtman and C. A. Johnson (2003). "Effects
of a social-network method for group assignment strategies on peer-led tobacco
prevention programs in schools." Journal Information 93(11): 1837-1843.
Valente, T. W. and D. Vlahov (2001). "Selective risk taking among needle exchange
participants: implications for supplemental interventions." American Journal of Public
Health 91(3): 406-411.
Valente, T. W., S. C. Watkins, M. N. Jato, A. Van Der Straten and L. P. M. Tsitsol (1997).
"Social network associations with contraceptive use among Cameroonian women in
voluntary associations." Social Science & Medicine 45(5): 677-687.
Vehovar, V., K. Lozar Manfreda, G. Koren and V. Hlebec (2008). "Measuring ego-centered
social networks on the web: Questionnaire design issues." Social networks 30(3): 213222.
Visser, P. S. and R. R. Mirabile (2004). "Attitudes in the social context: the impact of social
network composition on individual-level attitude strength." Journal of personality and
social psychology 87(6): 779-795.
Walters, S. T. (2000). "In praise of feedback: An effective intervention for college students
who are heavy drinkers." Journal of American College Health 48(5): 235-238.
Walters, S. T. and C. Neighbors (2005). "Feedback interventions for college alcohol misuse:
What, why and for whom?" Addictive Behaviors 30(6): 1168-1182.
Walters, S. T., A. M. Vader and T. R. Harris (2007). "A controlled trial of web-based feedback
for heavy drinking college students." Prevention Science 8(1): 83-88.
Wang, M. Q., E. C. Fitzhugh, J. M. Eddy, Q. Fu and L. Turner (1997). "Social influences on
adolescents' smoking progress: A longitudinal analysis." American Journal of Health
Behavior 21(2): 111-117.
Wasserman, S. and K. Faust (1994). Social network analysis: Methods and applications,
Cambridge Univ Pr.
Wasserman, S. and K. Faust (2007). Social Network Analysis: Methods and Applications,
Cambridge Univ Pr.
Watson, D. and S. Parsons (2005). "Domestic abuse of women and men in Ireland." Report on
the National Study of Domestic Abuse. National Crime Council, Dublin.
396
References
Watts, D. J. (2003). Small worlds: the dynamics of networks between order and randomness,
Princeton University Press.
Wechsler, H., A. Davenport, G. Dowdall, B. Moeykens and S. Castillo (1994). "Health and
behavioral consequences of binge drinking in college." JAMA: the journal of the
American Medical Association 272(21): 1672-1677.
Wechsler, H. and M. Kuo (2000). "College students define binge drinking and estimate its
prevalence: Results of a national survey." Journal of American College Health 49(2):
57-64.
Wellman, B. (1982). "Studying personal communities." Social Structure and Network
Analysis: 61-80.
Wellman, B. (1983). "Network analysis: Some basic principles." Sociological theory 1(1): 155200.
Wellman, B. and K. Frank (2001). Network capital in a multi-level world: Getting support
from personal communities. Social capital: Theory and research. N. Lin and K. Cook.
Chicago, Aldine DeGruyter: 233-273.
Wellman, B. and S. Wortley (1990). "Different strokes from different folks: Community ties
and social support." American Journal of Sociology 96(3): 558-588.
Werch, C. E., D. M. Pappas, J. M. Carlson, C. C. DiClemente, P. S. Chally and J. A. Sinder
(2000). "Results of a social norm intervention to prevent binge drinking among firstyear residential college students." Journal of American College Health 49(2): 85-92.
White, A. M., C. L. Kraus and H. S. Swartzwelder (2006). "Many college freshmen drink at
levels far beyond the binge threshold." Alcoholism: Clinical and Experimental
Research 30(6): 1006-1010.
White, H. C., S. A. Boorman and R. L. Breiger (1976). "Social structure from multiple
networks. I. Blockmodels of roles and positions." American Journal of Sociology
81(4): 730-780.
White, H. D., B. Wellman and N. Nazer (2004). "Does citation reflect social structure?:
Longitudinal evidence from the Globenet interdisciplinary research group." Journal
of the American Society for Information Science and Technology 55(2): 111-126.
White, H. R., M. E. Bates and V. Johnson (1991). "Learning to drink: Familial, peer, and
media influences." Society, Culture and Drinking Patterns Reexamined: 177-197.
White, K. M., D. J. Terry and M. A. Hogg (1994). "Safer sex behavior: The role of attitudes,
norms, and control factors." Journal of applied social psychology 24(24): 2164-2192.
WHO (2011). Global status report on alcohol and health. Geneva, Switzerland, World Health
Organization.
397
References
Wicki, M., E. Kuntsche and G. Gmel (2010). "Drinking at European universities? A review of
students' alcohol use." Addictive behaviors 35(11): 913-924.
Willis, G. B. (1994). Cognitive interviewing and questionnaire design: A training manual, US
Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Health Statistics.
Willis, G. B. (2004). Cognitive interviewing revisited: A useful technique, in theory? Methods
for testing and evaluating survey questionnaires, published online: 23-43.
Willis, G. B., T. J. DeMaio and B. Harris-Kojetin (1999). "Is the bandwagon headed to the
methodological promised land? Evaluating the validity of cognitive interviewing
techniques." Cognition and Survey Research. New York: Wiley: 133-153.
Windle, M. (2000). "Parental, sibling, and peer influences on adolescent substance use and
alcohol problems." Applied Developmental Science 4(2): 98-110.
Wolfson, S. (2000). "Students' estimates of the prevalence of drug use: Evidence for a false
consensus effect." Psychology of Addictive Behaviors 14(3): 295-298.
Wood, M. D., C. T. Nagoshi and D. A. Dennis (1992). "Alcohol norms and expectations as
predictors of alcohol use and problems in a college student sample." The American
Journal of Drug and Alcohol Abuse 18(4): 461-476.
Wood, M. D., J. P. Read, R. E. Mitchell and N. H. Brand (2004). "Do parents still matter?
Parent and peer influences on alcohol involvement among recent high school
graduates." Psychology of Addictive Behaviors 18 (1): 19-30.
Wood, M. D., J. P. Read, T. P. Palfai and J. F. Stevenson (2001). "Social influence processes
and college student drinking: The mediational role of alcohol outcome expectations."
Journal of Studies on Alcohol 62(1): 32-43.
398
Appendices
Appendix 1: Alcohol in Ireland: consumption, harm and policy
399
Appendix 2: Variable details
407
Appendix 3: General characteristics of pragmatism
413
Appendix 4: Pilot questionnaire
414
Appendix 5: Questionnaire for the web survey
439
Appendix 6: Results of independent t-test for early/late respondents
460
Appendix 7: Assumption testing
463
Appendix 8: Interview guide
480
Appendix 9: Glossary of common network terms
484
Appendix 10: Tie strength scores
486
Appendix 1: Alcohol in Ireland: Consumption, Harm and Policy
Percentage of Market Share
Market Share by Beverage Type

80
70
60
50
40
30
20
10
0
1986
1991
1996
2001
2006
Beer
69
69
65
55
51
Spirits
23
21
19
21
19
Wine
10
14
21
Cider
Figure 1: Market Share by Beverage Type (Hope, 2007)
Figure 2: Binge Drinking 1 - European Comparison (Ramstedt and Hope, 2005)

1
Binge drinking conceptualized as consumption of alcoholic beverages corresponding to at least one bottle of wine,
25 centilitres of spirits or 4 pints of beer, or more, during one drinking occasion in the past 12 months
- 399 -
Figure 3: Alcohol consumption in a single sitting (EC, 2011)
Figure 4: Age of drinking onset in Ireland (Unicef, 2011)
- 400 -
Figure 5: Summarized results from ESPAD (2011) (Hibell et al., 2012)
Figure 6: Proportion of school children reporting having ever been drunk by age and gender (Doyle et al.,
2009)
- 401 -
Figure 7: Alcohol consumption and related harm indicators in Ireland (Hope and Butler, 2010)
402
Figure 8: Harm to self by drinking patterns (Mongan et al., 2009)
Figure 9: Harm to others by drinking patterns (Mongan et al., 2009)
403
Figure 10: Trends in drunkenness, public order and assault offences among minors (Mongan et al., 2009)
Figure 11: Costs to society of problem alcohol use in Ireland (Byrne, 2010)
404
1990
1994
1994-01
1996
2000
2002
2003
2004
2005
2006
2008
2009
2011
Minister of Health requested the development of a National Alcohol Policy

BAC reduced to 0.80mg
No increase in alcohol taxes
National Alcohol Policy Published by Government
Intoxicating Liquor Act 2000
- Longer opening hours
- Free movement of Licences
- Lifting of restrictions for granting certain licences
- Temporary closure for selling to minors (under 18 years of age)
Tax increase on cider (Dec 01 Budget)
Strategic Task Force Interim Report
Tax increase on spirits (Dec 02 Budget)
Intoxicating Liquor Act (2003)
- Revert to earlier closing time on Thursday nights
- Temporary closure for serving to drunken customers
- Ban on happy hours
- Ban on children from pubs after 9 pm (extended to 10 pm later)
Proposed legislation to restrict alcohol marketing
Establishment of Central Copy Clearance Ireland (CCCI)
Strategic Task Force on Alcohol Second Report
Alcohol marketing legislation shelved in favor of industry self regulation
ASAI Voluntary Code of Practice came into effect
- covers alcohol advertising on TV, radio, cinema and outdoor/ambient media
- 33% audience profiling system introduced
AMCMB established to oversee implementation/adherence to voluntary codes
Mandatory alcohol testing (similar to Random breath testing)
Below unit cost selling of alcohol allowed (Abolition of Groceries Order)
Report of the Government Alcohol Advisory Group
Intoxicating Liquor Act (2008)
- Earlier closing time for off-licences
- Tougher public order provisions
- Court procedure to secure a new wine-only off-license
- New grounds for objection to granting of an off-license allowing Gardai and public to intervene
- New conditions attaching to the granting of a special exemption order
- Regulation to restrict promotions, changed to Industry self regulation
Revised Voluntary Code of Practice
- Audience Profile cut off reduced to 25%
- Ban on alcohol advertising during breakfast time (6 to 10 am)
Tax Increase on wine
Excise duty reduced by almost 20% on all alcoholic products (Budget 2010)
RRAI established
RRAI Code of Practice on the Display and Sale of Alcohol Products in Mixed Trading Premises
came into effect
RRAI code under review following the third compliance report
405
2012
Steering Group report provided recommendations on National Substance Misuse Strategy

