RELATIONSHIP TRANSITIONS
Proceedings of the
Australian Psychology Society’s
Psychology of Relationships Interest Group
4th Annual Conference
Australian Catholic University
Melbourne Australia
th
13 – 14th November 2004
EDITOR
Terry Bowles
ASSISTING EDITOR
Zoë J. Pearce
The Australian Psychological Society
Melbourne Australia
Published by
The Australian Psychological Society
ABN 23 000 543 788
PO Box 38
Flinders PO Melbourne 8009 Australia
Printed by Australian Catholic University Printers
Copyright © 2004 the authors
Title: Proceedings of the Australian Psychology Society’s
Psychology of Relationships Interest Group
4th Annual Conference
Author/Contributor: Bowles, Terry
Date of Publication: 08 November 2004
Format: PB
Size: 200x300
No. of pages: 155
Available from:
The Australian Psychological Society
Level 11
257 Collins Street
Melbourne VIC 3000
Australia
Tel: +61 3 8662 3300
Fax: +61 3 9663 6177
ISBN 0-909881-26-X
Price: AUD$15.00 (Incl. GST)
Proceedings of the
Australian Psychology Society’s
Psychology of Relationships Interest Group
4th Annual Conference
Conference Organising Committee
Barry J. Fallon Australian Catholic University (Chair)
Terry Bowles Australian Catholic University
Brendan Pryor Australian Catholic University
Alexandra West Australian Catholic University
Conference Proceedings
Terry Bowles Australian Catholic University (Editor)
Zoë J. Pearce Griffith University (Assisting Editor)
Psychology of Relationships Interest Group Committee
Terry Bowles
Barry Fallon
Julie Fitness
Kate Moore
Zoe Pearce
Lina Ricciardelli
Neil Welch
Ross Wilkinson
Interest Group Convenor
Interest Group Awards Coordinator
Conference Proceedings Editor
Conference Coordinator
Committee Secretary
Newsletter Editor & Coordinator
Conference Proceedings Assisting Editor
Treasurer
State Activities Coordinator
Website Coordinator
Australian Catholic University
Australian Catholic University
Macquarie University
Deakin University
Griffith University
Deakin University
Flinders University
Australian National University
The papers contained in these proceedings have been subject to a double blind peer-review
process.
Welcome from the Chair of the Conference Committee
One of the highlights for those interested in research and practice to do
with relationships is the APS Interest Group on the Psychology of
Relationships Annual Conference. This is our 4th Conference and
Melbourne is where the Relationships Conference happens.
For our 4th Conference the theme chosen is Relationships Transitions. A
theme such as this is very inclusive—it is difficult to imagine a relationship
which is static and unchanging, and as such all of our relationships are in a
state of Transition. It does so happen that some transitions are more
dramatic and more public than other. One of the outcomes of such a theme
is the considerable range in the topics of the papers being presented at this
year’s Conference. As a result of the reviewing of the abstracts, we have 39
individual paper presentations over the two days. We have designed these
sessions such that there should be time for plenty of discussion—3 papers
in 1.5 hours. In addition to the individual papers we have three special
addresses - Keynote Addresses by Kim Halford and Kate Moore and an
invited address by the recently elected President of the APS—Amanda
Gordon.
In addition to listening to the addresses and the individual papers, it is
hoped that you will enjoy the opportunity to socialise and network with
others who are interested in relationships.
On behalf of the Conference Organising Committee thank you for attending
the Conference. The Committee has done all it can so it is now up to you
the attendees to ensure that the Conference is all that you would want it to
be.
Regards
Barry J Fallon PhD FAPS
Chair, Conference Organising Committee
TABLE OF CONTENTS
The Effect of Advocacy and Adaptive Functioning on Student Selfesteem, Self-efficacy, School Life, Academic Performance and
Attendance
Terry Bowles & Ailsa Brown....................................................................................................7
The Effect of Selection, Training, and Participation on Student
Counsellor Self-description and Coping
Terry Bowles & Carly McKinnis ...........................................................................................14
Client Satisfaction with Relationship Counselling
Adele Carmady, Ann Knowles & Andrew Bickerdike ..........................................................20
Selves in contact: Multiple perspectives on the role of relationships in the
life of the individual with Dissociative Identity
Kymbra Clayton .....................................................................................................................24
The Influence of Interpersonal Relationship on Conflict Resolution Style
John A. Davidson ...................................................................................................................31
Do Adolescent and Parental Perceptions of Interparental Conflict Differ?
Melanie T. Davern, Petra K. Staiger & Ernest S.L. Luk ........................................................35
Theoretical Frameworks for Relationship Transitions and the Predictors
of Successful Transitions
Amanda E. Ferguson (amanda@lifethatworks.com) ..............................................................40
The relationship between love styles and relationship dependent and
independent outcome variables: A within couple analysis
Barry J. Fallon & Terry Bowles .............................................................................................46
Is The Extent Of My Lying And Deception With The One I Love Related
To My Attachment Style?
Barry J Fallon & Christopher Holt .........................................................................................52
Adoption, Attachment, and Relationship Experiences
Judith A. Feeney, Nola L. Passmore & Candida C. Peterson .................................................58
Favouritism and Rejection in Families: Black sheep and Golden-haired
Children
Julie Fitness .............................................................................................................................63
Excessive Internet Use as a Behavioural Addiction
Carina J. Henry & Cecelia Winkelman ..................................................................................68
Bullying Relationships in High Schools
Dorothy Lenthall .....................................................................................................................73
The healthy balance among work, family, and personal relationships: fact
or fiction?
Kathleen A. Moore.................................................................................................................. 79
Melancholic Mothers and their Mourning Infants
Clare O’Bree, Lisa N. Eisen, Zita Marks, Jeannette Milgrom & Kate Neilson...................... 85
The disparate olfactory determination of sex and age in humans,
reflecting the reproductive and relationship characteristics of the species.
Robert N. Paddle & Jacqui Robert.......................................................................................... 91
Parental Decision-Making surrounding Complicated Births: The Impact
on Parent-Infant and Infant-Marital Relationships
Sarah J. Phillips & Gregory A. Tooley .................................................................................. 96
Attachment Security and Intimate Relationship Satisfaction: The
Mediational Role of Maladaptive Attributions
Zoë J. Pearce & W.K. Halford .............................................................................................. 102
The Utility of Work and Family Stressors and Perceived Social Support
in the Prediction of Counselling Utilisation
Andrea Quah & Terry Bowles .............................................................................................. 109
Improving family and peer relationships with children who engage in
problem sexual behaviour
Dr. Petra Staiger, Dr Nicolas Kambouropoulos, Jari Evertsz, Joe Tucci & Janise
Mitchell ................................................................................................................................ 117
The Role of Interactions Between Parents and Children in Social Skills
Development of Prep Children
Gill Terrett & Michelle Puttick............................................................................................. 124
Attraction for Out-Group Members: Is There a Role for Attitude
Importance and Attitude Similarity ?
Gill Terrett, Cecelia Winkelman & Barry Fallon ................................................................. 132
Lying in relationships: Protecting ourselves or protecting our partners?
Alexandra E.West & Barry J.Fallon ..................................................................................... 137
Factors Affecting the Acquisition of Skills in PET (Parent Effectiveness
Training): Language, Gender and Education
Christine D. Wood & John A. Davidson ............................................................................. 141
Social Disconnection and Depression
Neil Welch ............................................................................................................................ 145
Adolescent Psychological Health and School Attitudes: The Impact of
Attachment Relationships
Ross B. Wilkinson & Monika Kraljevic ............................................................................... 150
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 7-13
The Effect of Advocacy and Adaptive Functioning on Student Self-esteem,
Self-efficacy, School Life, Academic Performance and Attendance
Terry Bowles (t.bowles@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, St Patrick’s Campus, Melbourne Vic 3065 Australia
Ailsa Brown (brown.ailsa.m@edumail.vic.gov.au)
Institute of Human Development and Counselling
Monash University, Melbourne Vic 3800 Australia
student compliance in the least restrictive
environment, serve a remedial, preventative function,
and follow an ecological view of student problems as
arising from the environments in which students live
(Rathvon, 1999). Mentoring programs are designed to
promote a young person’s identification with an adult
who possesses and displays a set of positive
characteristics (Roberts & Cotton, 1994). Adolescents
are presumed to acquire the attitudes and
characteristics of the mentor through the process of
observational learning as discussed in social learning
theory (Bandura, 1969; 1990). Improvement in
Academic Performance and enhanced Self-esteem are
the goals of many mentoring programs, however the
limited research on these projects does not indicate a
strong link between mentoring and its desired
outcomes. Roberts and Cotton (1994) found no
significant effects of participation in a mentor
program for African-American high school students
on grade point averages and Self-esteem. Slicker and
Palmer (1993) found that mentoring had no
significant effect on self-concept. Similarly, Royse
(1998) found no effects on Self-esteem, attitudes to
drugs and alcohol, grade point averages, school
absences, or disciplinary infractions when he
evaluated the Brothers Project, a mentoring program
for African-American adolescents. In contrast, Hill
(1993) found a positive effect for mentoring on
communication and self-concept, concluding that the
quality and function of the mentor relationship
contributed significantly to the effectiveness of the
mentoring program. An evaluation of Project RAISE
(McPartland & Nettles, 1991) found mentoring was
associated with a reduction in absence rates and grade
scores.
The advocacy project is a type of support-based
mentoring program that has been introduced into
selected Victorian secondary schools to address the
problems of poor Academic Performance, poor
Attendance, and risk of dropping out of school before
completing the VCE (Ocean & Jones, 2001b). The
advocacy project facilitates supportive relationships
between individual students and teachers, and
encourages appropriate interdependence (Gilligan,
1982). It derives from the ethic of care (Noddings,
1992), which is concerned with students functioning
positively through goal setting, decision making and
having belief in the future (Ocean & Jones, 2001a).
The role of the advocate, who is a teacher, is to be a
friend to help, guide, and support the student. The
recommended frequency of meetings is once per
Abstract
This research examines the influence of a teacher
advocacy program and adaptability on the outcome
measures of Self-esteem, Self-efficacy, Perception of
School Life, Academic Performance, and School
Attendance of a year 10 student group. It was expected
that students receiving the advocacy of teachers (n=39)
who were classified as Adaptors, would have a
significantly higher self-rating on the outcome
measures than participating Stabilizers followed by the
non-participating Adaptors and Stabilizers (n=43) after
the six month intervention. A mixed-between-withinsubjects ANOVA was used to measure the influence of
advocacy and adaptability. None of the expected
interaction effects were found. A main effect showed
Adaptors had higher Self-esteem, Self-efficacy and
School Life scores compared with Stabilizers. The
results are discussed in reference to improving the
advocacy program and analysis of such interventions in
future research.
Social support in the form of mentoring programs is
part of a movement to provide non-classroom
interventions for students (Carter & Sugai, 1989;
Graden, Casey, & Bonstrom, 1985). Mentoring
programs have been shown to have a positive effect
on student performance by improving attendance,
academic results, and reducing acting out behaviour
(Rhodes, Grossman, & Resche, 2000; Roberts &
Cotton, 1994; Royse, 1998). Advocacy programs
based on similar procedures to the mentoring program
have also been effective (Ocean, 2001a; Ocean,
2001b). The problem is no previous Australian
research has investigated the effectiveness of
advocacy programs using a questionnaire based preand post-test method and no previous research has
investigated the adaptability of adolescents involved
in an advocacy program. The aim of the present study
was to extend research into advocacy programs by
evaluating the effectiveness of one program provided
to a year 10 group of students who were defined in
terms of their adaptive functioning. The dimensions
used in the evaluation were the students’ Self-esteem,
Self-efficacy, attitude to school, grade scores in Math
and English, and Attendance.
Advocacy and Mentoring Programs
Mentoring programs have developed from
Intervention Assistance Programs and were designed
to enhance collaborative problem solving, facilitate
7
8
BOWLES & BROWN
fortnight (Ocean, 2001b). It is desirable that students
are able to choose their own advocates so that they are
paired with a person they like while taking account of
gender and cultural issues. Ideally advocates are
volunteers who rate highly on emotional competence
and avoid making insensitive or inappropriate
responses to student concerns. Being a counsellor
(Egan, 1994) is one of the functions of the advocacy
role (Ocean, 2001b).
If mentoring programs influence Academic
Performance, general attitudes, and Attendance it
follows that this will be associated with changes in
Self-efficacy and attitude to school. By extension, it is
expected that Self-esteem would be influenced by
successful
educational
outcomes,
through
participation in a successful mentoring relationship.
1) Students involved in the advocacy program will
have higher scores at the second testing, time two, on
adaptive functioning, Self-esteem, Self-efficacy,
School Life, Academic Performance and improved
Attendance in comparison with the control group.
2) The ACQ factors will form two clusters, a high
cluster of Adaptors and a low cluster of stable
students.
3) The scores on Self-esteem, Self-efficacy, School
Life, Academic Performance and improved
Attendance will be highest for the Adaptors in the
advocacy program, followed by the Stabilizers in the
program and the Adaptors not in the program, with
the lowest scoring expected to be from Stabilizers not
in the advocacy program.
Method
Adaptive Functioning
The aims of advocacy programs of setting goals,
making decisions, having a belief in the future,
remaining committed, and attending to the task
correspond closely with the aims of the theory of goal
setting and task motivation (Locke & Latham, 2002).
In the theory of goal setting and task motivation,
setting goals, capability, commitment, and managing
factors that influence success like Self-efficacy and
receiving feedback facilitate the development of
strategies toward a goal. A comprehensive model
entitled the Adaptive Change Model (ACM) is a
representation of the theory of goal setting and
motivation (Locke & Latham, 2002) and has been
developed to assist individuals to achieve their goals
in therapeutic and educational domains (Bowles,
under review1; under review2). These factors comprise
Openness to Opportunity, Visualization, Planning,
Action and Closure. When summed they are called
the Process factors. Influencing each of these factors
throughout the process of adapting are three Support
factors: Social Support, management of Negative
Emotions, and Inner Drive. When summed they are
called the support factors (Bowles & Fallon, 2003).
Previous research using the Process and Support
factors shows that tertiary students’ Self-esteem was
predicted by Support factor but not the Process factors
(Bowles, Wong, Aranjuez, & Hamid, 2004). Research
has also shown that adults attending therapy are less
adaptive than a non-attending control group, with the
clinical respondents scoring consistently and
significantly lower on the eight scales of adaptive
change (Bowles, under review1).
To date no previous Australian research has
systematically investigated the relationship between
adolescent Self-esteem, Self-efficacy, School Life and
Academic Performance. Further, no research
involving adolescents has explored the link between
adaptive functioning as operationalized by the
Adaptive Change Questionnaire (ACQ), individual
difference, and performance factors. In this research
the general aim is to investigate the influence of
involvement in an advocacy program on students with
different levels of adaptability in reference to
individual difference and performance factors. From
the above research the following hypotheses are
proposed.
Participants
Student participants in the Advocacy group were 39
Year 10 students (20 female and 19 male; mean age,
15.47; SD=0.41) from a Secondary College from
regional Victoria who were randomly selected for
participation in the advocacy program. Participants in
the control group were 43 Year 10 students (23
female and 20 male; mean age, 15.38; SD=.40) from
the same school. The 82 participants included in the
analysis were derived from the 90 students in the
whole year level.
Materials
The Coopersmith Self-Esteem Inventory, school form
(Coopersmith, 1989; SEI) was used in this research.
The SEI is designed to measure evaluative attitudes
toward the self in social, academic, family, and
personal areas of experience. In relation to the SEI,
the term Self-esteem refers to the evaluation a person
makes, and customarily maintains, of him- or herself.
The single factor school form consists of 50 Selfesteem items.
The Multidimensional Scales of Perceived SelfEfficacy (MSPSE) (Bandura, 1990) were used to
measure change in Self-efficacy, which is defined as
an individual’s evaluation of his or her ability to
accomplish a given task or activity (Choi, Fuqua, &
Griffin, 2001). The scale contains 57 items that
measure a range of Self-efficacious themes rated on a
Likert scale from 1 = not at all well to 7 = very well.
The School Life Questionnaire (Ainley, 1986) was
used to measure the respondent’s experience of
school. It consists of 40 statements about the student’s
experience of school to which students are asked to
indicate their level of agreement on a four-point
Likert scale from “definitely agree” to “definitely
disagree”.
Academic Performance was measured as an average
percentage of English results, and an average
percentage of Mathematics results at the end of
semester one. This data was compared with academic
data in school records from December of the previous
year. Attendance was measured as the total number of
days absent in semester one of the year of the
research. This was compared with total number of
days absent in semester two of the year before.
Advocacy and Adaptive Functioning
Procedure
In January of the year of the research, after students
had been allocated to advocacy and non-advocacy
groups, but before they had their first meeting with
their advocate, students completed the Coopersmith
SEI, the School Life Questionnaire, and the
Multidimensional Scales of Perceived Self-efficacy.
The same instruments were employed again in the
July post-testing.
Advocates were twelve volunteer teachers, 8 female
and 4 male. In selection they were judged by a
committee of three peers, as to their warmth,
empathy, the ability not to be judgmental or
authoritarian, and having suitable interpersonal skills
to implement the program in the spirit in which it was
intended. Not all applicants were accepted as
advocates. Advocates undertook a two-hour training
session, including basic counselling skills.
Students were given a list of all twelve advocates
and selected three with whom they would be happy to
work. The coordinator then matched students with one
of the advocates they nominated. Students who felt
that they were not able to work well with their
advocate could ask the coordinator to change
advocates, without giving a reason, within the first six
weeks of the program, but no student took up this
option. Gender was not considered in the matching
9
process, except in so far as the student selected an
advocate of a particular gender.
Advocates were to meet with their students at least
once every fortnight for approximately twenty
minutes. The focus of the meeting was to be agreed
between the advocate and the student. The advocates
were supported by a coordinator and had a reference
support group comprising a psychologist, social
worker, careers counsellor, school nurse, student
welfare coordinator, year level coordinator and the
assistant principal. The advocate recorded the date
and focuses of each meeting with the student and
provided this information to the coordinator. Other
records of meeting content, such as goal-setting or
problem solving, remained confidential between the
advocate and student.
Results
Tests for normality were conducted and considered to
be acceptable. Correlations and descriptive statistics
were obtained for the variables of interest: the ACQ,
SEI, MSPSE, School Life, Mathematics and English
scores and Absence are reported in Table 1. Time one
with time two correlations indicate that the ACQ
factors are weak to moderate and the reliability of the
other variables of interest are typically moderate.
Generally, the ACQ factors share weak to moderate
correlations.