- Maintain high excise duties
- Introduce a legislative basis for minimum pricing per gram of alcohol
- Introduce a Statutory Code of Practice on the sale of alcohol in off license sector
- Strengthen legislation on distance sales
- Develop a system to monitor the enforcement of the intoxicating liquor legislation
- 9pm watershed for all alcohol advertising on TV and radio
- Prohibit all outdoor advertising
- Age authentication controls on websites advertising alcohol
- Phase out drink industrys sponsorship of sports and other large events by 2016.
- Introduce a social responsibility levy through which the drinks industry would contribute to the
cost of social marketing and awareness campaigns relating to social and health harms caused by
alcohol
Table 1: Alcohol Policy Activity in Ireland: 1990-2012 (Hope and Butler, 2010; DoH, 2012)
Byrne, S. (2010). Costs to Society of Problem Alcohol Use in Ireland. Health Service Executive.
Dublin.
DoH (2012). Steering Group Report on a National Substance Misuse Strategy.
Doyle, P., M. Molcho and S. Gabhainn (2009). HSBC Ireland: Age related patterns in alcohol
consumption and cannabis use among Irish children between 1998-2006: Short report.
Galway, Health Promotion Research Centre: National University of Ireland.
EC (2011). "Youth attitudes towards drugs: Eurobarometer."
Hibell, B., U. Guttormsson, S. Ahlstrom, O. Balakireva, T. Bjarnason, A. Kokkevi and L. Kraus
(2012). "The 2011 ESPAD report: substance use among students in 36 European
countries."
Hope, A. (2007). Alcohol consumption in Ireland 1986-2006. Report for the Health Service
Executive-Alcohol Implementation Group, Dublin: Health Service Executive.
Hope, A. and S. Butler (2010). "Changes in consumption and harms, yet little policy progress.
Trends in alcohol consumption, harms and policy: Ireland 1990-2010." Nordic Studies on
Alcohol and Drugs 27(5): 479-496.
Mongan, D., A. Hope and M. Nelson (2009). Social consequences of harmful use of alcohol in
Ireland. HRB Overview Series 9. Dublin: Health Research Board.
Ramstedt, M. and A. Hope (2005). "The Irish drinking habits of 2002-Drinking and drinkingrelated harm in a European comparative perspective." Journal of Substance Use 10(5):
273-283.
Unicef (2011). Changing the Future : Experiencing Adolescence in Contemporary Ireland,
Report 3: Alcohol and Drugs Report
406
Variable
Rationale for inclusion
Source
Original Study
Ammended version for present study
Rationale for Ammendments
Pilot analysis revealed that using a specific value

for age such as an open ended numerical
variable will be more practical and realistic for
analysis.
Age
Control variable
McAlaney and McMahon (2007)
How old are you?

18-20, 21-24, 25-34, 35-44, 45-54, 55-64
65 or above
Gender
Control variable
Are you male or female?

Male
Female
Year in college
Control variable
Delaney et al (2008)
What year of the course are you currently in?

1, 2, 3, 4, 5, 6+
How many years have you been in DIT?

Less than 1 year, 1 year, 2 years, 3 years
4 years, 5 or more years
Ammendements based on pilot feedback. The

original question did not capture the total
number of years spent in DIT for students who
might have changed courses.
Real and Rimal (2007)

Rimal and Real (2005)
The original study provided the following response

options to age of drinking onset.
younger than 11 years
betweeen 11 and 12 years
between 13 and 14 years
never
How old were you when you had your first drink? (other
than just a sip)?
Never had a drink
11 years or younger, 12 years, 13 years
14 years, 15 years, 16 years, 17 years
18 years or above
Explanatory line 'other than just a sip', specific

ages and the option '18 or above' in response
options provided based on feedback from
cognitive interviews.
Darmody et al (2005)
How much money is available to you in a typical week for

During term time, approximately how much do you drinking? (related costs such as food and taxis available)
spend on average each month on alcohol?
I dont drink, 20 or less, 21-40, 41-60
________________________
61-80, 81-100, 101-120, 121-140
More than 140
Which of the following best describes your living
arrangment?
I am living with parents, I am living independently and
alone, I am living in a shared accomodation, I am living with
partner and/or dependent others (children)
Ammendements (response options) based on

feedback from cognitive interviews. The final
response options reflect various living
arrangments in an Irish context (no residential
halls)
How many people do you usually drink with on a typical

occasion?
I dont drink, alone, 1, 2-3, 4-5, 6-7, 8-9, 10 or above.
Ammendements (response options 'I don't drink'

and 'alone') based on feedback from cognitive
interviews
Age of first drink
Money available for

drinking
Control variable
Control variable
How old are you?

___________________
Living Arrangement
Control variable
Jones et al (1992)

options for living arrangments
living in a hall, not living in a hall but living alone or
with non relatives, living with family or spouse
Drinking group size
Control variable
Demers et al (2002)

options to group size
one, 2-3, 4-9, 10 or more
407
Ammendments (question wording, weekly

instead of monthly expenditure and addition of
explanatory line) based on feedback from
cognitive interviews
Variable
Communication
about alcohol
Control variable
Source
Dorsey et al (1999)
Real and Rimal (2007)
Original Study
Dorsey et al (1999) assesed frequency of

communication about 6 topics with friends, family
members, RA's and professors
1- safer sex practices, 2- effects of drinking too
much alcohol, 3- unwanted sexual advances caused
by drinking, 4- binge drinking, 5- physical violence,
injuries or fights, 6- feeling sick as a result of
drinking
Real and Rimal (2007) asked
1. Over the past 2 weeks, how often have you
talked with your friends or siblings about your
drinking alcohol? 7 point measure (never to more
than 6 times)
2. How often do you normally talk with your friends
or siblings about alcohol consumption? 10 point
measure (never to all the time)
Pilot questionnaire tested a version of Dorsey et al (1999)

(Refer to section in chapter 5). The final questionnaire used
the following ammended version from Real and Rimal
(2007)
1. How frequently does alcohol/drinking come up in your
conversation with others?
Several times a day, 2-3 times a day, about once a day, a
few times a day, once a week, never
2. How often in a typical week does alcohol/drinking come
up in your conversation with others?
very frequently, frequently, occasionally, rarely, very rarely,
never
* Explanation added: Think about casual and informal
conversations you have with people on day to day basis. For
example, After a night out with friends, setting status
messages or exchanging comments on facebook or bebo,
exchanging emails or text messages, twitter, over dinner at
home etc
Ammendments (question wording and response

options) based on feedback from cognitive
interviews and pilot survey. Dorsey et al (1999)
utilized a pen/paper survey in contrast to a web
survey employed in this study. The web format
of the original set of questions was found to
cause confusion in participants which often led
to incomplete/unanswered responses. The
explanation was added to clarify the meaning of
conversation.
Nationality
Filter variable to identify non

Irish students (not relevant
Delaney et al (2008)
for this study) in the final
sample
which of the following best describes your

situation?
Irish national, Foreign national studying for a full
qualification in Ireland, Foreign national studying as
part of an exchange program
Full/Part time
Filter variable to identify

part time students (not
McAlaney (2007)
relevant to this study) in the
final sample
Are you currently a full time/part time student?

Full time, Part time
Are you a fill time/part time student at DIT?

Full time, Part time
Under/post grad
Filter variable to identify

post graduate students (not
McAlaney (2007)
relevant to this study) in the
final sample
The measure on 'year in college' also included this

aspect in the original study
Please choose the one that applies to you.

I am an undergraduate student, I am a post graduate
student
Campus
Filter variable used later for

campus wise analysis of data
Which DIT campus are you based at?

DIT Mountjoy Square, DIT Cathal Brugha, DIT Bolton Street,
DIT Aungier Street, DIT Kevin Street, DIT Rathmines Road
408
ammendment (question wording) based on

feedback from cognitive interviews.
Variable
Frequency of
Drinking (Typical
month)
Dependant Variable,
measure
Source
Original Study
How many days in a typical month do you normally

drink alcohol?
Never or very rarely, less than once a month, once a
month, 2-3 days a month, once a week, twice a
week, 3-4 days a week, 5-6 days a week, everyday
How often in a typical month do you normally drink alcohol?