Table 1
Descriptive Statistics and Correlations of Variables in the Analyses
1. Open. to Opp’y
2. Visualization
3. Planning
4. Action
5. Closure
6. Neg. Emotions2
7.Inner Drive
8.Social Support
9. Self-esteem
10. Self-efficacy
11. School life
12. Math score
13. English score
14. Absence
Time 1 Mean
Time 1 SD
Time 2 Mean
Time 2 SD
11
2
3
4
5
6
7
8
9
10
11
12
13
14
343** 624** 58** 42** 61** 46** 70** 57** 26*
51** 29** 12
06 -01
65** 20** 56** 39** 51** 43** 50** 37** 28*
42** 33** 08
-09
00
**
**
**
**
**
*
**
**
*
75
47
41
58
25
66
37
26
47** 37** 14
-01 -05
77
30** 21
27*
47** 33** 16
06 -21
28*
27** 44** 46** 51** 28*
**
**
**
**
**
**
**
**
61
72
22
44
41
67
53
33
58** 36** 15
01 -09
79
31** 29** 25*
38** 34** 61** 40** 28*
49** 40** 27** 22*
21 -10
72** 61** 67** 44** 70** 32** 34** 46** 36** 57** 39** -05
-04 -23*
07 -19
70** 56** 62** 30** 75** 33** 67** 33** 37** 43** 33** 19
17
17
17
16
27*
27*
29** 21** 68** 48** 50** 26** 21 -14
63** 53** 23*
12 -21
62** 55** 64** 33** 67** 41** 64** 61** 22*
49** 42** 49** 06
55** 28*
55** 51** 29** 72** 58** 31** 31** -01
31** 15
19
20
09
29** 49** 45** 05
15
03
15
07
26*
11
04
12
-01
16
32** 00
17
04
08
26*
60** 54** 16
-01
07
-02
-05
03
09
00
-07
-06
02
-02
-19
-06
48**
4.16 3.92 3.96 4.73 4.29 3.83 4.00 4.34 72.48 279.67 113.31 54.59 70.14 10.19
.65
.64
.67
.69
.92
.77
.69
.85 14.86 43.50 11.37 29.39 21.06 7.59
4.20 4.09 3.88 3.78 4.32 3.96 4.08 4.39 73.62 281.86 111.41 57.94 73.30 7.13
.84
.80
.86
.75 1.04
.70
.80
.85 16.59 49.89 14.29 26.47 24.61 5.81
1
Note: Above the diagonal is time one, below is time two. 2 Note: The Negative Emotions scores have been reversed. 3 Note: Diagonals,
underlined, represent the correlation of the same variable at times one and two. 4 Note: Decimals have been removed from the correlations in
the table. * Indicates correlation is significant at the 0.05 level (2-tailed); ** indicates correlation is significant at the 0.01 level (2-tailed);
To investigate the first hypothesis that students
involved in the advocacy program will have higher
scores at time two on adaptive functioning, Selfesteem, Self-efficacy, School Life, Academic
Performance, and improved Attendance in
comparison to the control group, a mixed-betweenwithin-subjects ANOVA (2 levels of intervention by 2
times of testing) was completed. The results indicated
that there was no significant difference between
scores obtained before and after the intervention other
than for Attendance. That is, student Absences
reduced between time one and two (F(1,79) = 10.52,
p = .002, η2 = .12; Table 1) regardless of group
membership. This effect shows a decrease in Absence
from a mean of 10.19 days absent at time one, to 7.13
days absent at time 2.
BOWLES & BROWN
Table 2
Means, Standard Deviations, and Effects of Cluster
Structure on Factors of the Adaptive Change
Adaptors
Mean
(SD)
Opportunity
4.75
(.49)
Visualization
4.45
(.51)
Planning
4.51
(.67)
Action
4.15
(.67)
Closure
5.14
(.53)
Neg. Emotions
4.33
(.79)
Inner Drive
4.59
(.58)
Social Support
4.89
(.60)
Stabilizers
Mean
(SD)
F
3.82
71.15
(.47)
3.61
54.98
(.48)
3.64
51.70
(.42)
3.48
22.54
(.58)
3.79
81.09
(.71)
3.54
26.94
(.58)
3.66
59.96
(.49)
4.03
54.98
(.81)
.001
η2
.47
.001
.41
.001
.39
.001
.22
.001
.50
.001
.25
.001
.43
.001
.24
The relationship between the clusters of Adaptors (n
= 30) and Stabilizers (n = 52) is shown in Figure 1 in
reference to pre- and post-test scores. These results
illustrate in greater depth the magnitude of the
difference shown in Table 2. There was no difference
in the scores obtained for the Adaptors or the
Stabilizers as a result of involvement in the advocacy
program. Although not significant there was one
noteworthy pattern that emerged from this analysis.
The Stabilizers (darker two columns) increased their
scores for each of the eight ACQ factors, whereas the
Adaptors (lighter two columns) were less adaptive at
time two compared with time one for each of the
ACQ factors.
T1 Adaptors
T2 Adaptors
T1 Stabilizers
T2 Stabilizers
6
5
4
3
2
n
eg
C
lo
at
su
iv
e
Em re
ot
io
ns
In
ne
rD
So
riv
ci
al
e
Su
pp
or
t
ct
io
A
N
Vi
s
or
tu
m
ity
ua
liz
at
io
n
Pl
an
ni
ng
1
pp
To test the second hypothesis that the student’s
ACQ factors scores could be clustered to form two
groups of students, a k-means cluster procedure was
applied. The eight ACQ factors scores from time one,
with centers derived from the data, were loaded into
the analysis. Inspection of the dendograms and the
agglomeration schedule revealed that the best solution
contained two clusters. These two clusters are defined
as Adaptors who rate each of the ACQ variables
significantly higher than do the Stabilizers. Table 2
shows the mean difference between the groups on
each ACQ factors indicating that the Adaptors scored
significantly and consistently higher for each of the
ACQ factors. Adaptors are defined as being agentic,
preferring to change strategically and using their
supports to change. By comparison, the Stabilizers are
defined as preferring not to change and remain in a
more stable state.
O
10
Figure 1
Adaptors and Stabilizers scores on the ACQ factors
at time one and two
As a validity check to establish that the random
distribution of the students into each group had
resulted in a nonsystematic classification of students
on the basis of Adaptability clusters, a chi-square
analysis was completed. The frequency of students in
each of the cluster by group cells is shown in Table 3.
The result of a chi-square analysis indicated that there
was no systematic association between advocacy
group membership and membership of cluster of
adaptability at time one (χ2 (1,82) = 113, p =.457; η =
0.04).
Table 3
Frequency of Students in Advocacy Groups and
Adaptive Clusters
Group
Non-advocacy Advocacy
Total
Cluster
(SD)
(SD)
Stabilizers
28
24
Expected
27.3
24.7
Adaptors
15
15
Expected
15.7
14.3
Total
43
39
52
52
30
30
82
The analysis of the mean scores shown in Table 4
indicates that there was consistently little difference in
Self-esteem, Self-efficacy, School Life, Math,
English, and Attendance between time one and time
two scores (the within analysis). There was greater
variability in the differences between the Adaptors
and Stabilizers (the between clusters analysis) with
the Adaptors scoring consistently higher on Selfesteem, Self-efficacy, and School Life, at both time
one and time two. There was no difference between
Adaptors and Stabilizers, at either time one or time
two, on Math or English scores or on the amount of
Absence from school.
Advocacy and Adaptive Functioning
The final test of whether the Adaptors or the
Stabilizers responded differently to involvement in the
advocacy intervention was tested through a series of
mixed-between-within-subjects
ANOVAs
(six
individual difference and performance variables by
two times of testing by 2 ACQ clusters) were
completed. The results are in Table 4.
Table 4
Means, Standard Deviations, and Effects of FOCQ
Cluster Structure on Self-esteem, Self-efficacy, School
Life, Academic Performance and Attendance
Time
1. Self-esteem
2. Self-esteem
1. Self-efficacy
2. Self-efficacy
1. School life
2. School life
1. Math
2. Math
1. English
2. English
1. Absence
2. Absence
Adaptors
Mean
(SD)
79.47
(14.40)
81.77
(13.27)
307.84
(32.12)
312.65
(35.48)
118.74
(9.89)
118.04
(10.98)
57.33
(28.12)
61.37
(27.93)
70.38
(18.80)
77.23
(21.69)
8.90
(5.93)
7.17
(5.31)
Stabilizers
Mean
(SD)
F
68.46
11.83
(13.21)
68.92
13.12
(16.59)
263.12
25.93
(41.03)
264.09
22.89
(48.56)
107.58
12.27
(14.66)
107.58
11.51
(14.66)
53.00
0.41
(30.27)
55.96
0.79
(25.65)
69.70
0.02
(24.51)
71.04
0.18
(26.07)
10.94
1.38
(8.36)
7.09
0.01
(6.07)
p
.001
η2
.12
.001
.13
.001
.24
.001
.21
.001
.13
.001
.12
.524
.01
.376
.02
.888
.01
.668
.01
.243
.02
.995
.01
Discussion
In conclusion, the analysis investigating the influence
of the advocacy program showed that there was no
significant effect on student Self-esteem, Selfefficacy, School Life, and Academic Performance.
There was an improvement for Attendance but as this
was not involving an interaction with group
membership, it does not show an effect for the
program. The second hypothesis that two clusters
based on the factors of the ACQ would emerge from
the adolescent’s rating of their adaptability was found.
A cluster analysis showed a more adaptive group
called Adaptors and a group less likely to adaptively
change, the Stabilizers, emerged. The third hypothesis
linking the influence of the advocacy program and
adaptive change showed that there was no relationship
between the two factors on student Self-esteem, Selfefficacy, School Life, Academic Performance or
improvement in Attendance. By contrast, there were
differences in Self-esteem, Self-efficacy, and School
Life with the Adaptors consistently scoring higher
than the Stabilizers on Self-esteem, Self-efficacy, and
School Life, however there was no relationship
between Mathematics scores, English scores or
Absence.
These findings are in line with previous research
that shows little effect between mentoring and Self-
11
esteem (Slicker & Palmer, 1993), grade scores, and
absences (Roberts & Cotton, 1994; Royce, 1998) but
are contrary to the findings of Hill (1993). These
findings raise a number of questions related to
previous conclusions about advocacy and mentoring
programs. In particular, what explains the absence of
the anticipated influence of involvement in the
advocacy program? In short, the best explanation is
likely to be the result of method effect, content dealt
with in the program, frequency versus optimal
availability, and finally, a school effect. Firstly, this
research was a quasi-experimental design with
random assignment of students, which is very
different to previous Australian research, such as
Ocean’s (2001a; 2001b). The presumption that
adolescents acquire the attitudes and characteristics of
the mentor through the process of observational
learning as discussed in social learning theory
(Bandura, 1969; 1990), was not shown. Similarly,
improved Self-concept was not shown as a result of
involvement in the advocacy program (Hill, 1993).
The approach to mentoring through observational
learning is likely to be influential when the
association requires common activity and behaviour,
such as in Big Brother/Big Sister Programs over
prolonged periods of contact. The advocacy program
is unlike such programs in that the relationship
between the Advocate and student is more likely to be
characterized as student-directed support based,
solution-focused or student- (client-) centred
counselling, which may be more remedial of students
experiencing stress or difficulty. The outcome of such
interventions is more likely to result in alleviation of
the specific problem and associated stress. Measuring
this would require a different set of outcome measures
more relevant to the content of the discussions
between advocate and student.
Two previously unconsidered explanations of the
absence of effect are to do with the frequency of
meetings and general effect of the program on the
school. The frequency of the meetings has been noted
as an important contributing factor. It was expected
that the meetings would be once per fortnight (Ocean,
2001b) however, such regularity may not correspond
to the needs of the student or the impact of the
distressing event or issues requiring attention. Finally,
the general effect of the program on the school may
neutralize the effect of a specific advocate. That is,
the existence of the program may generalize
awareness and practice across the whole staff
coincidentally, which extends the effectiveness of all
staff and correspondingly diminishes the effect of the
specific advocate on the student. This makes
measuring the effect of a small number of advocates
problematic.
The differences in the student cohort attending year
10 in one secondary college was sufficient to generate
definition of two types of students, Adaptors and
Stabilizers. These findings show that the ACQ
meaningfully defines adolescent behaviour related to
Self-esteem, Self-efficacy and School Life, which
provides evidence of the potential effect of adaptive
functioning and support factors. These findings also
indicate the relevance of the theory of goal setting and
12
BOWLES & BROWN
task motivation in schools, in line with Locke and
Latham (2002). The level of adaptability of the
students did not have a systematic influence on the
students Mathematics or English scores or their
Attendance. The difference between these two sets of
dependent variables is that the former set is associated
with the self in context and the latter set is an outcome
and behaviour specific and related directly to school
performance. The test, retest correlations showed that
there was a low to moderate relationship indicating
that the factors of the ACQ are not highly stable. The
ACQ is therefore likely to be able to be readily useful
in therapy and educational settings if used as a model
of direct intervention that the participant may learn to
manipulate for themselves.
There are a number of limitations of this study.
Firstly, the sample was limited to a year 10 group in a
regional college, so generalizing from this research is
questionable. The number of respondents in the
research reinforces this caution. The use of cluster
procedure is advantageous to identification of
typologies but is based on the group definition, which
may vary from sample to sample, making generalizing
from this research more questionable. Finally, the
test- retest procedure measured the effect of the
program immediately before and after the semester of
the program. There may have been a residual or
developmental effect in which the effect may have
become apparent in the subsequent and further
semesters.
Future research needs to attend to refining the aims
and processes involved in the advocacy program.
Introduction of specifically targeted behaviours would
assist measuring effectiveness. Care would need to be
taken not to reduce the informal nature of the program
in so doing. Revision of the assumed elements of the
program may also assist improvement of the program
For example, more training of staff, limiting the aims
of the intervention, clearer record keeping of aims and
outcomes of students and greater emphasis on
changing behaviour as well as focusing on support
for, and buoyancy of, the student.
In reference to the ACQ, there is a need to replicate
this research and develop a clearer understanding of
the impact of informal use of, and direct teaching of
the eight factors of the ACQ in an attempt to improve
Self-esteem, Self-efficacy, and School Life.
Attempting to improve Mathematics and English
grade scores and Attendance from school is more
likely to eventuate through direct intervention based
on teaching the ACQ model.
In conclusion, this research provides insight into
new ways of conceptualizing student support and
measuring effectiveness of student support programs.
It has assisted in refining the understanding of the
effect of advocacy and the ongoing desire to improve
the programs. It has differentiated the advocacy
program from other mentoring programs and shown
that specific forms of measurement are more likely to
be advantageous in further refinement of such
programs. The research has also demonstrated that the
ACQ has merit and promise as a means of facilitating
the self and may be useful through direct intervention
in enhancing student academic management and selfmanagement in future.
References
Ainley, J. (1986). The ACER School Life
Questionnaire:
Secondary
school
version.
Hawthorn, Vic: ACER.
Bandura, A. (1969). Social learning theory of
identification process. In D. A. Gaslin (Ed.),
Handbook of socialization theory and research (pp.
214 - 262). Chicago, IL: Rand McNally.
Bandura, A. (1990). Multidimensional scales of
perceived self-efficacy. Stanford, CA: Stanford
University.
Bowles, T., Wong, Y. Y., Aranjuez, E., & Hamid, P.
(2004). Investigation of a model explaining the
retention of international students, part I: A crosscultural comparison of adaptive change, selfefficacy,
procrastination
and
self-esteem.
Proceedings of the 2004 Hawaii International
Conference on Education. Hawaii, USA.
Bowles, T., & Fallon, B. (2003). Validation of the
Adaptive Change Questionnaire and definition of
adaptive and maladaptive functioning of a clinical
and nonclinical group. Proceedings of the 2003
Hawaii International Conference on Social
Sciences. Hawaii, USA.
Bowles, T. (Under review1). Development of the
adaptive change model: Comparison of clinical and
nonclinical adult respondents.
Bowles, T. (Under review2). The stability of the
Adaptive Change Questionnaire and advancement
of the transtheoretical and transdomain relevant
model of adaptive change.
Carter, J., & Sugai, G. (1989). Survey on prereferral
practices: Responses from state departments of
education. Exceptional Children, 55, 298-302.
Choi, N., Fuqua, D. R., & Griffin, B. W. (2001).
Exploratory analysis of the structure of scores from
the multidimensional scales of perceived selfefficacy.
Educational
and
Psychological
Measurement, 61, 475 - 489.
Coopersmith, S. (1989). Self-Esteem Inventories. Palo
Alto, CA: Consulting Psychologists Press, Inc.
Egan, G. (1994). The skilled helper: A problem
management approach to helping. Pacific Grove,
CA: Brooks Cole.
Gilligan, C. (1982). In a different voice:
Psychological theory and women's development.
Cambridge, Mass.: Harvard University Press.
Graden, J., Casey, A., & Bonstrom, O. (1985).
Implementing a preferred intervention system: Part
II. The data. Exceptional Children, 51, 781-791.
Hill, R. L. (1993). An investigation of the effects of
mentor
relationships
on
motivation,
communication, and self-concept of secondary
students
at-risk.
Dissertation
Abstracts
International, 54(4-A), 1240.
Locke, E. & Latham, G. (2002). Building a practically
useful theory of goal setting and task motivation.
American Psychologist, 57, 705-717.
McPartland, J. A., & Nettles, S. M. (1991). Using
community adults as advocates or mentors for atrisk middle school students: A two-year evaluation
Advocacy and Adaptive Functioning
of Project RAISE. American Journal of Education,
99, 568-586.
Noddings, N. (1992). The challenge to care in
schools: An interactive approach to education. New
York: Teachers College Press.
Ocean, J. (2001a). Advocacy 2000: A visual executive
summary (Executive summary). Melbourne:
Department of Education, Employment and
Training.
Ocean, J. (2001b). Advocacy 2000: Evaluation report.
Melbourne: Department of Education, Employment
and Training, Victoria.
Ocean, J., & Jones, L. (2001). Corio Bay senior
college advocacy program manual 2001. Geelong:
Corio Bay Senior College.
Rathvon, N. (1999). Effective school interventions:
Strategies for enhancing academic achievement and
social competence. New York: The Guilford Press.
Rhodes, J. E., Grossman, J. B., & Resche, N. L.
(2000). Agents of change: Pathways through which
mentoring relationships influence adolescents'
academic adjustment. Child Development, 71,
1662-1671.
Roberts, A., & Cotton, L. (1994). Note on assessing a
mentor program. Psychological Reports, 75, 13691370.
Royse, D. (1998). Mentoring high-risk minority
youth: Evaluation of the Brothers project.
Adolescence, 33, 145-158.
Slicker, E. K., & Palmer, D. J. (1993). Mentoring atrisk high school students: Evaluation of a schoolbased program. The School Counselor, 40, 327-334.
Correspondence to:
Terry Bowles
School of Psychology
Australian Catholic University, St Patrick’s Campus,
115 Victoria Pd
Melbourne Vic 3065 Australia
t.bowles@patrick.acu.edu.au
13
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 14-19
The Effect of Selection, Training, and Participation on Student Counsellor
Self-description and Coping
Terry Bowles (t.bowles@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, St Patrick’s Campus, Melbourne Vic 3065 Australia
Carly McKinnis (carly_mckinnis@health.qld.gov.au)
Institute of Human Development and Counselling
Monash University, Melbourne Vic 3800 Australia
students live (Rathvon, 1999). Despite the variability
of peer support and mentoring programs all such
programs are based on, or conform to, the theory of
social support which suggests that social support in
the form of significant others, peers, or family
members, mitigates the impact of adversity of
stressful life events (Barrera, Chassin, & Rogosch,
1993; Cohen & Wills, 1985; Newcomb, & Bentler,
1988).
Some mentoring programs are designed with adults
as mentors (Roberts & Cotton, 1994). When the
mentoring is provided by students it has been defined
as, “having two or more individuals willingly form
mutually respectful, trusting relationships focused on
goals that meet the needs and focus the potential of
the mentee, while considering the needs of the mentor
and the context in which they both function” (Kochan,
2002, p. p. 284). A subset of such mentoring
programs are peer support programs and a subset of
these are peer mediation programs. The former are to
generally assist students in a mentee-centered fashion.
The latter are more specifically designed to address
the growing problem of disruptive and violent student
behavior and to provide effective preventative
programs to ensure school safety and increase
appropriate student social interaction (Smith, Daunic,
Miller, & Robinson, 2002).
The evidence suggests that peer support programs
are effective in assisting students (Forouzesh, Grant,
& Donnelly, 2001). Peer mediators describe the
process as effective in assisting changes to the
behavior of students who seek help (Gerber & TerryDay, 1999). Research into peer counselling programs
has shown that helpers also experience positive
outcomes. Helping others has resulted in positive
changes
to
self-esteem,
relationships,
and
communication skills for helpers (Douglas, 1998).
Student participants and student leaders, mentors, and
mentees consistently articulate a range of benefits of
peer programs (Forouzesh, Grant, & Donnelly, 2001).
However, two important questions have yet to be
sufficiently answered related to mentor behavior.
Firstly, are student counsellors different from, or
similar to, the students that they seek to assist?
Secondly, do students who act as mentors or
counsellors gain from the training for, and experience
of mentoring?
Results from a very large national survey of
American student counsellors provided some insight
into these questions. Eighty-five percent of the 862
students surveyed indicated that participation in the
program had helped them at school and 72% indicated
Abstract
The aim of this research is to examine the effects of the
selection and training of student counsellors. Previous
research suggests that enhancement of self-esteem,
relationships, and communication skills result from
participation as a peer counsellor. In this research, peer
counsellors (n = 20) were selected and matched with
same age and sex peers (n = 24) from the same school.
All participating students completed a pre-test and
post-test, seven months later. It was expected that the
counsellors would have higher scores on the selfdescription questionnaire and coping scales at time one
due to the effect of the selection. It was also expected
that the self-description and coping scores would
increase due to peer support training and experience in
relation with the comparison group. Between-subjects
analysis at time one showed effects for a range of the
self-description questionnaire and coping scales.
Mixed-between-within-subjects ANOVA showed no
increase for the counsellors or the comparison group
over time. The implications will be discussed in
reference to the peer counsellors, their training, and
improvements to the program.
Counselling is widely accepted as a successful tool for
helping individuals in their journey through life.
Research suggests that while access to professional
counsellors has increased, adolescents in particular
are still more likely to turn to peers for help before
seeking assistance from adult counsellors (Fallon &
Bowles, 1999; Rickwood, 1992). This has challenged
professionals working with adolescents to diversify
support services, provide alternatives, and to
encourage increased access to professionals.
Consequently, a number of practical programs have
been developed and implemented, particularly within
the educational setting, to provide support by utilizing
the resources of adolescents (Douglas, 1998). Such
programs have a long history informed by many
influences. The purpose of current student support
programs range from facilitating harmonious student
relations and reduction of school related violence, to
replacement of the more formal therapeutic
counselling and tuition of students with difficulties
learning. Such programs have their philosophical and
practical origins in early mentoring programs which
evolved from Intervention Assistance Programs that
were designed to enhance collaborative problem
solving, facilitate student compliance in the least
restrictive environment, serve a remedial, preventative
function, and follow an ecological view of student’s
problems arising from the environments in which
14
Student Counsellor Self-Description
that participation had helped them at home
(Forouzesh, Grant, & Donnelly, 2001). Specifically,
participants reported improvements in friendships,
class attendance, contribution to, and connection to
school, communication with family and teachers,
conflict resolution and decision-making. The students
also reported a resistance to peer pressure, and greater
propensity to refuse drugs, cigarettes, and alcohol as a
result of involvement in the program. Such evidence
suggests that involvement in peer programs is
favorable.
By logical extension counselors would be expected
to have different attributes of coping skills and selfperception compared with other students, prior to
participating in the training to become a counselor. It
would also be logical to expect that the student
counsellors would increase their coping capabilities
and strengthen their self-perception as a result of
participation in the training program. Thus, the
research will also examine the effects of counselling
on the student’s coping and self-description over time.
As yet there has been no known quantitative
research using psychologically designed measures of
coping and self-description into the experience of the
student counsellor. This research will examine
differences between students chosen for inclusion in a
peer-counselling program as mentors, and students
not participating in the program as counsellors. The
research will also focus on the effects on coping and
self-description for students who participate in a peercounselling program.
In summary, this research investigated the attributes
of student counsellors at entry to participation in a
program based on the Kids Help Line "Peer Skills"
(Kids Help Line Program Peer Skills, 2003) prior to
training. The focus of this research is not on the
effectiveness of peer mentoring counselling but on the
student counsellors. The absence of research into the
experience of, and direct and coincidental benefit of,
participation in student counsellor programs is the
problem being addressed in this research. Particularly,
the absence of fine grained information about the
coping strategies used and the self-description of
students trained as peer counsellors from the
Australian context is the problem under consideration
in this research.
The first general aim of this research is to establish
whether the entry process into student counselling
programs identifies students with particular
characteristics appropriate to the task demands. That
is, the Counselling and Training Group will have
higher positive and lower negative frequencies of
coping and rate factors defining their self-description
higher. The second aim is to investigate the gain in
coping and self-description for the Counselling and
Training Group, as a function of participation in the
training program. Therefore it is expected that the
Counselling Group will have significantly more
frequent positive coping and less frequent negative
coping and rate their self-description higher at time
two in comparison to the Comparison Group who are
not expected to alter their frequency scores of coping
and ratings of self-description from time one to time
two.
15
Method
Participants
Forty-four secondary students (aged 12-17) at a
private, single sex, male school in Melbourne,
Australia, volunteered to participate in the research.
Twenty of the students volunteered to be trained as
peer counsellors (Counselling and Training Group)
and completed the questionnaires. A further 24
students participated in the research as a comparison
group and received no training but completed the
questionnaires.
Students included in the Counselling and Training
Group were selected in a two-stage process. The first
stage required the students to nominate themselves or
be nominated by other students to participate in the
peer counselling training. During the second stage of
the selection process the school counsellor and a
teacher interviewed the nominated students. Students
included in the Comparison Group volunteered to
participate in this research after responding to
ageneral request to students. The comparison group
was matched on the basis of student sex and year
level. Below is a table (Table 1) summarizing the data
included for analysis in this research project.
Table 1
Distribution of Participants by Age and Comparison
Group
Year Level
CT1
C
Total
8
9
10
11
Total
9
6
4
5
20
7
3
4
5
24
16
9
8
11
44
Note: 1 CT represents the Counselling and TrainingGroup; C
represents the Comparison Group.
Materials
Two psychometrically sound questionnaires defining
psychological constructs were used in this research.
The two questionnaires were the Adolescent Coping
Scale (ACS; Frydenberg & Lewis, 1993) and the SelfDescription Questionnaire II (SDQ-II; Marsh, 1990).
The ACS is an 80-item self-report inventory that
assesses 18 different coping strategies. Scores on the
scales of the ACS were standardized in line with the
instructions of the authors (Frydenberg & Lewis,
1993). Each item reflects a different coping response.
All scales are reliable with alphas ranging from 0.54
to 0.85 (Frydenberg, 1994; 1999) with a median of
0.70. For a comprehensive description of the coping
scales development, see Frydenberg and Lewis
(1993).
The SDQ-II (Marsh, 1990) is a 102 item self-report
inventory that assesses three areas of academic selfconcept, seven areas of non-academic self-concept,
and general self-concept. The coefficient alpha
estimate of reliability for each SDQ-II scale is
generally high, varying from .83 to .91 (median = .87)
whereas the average correlation among the factors is
modest (mean r = .18) (Marsh, 1990). The SDQ-II
BOWLES & MCKINNIS
16
measures distinct facets of self-concept, underscoring
the utility of this questionnaire usefulness as both a
clinical and research instrument (Marsh, 1990). For
further information on the reliability and validity of
the SDQ-II see Marsh, Relich, and Smith (1983),
Marsh, Parker, and Barnes (1985), and Marsh and
Hocevar (1985). For a comprehensive description of
the SDQ-II development, scale and items see Marsh
(1990).
Design and Procedure
Testing occurred in three separate sessions. Two tests
at the beginning of the first school term and one at the
beginning of the fourth school term were completed.
The pre-testing of the counselors was completed on
the first day of the peer counselling training. The pretesting for the Comparison Group was held separately
during school lunchtime. Post-testing was concurrent
for the Counselling and Training and Comparison
Groups at the beginning of term four. On all occasions
of testing students were given verbal instructions in
completing the questionnaires and the researcher and
the school counsellor were present to answer
questions during testing.
Table 2
Name, Means and Standard Deviations of the 18
Coping Scales Used
Time 1
Scale
CT1 C
Seek Social Support
59.00 55.33
Focus-Solving the Problem 67.00 65.17
Work Hard and Achieve 72.20 74.33
Worry
50.40 52.83
Invest in Friends
59.60 67.33
Seek to Belong
55.60 65.17
Wishful Thinking
49.40 65.00
Not Cope
36.20 48.33
Tension Reduction
49.00 44.64
Social Action
33.75 43.33
Ignore the Problem
45.50 56.04
Self-Blame
47.25 9.17
Keep to Self
48.00 65.20
Seek Spiritual Support
32.00 42.50
Focus on the Positive
59.25 66.88
Seek Professional Help
40.00 42.92
Seek Relaxing Diversions 78.05 82.54
Physical Recreation
72.80 80.20
Time 2
CT C
53.80 47.50
63.40 61.33
73.00 71.17
48.80 50.17
57.40 63.00
52.60 58.83
50.20 63.50
34.20 45.83
43.33 41.00
33.50 35.63
44.75 53.13
43.50 53.75
51.50 63.13
33.00 38.33
58.50 62.50
39.50 41.87
81.90 82.54
73.15 76.13
Note.1 CT represents the Counselling and Training Group; C
represents the Comparison Group.
Training
The peer counselling program run by the school
counsellor was based on Kids help Line program
"Peer Skills" (Kids Help Line Program Peer Skills,
2003). In line with the program the training is
designed to enhance participant’s natural helping
skills by including consideration of values and
attitudes; listening and responding skills, problem
solving skills, self-care and information on the local
health resource network, referral processes and
understanding your limits.
Results
Four of the ASC scores failed to meet standard tests
of normality and were transformed. Analyses on both
the transformed and untransformed data provided
similar results so the untransformed item scores were
used in the analyses. The means and standard
deviations of the scale scores of the ACS at times one
and two are represented in Table 2.
To test whether the expected difference between the
Counsellor and Training Group and Comparison
Group differed on scores of coping and selfdescription, a series of between-subjects analyses at
time one were completed. The ANOVA tests showed
significant effects wherein six of the 18 variables
measuring coping were significantly different
between groups. That is, the Counselling and Training
Group was significantly lower rating on Seeking to
Belong (F(1,42) = 5.59, p = .023, η2 = .12); Wishful
Thinking (F(1,42) = 10.11, p = .003, η2 = .19); Not
Coping (F(1,42) = 7.19, p = .01, η2 = .15); Social
Action (F(1,42) = 4.44, p = .041, η2 = .10); Ignore the
Problem (F(1,42) = 6.58, p = .014, η2 = .14); and,
Keep to Self (F(1,42) = 15.09, p = .001, η2 = .26).
Figure 1 shows the standard scores of means for the
six scales and indicates that the difference between
groups was consistent across the six scales. The three
coping strategies used most frequently by both groups
were related to Belonging, Wishful Thinking, and
Keeping to Self.
Counselling Group
Comparison Group
70
60
50
40
30
20
10
0
Belonging Wishful Not
Thinking Cope
Social Ignore Keep to
Action Problem Self
Figure 1
Standardized coping scores for the counselling and
training group with the comparison groups at time
one
Similar analyses were completed using the eleven
scales measuring adolescent self-description. The
scales were screened and all scales met criteria for
normality and adequacy for the paired analyses. The
means and standard deviations of the scale scores of
the SDQ-II scales at times one and two, are
represented in Table 3.
Student Counsellor Self-Description
Table 3
Name, Means, and Standard Deviations of the Selfdescription Scales
Scale
Math
Physical Appearance
General Self
Honesty-Trustworthiness
Physical Abilities
Verbal
Emotional Stability
Parent Relations
General School
Same-Sex Relations
Opposite Sex Relations
Time 1
CT1 C
38.40 39.63
35.77 33.63
51.60 47.54
47.74 40.04
39.80 37.58
48.90 41.13
40.95 38.42
41.00 35.33
48.52 44.95
51.37 50.54
38.85 36.21
Time 2
CT
C
37.40 42.21
35.94 32.12
52.55 45.54
50.15 43.29
39.00 35.58
47.05 42.00
43.50 39.55
42.00 36.13
49.10 44.38
51.40 46.70
37.65 33.54
Note. 1 CT represents the Counselling and Training Group; C
represents the Comparison Group.
Of the eleven scales, four were found to vary
significantly between groups at time 1 (Figure 2). The
Counselling and Training Group was significantly
higher scoring than the Comparison Group on General
Self (F(1,42) = 5.00, p = .031, η2 = .10); HonestyTrustworthiness (F(1,42) = 6.46, p = .015, η2 = .14);
Verbal (F(1,42) = 7.53, p = .009, η2 = .15); and,
Parent Relations (F(1,42) = 4.88, p = .033, η2 = .10).
To test whether the ACS scores varied as a function
of the training and experience as a student counsellor,
a mixed-between-within-subject ANOVA was
completed using the 18 ACS scores. There was no
effect involving time one and time two score for
either the Counselling and Training or Comparison
Groups. Similarly there were no interactions
involving time and between group effects.
Counselling Group
Comparison Group
50
40
30
20
10
0
HonestyTrustworthiness
Verbal
Parent
General School
Relations
Figure 2
Self-description scores for the counselling and
training group with the comparison groups at time
one
To establish whether the SDQ-II scale scores varied
as a function of the training and experience as a
student counsellor, a mixed-between-within-subjects
ANOVA was completed using the 11 SDQ-II scales.
Effects were found in line with the tabulated findings
(Table 3). Like the ACS there was no effect involving
17
time one and time two score when comparing the
Counselling and Training Group and there was no
interaction.
Discussion
In summary, the findings show that as expected, there
were significant differences between the Counselling
and Training Group and the Comparison Group at
entry to the program on 10 of the 29 scales
considered. There was no change in coping or selfdescription as a function of participation in the
program. These results will be discussed in turn.
The Counselling and Training Group used the
coping strategies of Seeking to Belong; Wishful
Thinking; Not Coping; Social Action; Ignore the
Problem; and Keep to Self significantly less than the
Comparison Group. None of the 12 other scales
revealed a significant difference between groups at
time one. Similarly, four of the SDQ-II scales were
higher for the Counselling and Training Group
suggesting that the process of selection identified a set
of attributes that differentiated the groups. Closer
analysis of the scales suggests that the selection
criteria, whether unintended or by design, identified
appropriate attributes for the counsellors. These
attributes indicate a disposition in the counsellors
favouring independence, rational thinking, problem,
and solution focused behaviour, and a likelihood of
discussing and working through presenting issues in
preference to keeping them to oneself. Combining the
SDQ-II scores broadens the profile of the counsellors
indicating that they also reported that they were
honest and trustworthy, verbally capable, and secure –
that is connected to school and parents more than the
Comparison Group. Thus, the selection process of the
students was successful in identifying students with a
profile that could be considered appropriate for
student counsellors.
It was anticipated that the Counselling and Training
Group would use positive coping more, and negative
coping less frequently, and have higher selfdescription scores at time one in comparison to the
Comparison Group. It was expected that the
Counselling and Training Group’s self-description
would be higher at time two in comparison to time
one, and would be higher than the Comparison Group
at time two. It was also expected that the Counselling
and Training Group would use more frequently
positive and less negative coping at time two in
comparison to time one and in comparison to the
Comparison Group. Some evidence for difference
between groups at time one exists. No main effects or
interactions involving comparison of scores at times
one and two, were found.
While twelve of the eighteen coping scale scores
were not significantly different, six factors did differ
significantly between groups. Of these six scales, four
different scales could be considered unhelpful to the
individual when coping: Wishful Thinking, Not Cope,
Ignore the Problem and Keeping to Self. This
indicates that the Counselling and Training Group are
using unconstructive means of coping less frequently
than the Comparison Group, which does support the
18
BOWLES & MCKINNIS
expectations of difference between groups at time
one.
Similarly, four of the eleven SDQ-II scale scores
were different between the groups at time one. This
shows that the Counselling and Training Group was
more likely to consider that they were more honest,
more expressive, maintained better relations with
parents and a better attitude to school. These are
dispositions appropriate for students advising others.
The differences between groups at time one on the
coping and self-description scales gives some support
for process used to select students counsellors which
identified students with particular characteristics
appropriate to the task demands. The counselors
reported using unhelpful coping strategies less and
viewed themselves as appraising situations and
themselves more honestly, communicating more and
maintaining better relations with parents and their
school.
The expected difference between time one and two
scores for the Counselling and Training Group was
not found nor was the expected difference between
groups at time two. What does it mean that there was
no difference between time one and two for the
Counselling and Training Group and between groups
at time two? Does this nullify the noted value of such
program from other research (Forouzesh, Grant, &
Donnelly, 2001). A number of explanations might
elucidate the absence of the effects. Firstly, method
effects are likely. Previous research (Forouzesh,
Grant, & Donnelly, 2001) shows that the programs are
effective by asking direct questions about how
programs are perceived to assist students. This is a
very different approach to anticipating a gain in
individual difference factors such as coping and selfdescription as a result of participation as a student
counsellor. Secondly, such expectations may be
unrealistic. Coping may be predicated on a stress and
the Counselling and Training Group may not have
been stressed or challenged to use or develop coping
strategies. Thirdly, the training may not have
internalized anything associated with coping or the
self, but may have developed skills to assist others not
the self. Fourthly, there may be a misconstrual and
measurement of the wrong effect. That is, there may
be no effect at the individual level but there may be at
the school or community level requiring different
measurement such as is appropriate for measuring the
perception of the general effect of the program, for
example, Forouzesh, Grant, and Donnelly (2001).
Lastly, it may be explained as a special case of role
effects. By that I mean that helpers help and give, they
don’t gain and to do so would reduce the authenticity
of the process and the unconditional nature of the
focus on the person requiring help. This last possible
explanation fits with elements of the theory of social
support (Barrera, Chassin, & Rogosch, 1993; Cohen
& Wills, 1985; Newcomb, & Bentler, 1988). Taken
together, the research suggests that there may be no
direct secondary gain for the individual participating
in student counselling or helping programs other than
knowing that a personal and generous contribution
has been made. However, more research is necessary
to establish this.
What is the best training for student counsellors and
what skills, knowledge, and attributes do they need?