Ammendments ( question wording and response
Never, less than once a month, once a month, 2-3 days a
option 'never' instead of 'never or very rarely')
month, once a week, twice a week, 3-4 days a week, 5-6
based on feedback from cognitive interviews
days a week, everyday
Perceived frequency
of drinking of most
students in DIT
(typical month)
Perceived frequency
of drinking of close
friends (typical
Perceived descriptive norms
month)
used as independent
variables, measure
perceived frequency of
Perceived frequency
drinking (typical month) in
of drinking of best
friend (typical month) peers from different social
groups
Perceived frequency
of drinking of mother
(typical month)
How often in a typical month do you think most students in

DIT/your close friends/your best friend/your mother/your
father/ normally drink(s) alcohol?
Never, less than once a month, once a month, 2-3 days a
How many days in a typical month do you think
month, once a week, twice a week, 3-4 days a week, 5-6
most of your closest friends/average student your
options 'never' instead of 'never or very rarely'
days a week, everyday.
age at your University/average person your age in
and 'Not applicable' in questions related to
McAlaney and McMahon (2007) the UK normally drink alcohol?
mother and father) based on feedback from
* an additional response option 'Not Applicable' provided
Never or very rarely, less than once a month, once a
cognitive interviews. The explanatory line 'Please
for questions related to mother and father and an
month, 2-3 days a month, once a week, twice a
select the last option if the question is not
explanatory line added 'Please select the last option if the
week, 3-4 days a week, 5-6 days a week, everyday
applicable to you' based on pilot feedback
question is not applicable to you'
* choice of distal and proximal social groups based on
feedback from cognitive interviews
Perceived frequency
of drinking of father
(typical month)
409
Variable
Number of drinks
(typical occasion)
Dependant Variable,
measure
Source
Original Study
As you are probably aware, different alcoholic drinks have

different quantities of alcohol.For the purpose of the
following questions (17-23), please estimate 1 pint of beer
or cider as the equivalent of 2 drinks and all other drinks
(glass of wine, spirits or shots) as the equivalent of 1 drink
each.
Thus, somebody who has consumed 3 pints of beer and a
glass of wine has consumed the equivalent of 7 drinks i.e.
each pint is two drinks (3x2) plus 1 glass of wine (1 drink)
leads to a total of 7 drinks.
Ammendments (provision of standard drink

How many drinks would you normally drink during a
measures, question wording and 'college night'
night out in a pub/club?
context) based on feedback from cognitive
Based on the above values, approximately how many
0, 1-2, 3-4, 5-6, 7-8, 9-10, 11-12, 13-14, 15 or more
alcoholic drinks do you normally drink on a typical drinking interviews and pilot study.
occasion? Remember, 1 pint is the equivalent of 2 drinks; all
other beverage is the equivalent of 1 drink each.
Same response options as in the original study
*an additional question assessed number of drinks
consumed on a college night. An explanatory line 'A college
night is when you have college/classes the next morning'
was added.
Perceived number of
drinks consumed by
most students in DIT
(typical occasion)
Perceived number of
drinks consumed by
close friends (typical
occasion)
perceived descriptive norms

used as independent
Perceived number of
variables, measure
drinks consumed by
perceived number of drinks
best friend(typical
(typical occasion) consumed
occasion)
by peers from different
social groups
Perceived number of
drinks consumed by
mother (typical
occasion)
Based on the above values, approximately how many

alcoholic drinks do you normally drink on a typical drinking
occasion? Remember, 1 pint is the equivalent of 2 drinks; all
other beverage is the equivalent of 1 drink each.
How many drinks do you think most of your closest
friends/average student your age at your
Same response options as in the original study
McAlaney and McMahon (2007) university/average person in the UK would normally
drink during a night out in a pub/club?
0, 1-2, 3-4, 5-6, 7-8, 9-10, 11-12, 13-14, 15 or more for questions related to mother and father and an
Perceived number of
drinks consumed by
father (typical
occasion)
410
Ammendments (provision of standard drink

measures, question wording and 'not applicable'
option for questions related to mother/father)
based on feedback from cognitive interviews and
pilot study. The explanatory line 'Please select
the last option if the question is not applicable to
you' based on pilot feedback
Variable
Frequency of
Dependant Variable,
drunkenness (typical Personal consumption
month)
measure
Source
Original Study
how many days in a typical month do you drink

enough alcohol to become drunk?
Same response options as in the measure for
frequency of drinking (typical month) used in the
original study
How often in the course of a typical month do you drink

enough alcohol to become drunk?
Same response options as in the measure for frequency of
drinking (typical month) used in the present study

option 'never' instead of 'never or very rarely')
based on feedback from cognitive interviews
Perceived frequency
of drunkenness of
most students in DIT
(typical month)
Perceived frequency
of drunkenness+A33
of close friends
perceived descriptive norms
(typical month)
used as independent
variables, measure
Perceived frequency
perceived frequency of
of drunkenness of
drunkenness (typical month)
best friend (typical
in peers from different social
month)
groups
Perceived frequency
of drunkenness of
mother (typical
month)
How many days in a typical month do you think

most of your closest friends/average student your
age at your University/average person your age in
the UK normally drink enough alcohol to become
drunk?
Same response options as in the measure for
frequency of drinking (typical month) used in the
original study
Perceived frequency
of drunkenness of
father (typical
month)
411
How often in a typical month do you think most students in

DIT/your close friends/your best friend/your mother/your Ammendments ( question wording and response
father/ normally drink(s) enough alcohol to become drunk? options 'never' instead of 'never or very rarely'
Same response options as in the measure for frequency of
and 'Not applicable' in questions related to
drinking (typical month) used in the present study
mother and father) based on feedback from
cognitive interviews. The explanatory line
'Please select the last option if the question is
not applicable to you' based on pilot feedback
Variable
Source
Original Study
Perceived
acceptablity/approva
l of drinking to get
drunk in most
students at DIT
Perceived
lof drinking to get
drunk in close friends
Perceived injunctive norms
Perceived
used as independent
variables, measure
perceived
drunk in best friend
acceptability/approval in
peers from different social
Perceived
acceptablity/approva groups about drinking to get
drunk
drunk mother
Perceived
drunk in father
Most students at DIT think that drinking to get drunk is

acceptable
strongly disagree, disagree, neither disagree nor agree,
agree, strongly agree
Marcoux and Shope (1997)

McAlaney (2007)
Marcoux and Shope (1997) asked: How do your

parents/friends/older brothers/older sisters feel
about kids your age drinking alcohol?
very bad idea, bad idea, neither good nor bad idea, My close friends/best friend would approve of me drinking
to get drunk
good idea, very good idea
strongly disagree, disagree, neither disagree nor agree,
McAlaney (2007) asked: How do you think your agree, strongly agree
friends/your family/Uk society as a whole feel about
people your age drinking enough alcohol to become My mother/father thinks that my drinking to get drunk is
very unacceptable, unacceptable, neither unacceptable nor
drunk?
strongly disapprove, disapprove, neither approve acceptable, acceptable, very acceptable
nor disapprove, approve, strongly approve
for questions related to mother and father and an
*an additional set of questions also assessed perceived
acceptability/approval of 'drinking to get drunk on a college
night' for each peer group. An explanatory line 'A college
night is when you have college/classes the next morning'
was added
412
Ammendments (question wording, use of the

term 'drinking to get drunk', response options
and 'college night' context) based on feedback
from cognitive interviews and pilot study. The
explanatory line 'Please select the last option if
the question is not applicable to you' for
questions related to mother/father based on
pilot feedback
Appendix 3: General Characteristics of Pragmatism (Johnson and Onwuegbuzie, 2004) p18
Johnson, R. B. and A. J. Onwuegbuzie (2004). "Mixed methods research: A research paradigm

whose time has come." Educational researcher 33(7): 14-26.
413
Appendix 4: Pilot Questionnaire

Confidentiality
1. This survey is part of an academic research project on drinking habits of young people.
2. Any information that you disclose here in shall be treated as confidential and used only
for academic research purposes.
3. It shall not be possible for the researcher to identify the respondents after the completion
of this survey either.
4. Please note that you are free to withdraw from this study
at any time
without giving a reason for withdrawing
without affecting your future relationship with DIT
5. If you have any queries or concerns pertaining to this study or wish to discuss it further,
please contact Sarah Samdani on sarahsamdani@hotmail.com.
6. Please read each question carefully and answer honestly. Please answer ALL questions
on the survey. This survey typically takes about 10 minutes to complete. The asterisk (*)
sign against each question indicates that it is required to be answered
Consent Form
1. Have you been made fully aware about the purpose of this study? *
Yes
No
2. Do you understand that you are free to withdraw from this study? *
- at any time
- without giving a reason for withdrawing
- without affecting your future relationship with the Institute
Yes
No
3. Have you been informed that any information you reveal in this survey will be treated as
confidential and used only for research purposes? *
Yes
No
No
4. Do you agree to take part in this online survey? *
Yes
- 414 -
1. Are you male or female? *
Male
Female
2. How old are you? *
18-21 years
22-25 years
26-29 years
30 years or above
3. Which of the following best describes your living arrangements? *
I am living with parents
I am living independently and alone
I am living in a shared accommodation
I am living with my partner and/or dependent others (children)
4. Which of the following best describes your situation *
Irish National
Foreign National studying for a full qualification in Ireland
Foreign National studying as part of an exchange program
5. How many years have you been in DIT? *
Less than one year
4 years
1 year
5 or more years
2 years
3 years
415
6. How much money is available to you in a typical week for drinking (related costs such as
food and taxis included)? *
20 or less
21-40
41-60
61-80
81-100
101-120
121-140
More than 140
7. How often on average in a typical month do you normally drink alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
416
8. How often on average in a typical month do you think most students in DIT normally
drink alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
9. How often on average in a typical month do you think your close friends normally drink
alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
417
10. How often on average in a typical month do you think your best friend normally drinks
alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
11. How often on average in a typical month do you think your mother normally drinks
alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
Not Applicable
418
12. How often on average in a typical month do you think your father normally drinks
alcohol? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
Not Applicable
13. How many alcoholic drinks would you normally drink on a typical drinking occasion? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
419
14. How many alcoholic drinks do you think most students at DIT would normally drink on a
typical drinking occasion? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
15. How many alcoholic drinks do you think your close friends would normally drink on a
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
420
16. How many alcoholic drinks do you think your best friend would normally drink on a
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
17. How many alcoholic drinks do you think your mother would normally drink on a typical
drinking occasion? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
421
18. How many alcoholic drinks do you think your father would normally drink on a typical
drinking occasion? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
19. How many alcoholic drinks would you normally drink on a typical weekend night?
(Thursday, Friday and Saturday night) *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
422
20. How many alcoholic drinks do you think most students at DIT would normally drink on a
typical weekend night? (Thursday, Friday and Saturday night) *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
21. How many alcoholic drinks do you think your close friends would normally drink on a
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
423
22. How many alcoholic drinks do you think your best friend would normally drink on a
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
23. How many alcoholic drinks do you think your mother would normally drink on a typical
weekend night? (Friday and Saturday night) *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
424
24. How many alcoholic drinks do you think your father would normally drink on a typical
weekend night? (Friday and Saturday night) *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
25. How many alcoholic drinks would you normally drink when you have college the next
day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
425
26. How many alcoholic drinks do you think most students at DIT would normally drink
when they have college the next day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
27. How many alcoholic drinks do you think your close friends would normally drink when
they have college the next day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
426
28. How many alcoholic drinks do you think your best friend would normally drink when
he/she has college the next day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
29. How many alcoholic drinks do you think your mother would normally drink when she
has work the next day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
427
30. How many alcoholic drinks do you think your father would normally drink when he has
work the next day? *
1-2
3-4
5-6
7-8
9-10
11-12
13-14
15 or more
Not Applicable
31. How often on average in a typical month do you drink enough alcohol to become drunk?
*
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
428
32. How often on average in a typical month do you think most students at DIT drink enough
alcohol to become drunk? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
33. How often on average in a typical month do you your close friends drink enough alcohol
to become drunk? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
429
34. How often on average in a typical month do you think your best friend drinks enough
alcohol to become drunk? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
35. How often on average in a typical month do you think your mother drinks enough alcohol
to become drunk? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
Not Applicable
430
36. How often on average in a typical month do you think your father drinks enough alcohol
to become drunk? *
Never
1-2 days a month
2-3 days a month
Once a week
Twice a week
3-4 days a week
5-6 days a week
Everyday
Not Applicable
37. How old were you when you had your first drink? *
11 years or younger
12 years
13 years
14 years
15 years
16 years
17 years
18 years or above
Never had a drink
431
38. How many people do you typically drink with on a weekend night? (Thursday, Friday
and Saturday night) *
Alone
2-3
4-9
10 or more
39. How many people do you typically drink when you have college the next day? *
Alone
2-3
4-9
10 or more
40. Please select all that apply. The people that I typically drink with mostly include my *
Boy friend/ Girl friend
Class mates
Flat mates
Friends from college
Friends from outside college
Work mates
Neighbors
Siblings
Parents
Other relatives
432
41. How frequently do you communicate with the people you typically drink with? Please select the
appropriate answer for all categories that apply to you. *
Less than
once a
month
Once a
month
Twice a
month
Once a
week
Twice a
week
3-4 times
a week
5-6 times
a week
Everyday
Boy Friend / Girl Friend

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
42. Think about the past month. Approximately how many times have you spoken with the people
you typically drink with about safer sex practices?
Please select appropriate answers for all categories that apply to you *
Never
1-2 times
3-4 times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
433
5-6 times
7-8 times
9-10 times
11 or more
times
you typically drink with about The Effects of Drinking too much Alcohol?
Please select appropriate answers for all categories that apply to you. *
Never
1-2 times
3-4 times
5-6 times
7-8 times
9-10 times
11 or more times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
you typically drink with about unwanted sexual advances?
Never
1-2 times
3-4 times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
434
5-6 times
7-8 times
9-10 times
11 or more times
you typically drink with about binge drinking?
Never
1-2 times
3-4 times
5-6 times
7-8 times
9-10 times
11 or more times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
you typically drink with about physical violence, injuries or fights caused by drinking?
Please select appropriate answers for all categories that apply to you *
Never
1-2 times
3-4 times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
435
5-6 times
7-8 times
9-10 times
11 or more times
you typically drink with about feeling sick as a result of drinking?
Never
1-2 times
3-4 times
5-6 times
7-8 times

Class Mates
Flat Mates
Work Mates
Neighbors
Siblings
Parents
Other relatives
48. Most students at DIT think that drinking to get drunk is acceptable *
Strongly disagree
Disagree
Neither disagree nor agree
Agree
Strongly agree
49. My close friends would approve of me drinking to get drunk *
Strongly disagree
Disagree
Agree
Strongly agree
436
9-10 times
11 or more times
50. My best friend would approve of me drinking to get drunk *
Strongly disagree
Disagree
Agree
Strongly agree
51. My mother would approve of me drinking to get drunk *
Very unacceptable
Unacceptable
Neither unacceptable nor acceptable
Acceptable
Very acceptable
Not Applicable
52. My father would approve of me drinking to get drunk *
Very unacceptable
Unacceptable
Acceptable
Very acceptable
Not Applicable
437
Your feedback on this survey will be greatly appreciated. Please take a few minutes to answer
the following questions
1. Did you have any difficult answering the questions on this survey?
2. Were there any particular questions that confused you?
3. Are there any comments or suggestions that you might have about the survey?
438
Appendix 5: Questionnaire for the Web Survey
DIT Drinking Survey - Your chance to win an Apple iPAD!!!
THIS SURYEY IS MEANT ONLY FOR DIT STUDENTS. PLEASE ANSWER

ALL QUESTIONS ON THE SURVEY IN ORDER TO ENTER THE DRAW
FOR AN APPLE iPAD
CONFIDENTIALITY
1. This survey is part of an academic research project on the drinking habits of young people. It
will ask you questions about your drinking behavior and what you think about the drinking
habits of relevant others such as your close friends, other students at DIT, etc.
2. This survey is confidential and will be used only for academic research purposes.
3. Please note that you are free to withdraw from this study at any time and without giving a
reason for withdrawing.
4. If you have any queries regarding this study or wish to discuss it further, please contact Sarah
Samdani on sarah.samdani1@student.dit.ie.
5. Please read each question carefully and answer honestly. There are no right or wrong answers.
6. Please answer ALL questions on the survey. This survey typically takes about 10 minutes to
complete. The asterisk (*) sign against each question indicates that it is required to be answered.
- 439 -
Page One
CONSENT FORM
1. Please provide your consent to participate in this survey
i.
I have been made fully aware about the purpose of this study
ii.
I understand that I am free to withdraw from this study at any stage
iii.
I have been informed that any information I reveal during this survey will be treated
as confidential and used only for research purposes
iv.
I agree to take part in this online survey
I agree
I do not want to participate
Page Two
2. Are you male or female?*
i.
Male
ii.
Female
3.
How old are you?*

________________
4. Which of the following best describes your living arrangements?*

i.
I am living with parents
ii.
I am living independently and alone
iii.
I am living in a shared accommodation
iv.
I am living with my partner and/or dependent others (children)
- 440 -
5. Which of the following best describes your situation?*

i.
Irish National
ii.
Foreign National studying for a full qualification in Ireland
iii.
Foreign National studying as part of an exchange program
6.
Are you a full time or a part time student at DIT?*
i.
Full Time
ii.
Part Time
7. Which DIT campus are you based at?*

i.
DIT Mountjoy Square
ii.
DIT Cathal Brugha Street
iii.
DIT Bolton Street
iv.
DIT Aungier Street
v.
DIT Kevin Street
vi.
DIT Rathmines Road

8. Please choose the one that applies to you*
i.
I am an undergraduate student
ii.
I am a post graduate student

9. How many years have you been in the DIT?*
i.
Less than one year
ii.
1 year
iii.
2 years
iv.
3 years
v.
4 years
vi.
5 or more years
441
Page Three
10. How much money is available to you in a typical week for drinking (related costs
such as food and taxis included)?*
i.
I dont drink
ii.
20 or less
iii.
21-40
iv.
41-60
v.
61-80
vi.
81-100
vii.
101-120
viii.
121-140
ix.
More than 140
11. How often in a typical month do you normally drink alcohol?*

i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
442
12. How often in a typical month do you think most students at DIT normally drink
alcohol?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
13. How often in a typical month do you think your close friends normally drink
alcohol?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
443
14. How often in a typical month do you think your best friend normally drinks
alcohol?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
15. How often in a typical month do you think your mother normally drinks alcohol?*
Please select the last option if this question is not applicable to you.
i.
Never
ii.
iii.
once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
x.
Not applicable
444
16. How often in a typical month do you think your father normally drinks alcohol?*
Please select the last option if this question is not applicable to you
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
x.
Not applicable
Page Four
As you are probably aware, different alcoholic drinks have different quantities of
alcohol.
For the purpose of the following questions (17-23), please estimate 1 pint of beer
or cider as the equivalent of 2 drinks and all other drinks (glass of wine, spirits
or shots) as the equivalent of 1 drink each.
Thus, somebody who has consumed 3 pints of beer and a glass of wine has
consumed the equivalent of 7 drinks i.e. each pint is two drinks (3x2) plus 1 glass
of wine (1 drink) leads to a total of 7 drinks.
445
17. Based on the above values, approximately how many alcoholic drinks do you
normally drink on a typical drinking occasion?*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
Remember, 1 pint is the equivalent of 2 drinks; all other beverage is the equivalent of 1
drink each.
18. Based on the above values, approximately how many alcoholic drinks do you
normally drink on a typical college night? A college night is when you have
college/classes the next morning*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
446
drink each.
19. Based on the above values, approximately how many alcoholic drinks do you think
most students at DIT normally drink on a typical drinking occasion?*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
drink each.
your close friends normally drink on a typical drinking occasion?*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
447
drink each.
your best friend normally drinks on a typical drinking occasion?*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
drink each.
your mother normally drinks on a typical drinking occasion?
Please select the last option if the question is not applicable to you.*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
x.
Not applicable
448
drink each.
your father normally drinks on a typical drinking occasion?
Please select the last option is the question is not applicable to you.*
i.
ii.
1-2
iii.
3-4
iv.
5-6
v.
7-8
vi.
9-10
vii.
11-12
viii.
13-14
ix.
15 or more
x.
Not applicable
Page Five
24. How often in the course of a typical month do you drink enough alcohol to become
drunk?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
449
25. How often in the course of a typical month do you think most students at DIT drink
enough alcohol to become drunk?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
26. How often in the course of a typical month do you think your close friends drink
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
450
27. How often in the course of a typical month do you think your best friend drinks
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
28. How often in the course of a typical month do you think your mother drinks enough
alcohol to become drunk?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
x.
Not applicable
451
29. How often in the course of a typical month do you think your father drinks enough
alcohol to become drunk?*
i.
Never
ii.
iii.
Once a month
iv.
2-3 days a month
v.
Once a week
vi.
Twice a week
vii.
3-4 days a week
viii.
5-6 days a week
ix.
Everyday
x.
Not applicable
Page Six
30. How old were you when you had your first drink (other than just a sip)?*
i.
Never had a drink
ii.
11 years or younger
iii.
12 years
iv.
13 years
v.
14 years
vi.
15 years
vii.
16 years
viii.
17 years
ix.
18 years or above
452
31. How many people do you usually drink with on a typical occasion?*
i.
I dont drink
ii.
Alone
iii.
iv.
2-3
v.
4-5
vi.
6-7
vii.
8-9
viii.
10 or above
32. Please select all that apply. The people that I typically drink with mostly include my*
i.
Boy friend/ Girl friend
ii.
Class mates
iii.
Flat mates
iv.
v.
vi.
Work mates
vii.
Siblings
viii.
Parents
ix.
I dont drink
453
33. How often in a typical week do alcohol/drinking come up in your conversations with
others?*
Think about casual and informal conversations you have with people on day to day basis.
For example, After a night out with friends, setting status messages or exchanging
comments on facebook or bebo, exchanging emails or text messages, twitter, over dinner
at home etc
i.
Several times a day
ii.
2-3 times a day
iii.
About once a day
iv.
A few times a week
v.
Once a week
vi.
Never
34. How frequently does alcohol/drinking usually come up in your conversations with
others?*
Think about casual and informal conversations you have with people on day to day basis.
For example, After a night out with friends, setting status messages or exchanging
comments on facebook or bebo, exchanging emails or text messages, twitter, over dinner
at home etc
i.
Very Frequently
ii.
Frequently
iii.
Occasionally
iv.
Rarely
v.
Very Rarely
vi.
Never
454
Page Seven
35. Most students at DIT think that drinking to get drunk is acceptable*
i.
Strongly disagree
ii.
Disagree
iii.
Neither disagree nor Agree
iv.
Agree
v.
Strongly Agree
36. Most students at DIT think that drinking to get drunk on a college night is
acceptable*. A college night is when you have college/classes the next morning.
i.
Strongly disagree
ii.
Disagree
iii.
iv.
Agree
v.
Strongly Agree
37. My close friends would approve of me drinking to get drunk*