Improvements to future student counselling programs
may include the refinement of the aims of the
program to include some anticipated change for the
student counsellors. Increasing the focus of the
program on coping and self-enhancing activities may
increase the coping skills and self-description of
counsellors. Providing time for training that is
programmed and includes the development of skills,
understanding, attitudes, and values associated with
care of self and other would broaden the program.
Specifically dealing with debriefing, vicarious
experiences, self-regulation of behavior, and personal
growth of the student counselor and the counseled
person may also assist the development of
counsellors. Limiting the range of experiences that
student counselors are to address, and clearly defining
the skills and processes within a problem-solving
model for student counsellors to follow may
coincidentally increase coping and self-description of
student counsellors.
Although not of major interest in this research the
difference between groups may be used to inform
future planned educational programs wherein greater
emphasis is given to enhancing students coping by
using unhelpful coping less frequently. Enhancing the
facility of students to be more honest and less
defensive, more appropriately expressive, and
maintain a more positive self in relation to parents and
school would be positive.
The conclusions and recommendations from this
research have to be regarded with caution, as there are
clear limitations. There is a small sample size,
restricted to one, single sex, school that reduces the
generalizability of the findings. Similarly, the
conclusions are based on two questionnaires that may
not represent the best factors for consideration when
measuring selection, training, and participation in
such a program. There is also scope for further
research to identify a model that best defines the
attributes of the student counsellor. More research is
required to identify whether there are secondary gains
to be made by participating in such programs, both by
designing them and coincidentally. Further research
could also address whether there are differences that
occur as a function of the sex of the students in the
counsellor and the counselled person.
References
Barrera, M., Chassin, L., & Rogosch, F. (1993).
Effects of social support and conflict on adolescent
children of alcoholic and nonalcoholic fathers.
Journal of Personality and Social Psychology, 64,
602-612.
Cohen, S., & Wills, T. A. (1985). Stress, social
support and the buffering hypothesis. Psychological
Bulletin, 98, 310-357.
Douglas, G. (1998). Peers as Social Support: A
Practitioner's Approach to Developing a SchoolBased Peer Helper Program. Australian Journal of
Guidance and Counselling, 10, 69-89.
Student Counsellor Self-Description
Fallon, B., & Bowles, T. (1999). Adolescent help
seeking for major and minor problems. Australian
Journal of Psychology, 51, 12-18.
Forouzesh, M., Grant, K., & Donnelly, M. (2001).
California evaluation of peer programs:
Comprehensive evaluation of peer programs.
Pasadena, Ca.: California Association of Peer
Programs.
Frydenberg, E., & Lewis, R. (1993). Adolescent
Coping Scale: Administrators Manual. Victoria:
Allenby Press Printers Pty Ltd.
Frydenberg, E., & Lewis, R. (1994). Coping with
difference concerns: Consistency and variation in
coping strategies used by adolescents. Australian
Psychologist, 29, 45-48.
Frydenberg, E., & Lewis, R. (1999). Things don't get
better just because you're older. British Journal of
Educational Psychology, 69, 81-94.
Gerber, S., & Terry-Day, B. (1999). Does peer
mediation really work? Professional School
Counselling, 3, 169-171.
Peer skills National Evaluation. Retrieved July 15,
http://www.
2004
from
peerskills.com.au/2003PeerSkillsInfosheet.pdf.
Kochan, F. K. (2002). Examining the organizational
and human dimensions of mentoring: A textual data
analysis. In F. K. Kochan (Ed.), The organization
and human dimensions of successful mentoring
programs and relationships. (pp. 269-286).
Greenwich, Conn.: Information Age Press.
Marsh, H. (1990). Self-Description Questionnaire II.
Sydney: University of Western Sydney.
Marsh, H., & Hocevar, D. (1985). The application of
confirmatory factor analysis to the study of self
concept: First and higher order factor structures and
their invariance across age groups. Psychological
Bulletin, 97, 562-582.
Marsh, H., Parker, J., & Barnes, J. (1985). Multidimensional adolescent self-concepts: Their
relationship to age, sex and academic measures.
American Educational Research Journal, 22, 422444.
Marsh, H., Relich, J.D., & Smith, I.D. (1983). Selfconcept: The construct validity of the
interpretations based upon the SDQ. Journal of
Personality and Social Psychology, 45, 173-187.
Newcomb, M. D., & Bentler, P. M. (1988). Impact of
adolescent drug use and social support on problems
of young adults: A longitudinal study. Journal of
Abnormal Psychology, 35, 64-75.
Rathvon, N. (1999). Effective School Interventions:
Strategies for enhancing academic achievement and
social competence. New York: The Guilford Press.
Rickwood, D. J. (1992) Help seeking for
psychological problems in late adolescence.
Unpublished PhD Thesis, Australian National
University, Canberra, Australia.
Roberts, A.., & Cotton, L. (1994). Note on assessing a
mentor program. Psychological Reports, 75, 13691370.
Smith, S. W., Daunic, A. P., Miller, M. D., &
Robinson, T. R. (2002). Conflict resolution and peer
mediation in middle schools: Extending the process
19
and outcome knowledge base. Journal of Social
Psychology, 42, 567-586.
Correspondence to:
Terry Bowles
School of Psychology
Australian Catholic University, St Patrick’s Campus,
115 Victoria Pd
Melbourne Vic 3065 Australia
t.bowles@patrick.acu.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 20-23
Client Satisfaction with Relationship Counselling
Adele Carmady (acarmady@swin.edu.au)
Psychology Discipline
Swinburne University of Technology, Melbourne Victoria 3122 Australia
Ann Knowles (aknowles@swin.edu.au)
Psychology Discipline
Swinburne University of Technology, Melbourne Victoria 3122 Australia
Andrew Bickerdike (abickerdike@rav.org.au)
Senior Researcher
Relationships Australia, Melbourne Victoria 3101 Australia
In non-couple counselling research variables found
to be associated with client satisfaction include: client
and therapist reports of improved ability to handle
presenting problems (Deane, 1993; LaSala, 1997);
attainment of goals for therapy (Deane); counsellor
qualities (Hampson, et al., 1999; Helmke, Bischof &
Ford Sori, 2002); and client expectations of
counselling (Sanders, Trinh, Sherman & Banks,
1998). Generally, clients who agree to participate in
such studies report moderate to high levels of
satisfaction and few differences in satisfaction have
been found between male and female clients (Johnson
& Lebow, 2000). The high levels of client satisfaction
typically reported have raised concerns about social
desirability and acquiescence (Deane, 1993; Gaston &
Sabourin, 1992). Gaston and Sabourin did not find a
significant relationship between client satisfaction and
social desirability but did find a low, but significant,
relationship between treatment length and satisfaction.
While the question of how much therapy is effective
depends on many factors (Shadish et al., 1995) this is
an important issue where agencies are increasingly
required to demonstrate effective client outcomes to
secure ongoing financial support.
Research suggests that for clients in individual
therapy, goal attainment is linked to client satisfaction
(Deane, 1993). However, whether this applies to
couples counselling in real life settings does not
appear to have been investigated previously. The
current research explored factors that influence client
satisfaction with couple counselling in community
settings within Australia. These included the issues
couples considered important to discuss in therapy,
strategies clients had tried to resolve their problems
before attending counselling and their counselling
goals, including the influence on satisfaction of
whether or not couples attended therapy with the same
goals. Gender differences were explored.
It was hypothesised that clients and their partners
who have more shared goals/reasons before attending
counselling would be more satisfied with the outcome
of counselling than clients who have different
goals/reasons than their partners for attending
counselling.
Abstract
While relationship counselling has been provided in
Australia for over 50 years there has been little research
that has investigated outcomes for couples who have
received relationship counselling in real life settings.
Participants were 316 heterosexual couples who
attended counselling together at one of the 78 branches
of a national Australian relationship counselling
organisation. Participants completed telephone
interviews with independent interviewers after their
counselling had finished. Results showed that both men
and women considered communication to be the most
important issue discussed in counselling. Respondents
also reported using various communication strategies to
solve their relationship problems before seeking
counselling, with women talking to a wider range of
people than did men. Men who attended counselling
with the same goal as their partner reported higher
levels of satisfaction with counselling than men who
had a different goal than their partner. Couples who
were no longer living together when surveyed were
significantly more likely to have come to counselling
with different goals to their partner than couples who
were still together. Results indicated that future
research should examine couples’ goals pre and post
relationship counselling.
The current study evaluated outcomes for couples
who had relationship counselling through a national
Australian organisation. Such studies are rare. A
meta-analysis of previous research into the efficacy of
marital therapy (Shadish, Ragsdale, Glasner &
Montgomery, 1995) concluded that while marital
therapy appeared to be at least as efficacious as other
forms of psychotherapy, few such evaluation studies
had been conducted in clinical settings. Vidler (2000)
reviewed 30 evaluations of couples’ satisfaction with
counselling dating from 1995-2000 and found that
only 3 of these studies were conducted in clinical
settings. One of these studies (Hampson, Prince, &
Beavers, 1999) found that 92% of couples attending 3
or more counselling sessions experienced an
improvement in their relationship. A German study
which evaluated the effectiveness of marital
counselling using pre and post measures found
moderate improvement in marital satisfaction due to
counselling, although the attrition rate was high with
only 51% of the original group participating in posttreatment assessment (Hahlweg & Klann, 1997).
20
Satisfaction with Counselling
Method
Participants
One hundred and fifty-eight male/female couples who
attended counselling together and who both responded
to the questionnaire participated (N= 316 individuals).
Clients were given a letter requesting their
participation in the survey when they attended one of
the agency branches over a six month period. They
were asked if they were willing to be contacted by
telephone after counselling had ceased and respond to
questions about their experience of counselling. The
mean number of sessions attended by participants in
the current study was 4, and 96% of participants
attended for 10 sessions or less.
Materials
The information sheet given to clients explained that
their questionnaire responses would be confidential
and anonymous. The purpose of the study was defined
as an evaluation of outcomes, including their rating of
the progress they had made on the issues that brought
them to counselling. Clients were informed that the
telephone survey would take around 10-15 minutes.
Questionnaire The questionnaire contained 52
questions including demographics such as age, gender
and marital status. Clients were asked their main
reason for attending counselling and whether they
believed their partner attended for the same reason.
They indicated from a range of nine strategies the
ones they had used to sort out their problems before
they contacted the agency for counselling and
answered questions about what actually happened in
counselling. Couples were also asked if they were still
living together when contacted by the interviewer.
A 5 item Satisfaction with Outcome Scale (SWOS),
based on La Sala (1997), was created to measure
satisfaction with the counselling outcome. All were
Likert-type items with responses ranging from 1 (very
dissatisfied) to 5 (very satisfied). Two items asked
clients how satisfied they felt: “Overall, how satisfied
are you with the outcome of counselling?” and
“Overall, how satisfied are you with the way
counselling was carried out?” These items, as well as
“How much do you think you have personally
changed as a result of counselling?” “My experience
of counselling has improved my ability to manage
other relationship issues”, “My experience of
counselling will help me manage my relationships in
the future” were summed to create the scale.
Cronbach’s alpha for this scale was .84 and the range
of possible scores was 5-24. The scale had a normal
distribution (M = 17.88, SD = 3.78). Higher scores on
the scale reflect higher levels of satisfaction.
Procedure
Due to the sensitive nature of the study the
questionnaire data were coded into a SPSS data file
by an independent specialist research institute and
then the file containing de-identified data was
forwarded to the national community organisation.
21
Results
Data were analysed using SPSS for Windows
statistical package version 11.
Importance of Issues Addressed in
Counselling
Respondents were asked about the importance of 17
issues which may have been discussed in their
counselling on a Likert-type scale from 1 (Not at all
important) to 4 (Very important). For 5 of the 17
issues the mean response was over 3, indicating that
these were perceived as the most important issues
discussed. They were: Communication Issues (M =
3.51; SD = 0.74); Concerns about Closeness (M =
3.36; SD = 0.86); Concerns about Commitment (M =
3.20; SD = 1.00); Parenting/Children’s Issues (M =
3.15; SD = 1.00); and Extra Marital Affairs (M =
3.03; SD = 1.11).
Problem Solving Strategies Tried by Clients
Prior to Counselling
Respondents were asked to indicate how many (if
any) of nine problem solving strategies they had tried
before coming to counselling. The frequency of “yes”
responses for males and females are given in Table 1,
together with chi square analyses examining gender
differences.
Table 1
Frequencies for Problem Solving Strategies Tried
Before Counselling with Gender Differences
Strategy
Males
Tried to talk to partner
151
Tried to solve it myself
143
Talked to friends
58
Read books/literature
59
Talked to family
46
Saw counsellor/psychologist 43
Went to a doctor
36
Other agency
11
Religious help
12
Females
148
152
109
96
74
70
55
22
20
χ2 (1)
ns
ns
33.27**
17.44**
10.58**
10.08**
5.59*
ns
ns
*p< .05, **p<.001
Table 1 shows that ‘Tried to talk to my partner’ and
‘Tried to solve it myself’ were strategies tried by
almost all respondents. Of the remaining seven
strategies, women reported trying five of them
significantly more often than did men.
Couples’ Goals and Satisfaction with the
Outcome of Counselling
Participants were given five possible reasons (goals)
for attending counselling and were asked to indicate
which one was the main reason for them. They were
also asked whether they thought their partner’s goal
was the same as their goal. The goals were: “To
improve your relationship”; “Discuss how to
separate”; Decide about the future of your
CARMADY, KNOWLES, & BICKERDIKE
22
relationship;
Discuss
parenting/stepparenting/children’s issues”; and “Discuss personal
issues around building relationships”. Frequencies for
the goals (reasons) couples gave for attending
counselling are given in Table 2.
Table 2
Frequencies of Main Goal for Attending Counselling
for Males and Females
Variable
Male
(n=158)
Improve relationship
Decide future of relationship
Personal issues
Discuss parenting
Discuss how to separate
Other
71
40
22
13
3
9
Same goal
Different goal
Females
Same goal
Different Goal
n
t
Together
Separated
%
n
%
n
Same
50
105
36
39
Different
50
104
64
68
χ2 (df)
5.43 (1)*
* p<.05
Satisfaction with Outcome Scale (SWOS)
(SD)
Variable
Goal
74
33
24
14
9
4
Table 3
T-test Comparisons on Satisfaction with Counselling
for Couples Commencing Counselling with the Same
or Different Goals
M
Table 4
Chi-Square Analyses for Couples Together versus
Separated on Same Goals or Different Goals for
Counselling
Female
(n=158)
Table 2 shows the most common goals were:
‘Improve the relationship’ and ‘Decide future of the
relationship’. Chi-square analyses showed no
significant gender differences.
Inspection of the data indicated that while 77% of
both men and women thought that their partner
attended counselling with the same goal as themselves
in fact only 46% of couples entered counselling with
the same goal. Independent samples t-test were
performed to examine gender differences in
satisfaction with the outcome of counselling for these
couples who had the same goal and couples who had
different goals. The results are shown in Table 3.
Males
had separated after counselling finished. Results are
presented in Table 4.
(df)
18.29 (3.24) 107
16.78 (4.50) 50
2.30*(155)
18.20 (3.65) 101
17.48 (4.13) 56
1.08 (155)
*p <.05
Table 3 shows that the SWOS score for men in
couples who had the same goal in counselling as their
partner was significantly higher than scores for men in
couples who had different goals to their partner in
counselling. Similar comparisons for women were not
significant
Taking couples who had agreed on goals of
counselling compared to couples who did not agree,
chi-square analyses were performed to examine
whether there was a difference on whether or not they
Results in Table 4 show that couples who were no
longer living together were significantly more likely
to have come to counselling with different goals than
their partner compared to couples who were still
together. Couples who were still living together had
an equal distribution of same goals or different goals
for counselling compared to their partner.
Discussion
Communication was reported by 83% of both men
and women as the most important focus of
counselling. Concerns about closeness (75%) were
identified as next in importance. This finding suggests
that teaching a couple how to communicate more
effectively should be a major focus of couples
counselling.
When asked about strategies tried before
counselling commenced, most men and women said
they had tried to talk to their partner and/or tried to
solve the problems themselves, which again
highlights the importance of developing successful
communication strategies. Women reported that they
were more likely than men to have talked to family,
friends, a counsellor or a doctor, and to have read
relevant literature. This would suggest that women
may make more attempts to communicate with a
range of other people about their relationship
problems than do men; reasons for this difference
should be explored in future research.
Results indicated that most couples attended
counselling with the goal of improving their
relationship or deciding the future of the relationship.
While most couples believed they had the same goals
for attending counselling, in fact 46% said they had
different goals and participants did not appear to be
aware of this discrepancy. Men in couples with the
same goal were more satisfied with the counselling
outcome although this was not found for women.
Results also showed that couples who had separated
since counselling had ceased were more likely to have
come to counselling with a different goal to their
Satisfaction with Counselling
partner. Previous research on individual therapy has
suggested that goal attainment is linked to client
satisfaction (Deane, 1993). The current study findings
suggest that similarity of goals for couples
counselling is important for satisfaction with the
outcome of counselling. This is an area therapists may
need to focus on, perhaps investigating early in
counselling whether couples share goals for
counselling and working sensitively with couples who
present with different goals. Research has shown that
for couples to feel satisfied with counselling they need
to feel that the needs and goals of both partners are
addressed (Sells et al., 1996). Also, if couples attend
counselling for different reasons, therapeutic
interventions that articulate and clarify differences
between the couple could still lead to satisfaction with
the process if not the outcome (Stacey, Allison,
Dadds, Roeger, Wood & Martin, (2002).
Results showed that satisfaction with outcomes was
related to commencing counselling with the same
goals for men but not for women. It is not clear why
this difference occurred and these differences should
be further addressed in future research.
In conclusion, the current study provided a rare
opportunity to examine couples data collected in a
real-life setting. Nevertheless, data were only obtained
from clients after counselling had finished. Future
research should focus on pre and post measures of
client outcomes in couples counselling. The high
financial, emotional and social costs of relationship
breakdown highlight the importance of ongoing
research in this area.
References
Bray, J. H., & Jouriles, E. N. (1995). Treatment of
marital conflict and prevention of divorce. Journal
of Marital & Family Therapy, 21, 461-473.
Christensen L. L., Russell, C. S., Miller, R. B., &
Peterson, C. M. (1998). The process of change in
couples therapy: A qualitative investigation.
Journal of Marital and Family Therapy, 24, 177188.
Deane, F. P. (1993). Client satisfaction with
psychotherapy in two outpatient clinics in New
Zealand. Evaluation and Program Planning, 16, 8794.
Dufrene, R. L. (2000). An evaluation of a patient
satisfaction survey: Validity and reliability.
Evaluation and Program Planning, 23, 293-300.
Gaston, L., & Sabourin, S. (1992). Client satisfaction
and social desirability in psychotherapy. Evaluation
and Program Planning, 15, 227-231.
Halweg, K., & Klann, N. (1997). The effectiveness of
marital counselling in Germany- A contribution to
health services research. Journal of Family
Psychology, 11, 410-421.
Hampson, R. B., Prince, C. C., & Beavers, W. R.
(1999). Marital therapy: Qualities of couples who
fare better or worse in treatment. Journal of Marital
and Family Therapy, 25, 411-424.
Helmke, K. B., Bischof, G. H., & Ford Sori, C. E.