i.
Strongly disagree
ii.
Disagree
iii.
iv.
Agree
v.
Strongly Agree
38. My close friends would approve of me drinking to get drunk on a college night*. A
college night is when you have college/classes the next morning
i.
Strongly disagree
ii.
Disagree
iii.
iv.
Agree
v.
455
Strongly Agree
39. My best friend would approve of me drinking to get drunk*

i.
Strongly disagree
ii.
Disagree
iii.
iv.
Agree
v.
Strongly Agree
40. My best friend would approve of me drinking to get drunk on a college night*. A
college night is when you have college/classes the next morning
i.
Strongly disagree
ii.
Disagree
iii.
iv.
Agree
v.
Strongly Agree
41. My mother thinks that my drinking to get drunk is* Please select the last option if the
question is not applicable to you.
i.
Very unacceptable
ii.
Unacceptable
iii.
iv.
Acceptable
v.
Very acceptable
vi.
Not Applicable
456
42. My mother thinks that my drinking to get drunk on a college night is* A college
night is when you have college/classes the next morning.
Please select the last option if the question is not applicable to you.
i.
Very unacceptable
ii.
Unacceptable
iii.
iv.
Acceptable
v.
Very acceptable
vi.
Not Applicable
43. My father thinks that my drinking to get drunk is* Please select the last option if the
question is not applicable to you.
i.
Very unacceptable
ii.
Unacceptable
iii.
iv.
Acceptable
v.
Very acceptable
vi.
Not Applicable
44. My father thinks that my drinking to get drunk on a college night is* A college
night is when you have college/classes the next morning. Please select the last option
if the question is not applicable to you.
i.
Very unacceptable
ii.
Unacceptable
iii.
iv.
Acceptable
v.
Very acceptable
vi.
Not Applicable
457
Page Eight
45. Please enter your email address in the space provided.*
Please enter an email address that you can be contacted on.
Your email address will not be passed on to anyone else and you will not be contacted for
any promotional schemes or sent any spam mail.
It is being asked so that you might be contacted in case you win the Apple iPAD after the
raffle. This is subject to presentation of a valid DIT student ID.
You might also be contacted for an interview relating to the second phase of this research.
However, you can choose not to participate at that stage if you so wish.
Winning an iPAD is not dependent upon taking part in the 2nd phase. You can still say 'No' if
you win the iPAD.
____________________________________________
46. Please enter your Student ID or your DIT email address.*
This survey and the related incentive are meant only for DIT students. Your student number/DIT
email address is required as a means to verify that you are a DIT student. It will not be passed on
to anyone else including other students, faculty or the administrative staff at DIT.
____________________________________________
47. Which course/programme are you enrolled into at DIT? e.g. Bachelor of
Engineering (Civil)*
____________________________________________
458
Thank You!
Thank you for taking our survey. Your response is very important to us.
459
Appendix 6: Results of Independent t-test for Early and Late Respondents
Control variables
Timeliness
Mean
Std. Deviation
Early
1453
.57
.495
Late
247
.51
.501
Early
1453
21.10
2.941
Late
247
21.31
2.521
Early
1453
1.68
1.004
Late
247
1.54
.948
Early
1453
1.59
1.562
Late
247
2.31
1.656
Early
1453
45.96
34.407
Late
247
52.69
34.667
Early
1453
14.63
2.826
Late
247
15.00
1.571
Early
1453
1.64
21.743
Late
247
4.61
13.536
Early
1453
2.43
1.253
Late
247
2.39
1.251
Early
1453
3.24
.974
Late
247
3.24
.949
Gender
Age
Living Arrangement
Year in college
Money available for

drinking
Age of first drink
Drinking group size
Weekly communication
about alcohol
General communication
about alcohol
- 460 -
Sig. (2-tailed)
1.856
0.064
-1.17
0.241
1.86
0.069
-1.16
0.24
1.406
0.177
-1.17
0.241
-1.175
0.241
0.466
0.641
-0.01
0.992
Frequency of Drinking (typical month)
Frequency of Drinking
(typical month)
Perceived Norm (most

students at DIT)
Perceived Norm (Close

friends)
Perceived Norm (Best

friend)
Perceived Norm (Mother)
Perceived Norm (Father)
Timeliness
Mean
Std. Deviation
Early
1453
4.90
3.740
Late
247
4.96
3.259
Early
1453
8.53
3.892
Late
247
8.37
3.744
Early
1453
7.35
4.055
Late
247
6.50
3.442
Early
1453
6.64
4.626
Late
247
5.99
3.843
Early
1424
5.15
6.547
Late
246
5.32
5.997
Early
1343
7.03
7.679
Late
229
7.17
7.193
Sig.(2-tailed)
-.225
.822
.614
.539
.839
.402
1.041
.299
-.391
.696
-.262
.793
Number of Drinks
Timeliness
Number of Drinks (Typical

occasion)
Number of Drinks (Typical

college night)

students at DIT)

friends)

friend)
Mean
Std.
Deviation
Early
1452
8.64
4.687
Late
247
8.98
4.319
Early
1452
6.22
5.042
Late
247
6.73
4.775
Early
1452
9.71
3.782
Late
247
9.93
3.858
Early
1452
9.56
4.235
Late
247
9.75
4.165
Early
1452
9.00
4.604
Late
247
9.17
4.509
Early
1404
2.96
2.644
Late
246
3.37
2.614
Early
1327
4.91
4.092
Late
229
5.39
4.054
- 461 -
Sig. (2tailed)
-1.067
.286
-1.464
.143
-.841
.401
-.665
.506
-.517
.605
-1.619
.106
-1.619
.106
Frequency of Drunkenness (Typical month)
Frequency of Drunkenness
(Typical month)

students at DIT)

friends)

friend)
Timeliness
Mean
Std. Deviation
Early
1452
3.23
3.137
Late
247
3.44
2.504
Early
1452
6.13
3.456
Late
247
6.20
3.388
Early
1452
5.05
3.428
Late
247
4.85
3.158
Early
1452
4.56
3.836
Late
247
4.32
3.244
Early
1407
1.06
3.167
Late
245
1.36
3.379
Early
1322
1.94
4.477
Late
228
2.01
3.718
Sig. (2-tailed)
-1.157
.248
-.284
.776
.839
.402
1.041
.299
-1.350
.177
-.221
.825
Perceived Injunctive Norms

Timeliness
Mean
Std. Deviation
Most students at DIT think

that drinking to get drunk is
acceptable
Early
1453
3.89
.872
Late
247
3.94
.863
My close friends would

approve of me drinking to
get drunk
Early
1453
3.79
.973
Late
247
3.81
.914
My best friend would

approve of me drinking to
get drunk
Early
1453
3.72
1.074
Late
247
3.75
1.001
My mother thinks that my

drinking to get drunk is
Early
1363
2.48
.916
Late
231
2.59
.928
Early
1278
2.61
.972
Late
216
2.53
.935
My father thinks that my

drinking to get drunk is
- 462 -
Sig. (2-tailed)
-.949
.343
-.294
.769
-.454
.650
-1.680
.093
1.085
.278
Appendix 7: Assumption Testing

Quantitative models always rest on assumptions about the way things work, regression models
are no exception. As explained in the main document, three sets of regression were performed in
this study one for each dependant variable. In each case, after having settled for a set of key
variables which were both statistically significant and intuitively meaningful from the knowledge
of the problem, the following steps were taken to test the adequacy of the regression
assumptions.
1. The assumption of multicollinearity
Multicollinearity is a condition in which the Independent variables (IVs) are very highly
correlated (.90 or greater) (Tabachnick and Fidell, 1989). It can be caused by high bivariate
correlations. Statistically, multicollinearity is unwanted because calculation of the regression
coefficients is done through matrix inversion. Consequently, if it exists the inversion is unstable.
Logically, multicollinearity is not wanted because if it exists, then the IVs are redundant with
one another. In such a case, one IV doesn't add any predictive value over another IV, but you do
lose a degree of freedom. As such, having multicollinearity can weaken the analysis (Tabachnick
and Fidell, 1989).
In SPSS, requesting collinearity diagnostics as part of standard regression output returns VIF
(Variance inflation factor) and tolerance values. High values of VIF and low tolerance values
indicate serious multicollinearity issues. As a rule of thumb, VIF values close to 10 and tolerance
values close to 0 are indicative of such a problem. The three tables that follow indicate these
values for the full models run for each dependant variable (DV). As can be verified from the
table, there were no multicollinearity issues.
- 463 -
Collinearity Diagnostics for DV: frequency of drinking in a typical month

Block 5
Tolerance
VIF
Gender
0.890
1.124
living in a shared accommodation
0.282
3.541
living with parents
0.276
3.629
0.912
1.097
Age of first drink
0.905
1.104
Drinking Group Size
0.894
1.119
0.603
1.659
Communication about alcohol in general
0.608
1.645
Injunctive norm for most students at DIT*
0.721
1.386
Injunctive norm for close friends*
0.347
2.880
Injunctive norm for best friend *
0.358
2.794
Injunctive norm for mother*
0.509
1.964
Injunctive norm for father*
0.415
2.411
Descriptive norm for most students at DIT*
0.802
1.247
Descriptive norm for close friends*
0.608
1.646
Descriptive norm for best friend *
0.648
1.543
Descriptive norm for mother*
0.582
1.720
Descriptive norm for father*
0.590
1.695
Interaction term for close friends (DN*IN)
0.795
1.258
Interaction term for best friend (DN*IN)
0.808
1.237
IN: Injunctive Norm, DN: Descriptive Norm, * variable centered around the mean
Table 1: Collinearity Diagnostics for DV: frequency of drinking in a typical month
- 464 -
Collinearity Diagnostics for DV: No. of drinks on a typical drinking occasion

Block 5
Tolerance
VIF
Gender
0.758
1.319
0.900
1.111
Age of first drink
0.893
1.120
0.627
1.596
Communication about alcohol in general
0.627
1.594
0.724
1.381
0.342
2.925
0.332
3.016
0.498
2.009
0.500
2.001
0.454
2.203
0.199
5.014
0.231
4.328
Descriptive norm for mother*
0.703
1.421
Descriptive norm for father*
0.666
1.501
0.541
1.849
0.552
1.812
Table 2: Collinearity Diagnostics for DV: No. of drinks on a typical drinking occasion
- 465 -
Collinearity Diagnostics for DV: Frequency of Drunkenness in a typical month

Block 4
Tolerance
VIF
Age
0.83
1.21
Gender
0.89
1.12
0.88
1.14
Age first
0.91
1.10
0.58
1.72
Communication about alcohol in gen
0.61
1.63
0.73
1.38
0.34
2.98
0.35
2.84
0.55
1.82
0.55
1.83
0.70
1.44
0.43
2.31
0.51
1.98
0.65
1.53
0.70
1.43
Table 3: Collinearity Diagnostics for DV: Frequency of Drunkenness in a typical month
2. The assumption of linearity

Standard multiple regression can only accurately estimate the relationship between the dependent
and independent variables if the relationships are linear in nature (Ostrom Jr, 1990). It is
imperative to examine linearity as there are many instances in social sciences where non-linear
relationships occur. If the relationship between the independent variables (IV) and the dependent
variable (DV) is not linear, the results of the regression analysis might under-estimate the true
relationship. This under-estimation carries two risks: an increased chance of a Type II error for
the IV being examined, and in the case of multiple regressions, an increased risk of Type I errors
(over-estimation) for other IVs that share variance with the IV being examined.
- 466 -
The literature suggests three primary ways to detect non-linearity (Berry and Feldman, 1990,
Cohen et al., 1983, Cohen et al., 2007, Pedhazur, 1997). The first method is related to the use of
theory or previous research to inform current analyses. The design of the current study is
theoretically informed by past research and the measures used in this study are validated by
similar studies such as (Beck and Treiman, 1996, McAlaney and McMahon, 2007, Perkins,
2007, Rimal, 2008, Mallett et al., 2009). Therefore the first method lends adequate confidence in
assuming linearity among the variables. However, it is possible that prior researchers might have
overlooked the possibility of non-linear relationships. Therefore, a preferable method of
detecting non-linearity is the examination of residual plots (plots of the standardized residuals as
a function of the standardized predicted values). The residual plots for each DV showing a linear
relationship are presented next.
Figure 1: Residual plot for the frequency of drinking in a typical month
- 467 -
Figure 2: Residual plot for the number of drinks on a typical drinking occasion
Figure 3: Residual plot for the frequency of drunkenness in a typical month
3. The Assumption of Homoscedasticity

Homoscedasticity means that the variance of the errors is the same across all levels of the
independent variables (Ostrom Jr, 1990). When the assumption is violated, it is called
- 468 -
hetroscedasticity. This assumption can be checked by visual examination of a plot of the

standardized residuals (the errors) by the regression standardized predicted values as has been
done in the preceding section. Ideally, residuals are randomly scattered around 0 (the line of total
fit) providing a relatively even distribution. Heteroscedasticity is indicated when the residuals
are not evenly scattered around the line. There are many forms heteroscedasticity can take, such
as a bow-tie or fan shape. When the plot of residuals appears to deviate substantially from
normal, more formal tests for heteroscedasticity should be performed. Possible tests for this are
the Goldfeld-Quandt test when the plot resembles a fan shape or the Glejser tests when the plot
resembles a bowtie (Berry and Feldman, 1990). That is not the case here. From the plots above
one might infer that slight hetroscadasticity is present. However, based on the literature, it is not
large enough to cause major distortion of results (Miles and Shevlin, 2001). The literature
suggests that when hetroscedasticity is present, the parameter estimates are correct i.e. not
biased. It is the standard errors and hence the p values that are incorrect. That being said if there
is no skew in the predicted scores, the p values can be considered to be reasonably accurate
(Miles and Shevlin, 2001). For confirmation the standardized predicted scores were saved as an
output of regression and examined for skewness which was well within the range of +/-2. As a
further step, the plots of standardized residuals with key independent variables (namely the
injunctive and the descriptive norms) were also examined. The plots suggested that they were not
hetroscedastic. According to Berry and Feldman (1990) and Tabachnick and Fidell (1989) slight
heteroscedasticity has little effect on significance tests; however, when heteroscedasticity is
marked it can lead to serious distortion of findings and seriously weaken the analysis thus
increasing the possibility of a Type I error. In fact a number of statisticians have frequently
found ANOVA and the procedures of regression to be robust to the assumptions of homogeneity
- 469 -
of variance (equal variance) and normality (Box, 1953, Lorenzen and Anderson, 1993, Van
Belle, 2008). Allison (1999) quotes
My own experience with heteroscedasticity is that it has to be pretty severe before it leads to
serious bias in the standard errors. Although it is certainly worth checking, I wouldnt get overly
anxious about it (p 128)
4. The Assumption of nonstochastic X
This assumption implies that the error terms should be uncorrelated with the individual
predictors (Ostrom Jr, 1990). This was tested by running correlations of saved standardized
residuals with all the predictors in the model for each dependant variable. The resulting
correlation coefficients were had zero values indicating no correlation.
5. The Assumption of zero mean
This assumption implies that the mean of the error term should be zero (Ostrom Jr, 1990). This is
not a problem because the least squares method of estimating regression equations guarantees
that the mean is zero. This was also verified from the residual descriptive for all DVs.
6. The Assumption of normality
One of the assumptions of regression is that residuals should be normally distributed at each
value of the dependant variable. There is a lack of consensus on how best to decide whether a
variable is normally distributed or not. Statisticians generally use one or more of the following
three criteria
Eye balling the histograms with superimposed normal curves
Examining the values for skewness and kurtosis with predefined acceptable boundaries
Employing formal normality tests

- 470 -
A range of formal tests such as the Shapiro-Wilk test and the Kolmogrov-Smirnov test are
available for testing normality. While these normality tests are useful they are not infallible.
Most of these tests are sensitive to the sample sizes. For large samples, for example 1000
observations or more, most normality tests become very sensitive and might regard a small
deviation from normality as being significant (Pedhazur, 1997). Therefore, as the sample size
increases so does the likelihood of rejecting a distribution that deviates only slightly from
normality. It is to be remembered that true normality is relatively rare in psychology (Micceri,
1989). Researchers also generally agree that the larger the sample size, the less deviations from
normality matter (Miles and Shevlin, 2001).
Therefore, the criterion used in this study to assess the normality of residuals relied upon
examining the histograms and skew and kurtosis values. Some texts suggest that you divide the
skew and kurtosis values by their standard errors to get z-scores and then examine if they fall
within the set boundaries. Unfortunately as the standard errors are directly related to sample size
in large samples most variables will fail these tests even though the variables may not differ from
normality by enough to make any real difference (Harrington, 2008). Given the large sample size
of this study, it was decided to be guided by the histogram and the absolute sizes of the skew and
kurtosis values (within +/-2) based on suggestions by Kendal and Stuart (1967) and Miles and
Shevlin (2001). Absolute values above 2 were decided to likely indicate substantial nonnormality.
- 471 -
DV: Frequency of drinking in a typical month
Figure 4: Histogram of standardized residuals (DV: Frequency of drinking in a typical month)
The histogram indicates that the underlying distribution is fairly normal. As a further check for
normality, P-P plot and Q-Q plot of normality were plotted. They indicated slight deviation from
normality at the upper end. However, the deviation is not huge. These values are not outliers.
These are genuine data and it was not deemed wise to remove it or manipulate it in an
unreasonable manner.
- 472 -
Figure 5: P-P plot and Q-Q plot of standardized residuals (DV: Frequency of drinking in a typical month)
- 473 -
The skewness and kurtosis values of the standardized residuals were 0.3 (Se: 0.064) and 1.8 (SE:
0.127) respectively well within the acceptable range.
In order to further assess this situation and bring the distribution further closer to normality,
square root transformation was applied to the DV. This did not bring a significant change in the
skewness but brought the kurtosis value down to 1.5. The P-P plot now indicates a very normal
distribution
Figure 6: P-P plot of standardized residuals (DV: Square rooted freqency of drinking in a typical month)
There wasnt much difference in the regression results after applying the transformation and
therefore non transformed values are reported in the main document for ease of interpretation.
- 474 -
DV: Number of drinks on a typical occasion