(2002). Dropping out of couple therapy: A
qualitative case study. Journal of Couple and
Relationship Therapy, 1, 51-73.
23
Johnson, S., & Lebow, J. (2000). The ‘Coming of
age’ of couple therapy: A decade review. Journal of
Marital & Family Therapy, 26, 23-38.
LaSala, M. C. (1997). Client satisfaction:
Consideration of correlates and response bias.
Families in Society, 78, 54-65.
Sanders, L. M., Trinh, C., Sherman, B. R., & Banks,
S. M. (1998). Assessment of client satisfaction in a
peer counselling substance abuse treatment program
for pregnant and postpartum women. Evaluation
and Program Planning, 21, 287-296.
Sells, S. P., Smith, T. E., & Moon, S. (1996). An
ethnographic study of client and therapist
perceptions of therapy effectiveness in a universitybased training clinic. Journal of Marital and Family
Therapy, 22, 321-342.
Shadish, W. R., Ragsdale, K., Glasner, R. R.,
&Montgomery, L. M. (1995). The efficacy and
effectiveness of marital therapy: A perspective from
meta-analysis. Journal of Marital and Family
Therapy, 21, 354-360.
Stacey, K., Allison, S., Dadds, V., Roeger, L., Wood,
A., & Martin, G. (2002). The relationship between
change and satisfaction: Parents’ experiences in a
child and adolescent mental health service.
Australian & New Zealand Journal of Family
Therapy, 23 (2), 79- 89.
Vidler, H. (2000). Credible couple therapy: A review
of the research literature 1995-2000. A Report for
Relationships Australia.
Correspondence to:
Associate Professor Ann Knowles
Psychology Discipline
Swinburne University of Technology
John Street
Hawthorn 3122
aknowles@swin.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 24-30
Selves in contact: Multiple perspectives on the role of relationships in the
life of the individual with Dissociative Identity
Kymbra Clayton (kclayton@psy.mq.edu.au)
Macquarie University
Psychology Department
Nth Ryde, NSW 2109 Australia
Kluft & Fine, 1993; Rivera, 1996; Ross, 1989, 1997)
and (2) interviews with Australian participants in my
PhD research project. The participants are people who
experience DI and supporters, and therapists of a
person with DI.
A search of the academic literature found very few
references to the relationships of individuals with DI.
In these few, the message was negative. For example,
Chu (1998) suggests that relationships bring little
gratification and perpetuate uncertainty and fear for
the individual with DI, and Ross (1997) provides a
very brief section on spousal relationships.
Abstract
This paper explores two main relationships which are
significant to the individual with dissociative identity
disorder (DID): the client –therapist relationship and
the individual’s relationship with significant others. By
examining recent research on DID and analysing
transcripts of interviews, discourses are identified that
challenge the traditional assumptions concerning the
therapist/client relationship and the capacity of those
with multiple selves to establish meaningful
relationships. They offer new ways of looking at the
role and significance of relationships in the life of the
individual with DI.
Individual Views to DI
Sitting on the fence
Come sit down beside me
I said to myself,
And although it doesn’t make sense
I held my own hand
As a small sign of trust
And together I sat on the fence
(Michael Leunig)
To introduce some of the interpersonal issues for a
person with dissociative identity, I’d like to pose
some questions to you. Look at the person sitting next
to you. What if I tell you that this person is multiple?
What does this mean to you? Does it mean they are
mad? Or dangerous, like the infamous Dr Jekyll/Mr
Hyde? Or delusional? Or making it up? Are they
likely to be dysfunctional or unreliable or
irresponsible?
People’s responses to these questions range from
the polarised position of viewing individuals with
multiplicity as insane, dangerous, delusional,
dysfunctional, unreliable and irresponsible through to
sane, safe, functional, genuine, reliable and
responsible. Some examples from my interviews:
“A lot of people I know think that my friends with
DID are just crazy.” (Terri, supporter)
“DID is not insanity; it is a normal reaction to an
abnormal situation. We missed our opportunities to
go insane as soon as we learned how to dissociate.”
(Sharon, DI)
“Well, I am not dysfunctional. In fact, I know I’m
extremely functional.” (Marta, DI)
“It is funny how many people either think DID is
invented to keep from accepting responsibility or
confuse us with schizophrenics.” (Johanna, DI)
What framework are you using to answer the
questions posed above? We are not taught how to
relate to individuals who experience multiple selves.
DI is invisible or kept secret. Most people seem to
work with the accepted notion that everyone has a
‘true’ or ‘core’ self, “a self which is unitary, rational,
consistent, and the origin of its own actions.”
(Neimeyer & Raskin, 2000). We may say such things
as ‘part of me feels’, but we see those parts as
encompassing a unified whole. What we do learn of
DI is often that which is promoted by the media, that
is, that individuals with DI are crazy and dangerous.
“I think that a lot of the world’s perspective is that
if you have dissociation, you have a split
People are described as having Dissociative Identity
Disorder (previously known as Multiple Personality)
if their lived experience is one of multiple selves or
different identities (Kluft & Fine, 1993; Rivera, 1996;
Ross, 1997). Each of these multiple selves may
function in thought and action as though they are the
only self, being totally amnesic to the other selves, or
may function as a ‘group’ when amnesia is not
present between all parts.
Depending upon which theory you read, DID is
either an extremely rare condition (Spanos, 1994) or
is experienced by 1% of western populations (Kluft &
Fine, 1993; Ross, 1997). DID is understood to be a
result of severe interpersonal trauma such as
emotional, physical and sexual abuse from early
childhood (Kluft, 1999; Rivera, 1996). The
perpetrators are often those closest to the child, (i.e.
parents, siblings, caregivers, family friends, or in
some cases church and community groups). Given
that dissociation is a common coping response to
trauma, the word “disorder” in the label Dissociative
Identity Disorder can be questioned. To avoid
assuming “disorder”, the term dissociative identity
(DI) will be used in this paper.
In any exploration of DI, an understanding of
interpersonal relationships is crucial. This paper
explores two main relationships: the therapist-client
relationship and the individual’s relationship with
significant others. The data were taken from two
sources: (1) prominent clinical texts on the etiology
and management of DI (Barach, 1994; Kluft, 1999;
24
Dissociative Identity
personality, and you have one that’s capable of
murder.” (Rhonda, therapist, integrated DI)
Recently a mainstream movie ‘Identity’ (2002/3
release) portrayed exactly that image.
What are the implications of your particular
positioning for how you would relate to a person with
DI? You may decide not to relate to the person. If you
do relate to them, will you acknowledge or reject DI?
For example
“Well, it’s a ‘fad’ thing to her, because what she
actually said to me was: ‘Oh, well, look, if you’re
going to go down that path, don’t get me involve’.”
(Marta, DI)
Or will you believe their experience of DI but
interpret their way of being as disordered and even
sick?
“My husband was really quite willing to see me as
mentally disordered. And that sort of, being under
that gaze, it was quite horrific actually.” (Anna, DI)
“I no longer have close personal friendships. It's
impossible to sustain a friendship. Rejection of
people with mental illnesses is alive and
well.”(Hope, DI)
In contrast, in relating to the person with DI, will you
try to learn about and understand their multiplicity?
“People fear what they don’t understand so yes,
gaining some true understanding of DID, of any
disorder, problem, that doesn’t fit ‘normal’ is vital
for that group’s safety and well being. I would love
for people to have open minds about things they
don’t understand/or know about instead of
prejudging.”(Qwy, DI)
Whichever of these options you choose, several issues
crucial to relationships with someone with DI must be
taken into account. Safety and stigma are mentioned
above. Another issue is trust. Do you trust your work
colleague to remember what was said at the meeting
and act on it? Do you trust your friend to mind your
child?
“At first my best friend was terrified to leave her
children in my care.”(Sharon, DI)
Another issue is who you relate to. Will you relate to
all the parts or will you set boundaries and only relate
to one or two? Will you request that certain parts
come out to interact with you at certain times or will
you just go with the flow of who is there at any given
moment? Will you seek out parts that serve your
purposes? A partner of an individual with DI, Laurie
recommends that you
“Try and be aware if you are manipulating parts
for your own purposes. That’s a bit of a minefield!”
(Laurie, DI)
“(For sex) my husband makes sure he is with an
adult and not a child from our system.” (Barbara,
DI)
How will you respond if the parts don’t behave as you
had envisaged? What if some parts want nothing to do
with you? This is a situation faced by Marta’s partner
Edward
“Before our son was born, it was more Matilda,
one of the others, that was around and, yes Edward
wants Matilda back. And now she’s just not going
to go with him.” (Marta, DI)
25
Some of these relationship issues are shown in Figure
1. This pictorial representation only offers a glimpse
of the possible relationship outcomes and does not
cover all the factors involved in such complex issues.
Knowing/Not Knowing: Relationship Impact
I actually made it easy for you at the beginning
because I told you the person next to you was
multiple. What if you hadn’t known or what if they
don't know now but are diagnosed later into a
relationship with you? Do you have a right to know?
Would it change how you related to them?
“I live in a tiny community. If I want to be totally
isolated from all humanity, telling these people
would accomplish all of that.”(Hope, DI)
“As for my family, they haven’t changed their
beliefs so I have decided not to have a lot to do with
them (and definitely not tell them I have DID) for
my own health and safety.” (Qwy, DI)
The questions posed in relating to a person with DI
are new, challenging and possibly confronting for
many people. Unfortunately most people have only
the negative attitudes from the media to inform their
response. Researchers have not explored the issue.
The Therapist and DI Client
To be nobody but yourself - in a world which is
doing its best, night and day, to make you like
everybody else means to fight the hardest battle
which any human being can fight, and never stop
fighting."
~E.E. Cummings
What are the implications of DI for the person’s
relationship with their therapist? The same issues
explored previously operate within the therapeutic
relationship. The therapist can view their client’s DI
as dysfunctional or as simply further along a normal
spectrum. Whereas the individual relating to a person
with DI has little framework to base their responses
on, the therapist has a wealth of clinical research
(mainly North American) and the International
Society for the Study of Dissociation (ISSD)
treatment guidelines to inform their practice. The
Diagnostic Statistical Manual (DSM-IV) presents
dissociative identity as a disorder (DID). The ISSD
treatment guidelines explicitly state that individuals
require “symptom stabilisation”, “control” of their
behaviour and “restoration of functioning.” (Barach,
1994).The assumption is that individuals with DI are
unstable, out of control and dysfunctional. Indeed,
dysfunction is assumed to the extent that, unlike other
disorders such as schizophrenia, “diagnosis can be
made in the absence of significant objective
dysfunction” (Summerfield, 2001). The dominant
discourse in the texts is also dysfunctionality:
CLAYTON
26
Individual knows they have DI
They give up
on having
relationships
DI individual doesn’t know
they have DI but finds out during
relationship
They try to
make a
relationship
Individual never learns they
have DI
Doesn’t tell other
Tells other
Other believes but it is
too difficult and the
relationship ends
Other believes and
tries to learn about
it and adapt
Other finds it
difficult to believe in
the notion of DI and
relationship suffers
Other believes and
sees it as disordered
and relationship
suffers
Rollercoaster
of moods and
difficulties
Other wonders
how to relate
to varying self
presentations
Other battles on
with relationship
No relationship
Relationship ends
Relationship grows
Figure 1
Impact of multiplicity on relationship with other person
“In adulthood, now that the patient is removed
from the precipitating traumatic situations, frequent
and easily activated dissociation becomes a
seriously maladaptive response to the normal
stresses of life.” (Putnam, 1989, p. 137)
Interviews carried out in my research with current
Australian therapists, however, indicate a broad range
of views with regard to disorder and dysfunctionality
in people with DI.
“It was an effective way of coping with an
overwhelming and un-copeable situation. It’s a
disorder in the sense that it is not an affective,
functional way of dealing with life as an adult.”
(Lindsay, therapist)
“In terms of having, you know, a very functioning
life, in terms of personal and
business
relationships, it’s a huge detraction.” (Robert,
therapist)
“It’s not a disorder; it’s actually a response to a
disordered thing that happens to them. It’s not a
‘disorder’ in itself.” (Lucy, therapist)
The dominant view within the clinical texts is that
clients with DI are delusional, seeking attention, or
even faking it, although this provokes frustration
amongst some therapists and individuals.
“MPD is elaborate pretending. The patient
pretends that she is more than one person, in a very
convincing manner. She actually believes it herself.
The alters are, put another way, devices.” (Ross,
1989, p.109)
“I get frustrated with that concept of the medical
profession or the psychiatric profession, the mental
health profession, saying it’s ‘made up.” (Fleur,
therapist)
“But none of the current therapists believe I have
DID” (Jessica, DI)
It is important to remember that much of the literature
on dissociative identity is based on clinical samples
and hospital-based research.
“There may be in the general population a large
number of people with MPD who are highfunctioning,
relatively
free
of
overt
psychopathology, and no more in need of treatment
than most of their peers. They may not have abuse
histories and may have evolved a creative and
adaptive multiplicity. If these people exist, virtually
nothing is known about them.” (Ross, 1989, p. 97)
In contrast to the dominant view of multiplicity
equalling dysfunction, the therapists interviewed took
a more non pathologising perspective; with many
defining multiplicity as a different way of being (and
in some cases taking the position that it was a normal
way of being) rather than as psychopathology.
“They come to understand it as a healthy process, a
reaction to what has happened in their lives and a
protective process for them, okay? So I think that’s
quite useful for people.” (Lucy, therapist)
“Everybody has parts, and we all talk to ourselves,
we all argue with ourselves.” (Lindsay, therapist)
Goals and the Therapeutic Relationship
The position that therapists take with regard to DI
impacts on what would be seen as the ideal goal(s) of
therapy. Because the dominant discourse is that DI is
dysfunctional, the texts focus on the issue of ‘cure’.
The goal – ‘a script to order’- is generally seen as
integration of the ‘parts’ into a unitary self. The
relationship that arises out of such a medical focus is
one of doctor and patient.
Dissociative Identity
“The goal of treatment of MPD is not palliation. It
is cure.” (Ross, 1997, p. 204)
“She needs to integrate.” (Ross, 1997, p.294)
“Integration as an overall treatment goal.”
(Barach, 1994)
Many of the interviewed therapists, however,
mentioned that as they got to know their individual
clients better, the concept of a standard goal of
integration became questionable
“You see, one of the things, one of my big learnings
with this is that no two people are the same, in
terms of dissociation. I mean, I thought at first,
yeah, multiple personality, I can get a handle on
this, yeah, yeah, once I knew the nuts-and-bolts,
then just put those in place, which books like
Putnam kind of almost led you to believe.” (Libby,
therapist)
The individuals with DI and partners offered some
alternate views to integration as well, suggesting that
control and functionality were much more a focus.
“Yeah, in fact, you know, I like it this way. I don’t
want to change it. I don’t want to be integrated.
(but) I’d like to have some sort of more control.”
(Marta, DI)
“Provided I can function in society & do what ever
I set out to do (without negatively affecting myself
or other people), then I see no reason not to keep
my ‘family’. I see no reason to integrate.” (Qwy,
DI)
“I guess it would be more functional though if the
parts worked together as a team. Life would be
much easier for her.” (Laurie, partner)
So how do therapists position themselves in regard to
the dominant discourse of integration espoused by the
texts? The majority of the therapists interviewed
consciously placed themselves outside this discourse
“I’m not clear in my mind that integration is
necessary. I think once everybody’s cooperating,
that can lead to a functional person.” (Libby,
therapist)
Some of the therapists encapsulated their goals into
words such as upgrading, cooperation, and control.
“The parts are all upgraded to 2003. That
everybody’s in the present time. That the person is
able to engage in relationships, whether they be
intimate sexual ones or friendship ones.” (Fleur,
therapist)
“I generally talk “cooperation”, so the parts are
actually working together for a common goal.”
(Libby, therapist)
“I’m aiming that the client has control over her
own life and feels good about herself.” (Lindsay,
therapist)
The literature also instructs us to focus on the fact that
the individual with DI is to be seen as one self.
“The more deeply parts of the individual connect
with the therapist; the more important it is for the
therapist to remember that the client is one
person.” (Rivera, 1996, p. 122)
“The patient’s conviction that there is more than
one person in her is a dissociative delusion and
should not be compounded by a ‘folie a deux’ on
the part of the therapist.” (Ross, 1989, p.109)
27
What does this mean for the therapist’s relationship
with the client and the client’s selves? The concept of
one self sat well for only a couple of the therapists
interviewed, eg Lindsay, while for Marta, an
individual with DI, being discouraged from using the
word multiple was confusing.
“I don’t think of them as different personalities. I
prefer the ‘parts’ language and when the parts all
work well together, then it feels integrated.”
(Lindsay, therapist)
“The psychiatrist that I am seeing at the moment
doesn’t like me to use the word ‘multiple. Well it
makes me feel like I am making things up and its
very confusing if someone says there is one me and
it feels so obvious that there isn’t.” (Marta, DI)
For some other therapists it meant positioning
themselves in opposition to the traditional framework
of one self, and working from within the client’s
framework.
“Validate who they are. Validate their experience.
Appreciate it. I have tremendous compassion for
these parts.”(Fleur ,therapist)
“I think the biggest thing I’ve found is that they’re
accepted for who they are.” (Rhonda, therapist,
integrated DI)
One of the individuals with DI saw this alternative
positioning by her therapist as
“Well, it gave a chance for everybody to be... to be
heard, and for everybody to tell the
story as
they are. And to understand what on earth had been
going on. I was just so relieved and so joyous that
somebody’s hearing us!” (Rhonda, DI)
Therapist Knows Best
Within the medical/psychiatric field there exists an
accepted dominant discourse that ‘the therapist knows
best.’ The therapist is thereby given authority and
power.
“Treatment will be a collaboration but not
democratic. The patient is the patient and I am the
doctor. We are not friends, and I am the only one
getting paid.” (Ross, 1997, p.302)
“Although the patients feelings and preferences
need to be explored while devising and
implementing a treatment plan, the therapist not the
patient, ought to be the primary architect of the
treatment plan.” (ISSD guidelines, 1994, section
IIIB)
“There’s a real swing back to ‘You are the
therapist and you are the expert’ that’s what I’ve
found quite confronting.” (Lucy, therapist)
Interestingly, the majority of therapists interviewed
positioned their clients as the experts, the ones who
know best; a completely polarised view to the clinical
texts. Some of the therapists did mention that early in
their careers they too thought they were to be the
expert, that this was expected from the clients. Almost
all the therapists mentioned that the acceptance that
they didn’t need to know everything came with time
and experience.
“The client’s the expert on their life, and where
they want to go. But they can use the
tools that
you can bring and show them.” (Libby, therapist)
28
CLAYTON
“You need to be secure about people, like who they
are, and allow them to find who they are on their
terms. They need your help but not your trying to
give them your standards.” (Rhonda, therapist)
For many of the therapists and individuals some
important shifts occurred during the course of the
therapeutic relationship. Therapists and DI individuals
acknowledged that the relationship brought about a
mutual process of growth and learning. Often the
therapists remarked how they were in awe of the
survival of their DI clients, of their skills, and their
ability to remain ‘human’.