The histogram suggests that the underlying distribution is close to normal.
Figure 7: Histogram of standardized residuals (DV: Number of drinks on a typical occasion)
However as a further check, the normal P-P plot of standardized residuals was examined.
- 475 -
Figure 8: P-P plot of standardized residuals (DV: Number of drinks on a typical drinking occasion)
As can be seen from the plot, it follows the straight line very closely.
Though the histogram and the P-P plot of the standardized residuals looked quite reasonable, the
descriptive values for the residuals were also examined for skewness and kurtosis to make sure
that the distribution did not deviate substantially from a normal curve. The residuals were not
overly skewed (skewness: 0.205, SE= 0.065) however, the kurtosis value slightly exceeded the
set criteria of +/-2 (Kurtosis: 2.07, SE=0.128). In order to resolve this issue, the standardized
residuals were examined for cases beyond +/-4 standard deviations. 4 such cases were found.
These 4 cases were excluded from regression and the model was rerun. Given the same results
(except that R square=72.6 for model 2) these cases were kept excluded for the rest of the
analysis for this dependant variable. The value for kurtosis was now 1.4 (Se.0.128) and
Skewness was 0.09 (SE: 0.064). Both were now within the acceptable range.
- 476 -
DV: Frequency of Drunkenness

Initially, when the standardized residuals were examined, the skewness and kurtosis were noted
to be 0.6 (SE: 0.065) and 3.08 (SE: 0.129). The value for kurtosis exceeds the acceptable range
of +/-2. To rectify this, standardized residuals outside +/-4 standard deviations were excluded.
There were 6 such cases. Model 2 was rerun and it was noticed that removal of these cases
improved the R square value by 2.5%. The reexamination of the standardized residuals revealed
that the skewness and kurtosis values were now 0.3 (Se: 0.065) and 1.6 (SE: 0.129) respectively.
The histogram of standard residuals follow.
Figure 9: Histogram of standardized residuals wihtin +/- 4 st deviations (DV: Frequency of drunkenness in a
typical month)
A square root transformation of the dependant variable was also attempted to bring down the
value of kurtosis further. The values for skew and kurtosis were now -0.4 (SE: 0.065) and 0.5
- 477 -
(Se: 0.130) respectively. Given the same results and negligible increase in R square (0.9%) the
untransformed values for regression coefficients are reported in the main document for ease of
interpretation. The histogram, P-P plot and Q-Q plot for the residuals of the transformed DV are
presented next.
Figure 10: P-P plot, Q-Q plot and histogram of standardized residuals (DV: Square rooted frequency of
drunkenness in a typical month)
- 478 -
References for Appendix 4

ALLISON, P. D. (1999) Multiple regression: A primer, Pine Forge Pr.
BECK, K. H. & TREIMAN, K. A. (1996) The relationship of social context of drinking, perceived social
norms, and parental influence to various drinking patterns of adolescents. Addictive Behaviors,
21(5)633-644.
BERRY, W. D. & FELDMAN, S. (1990) Multiple regression in practice. Quantitative Applications in the
Social Sciences. Sage Publications, Newbury Park.
BOX, G. E. P. (1953) Non-normality and tests on variances. Biometrika, 403/4, 318-335.
COHEN, J., COHEN, P., WEST, S. G. & AIKEN, L. S. (1983) Applied multiple regression/correlation
analysis for the behavioral sciences. Hillsdale. NJ Eribaum.
COHEN, L., MANION, L., MORRISON, K. & MORRISON, K. R. B. (2007) Research methods in
education, Psychology Press.
HARRINGTON, D. (2008) Confirmatory factor analysis, Oxford University Press, USA.
KENDALL, M. G. & STUART, A. (1967) The advanced theory of statistics: inference and relationship,
Griffin.
LORENZEN, T. J. & ANDERSON, V. L. (1993) Design of experiments: a no-name approach, CRC.
MALLETT, K. A., BACHRACH, R. L. & TURRISI, R. (2009) Examining the unique influence of
interpersonal and intrapersonal drinking perceptions on alcohol consumption among college
students. Journal of studies on Alcohol and Drugs, 70(2)2, 178-185.
MCALANEY, J. & MCMAHON, J. (2007) Normative beliefs, misperceptions, and heavy episodic
drinking in a british student sample. Journal of studies on alcohol and drugs, 683, 385-392.
MILES, J. & SHEVLIN, M. (2001) Applying regression & correlation: A guide for students and
researchers, Sage Publications Ltd.
OSTROM JR, C. W. (1990) Time Series Analysis: Regression Techniques. 2d ed. University Paper series
on Quantitative Applications in the Social Sciences, no. 9. Newbury Park, CA: Sage.
PEDHAZUR, E. J. (1997) Multiple regression in behavioral research: Explanation and prediction.
PERKINS, H. W. (2007) Misperceptions of peer drinking norms in Canada: Another look at the "reign of
error" and its consequences among college students. Addictive Behaviors, 32(11)11, 2645-2656.
RIMAL, R. N. (2008) Modeling the Relationship Between Descriptive Norms and Behaviors: A Test and
Extension of the Theory of Normative Social Behavior (TNSB). Health Communication, 232,
103-116.
TABACHNICK, B. G. & FIDELL, L. S. (1989) Using multivariate statistics, Harper & Row (New
York).
VAN BELLE, G. (2008) Statistical rules of thumb. Medicine, 172525-2534.
- 479 -
Appendix 8: Interview Guide

Opening Statement: Now that we have a list of names here, I want you to tell me about these
people. You may start from anywhere you like and take your time. Ill listen first and try not to
interrupt you. I might take some notes
Areas to be Explored
Relationships with network members
1. Formation of Relationship
Cue: Tell me about how you got to know each other
Cue: Can you talk about your experiences with ___
Probes:
Can you tell me more about it?

Can you talk about these common contacts (if ego mentions having common friends)
Can you elaborate with an example?
2. Strength of Relationship (Formation, Duration, Communication, Intensity, Shared

activities, Meaning of important matters and Exchange of support)
Main Cue: Can you talk about your relationship with ____?
Sub cues: Can you talk about how long have you known this person?
Can you talk about how close you feel towards this person? (Probe for examples)
Can you talk about how you communicate?(Probe for examples)
Can you talk about what activities you do together?(probe for examples)
Can you talk about what you mean by the important matters you discuss with __
(Probe to explore if the exchange is reciprocal)
- 480 -
3. Ways of Socializing (Where and how)

Cue: Tell me what you do when you are together
Probe
Can you tell me what usually happens at these occasions?

Can you recall one of these occasions?
Can you tell me where and when it usually happens
Weekends/weekdays, after classes, scheduled/unscheduled, common meeting
places e.g. club meetings, society events etc., birthdays, class events, family events,
others?
- 481 -
Drinking behavior of network members

1. Low, moderate or heavy
Cue: Some people drink a lot, others not so much, what are your thoughts about the drinking
behavior of these people?
Probes:

Can you tell me what you mean by that?
Can you elaborate with an example?
Can you think of an occasion when you felt that happened?
How do you feel is he/she different/similar to you (if ego mentions being similar or different)
2. Development and practice of drinking behaviors

Cue: Tell me about the time when you started drinking with these people.
Probes:
Can you elaborate with an example to help me understand?

Do you recall an occasion when that happened?
Can you describe how you felt at the time?
3. Where, Why and How?

Probes:
Can you talk about where it used to happen? (home, pubs/clubs, at a friends place, sporting
events/concerts, elsewhere)
Tell me about what would usually happen at such an occasion?
Drinking contexts: Parties, class nights out, club or society meetings, sporting events,
general meet ups with friends, family events, special occasions
Can you tell me more about these _______?
4. Approval/Disapproval towards drinking to get drunk

Cue: How do you think she/he feels about drinking to get drunk
Probes:

Can you give an example of an occasion when.
Can you tell me similar episodes when it happened?
- 482 -
5. Normative pressures to drink/feeling motivated or encouraged

Cue: Can you talk about your experiences of drinking with ___
Probes
What happens when you drink together? How often?

In what ways do you think he/she encourages you (ask if ego says he feels encouraged)
How do you feel about it?
Why do you think it happens?
Can you recall an occasion when that happened?
6. Family history of alcoholism, health issues, moral/religious issues
Probe
Can you tell me what happened?

Can you give an example or tell an episode when it happened?
How do you think he/she feels about it?
How do you feel about it?(if the alter is a parent/sibling or someone very close)
- 483 -
Appendix 9: Glossary of Common Terms used in Social Network Analysis
1. Nodes or Actors: Network members that can be distinct individuals (for example clients
of a health service, residents of a neighbourhood), events, memberships or collective
units (for example health organizations within a community).
2. Ties: Linkages between actors within a network. These can be informal (friendships) or
formal (an organization funding another).
3. Ego: The focal individual or node whose personal network is under examination.
4. Alters: The nodes or actors an ego is directly connected to in an ego network.
5. Role Relationships: The relationship that an alter shares with the ego such as friend,
parent, sibling, class mate, flat mate, cousin etc.
6. Multiplexity: When actors have multiple ties. (For example, a friend can be a neighbour
and a class mate).
7. Name Generators: Asking a focal actor for the names of people to whom he or she is
connected in a particular way.
8. Name Interpreters: Questions designed to elicit information about the named actors, their
attributes, relations to focal actor and to other named alters.
9. Graph: Visual representation of actors showing actors as nodes and relational ties as
lines.
10. Sociogram: A drawing that permits visualizing a personal network containing each alters
name and the ties among them.
- 484 -
11. Clique: Maximum number of actors who have all possible ties present among
themselves, such as the network below
12. Liaison/Broker: An actor who lies between a direct between two others such as node B in
the example below
13. Isolate: An actor who is not connected to any other actor in a network such as the green
node in the example below
- 485 -
Appendix 10: Tie Strength Scores
1.
2.
3.
4.
Respondent: Mary (Female, 20 years old, High drinking cohort)