Responsibility
A separate issue that also impacts on the relationship
between therapist and client is responsibility. What
does responsibility mean to the therapist and to the
client with DI? If the therapist holds to the notion of
one self, how do they reconcile that their client may
not be aware of some actions that the therapist then
says their client is responsible for? What does this
mean for the therapeutic relationship? The texts are
explicit in their recommendation - and suggests that
often clients will resist being responsible.
“Therapists ought to hold the whole person to be
responsible for the behaviour of all the alternate
personalities.” (ISSD guidelines, 1994, Section
IIIA)
“Initially, confrontation should be kind, firm,
matter of fact, and incorruptible. Once it is clear
that the patient understands what is required but
behaves inappropriately nonetheless, more
forthright confrontation may have a role. This may
be especially forceful if the issue concerns
cooperation with therapy.” (Kluft, 1993, p. 43)
The therapists interviewed, however, found this to be
a much more complex issue and none were able to
give a definitive answer. For most, they felt it
depended upon the situation, and the individual client.
“I mean, my basic understanding of life is that
we’re each responsible for ourselves. I’m not clear
in my mind totally how responsible people are in
terms of another part, you know. My critical self
says: Of course they’re ruddy responsible, they’ve
got to get their act together and…But another part
of me says: I don’t know. You know, I see people
doing things they’re not aware that they’ve done
that. So that the jury’s out on that, I’m still not quite
clear. (Libby, therapist)
In contrast, the individuals, perhaps surprisingly, had
no such turmoil in relation to responsibility. All of
them mentioned the need to take responsibility for all
actions, regardless of which self committed the
actions (and whether they had memory of the actions).
“I know one thing that is important to me is taking
responsibility for my behaviour. And that is not easy
to do sometimes especially if I don't remember
saying or doing something.”(Celine, DI)
Being Multiple
What would happen if I now asked you to switch
places? You now are multiple. How do you feel about
being multiple?
“I’m in awe of it, (multiplicity as a way of being)
absolute awe. My whole system is just so clever, it’s
wonderful. So imaginative and creative, I love it.”
(Rhonda, DI)
“I cannot tolerate having this disorder as nothing
ever seems real.” (Jessica, DI)
“Well, it’s totally normal. It’s just me, that’s all
I’ve ever known.”(Marta, DI)
How do you experience the world? Does it feel OK
for you?
“I still find myself in "my own world" as I am
told by the people in my life- but its safe
there.” (Orion, DI)
“I think dissociative identity is very isolating in that
it’s not like something you put out there.” (Anna,
DI)
“I like being multiple! I don’t like all the problems
that can accompany having a dissociative disorder,
but I like who I am, all of me.” (Qwy, DI)
Is your being multiple OK for those around you? Are
they able to relate to your other selves?
“I want him better, it is an extreme challenge to be
supportive when sometimes I feel he has no control
or makes little effort to communicate with the
therapist, so we can take all necessary cautions to
prevent an episode. It greatly affects me--it has
turned life upside down.”(Tina, partner)
“It makes life a bit of a roller-coaster. You can
never take anything for granted because what is one
day is often different the next. In some ways that
makes life difficult but in other ways it keeps life
interesting and ensures you never take your partner
or the relationship for granted.”(Laurie, partner)
“I don’t have any proper confirmation from
anybody saying, “Yes, it’s okay to be like that. It’s
sort of like not even talked about.” (Marta, DI)
How do you relate to your other selves?
“Firstly accepting their existence which is so
obvious but has been rejected for so long and then
honouring them by giving them time and beginning
the connection”(Sharon, DI)
Who needs to change?
“Never lose sight of the fact that your partner is an
individual above all. Get familiar with her ways!”
(Laurie, partner)
“Well, I guess, what I, for me personally, I want to
be whole. I mean, I don’t have a problem with – if
someone can live and function as a multiple.”
(Rhonda, therapist, integrated DI)
What about if for others your being multiple is ok?
Does this help you? Do you feel ok, comfortable in
the world?
“Thank fully I have met a few who can see that
though I am DID I am not an idiot, mentally
impaired, sick or in need of incarceration or an
institution.” (Tribe, DI)
“When I’m with people who understand it, I feel
quite ok and normal.”(Marta, DI)
Whose world do you want to be in: the singleton or
the multiples?
“You know what…I’m fascinated …your own
experience is the only way you experience the world
I can’t imagine what it would be to experience the
world in a way other than my own. I didn’t think
Dissociative Identity
mine was different in the first place. Or that it could
be different. It’s a challenging thought.”(Anna, DI)
Issues to Question
An analysis of the research and transcripts raises some
important questions which impact on individuals with
DI and their relationship to others and to therapists.
Does the traditional view of dissociative identity
empower or pathologise individuals’ experiences?
“It’s just ludicrous, there’s no acceptance, there’s
no understanding, and people are treated not very
well. So I can understand why they don’t want to
come out and say… yeah, and I can understand the
levels of distrust and suspicion.”(Lucy, therapist)
Does the therapy offer space for clients to develop
new forms of subjectivity or does it confirm them
within positions of both being misunderstood and
needing to hide their multiplicity.
“They’re not listening to what is being said, for a
start, and it’s like turning it into nothing. Turning it
into absolutely nothing. Again, that “You just don’t
exist.” (Marta, DI)
“Psychiatrists just see it as a disorder. Something
wrong with the person, that’s as far as it ever goes.
So you can see how difficult it would be for anyone
with DID.” (Lucy, therapist)
Would it help if more therapists were willing to talk
about their experiences with other therapists without
fear of being isolated, or ridiculed?
“I think the biggest problem is the lack of
education. Yeah. …I think more therapists will then
to actually expand their mind to understand that
this exists, this is real, and you need to deal with it.
Not brush it under the carpet or think: “Well, okay,
I need to find someone, if this is what you have got,
I need to find someone else to send you to.”
(Rhonda, therapist, integrated DI)
Conflicting
experiences
Individual’s
SELVES
DSM-IV
& Clinical
research
Dominant
discourses
Media
Conflicting
inputs
OTHERS:
FRIEND, PARTNER
29
experiences, and that often they are unable to make
friendships due to stigma? Their other selves may also
need befriending, both from the individual themselves
and others outside.
Because most of the literature on DI is in the form
of clinical case studies or quantitative reports, the
experiential issues faced by the individual and the
therapist are lost. This brief qualitative exploration
demonstrates the complexity of a few of the many
issues (i.e. trust, responsibility and knowing) that are
involved in relating with an individual with DI. No
longer can we assume a simple two-way relationship.
Now we have multiple interpersonal relationships (Fig
2), each affected by the dominant discourses in
society, the media, and if in therapy, the DSM-IV.
Individuals with DI are still required to ‘hide’ or
pass as singletons, otherwise they are pathologised or
seen as ‘freaks’. By taking another perspective
towards DI, and focusing more on the day to day
issues, (i.e. coconsciousness, communication and
cooperation between the selves; being able to work,
relating to others), a different approach to therapy
could also develop. Issues such as the partners
involvement in therapy and the partners own needs
and support are urgently in need of further
exploration. The issue of children and DI parents is
conspicuously absent from most dissociative
literature. In that which is written, it is emphasised the
individuals with DI should integrate before having
children (Putnam, 1989).
There is no literature however which looks at this
issue from an experiential framework. It is therefore
of real importance to gain further insight to the issue
of parenting. This was further emphasised to this
author by the fact that the majority of the DI
individuals in this paper had children who were totally
aware of their parents DI, were quite okay with it, and
as Rhonda (DI) mentioned “They were great. They
were terrific.” From the childrens’ perspective it is
important to find out what helped them understand,
what did or do they find difficult, what do they think
would be useful to know?
Other issues such as how the individual with DI
deals with the loss of friends/ partners, and the
everyday issues that we all face are also in need of
further exploration from a qualitative framework. All
of these questions help to create possible paths, our
directions depend upon positions taken; the route we
take is ours to choose. The way from here is to
attempt a greater understanding by creating more
room for research that describes the experiential
world of DI.
Summing up... from ‘Kevin & the crew’
Figure 2
Multiple Interpersonal Relationships
Is there enough support and education available for
partners and friends of individuals with multiple
selves? What about the issues that many people with
DI may not have relationships at all that often they
may have had to cut off the family for safety reasons,
they cannot trust potential friends due to past
“To answer some of your questions: I hold down an
office job, where I work as a computer programmer
and web designer. I'm also an artist. I have a
driver's license and a car; I have friends,
girlfriends, and a regular schedule of social events
and contacts. I have a sense of humour, enough so
that I'm infamous among my friends for my twisted
humour and bad puns. (Is there any such thing as a
good pun?) There are many things that I enjoy and
have fun with. For all this, I don't know that I'll ever
CLAYTON
30
be "fully recovered." I don't know that I need to be
"fully recovered." One of my pet peeves with the
mental health establishment in our culture is that
we're all tied up with the concepts of illness,
disorder and wellness. It's the assumption that if
we're different, we are "wrong" and need to be
"cured." In traditional Western psychiatry and
psychology there doesn't seem to be a whole lot of
room for just being "different" if that involves living
outside the accepted parameters of "normal." It's
taken me a long time and some hard struggles to
reach this point, but more and more I don't look at
myself as sick, crazy or "unwell." I try not to regard
The Crew as a problem to be fixed (especially
because I'm one of them!). This is just how I live,
and I'm making the best of it I can.”
References
American Psychiatric Association. (1994). Diagnostic
and statistical manual of psychiatric disorders,
fourth edition (DSM-IV). Washington DC:
American Psychiatric Association
Barach, P. (1994). ISSD Guidelines for treating
dissociative identity disorder (multiple personality
disorder) in adults (1994). Skokie, IL: The
International Society for the Study of Dissociation.
Chu, J. A. (1998). Rebuilding shattered lives: The
responsible treatment of complex posttraumatic and
dissociative disorders. New York: Wiley.
Kluft, R. P. (1999). An overview of the psychotherapy
of dissociative identity disorder. American Journal
of Psychotherapy, 53, 289-317.
Kluft, R. P. (1993). Basic principles in conducting the
psychotherapy of multiple personality disorder. In
R. P. Kluft & C. G. Fine (Eds.), Clinical
Perspectives on Multiple Personality Disorder, (pp.
19-50). Washington: American Psychiatric Press.
Neimeyer, R. A., & Raskin, J. D. (2000). On
practicing postmodern therapy in modern times. In
R. A. Neimeyer & J. D. Raskin (Eds.),
Constructions of disorder: Meaning-making
frameworks for psychotherapy (pp. 1-14).
Washington,
DC:
American
Psychological
Association Press.
Putnam, F. W. (1989). Diagnosis and treatment of
multiple personality disorder. New York: Guilford
Publications Inc.
Rivera, M. (1996). More alike than different: Treating
severely dissociative trauma survivors. Toronto:
University of Toronto Press.
Ross, C. A. (1989). Multiple personality disorder:
Diagnosis, clinical features, and treatment. Canada:
John Wiley & Sons, Inc.
Ross, C. A. (1997). Dissociative identity disorder:
Diagnosis, clinical features, and treatment. (2nd
Ed.). Canada: John Wiley & Sons, Inc.
Spanos, N. (1994). Multiple identity enactments and
multiple personality disorder: A sociocognitive
perspective. Psychological Bulletin, 116, 143-165.
Summerfield, D. (2001). The invention of posttraumatic stress disorder and the social usefulness
of a psychiatric category. British Medical Journal,
332, 95-98.
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 31-34
The Influence of Interpersonal Relationship on Conflict Resolution Style
John A. Davidson (John.Davidson@utas.edu.au)
School of Psychology
University of Tasmania, Hobart, Tas 7001 Australia
contending, with three of the four being statistically
significant. That study, however, did not explore the
effects of different interpersonal relationships.
In addition to finding this moderate degree of
individual consistency in the use of conflict resolution
styles, Davidson and Biffin (2003) also found
substantial effect sizes for situations and the situation
x person interaction. Substantial interaction effects
have been found previously in a number of other
studies attempting to partition the response variance
between personality and situational factors (Endler &
Hunt, 1966, 1968; Moos, 1969).
If contextual factors such as relationships are
included as part of the situation effect it becomes
important to explore more thoroughly the effects of
different relationships on the choice of conflict
resolution styles. The purpose of the present study
was to extend the Sternberg and Dobson work by
including relationships with subordinates or younger
siblings as well as with parents and peers, then
comparing mean levels of strategy use across the
different types of relationship, and assessing by
correlations the consistency of use of each strategy
between pairs of relationships.
Abstract
Sternberg and Dobson (1987) reported strong
consistencies in conflict resolution style and
widespread individual differences across conflicts
involving various types of interpersonal relationships.
The present study used the same 16 conflict resolution
styles but reduced them to the four primary strategies
of problem-solving, contending, conceding and
avoiding postulated in the dual concern model (Rubin,
Pruitt, & Kim, 1994). Ratings were obtained for these
strategies from 81 students in a course on the
psychology of peace and conflict. Students recorded
and rated conflicts with a parent or authority figure,
with a same or opposite sex peer, and with a
subordinate or younger sibling. Conceding was used
significantly less than the other styles in conflicts with
an authority figure or peer. In conflicts with a
subordinate conceding and avoiding were both used
significantly less than contending and problem-solving.
There were low to moderate correlations for a single
strategy across two relationship types, ranging from .13
to .47. These results suggest that cross-situational
consistencies in conflict resolution styles have been
overestimated by Sternberg and Dobson, and they
accord with the findings of Davidson and Biffin (2003)
on the importance of contextual factors.
Method
Sternberg and Dobson (1987) conducted three
experiments to assess the consistency of conflict
resolution styles across a variety of interpersonal
relationships. They reported high levels of
consistency across the various relationships, though
correlations appear to have been elevated by
correlating means, and the sample sizes of only 40 are
somewhat problematic. Relationships included those
with parents, same and opposite sex peers, teachers,
room-mates and romantic partners.
One of the contributions of the Sternberg and
Dobson research is to identify a list of 16 methods of
conflict resolution that are frequently employed.
These methods can be categorized into the more
theoretically useful strategies of contending, problemsolving, conceding or avoiding in the dual concern
model (Blake & Mouton, 1964; Pruitt & Rubin, 1986;
Rubin, Pruitt, & Kim, 1994). In the dual concern
model these strategies are predicted to be more likely
to be elicited in situations where concern for self and
for the other are respectively high and low
(contending); high and high (problem-solving) low
and high (conceding); or low and low (avoiding).
A more recent study of levels of individual
consistency by Davidson and Biffin (2003) found
correlations between the four strategies based on the
Sternberg and Dobson items and the same strategies
measured by a specially constructed Dual Concern
Model Questionnaire to be low to moderate, ranging
between .09 for problem-solving and .45 for
Participants
Participants for the study were 81 second or third year
undergraduate psychology students (72 female, 9
male) enrolled in a unit on the psychology of peace
and conflict.
Materials
Instruction sheet for recording conflicts
Instructions included describing conflicts with a same
or opposite sex peer, a parent or authority figure, and
a younger sibling or other person over whom the
participant was exercising some degree of authority.
A conflict was defined as any interpersonal
interaction in which the person did not feel OK, or
was conscious of having incompatible wishes or
aspirations with those of another person where it
appeared both could not be fulfilled.
Scoring Instructions and Recording Sheet The
scoring instructions included a list of the 16 methods
of conflict resolution with definitions as used by
Sternberg and Dobson (1987). The sheet also
provided a rating scale with scores of 0 (not at all) to
10 (high) for the use of each method. Any number of
methods could be rated for each of the three conflicts.
Procedure
Participants wrote down a conflict of each type in
which they had been personally involved and then
31
DAVIDSON
32
scored the conflicts themselves using the scoring
instructions and recording sheet provided.
Results
The 16 conflict resolution methods were grouped into
the following categories representing the four basic
strategies:
Contending: Economic action, physical force,
undermine, withhold support, manipulate, verbal
force, confrontational discussion;
Avoiding: Wait and see, avoidance, separation;
Conceding: Accept, back down, give in;
Problem solving: Third party, bargain/compromise,
mutual discussion.
The score for each strategy was determined as the
maximum score of any of the methods falling within
that category. The .05 significance level was used for
all statistical tests, unless otherwise specified.
Differences in Strategy Use
A repeated measures analysis of variance was
performed with the factors strategy with four levels
(contending, problem-solving, conceding, avoiding),
and type of conflict with three levels (parents, peers,
subordinate). As shown in Figure 1 there was a
significant interaction between strategy and type of
conflict, F(5,313) = 5.87, p < .001, following a
Greenhouse-Geisser correction to the degrees of
freedom. Follow-up tests for the three relationships
showed that for authority figures the only significant
difference was between avoiding, the most-used
strategy, and conceding, the least used. For peers
conceding was used significantly less than all three
other strategies which were not significantly different
from each other. For subordinates contending and
problem solving were both used significantly more
than avoiding or conceding. The remaining
differences were not statistically significant.
Consistencies in the Use of Particular
Strategies
To assess the degree of consistency in the use of
strategies across different types of conflict
correlations were determined for the use of each
strategy across the different pairs of relationships. The
results are shown in Table 1. The conflict resolution
strategy which had the highest level of consistency
was contending with correlations of .41 to .47 across
the pairs of relationships. Next highest was problemsolving with correlations of .24 to .42. The only
significant correlation for conceding was .42 between
conflicts with peers and subordinates. There were no
significant consistencies in the use of avoiding across
pairs of relationships.
Discussion
The finding of a significant interaction between the
type of interpersonal relationship and choice of
conflict resolution strategy is consistent with the
importance of contextual or situational factors in
influencing the use of particular methods of conflict
resolution (Davidson & Biffin, 2003). The finding
also complements the work by Sternberg and Dobson
(1987) by showing that the participants in this study
handled conflicts with subordinates differently from
those with parents or peers by a substantially lower
use of avoidance.
Conceding was uniformly the least used conflict
resolution method. The use of the other three methods
was not significantly different for parents or peers,
though avoidance was admittedly the most used in
conflicts with parents.
In relation to consistencies in the use of particular
conflict resolution strategies across conflicts
involving different interpersonal relationships, the
results qualify rather than negate the findings of
Sternberg and Dobson (1987). The correlations are
lower, being based on individuals as opposed to group
means, but they are still moderately high for
contending across all pairs of relationships. Low to
moderate correlations were obtained for problem
solving and conceding. Only in the case of avoiding
were all the correlations non-significant. Avoiding is
the most problematic of the conflict resolution
strategies, and often yields inconsistent results within
the framework of the dual concern model (Davidson
& Biffin, 2003).
In terms of general social relevance two other
questions emerge from a consideration of the results.
Given that problem solving is the method most likely
to lead to a mutually beneficial solution, why has it
not been adopted as clearly superior to dominating
across all of the relationships in which the present
participants have been involved? Secondly, why is it
that conflicts with parents are currently characterized
by the use of avoidance rather than problem solving
as the most typical method of resolving conflict?
There is an accumulating body of evidence that
conflict resolution skills related to communication and
problem solving can be successfully taught (Davidson
& Wood, 2004), and that they are particularly
effective when integrated into an institutional culture
that fosters cooperation and respect for diversity
(Coleman & Fisher-Yoshida, 2004; Johnson &
Johnson, 2004). These findings now need to be reevaluated and applied in family and educational
settings. A child who grows up without learning how
to listen, or to be able to express a contrary view
without blaming, or who cannot work through a
disagreement with someone else to produce a
mutually beneficial outcome is as socially deprived as
a child that has never been taught to share.
Resolving a conflict with a parent or authority
figure presents the added difficulty of a power
differential. In a recent study based on a workplace
problem, Davidson, McElwee, & Hannan (2004)
found that while trust in a relationship had direct and
indirect effects on satisfaction with outcome, being in
part mediated by the use of the problem solving
strategy, the effect of power was direct, and reduced
outcome satisfaction irrespective of strategy choice.
In this regard approaches to parenting that build on
mutual respect and good communication are clearly
advantageous in facilitating constructive adult
relationships (Wood & Davidson, 2003).
Interpersonal relationship and conflict resolution style
33
Strategy
Problem solve
Contend
6
Concede
Mean Strategy Ratings
Avoid
5
4
3
Parent
Peer
Subordinate
Relationship
Figure 1
Ratings for mean conflict resolution strategy use for each relationship
Table 1
Correlations for ratings of strategy use between pairs of interpersonal relationships
___________________________________________________________________________________
Strategy
Relationship
___________________________________________________________________________________
Parent-Peer
Parent-Subordinate
Peer-Subordinate
N=76
N=62
N=63
Problem Solving
.25*
.24
.42**
Contending
.43**
.41**
.47**
Conceding
.14
.20
.42**
Avoiding
.13
.25
.17
____________________________________________________________________________________
* p<.05, ** p<.01
In conclusion the study has demonstrated the
importance of interpersonal relationships in the
determination of conflict resolution strategy. Future
studies might profitably attempt to characterize these
effects and relationships within a broader theoretical
context such as social interdependence theory
(Johnson & Johnson, 2003).
References
Blake, R.R., & Mouton, J.S. (1964). The managerial
grid. Texas: Gulf Publishing.
Coleman, P.T., & Fisher-Yoshida, B. (2004). Conflict
resolution at multiple levels across the lifespan: The
34
DAVIDSON
work of the ICCR. . Theory into Practice, 43(1), 3138.
Davidson, J. A., & Biffin, F. E. (2003). The influence
of person and situation factors in the choice of
conflict resolution style. In K.A. Moore (Ed.),
Proceedings of the 3rd Australasian Psychology of
Relationships Conference: Relationships: Family,
work and community (pp. 20 - 26). Melbourne:
Australian Psychological Society.
Davidson, J. A., McElwee, G., & Hannan, G. (2004).
Trust and power as determinants of conflict
resolution strategy and outcome satisfaction. Peace
and Conflict: Journal of Peace Psychology, 10,
275-292.
Davidson, J., & Wood, C. (2004). A conflict
resolution model. Theory into Practice, 43(1), 6-13.
Endler, N.S., & Hunt, J.M. (1966). Sources of
behavioural variance as measured by the S-R
inventory of anxiousness. Psychological Bulletin,
65, 336-346.
Endler, N.S., & Hunt, J. M. (1968). S-R inventories of
hostility and comparisons of the proportions of
variance from persons, responses, and situations for
hostility and anxiousness. Journal of Personality
and Social Psychology, 9, 309-315.
Johnson, D. W., & Johnson, R. T. (2003). Field
testing integrative negotiations. Peace and Conflict:
Journal of Peace Psychology, 9, 39-68.
Johnson, D. W., & Johnson, R. T. (2004).
Implementing the “Teaching Students to be
Peacemakers Program”. Theory into Practice,
43(1), 68-79.
Moos, R.H. (1969). Sources of variance in responses
to questionnaires and in behaviour. Journal of
Abnormal Psychology, 74, 405-412.
Pruitt, D.G., & Rubin, J.Z. (1986). Social conflict:
Escalation, stalemate, settlement. New York:
Random House.
Rubin, J. Z., Pruitt, D. G., & Kim, S.H. (1994). Social
Conflict: Escalation, Stalemate, and Settlement
(2nd ed.). New York: McGraw-Hill.
Sternberg, R. J., & Dobson, D. M. (1987). Resolving
interpersonal conflicts: An analysis of stylistic
consistency. Journal of Personality and Social
Psychology, 52, 794-812.
Wood, C., & Davidson, J. (2003). Helping families
cope: A fresh look at parent effectiveness training.
Family Matters, 65, 28-33.
Correspondence to:
John Davidson
School of Psychology
University of Tasmania
Private Bag 30
Hobart 7001
John.Davidson@utas.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 35-39
Do Adolescent and Parental Perceptions of Interparental Conflict Differ?
Melanie T. Davern (mtadavern@deakin.edu.au) and
Petra K. Staiger (pstaiger@deakin.edu.au)
School of Psychology, Deakin University, Burwood, Victoria, 3125 Australia
Ernest S.L. Luk (slluk@netvigator.com)
Department of Psychological Medicine, Monash University, Clayton, Victoria, 3168 Australia
parental reports of interparental conflict. Hence, it is
parental perception of interparental conflict that has
been associated with adolescent psychopathology
rather than adolescents’ perceptions of such conflict.
It is argued in this paper that consistent with clinical
practice, research into family functioning needs to
take into account adolescents’ perceptions as well as
that of the parents. The majority of studies on family
conflict focus on parental perceptions, and do not
assess adolescents’ or children’s points of view.
Greater understanding of the perceptions of conflict
of both parents and adolescents will lead to better
understanding of the choice, and successes, of
different types of coping strategies employed by
adolescents. For example, it is possible that
adolescents perceive interparental conflict as being
severe in a situation where parents rate conflict as
being low. The volume of an argument might be a key
indicator of conflict to an adolescent irrespective of
the content. In contrast, parents may perceive the
content of an argument to be a more significant
indicator of the level of conflict intensity. Adolescents
uncomfortable with emotional expression may
perceive any type of high emotional expression to be
conflict, whereas their parents may be engaging in
this behaviour to effectively deal with disagreements.
Further, emotional expression in the family is a
predictor of illness outcomes in psychological
disorders (Van Humbeeck, Van Audenhove, De Hert,
Pieters, & Storms, 2002). Past research has found that
level of hostility, presence of physical aggression, and
children’s age are the best predictors of children’s
appraisals of interparental conflict (Grych, 1998).
Few studies have investigated differences in
perceptions held by adolescents and their parents,
particularly in relation to interparental conflict. Much
of the research conducted on perceptions of family
conflict has been focused on younger children, older
adolescents and/or has relied on different conflict
measures for parents and adolescents (Aquilino, 1986;
Grych, 1998; Grych, Seid, & Fincham, 1992; Joshi,
2001; Tschann et al., 2002). Young adolescent
perceptions of interparental conflict have contributed
unique variance to adolescent adjustment beyond
what was accounted for by parental reports of conflict
(Wierson, Forehand, & McCombs, 1988). Other
studies that have assessed only the influence of
interparental conflict in adolescents have relied on
parental reports of conflict (Neighbours et al., 1997),
while those that have assessed both adolescent and
parental perceptions have relied on non-standardised
measures of conflict assessment (Harold & Conger,
1997).
Abstract
Research indicates adolescent adjustment problems are
associated with high levels of interparental conflict.
This study examined parents’ and adolescents’
perceptions of interparental conflict and the
relationship of these perceptions to adolescent
adjustment through a survey of adolescent secondary
students and their parents. Sixty-two adolescents and
62 parents participated in the study. Adolescents
completed three self-report measures of psychological
adjustment, a demographic questionnaire and the
Consensus and Cohesion subscales of the Dyadic
Adjustment Scale (DAS). Parents completed the same
subscales of the DAS and a demographic questionnaire.
There were no significant differences between the
levels of interparental conflict perceived by adolescents
and parents. However, adolescent perceptions of
interparental conflict were a better predictor of
adolescent psychological adjustment when compared to
parental perceptions of conflict. The results of this
study emphasise the importance of adolescent
perceptions of interparental conflict, and provide
support for both parental and adolescent reports of
family functioning to be taken into account in future
clinical studies.
The link between interparental conflict and adolescent
maladjustment was established in the divorce
literature of the 1970s (Emery, 1982). The
examination of such conflict in the family
environment led to the consensus that interparental
conflict was more detrimental to adolescent
adjustment than a divorced family structure (Emery,
1982; Forehand, Long, & Brody, 1988; Forehand,
Thomas, Wierson, Brody, & Fauber, 1990; Lussen,
1988; Mechanic & Hansell, 1989; Neighbours,
Forehand, & Bau, 1997).
Interparental conflict has been associated in the
literature with several different expressions of
adolescent psychological health issues, including
social withdrawal (Forehand, McCombs, Long,
Brody, & Fauber, 1988), lowered self-esteem,
lowered self-perceived competence, increased
anxiety, less internal control, and higher levels of
anger and aggression (Enos & Handal, 1986; Kelly,
1998; Mechanic & Hansell, 1989), increased maturity,
pseudoadolescence, and increased vulnerability
(Lussen, 1988; Wallerstein & Kelly, 1976),
depression and physical symptomatology (Kelly,
1998; Krishnakumar & Buehler, 2000; Mechanic &
Hansell, 1989), and poorer academic functioning and
problem solving skills (Long, 1987). Such research
has concentrated on associating adolescent
psychopathology with interparental conflict using
35
DAVERN, STAIGER, & LUK
36
Both adolescent and parental perceptions of body
image and the family environment have been
investigated in eating disorders research. Congruence
between mother and daughter perceptions of body
image was assessed by Rupp and Jurkovic (1996)
within the families of bulimic daughters and control
families. Bulimic daughters were less able to predict
how their mothers would describe their personality
than were daughters in the control families. Further,
the descriptions that mothers of bulimic daughters
gave of their daughters were significantly different to
their
daughters’
self-perceptions.
Similar
misperceptions have been found by Geller,
Srikameswaran, Zaitsoff, Cockell and Poole (2003) in
body image ratings provided by adolescent daughters
and their parents. Greater discrepancies between
parents’ and daughters’ ratings of daughters’ feelings
about their bodies were associated with greater levels
of daughter body dissatisfaction. Daughters who were
most unhappy with their bodies had parents who were
not aware of their daughters’ feelings regarding their
bodies. Thus, in eating disorder research, a lack of
congruence in parental and adolescent perceptions has
been associated with mental health issues in
adolescence. The eating disorder literature emphasises
the importance of different perceptions between
parents and adolescents, and the importance of such
perceptions in relation to mental health. It is also
possible that differences in adolescent and parental
perceptions of interparental conflict also exist, and
may produce different associations with adolescent
mental health
Existing literature in interparental conflict has been
limited by a reliance on parental perceptions of
conflict, the effect of conflict on children rather than
adolescents, and the use of separate conflict measures
for parents and adolescents. The present study
investigated the relationship between parental and
adolescent perceptions of interparental conflict and
adolescent adjustment using the same standardised
conflict measure. It was hypothesised that adolescent
perceptions of conflict would be more important to
adolescent adjustment than parental perceptions of
conflict.
(Coopersmith, 1967), the Nowicki-Strickland Locus
of Control Scale (Nowicki & Strickland, 1973), and
the Comprehensive Quality of Life Scale for Students
in Grades 7–12 (Cummins & McCabe, 1994). The
adolescents also completed a demographic
questionnaire and the Consensus and Cohesion
subscales of the DAS (Spanier, 1976). All measures
have sufficient reliability and validity (Cummins &
McCabe, 1994; Nowicki & Strickland, 1973; Omizo
& Omizo, 1988; Robinson & Shaver, 1975; Spanier,
1976). The three separate self-report measures of selfesteem, locus of control, and QOL were aggregated to
provide an indication of adolescent adjustment.
Parents completed the same two subscales of the DAS
and a demographic questionnaire. The information
obtained from the demographic questionnaire related
to age, gender, and absence of significant family
members from the family home. Parents were also
asked to estimate the household’s current income.
Method
Adolescents perceived more extreme levels of
interparental conflict (score range = 25.00–82.00, M =
60.62, SD = 11.85) than did their parents (score range
= 40.00–76.00, M = 61.76, SD = 7.71), as indicated in
the greater score range and standard deviation of
adolescent perceptions of conflict. A one-way
analysis of variance revealed no statistically
significant differences between the amount of
interparental conflict perceived by adolescents or by
parents, F (1,115) = .38, p > .05.
Initial investigation suggested a significant
relationship between adolescent perceptions of
interparental conflict and parental perceptions of
interparental conflict (r = .43, p= <.001). There was a
stronger relationship between adolescent perceptions
of interparental conflict and adolescent health (r = .41,
p= <.01) than there was between parental perceptions
of conflict and adolescent health (r = .20, p > .01).
Participants
Adolescent participants and one of their parents were
recruited from two independent secondary schools.
Sixty-two adolescents and 62 parents participated in
this study. Each adolescent participating in the study
was aged between 14 and 16 years, and was in either
Year 9 or Year 10 in secondary school. Mothers
comprised 89% of the parent population, while 11%
of fathers participated. Adolescents and parents were
predominantly from an Anglo-Saxon background in a
middle class area of Melbourne, Australia. Of the
adolescents, 58% were females and 42% were male
(M = 14.5 years, SD = 0.6).
Measures
Adolescents completed three self-report measures of
adjustment: the Coopersmith Self-Esteem Inventory
Design and Procedure
Students in their middle years of secondary education
and their parents were invited to participate in the
study. These students were recruited from two
independent secondary schools in metropolitan
Melbourne, Australia. Participation was required from
an adolescent and a parent in each family.
In order to include single parent families, only one
parent from each adolescent’s family was required to
participate. Parents were asked to complete the
shortened version of the DAS. The instructions on this
measure stipulated that the questionnaire was to be
answered according to the present situation between
the adolescents’ parents, irrespective of whether these
parents
were
living
together
or
were
separated/divorced. This is because interparental
conflict has been found to be detrimental to
adolescent health regardless of family structure
(Wallerstein & Kelly, 1980). After the researcher
received the completed and returned parent
questionnaires, adolescents completed the four selfreport measures during school hours.
Results
Perceptions of interparental conflict
The aim of the present investigation was to examine
adolescent perceptions of interparental conflict and
parental perceptions of interparental conflict and their
relationship to adolescent psychological adjustment.
To examine this, a standard multiple regression
analysis was performed. The model using adolescent
perceptions of conflict (β = .45, p = <.01) and parental
perceptions of conflict (β = –.08) as predictors
explains a significant proportion of adolescent
adjustment (R2 = .17, p < .005) as indicated by the
significant F-ratio, F (2, 48) = 5.08, p < .01.
To test for the unique contributions of adolescent
perceptions of interparental conflict, sr2 significant
contributions were calculated. Of the two independent
variables, adolescent perceptions of interparental
conflict was the only variable that contributed
significantly to adolescent adjustment (sr2 = .40).
Parental perceptions of conflict did not contribute
significantly to adolescent adjustment (sr2 = –.08).
The two variables in combination contributed another
.01 in shared variability. In total, 17% of the
variability of adolescent adjustment was predicted by
knowing adolescent and parental perceptions of
interparental conflict. The results of this standard
multiple regression suggest that adolescent
perceptions of interparental conflict are a better
predictor of adolescent health than are parental
perceptions of interparental conflict.
Discussion
Adolescent perceptions of interparental conflict were
found to be a better predictor of adolescent adjustment
than were parental perceptions of interparental
conflict. However, adolescent and parental
perceptions of interparental conflict were not
significantly different.
These results suggest that there is little rationale for
the sole reliance on parental reports of conflict and/or
adolescent health in earlier studies of adolescent
functioning. Perhaps in the past, parental reports of
adolescent functioning were thought to be more
objective than adolescent self-reports. Alternatively,
parental perceptions of the family environment may
have been assumed to be similar enough to adolescent
perceptions of the family environment (Neighbours et
al., 1997). The results of this study suggest that
adolescent perceptions of the family environment may
be more important than parental perceptions of the
family environment in predicting adolescent
psychological adjustment. The importance of
adolescent perceptions of conflict is consistent with
Wierson et al. (1988) who also found that adolescent
perceptions of conflict contributed unique variance in
the prediction of adolescent adjustment.
Careful examination of the data suggests that
adolescent perceptions are more extreme when
compared to parental perceptions, even though both
are moderately correlated. Parents and adolescents
reported similar levels of conflict when analysed as a
group, as indicated by non-significant differences in
the group data of parents and adolescents. However,
important information can be drawn from the
individual differences within the adolescent group,
with more variation in the level of discord reported by
37
adolescents. The increased variation in the results of
the adolescents does suggest that there were
individual differences in what parents and adolescents
reported. However, these differences were not large
enough to produce a statistical difference in the group
data. The individual differences between what parents
and adolescents reported may have been a result of
adolescents having different perceptions of, and
attributing different meanings to, the same event
experienced by themselves and their parents.
Differences in adolescent–parent perceptions are
not unique to interparental conflict. They have also
been found in body image perceptions of parents and
adolescents in families where a daughter had been
diagnosed with an eating disorder (Geller et al., 2003;
Rupp & Jurkovic, 1996). These studies also provide
support for the idea that sole reliance on parental
perceptions of the family environment may not
provide an accurate picture of the actual amount of
conflict present in the family. The findings of the
current study add to the literature emphasizing the
importance of collecting both parental and child
perceptions of family environment or specific
psychological disorders such as bulimia nervosa.
The perceptions of a stressful event are related to
its impact on an individual (Grych & Fincham, 2000).
If adolescents perceive more or less interparental
conflict than their parents, or associate it with a
different meaning, it could explain why adolescent
reports of conflict are a better predictor of adolescent
psychological health. Grych and Fincham (1993)
suggested that children’s perceptions of the threat
posed by conflict, their ability to cope effectively with
conflict, and attributions regarding the cause of
conflict were particularly important for shaping
emotional and behavioral responses. Neighbours and
colleagues (1997) also suggested that the young
adult’s perspective of interparental conflict may be
most pertinent to his or her functioning. Future studies
could investigate adolescents’ appraisals of perceived
conflict and their attributions regarding its cause.
The findings emphasise the importance of
adolescent perceptions of interparental conflict. It is
common practice for clinicians to take into account
differences in perceptions of the family environment
during assessment. The results of this study provide
support for this practice, but research continues to rely
largely on parental reports of conflict. The current
results provide support for both parental and
adolescent reports of family functioning to be taken
into account in future clinical studies. These results
may also be of interest to those developing assessment
guidelines in child protection and community services
policy.
DAVERN, STAIGER, & LUK
38
References
Aquilino, W. S. (1986). Children’s perceptions of
marital interaction. Child Study Journal, 16, 159172.
Coopersmith, S. (1967). The antecedents of selfesteem. San Francisco, CA: W. H. Freeman.
Cummings, E. M., & Cummings, J. L. (1998). A
process-oriented approach to children’s coping with
adults’ angry behavior. Developmental Review, 8,
296-321.
Cummins, R. A., & McCabe, M. P. (1994). The
Comprehensive Quality of Life Scale: Instrument
development and psychometric evaluation on
tertiary staff and students. Educational and
Psychological Measurements, 54, 372-382.
Emery, R. E. (1982). Interparental conflict and the
children of discord and divorce. Psychological
Bulletin, 92, 310-330.
Enos, D. M., & Handal, P. J. (1986). The relation of
parental marital status and perceived family conflict
to adjustment in white adolescents. Journal of
Consulting and Clinical Psychology, 54, 820-824.