Tie Strength Scores
Alters
6
31, 32, 33, 34, 35, 36
5
37, 38, 312, 313, 315, 317
4
39, 310, 311, 314, 316, 318, 319
3
320, 321, 322, 323, 324, 325, 326, 328
2
327, 329, 330, 331, 332
Respondent: Emma (Female, 20 years old, High drinking cohort)

Tie Strength Scores
Alters
6
41, 43, 44, 414
5
42, 411, 412, 415, 416
4
45, 46, 47, 48, 49, 410
3
413, 418, 419
2
417, 420
Respondent: Helen (Female, 20 years old, High drinking cohort)

Tie Strength Scores
Alters
6
51, 52, 53, 55, 57, 59, 520
5
54, 517, 519
4
56, 58, 510, 511, 514, 515, 516
3
512, 513, 518
Respondent: Ruth (Female, 22 years old, Low drinking cohort)

Tie Strength Scores
Alters
6
71, 72, 73, 74, 75, 76, 710, 711
4
77, 78, 79
- 486 -
5.
6.
7.
8.
Respondent: Linda (Female, 19 years old, High drinking cohort)

Tie Strength Scores
Alters
6
81, 82, 83, 85
5
84, 86, 87, 89
4
88, 812, 813, 816, 819
3
810, 811, 814, 815, 820
2
817
Respondent: Edel (Female, 19 years old, High drinking cohort)

Tie Strength Score
Alters
6
91, 92
5
93, 94, 96, 97
4
95, 98, 99, 910, 911
3
912, 913
Respondent: Sue (Female, 18 years old, High drinking cohort)

Tie Strength Score
Alters
6
1101, 1102, 1107
5
1103
4
1105, 1106
2
1104
Respondent: Pam (Female, 19 years old, High drinking cohort)

Tie Strength Scores
Alters
6
101, 102, 103, 107
5
104, 105, 106, 1014
4
1012, 1016, 1017
3
108, 109, 1011
2
1015
1
1010, 1013
- 487 -
9.
10.
11.
12.
Respondent: Debbie (Female, 21 years old, High drinking cohort)

Tie Strength Scores
Alters
6
1201, 1202, 1203, 1205
5
1204, 1207, 1208
4
1206, 1210
3
1209, 1211, 1212, 1213, 1216
2
1215
1
1214, 1217
Respondent: Meg (Female, 19 years old, Moderate drinking cohort)

Tie Strength Scores
Alters
6
1303, 1308
5
1301, 1302, 1303, 1305, 1312
4
1306, 1307, 1309, 1310, 1311
3
1313, 1314, 1315, 1316, 1317, 1318
2
1319
Respondent: Fay (Female, 21 years old, High drinking cohort)

Tie Strength Scores
Alters
6
1501, 1502, 1503, 1504, 1515, 1517
5
1505, 1506, 1510, 1511, 1512, 1513
4
1507, 1509
3
1508
2
1518, 1519
1
1514, 1516, 1520
Respondent: Bella (Female, 21 years old, High drinking cohort)

Tie Strength Scores
Alters
6
1701, 1703, 1705
5
1702, 1706
4
1704, 1707, 1708, 1709
3
1710
- 488 -
13.
14.
15.
16.
Respondent: Katie (Female, 19 years old, High drinking cohort)

Tie Strength Scores
Alters
6
1801, 1802
5
1803, 1804, 1805, 1809
4
1806, 1807, 1808, 1810, 1811, 1812, 1817
3
1816
2
1814, 1818
1
1813, 1815
Respondent: Amy (Female, 20 years old, Moderate drinking cohort)

Tie Strength Scores
Alters
6
1901, 1902, 1903, 1904
5
1910, 1911, 1912, 1915
4
1905, 1913, 1914, 1918, 1919
3
1906, 1907, 1908, 1909, 1917
2
1916, 1920
Respondent: Lisa (Female, 20 years old, High drinking cohort)

Tie Strength Scores
Alters
6
2601, 2602, 2603, 2604, 2605, 2606, 2607, 2608, 2618
5
2611, 2617, 2619
4
2610, 2612, 2613, 2620
3
2614, 2616, 2621, 2623
2
2615, 2622, 2624
1
2609
Respondent: Tom (Male, 21 years old, High drinking
cohort)
Tie Strength Scores
Alters
6
11, 12, 14, 15, 18
5
13, 111, 115
4
110, 112, 114, 118, 119, 120, 121
3
16, 17, 19, 113, 116, 117, 122, 123
- 489 -
17.
18.
19.
20.
Respondent: Peter (Male, 20 years old, Moderate drinking cohort)

Tie Strength Scores
Alters
6
21, 22, 23
5
210, 213
3
24, 211
2
27, 29
1
25, 26, 28, 212
Respondent: Sam (Male, 23 years old, High drinking cohort)

Tie Strength Scores
Alters
6
61,62, 63, 64, 65, 66, 611, 618
5
67, 68, 69
4
69, 613, 620, 625, 626
3
614, 615, 616, 619, 621, 624
2
610, 622, 623
1
612
Respondent: Alex (Male, 20 years old, Moderate Drinking Cohort)
Tie Strength Scores
Alters
6
1403
5
1404
4
1402
3
1401
2
1405
Respondent: John (Male, 22 years old, High drinking cohort)

Tie Strength
Scores
Alters
1601, 1602, 1603, 1604, 1605, 1609, 1610, 1611,
6
1615, 1622
5
1606, 1607, 1608, 1612, 1619, 1620
4
1613, 1614, 1616, 1617
3
1618, 1621, 1623
- 490 -
21.
22.
23.
24.
Respondent: Adam (Male, 20 years old, Moderate drinking cohort)

Tie Strength Scores
Alters
5
2001, 2003, 2005
4
2002, 2004, 2006, 2008, 2009, 2013, 2014
3
2007, 2012, 2015, 2016, 2017, 2018
2
2010, 2011
Respondent: Brian (Male, 22 years old, High drinking cohort)

Tie Strength Scores
Alters
6
2102
5
2101, 2103, 2106
4
2105, 2107, 2108, 2109
3
2104, 2110, 2111, 2113
2
2112
1
2114
Respondent: Cormac (Male, 21 years old, Moderate drinking Cohort)

Tie Strength Scores
Alters
6
2201, 2202, 2217
5
2203, 2209, 2213
4
2206, 2207, 2208, 2216, 2218, 2219
3
2204, 2205, 2209, 2210, 2220, 2221
2
2212, 2214
1
2211, 2215
Respondent: Gary (Male, 23 years old, High drinking Cohort)

Tie Strength Scores
Alters
6
2301, 2304, 2307
5
2308, 2309
4
2310
3
2302, 2312
2
2306, 2311
1
2303, 2305
- 491 -
25.
26.
Respondent: Rob (Male, 21 years old, Heavy drinking cohort)

Tie Strength Scores
Alters
6
2401, 2402, 2403, 2417
5
2404, 2406, 2407, 2408, 2412
4
2405, 2409, 2410, 2411, 2413, 2415
3
2416
1
2414
Respondent: Ken (Male, 20 years old, Moderate drinking cohort)

Tie Strength Scores
Alters
6
2501
5
2504, 2512, 2513, 2514
4
2502, 2505, 2506
3
2503, 2508, 2510, 2515
2
2507, 2509, 2511, 2517
1
2516
- 492 -

Social Norms Marketing Social Networks and Alcohol Consumption

Uploaded by

Copyright:

Available Formats

Social Norms Marketing Social Networks and Alcohol Consumption

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Social Norms Marketing Social Networks and Alcohol Consumption

Uploaded by

Copyright:

Available Formats

Dublin Institute of Technology

Social Norms Marketing, Social Networks and

Follow this and additional works at: http://arrow.dit.ie/busdoc

Social Norms Marketing, Social Networks and

Engr. Sarah Samdani (MSc.)

School of Marketing, Faculty of Business

Alcohol Action Ireland

Alcohol Beverage Foundation of Ireland

Alcohol Marketing Communications Monitoring Body

Advertising Standards Authority for Ireland

Blood Alcohol Content

Broadcasting Authority of Ireland

Broadcasting Complaints Commission

Broadcasting Commission of Ireland

Central Copy Clearance Ireland

Drinks Industry Group of Ireland

Dublin Institute of Technology

Department of Health and Children

European School Survey Project on Alcohol and Other Drugs

Government Alcohol Advisory Group

Health Service Board

Health Service Executive

Intoxicating Liquor Act

Mature Enjoyment of Alcohol in Society

National Drug Treatment Reporting System

National Drugs Treatment Reporting System

Organization for Economic Cooperation Development

Office of Tobacco Control

Pay Related Social Insurance

Police Using Leading Systems Effectively

Research and Development Corporation

Responsible Retailing of Alcohol in Ireland

Responsible Serving of Alcohol Training

Road Safety Authority

Social Network Analysis

Strategic Task Force on Alcohol

University College Dublin

World Health Organization

The Social Norms (SN) Theory and SN Marketing Interventions .............................. 19

Scope and Limitations of the Study ............................................................................ 20

Thesis Plan ................................................................................................................. 22

Research Context .............................................................................................................. 25

Alcohol Consumption in Ireland ................................................................................ 25

Alcohol Related Health and Social Consequences in Ireland .................................... 26

Irish Policy Interventions ........................................................................................... 28

Literature Review: The Social Norms Approach .......................................................... 30

What are Norms? ........................................................................................................ 31

Descriptive Norms .................................................................................................. 31

Injunctive Norms .................................................................................................... 32

Distinction between Injunctive Norms and Subjective Norms .............................. 33

Characteristics of Norms ........................................................................................ 34

Theory of Social Ecology and Importance of Norms ................................................. 35

Social Influence Research .......................................................................................... 36

Direct Peer Influence .............................................................................................. 37

Indirect Peer Influence ........................................................................................... 38

Normative Beliefs ........................................................................................... 40

Students and Susceptibility of Social Norms .............................................................. 42

Social Norms (SN) Theory .......................................................................................... 43

Causes and Types of Misperceptions ..................................................................... 44

Pluralistic Ignorance (I think they differ, when they dont) ........................... 45