Forehand, R., Long, N., & Brody, G. (1988). Divorce
and marital conflict: Relationship to adolescent
competence and adjustment in early adolescence. In
E. M. Hetherington & J. D. Arasteh (Eds.), Impact
of Divorce, Single Parenting, and Stepparenting on
Children. NJ: Lawrence Erlbaum Associates.
Forehand, R., McCombs, A., Long, N., Brody, G., &
Fauber, R. (1988). Early adolescent adjustment to
recent parental divorce: The role of interparental
conflict and adolescent sex as mediating variables.
Journal of Consulting and Clinical Psychology, 56,
624-627.
Forehand, R., Thomas, A., Wierson, M., Brody, G., &
Fauber, R. (1990). Role of maternal functioning and
parental skills in adolescent functioning following
parental divorce. Journal of Abnormal Psychology,
99, 278-283.
Geller, J., Srikameswaran, S., Zaitsoff, S. L., Cockell,
S. J., & Poole, G. D. (2003). Mothers’ and fathers’
perceptions of their adolescent daughters’ shape,
weight and body esteem: Are they accurate?
Journal of Youth and Adolescence, 32, 81-87.
Grych, J. H. (1998). Children's appraisals of
interparental conflict: Situational and contextual
influences. Journal of Family Psychology, 12, 437453.
Grych, J. H., & Fincham, F. D. (1993). Children’s
appraisals of marital conflict: Initial investigations
of the cognitive-contextual framework. Child
Development, 64, 215-230.
Grych, J. H., & Fincham, F. D. (2000). Interparental
conflict and child adjustment: Testing the
mediational role of appraisals in the cognitivecontextual framework. Child Development, 71,
1648-1661.
Grych, J. H., Seid, M., & Fincham, F. D. (1992).
Assessing marital conflict from the child’s
perspective: The Children’s Perception of
Interparental Conflict Scale. Child Development,
63, 558-572.
Harold, G. T., & Conger, R. A. (1997). Marital
conflict and adolescent distress: The role of
adolescent awareness. Child Development, 68, 333350.
Joshi, A. (2001). Parents’ and children’s perceptions
of interparental conflict resolution. Psychological
Reports, 88, 943-946.
Kelly, J. (1998). Marital conflict, divorce, and
children’s adjustment. Child Custody, 7, 259-271.
Krishnakumar, A., & Buehler, C. (2000). Interparental
conflict and parenting behaviours: A meta-analytic
review. Family Relations, 49, 25-44.
Larson, R. (1989). Is feeling “in control” related to
happiness in daily life? Psychological Reports, 64,
775-784.
Long, N. (1987). Self-perceived and independently
observed competence of young adolescents as a
function of parental marital conflict and recent
divorce. Journal of Abnormal Child Psychology, 15,
15-27.
Lussen, L. B. (1988). The female adolescent’s
unconscious experience of parental divorce. Smith
College Studies of Social Work, 58, 101-121.
Mechanic, D., & Hansell, S. (1989). Divorce, family
conflict and adolescents’ well-being. Journal of
Health and Social Behaviour, 30, 53-63.
Neighbours, B. D., Forehand, R., & Bau, J. (1997).
Interparental conflict and relations with parents as
predictors of young adult functioning. Development
& Psychopathology, 9, 169-187.
Nowicki, S., & Strickland, B. R. (1973). A locus of
control scale for children. Journal of Consulting
and Clinical Psychology, 40, 148-154.
Omizo, M. M., & Omizo, S. A. (1988). The effects of
participation in group counseling on self-esteem
and locus of control among adolescents from
divorced families. The School Counselor, 36, 54-60.
Robinson, J. P., & Shaver, P. R. (1975). Measures of
social psychological attitudes. MI: Institute for
Social Research.
Rupp, G., & Jurkovic, G. (1996). Familial and
individual
perspective-taking
processes
in
adolescent females with bulimic symptomatology.
American Journal of Family Therapy, 24, 75-82.
Spanier, G. B. (1976). Measuring dyadic adjustment:
New scales for assessing the quality of marriage
and similar dyads. Journal of Marriage and the
Family, 38, 15-28.
Tschann, J. M, Flores, E., Martin, B. V., Pasch, L. A.,
Baisch, E. M., & Wibbelsman, C. J. (2002).
Interparental conflict and risk behaviors among
Mexican American adolescents: a cognitiveemotional model. Journal of Abnormal Child
Psychology, 30, 373-385.
Van Humbeeck, G., Van Audenhove, C., De Hert, M.,
Pieters, G., & Storms, G. (2002). Expressed
emotion: A review of assessment instruments.
Clinical Psychology Review, 22, 321-341.
Wallerstein, J. S., & Kelly, J. B. (1976). The effects of
parental divorce: Experiences of the child in later
latency. American Journal of Orthopsychiatry, 46,
256-269.
Wallerstein, J. S., & Kelly, J. B. (1980). California’s
Children of Divorce. Psychology Today, 67-76.
Wierson, M., Forehand, R., & McCombs, A. (1988).
The relationship of early adolescent functioning to
Perceptions of interparental conflict
parent-reported
and
adolescent-perceived
interparental conflict. Journal of Abnormal Child
Psychology, 16, 707-718.
Correspondence to:
Dr. Petra Staiger
School of Psychology, Deakin University,
Burwood 3125. Australia.
39
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 40-45
Theoretical Frameworks for Relationship Transitions and the Predictors of
Successful Transitions
Amanda E. Ferguson (amanda@lifethatworks.com)
PhD student, Department of Psychology
Macquarie University, Sydney NSW 2052 Australia
term relationships is here defined as one in which a
couple is highly satisfied over longitudinal time,
(Lerner, 2003). ‘Success’ in transitions between
relationships, is here defined as leading to personal
growth (Schneller, 2003). Thus transitions within and
between relationships require individuals to become
increasingly competent if they are to negotiate these
changes successfully.
Romantic relationships have a significant impact on
adult development and adaptation; ‘developmental
trajectories are enmeshed with relationship
trajectories’ (Laursen, 1997, p. 641). Following Freud
(1905/1962, cited in Schneller, 2003), both Erikson
(1950, cited in Schneller, 2003) and Sullivan (1953,
cited in Schneller, 2003) created elaborate theoretical
frameworks where psychological competencies were
organised according to social tasks and interpersonal
relationships. Erikson’s influential theory of life tasks
over the life time found the pursuit of intimacy and
identity especially important during later adolescence.
The task to achieve intimacy is particularly important
to young women yet the task to achieve identity goals
such as tertiary qualifications and career, can present
a conflict. Facing a new life task involves making a
transition into a new life period and anxiety and
conflict is associated with this (Cantor, Acker, &
Cook-Flannagan,
1992).
Hence
relationship
transitions have a significant impact on individuals.
Abstract
Increasing rates of divorce and a decline in the
traditional marriage for life is seen particularly in the
X-generation (1961-1976). Yet a return to traditional
marriage values has also been found in the younger Ygeneration (1976-1991). Both X and Y – generation
have high hopes for the future of their relationships.
Despite the various trends towards and away from long
term relationships, there is an increasing incidence of
transitions, both within relationships and between
relationships. Individuals must become increasingly
sophisticated in relationship knowledge, skills and
attitudes if they are to negotiate these changes
successfully. Transitions are considered from four
theoretical bases: social constructionism, attribution
theory, attachment theory and theories of loss and
renewal. A review of the literature reveals the main
factors required for the success of long term
relationships as well as successful transitions between
relationships. Communication training is perhaps the
one skill found universally in marital therapies, yet
requires ongoing qualitative research to identify the
factors
that
couples
find
improves
their
communication. Finally, suggestions are made for
future research and approaches to relationship
counselling.
Globalisation, and an accelerating pace of
technological and social change has brought with it
increasing rates of divorce and a decline in the
traditional marriage for life. This is particularly seen
in the X-generation (1961-1976) (Bumpass, Sweet, &
Martin, 1990; Glick, 1989; Martin & Bumpass, 1989).
Yet a return to traditional marriage values has also
been found in the younger Y-generation (1976-1991).
Both generations have high hopes of relationships and
marriage as the foundation of their future happiness.
Yet the expectations of these generations do not
match actual experiences. Research shows that these
relationships will probably not last and provide a
future foundation (Arnett, 2000; Larson, Wilson,
Brown, Furstenberg & Verma, 2002). The divorce
rate of first marriages in the USA has increased to
52% and for second marriages is 10% higher
(Goldberg, 2003), making relationship transitions a
major social issue.
Despite the various trends towards and away from
long term relationships, there is an increasing
incidence of transitions, both within relationships and
between relationships. Yet there continues to be a lack
of research on the process that adults traverse when
making transitions between significant relationships,
particularly
when
re-establishing
intimate
relationships after divorce (Schneller, 2003).
Longitudinal research has shown how marriages also
change over time (Goldberg, 2003). ‘Success’ in long
Theoretical Frameworks
Recent literature has begun to discuss the complex
transition through divorce and the establishment of
post-divorce relationships. Much of the research on
adjustment to divorce operationalises a stress and
coping framework and focuses on the distressing
aspects, such as loss of social and financial structures,
which applies mostly to initial stages of divorce.
Recent findings show that individuals eventually
adjust better to divorce depending on their personal
perceptions and formation of new relationships (Ross,
1995; Wang & Amato, 2000). Yet research still
neglects to explore the process by which individuals
move from martial breakdown to new relationships
(Schneller, 2003). Schneller (2003) conceptualised
divorce as an emotional and cognitive process that can
potentially promote individual development and
renewal. Research that does extend analysis beyond
the initial stages of stress focuses on the structures of
new marriages and families. Notable exceptions
include Furstenberg & Spanier (1987) who found that
divorced individuals perceived relationships less
romantically and more practically than those in their
first marriages. Attachment theorists have found
tentative evidence for some changes in attachment
representations (Schneller, 2003). Divorced people
40
Relationship transitions
have been found to make diverse choices that may
reflect a newfound sense of freedom, such as dating,
cohabitation as well as commitment (Weiss, 1975).
Yet there has remained a lack of theoretical
framework to explain the transitional process and why
some people reinvest in close relationships and others
do not (Schneller, 2003).
Scheneller (2003) has drawn together four
theoretical frameworks to explain findings. Firstly,
social constructionism provides an explanation for
how the reaction to divorce may be impacted by
language, as seen by the ‘explanations an individual
makes, by social interchange with others, and by the
cultural meanings of marriage and divorce that have
influenced a person’s thoughts and perceptions’.
Social constructionism emphasises how people create
knowledge and meaning through ‘languaged
interaction’ (Schneller, 2003, p.6). This is highly
relevant to research on relationship transitions and
possible changes in regard for and meaning found in
intimate relationships. Social constructionism helps us
to understand how reaction to a divorce can be
influenced by explanations given, communication
with others and cultural influences in the meanings of
marriage and divorce.
Secondly, attribution theory may explain why
individuals make different choices about post-divorce
intimacy. Attribution theory contributes a systematic
approach to understanding how people explain the
causes of life events (Bensen, Arditti, Reguero de
Atiles, & Smith, 1992, cited in Schneller, 2003).
Individuals may construe their divorce in ways that
are positive or negative: either damaging trust,
promoting a sense of mastery and optimism regarding
future relationships, or encouraging personal change.
The attributions that individuals develop about the
breakdown of relationships have been found to
influence their cognitions, affect and behaviour about
future relationships (Grych & Fincham, 1992, cited in
Scnheller, 2003). Research has found that individuals
who make interactive attributions such as a lack of
communication, lifestyle differences or values were
associated with better post-divorce adaptation than
those who blamed themselves or their ex-partners
(Newman & Langer, 1981, cited in Schneller, 2003;
Ty & Frazier, 2003). These individual attribution
patterns are about the cause of marital failure and are
related to decisions about whether to engage in new
intimate relationships. Thus, attribution theory shows
how an optimistic or pessimistic attitude to new
relationships is linked to attributions, conceptions and
relationship transitions.
Thirdly, attachment theory provides a conceptual
basis for understanding interactions between security
and change in relationship transitions. Attachment
theory emphasises that our earliest experiences with
caretakers teaches us what to expect in intimate
relationships, from which we form mental
representations or working models of relationships
(Van IJzendoorn, 1995, cited in Schneller, 2003).
While these mental representations are consistent and
stable components for daily functioning, they are also
flexible structures open to change. Relationship
break-ups are among the experiences frequently cited
41
as being able to cause fundamental changes in
attachment styles (Bakermans-Kranenburg & van
IJzendoorn, 1997; Feenney, 1999, cited in Schneller,
2003). One four-year study found that break-ups
consistently predicted change from secure to insecure
attachment (Kirkpatrick & Hazan, 1994). However
researchers have not examined how these changes to
attachment styles occur or the change to individuals’
feelings, opinions and desires for future intimate
relationships.
Finally, loss and renewal theories offer a conceptual
basis for understanding how our experiences of loss
evolve and transfer to post-divorce relationships.
Theorists have focused on the processes of recovering
from the loss of a primary relationship, applying
Bowlby’s (1961, cited in Schneller, 2003) theory of
mourning. Three phases of mourning must be
negotiated: (1) the urge to recover the lost object, (2)
disorganisation, and (3) reorganisation (Gray &
Shields, 1992 cited in Schneller, 2003). This shows
how cognitions and feelings change over the
mourning period, or the transitions that an individual
makes. Recent findings extend this stage theory to a
more interpretive process of how the individual finds
meaning from loss (Riessman, 1990, cited in
Schneller, 2003), and possibly growth and renewal as
well, as divorce is an opportunity to change negative
patterns. Traditional mourning theories implied a
necessity of relinquishing the lost loved one in order
to begin a new relationship. However, individuals are
now thought to experience a continuity of relationship
with the lost one as an integrated internal
phenomenon (Baker, 2001, cited in Schneller, 2003).
The lost relationship may continue to be re-interpreted
and inform new experiences as part of positive
adjustment (Madden-Derdich & Arditti, 1999, cited in
Schneller, 2003). Thus, it is the interpretation of
relationships rather than the loss itself that is critical
to transitions (Schneller, 2003).
Scheneller’s studies contribute to understandings
about the connection between divorce experiences and
post-divorce intimacy. Scheneller (2003) draws three
conclusions from her study. First, divorce serves
consistently as a catalyst for self-analysis, or
‘interpretation and personal growth’ (Schneller, 2003,
p.iii), and many authors are now focusing on how
people can reconstruct a higher quality of life
following break-ups (Robbins, Caspi & Moffitt, 2002;
Schneller, 2003; Tashiro & Frazier, 2003). These
sources have helped people to understand divorce
responses and differences such as gender and age. The
social context within which divorce occurs is an
important influence on the interpretative process, and
the stigma that divorce still carries in our society
makes this process more of a challenge. Second, the
interpretive process shapes ‘adult’s post-divorce
perceptions and experiences in intimate relationships’
in positive ways (Schneller, 2003, p,iii). Deliberate
changes are made in communication patterns,
interactions,
attitudes
and
expectations
in
relationships. Third, fundamental changes in mental
representations occur. These changes are linked to
gender: women come to view themselves as more
42
FERGUSON
assertive and men as more egalitarian and responsible
for relationship maintenance (Schneller, 2003).
Predictors of Transitions
Transitions within and between relationships have
been considered in terms of changes in levels of
intimacy. Changes in intimacy have effects on passion
such that increases in intimacy produces stronger
passion, whereas stable intimacy (high or low)
produces low passion (Baumeister & Bratslavsky,
1999). Romantic beliefs have been found not to be
linked to longevity of relationships, yet scores on the
romanticism scale tend to decrease over time. Couples
who break up have been found to experience a
substantial decrease in their romantic beliefs from
before to after the break-up (Sprecher & Metts, 1999).
Some researchers argue that degree of dependence
within the relationship is the primary issue in
understanding break-up decisions. Dependency is
found to increase ‘when important outcomes in the
current relationship are not available elsewhere’, and
to predict those who stay in relationships, no matter
how dissatisfying the relationship might be (Drigotas
& Rusbult, 1992, p.1).
Pre-marital relationship break-ups have been
attributed to sources within the individual, the
relationship and the social network environment
(Felmlee, Sprecher & Bassin, 1990). Significant
predictors included the level of comparison for
alternatives, the amount of time spent together, racial
differences, support from partner’s social network,
and duration of the relationship(Felmlee, Sprecher &
Bassin, 1990). Social exchange, similarity and social
network theories all contribute towards an explanation
of pre-marital break-ups (Felmlee, Sprecher & Bassin,
1990). Other factors include self-control, partner’s
control, control external to the relationship, partner’s
lack of caring, instability and lack of ability to commit
(Hortacsu & Karanci, 1987), unequal involvement in
the relationship and discrepant age, educational
aspirations, intelligence and physical attractiveness.
The desire to break-up was seldom mutual. Women
are more likely to find problems in these relationships
and are somewhat more likely to end them (Hill,
Rubin & Peplau, 1976).
People who break-up because of affairs have
reported higher dissatisfaction, attribute their own and
their partners’ extradyadic relationships more to
motives of aggression and deprivation, and cite a
higher level of conflict generated by these
relationships. These people tend to overlook their own
extradyadic involvements as contributing significantly
to the break-ups. Men, much more often than women,
tend to blame their break-ups on their partners’
extradyadic relationships. Some evidence has been
found that partners’ extradyadic involvement had a
stronger influence on the decision to break-up
(Buunk, 1987).
Findings show that the very ‘qualities that are
disliked in a partner, and that are implicated in a
break-up are often very similar to those that were
found to be initially attractive’ (Felmlee, 1998). This
phenomenon has been termed ‘fatal attraction’, where
the relationships are doomed from the beginning.
These ‘disenchanting attractions’ occur because of
contradictory dilemmas faced by those in intimate
relationships, for instance, a desire for intimacy
combined with a need for independence (Felmlee,
1998). Brickman (1987) suggested that the integration
of negativity is the greatest challenge of intimacy
(cited in Thompson, 1995). Ambivalence is the
presence of at least moderate amounts of positive and
negative attitudes regarding partner attributes and has
been found to be a predictor of break-ups, over and
above feelings of love for a partner, or the incidence
of conflict in the relationship. Its effect has been
found to be moderated by individuals’ commitment to
their relationships. Ambivalence acts as a catalyst
either facilitating or impeding the growth of intimacy
(Thompson, 1995). Other theorists concur on the need
for couples to learn to live with negativity and
ambivalence, and even to see virtue in faults (Harvey
& Weber, 2001).
Predictors of Success
Little research has been conducted on factors that
determine the development and maintenance of longterm relationships. For instance, there has been
relatively little research into commitment compared to
other relationship constructs such as love and
attachment, trust and satisfaction. Research into the
predictors of sustained intimacy in marriage shows
that the partners must each have attained an adequate
degree of individuality and can also allow them selves
to become physically and emotionally close. Changes
by one must be accommodated by the other. Hence
changes should be of a complimentary nature for
which both partners are carefully prepared (Birtchnell,
1986).
Findings have shown that positive couple
agreement in marriages significantly decreases over
time in five aspects: personality issues,
communication, conflict resolution, leisure activities
and sexual expectations. However, agreement
increases significantly in financial management,
marriage expectations, children and parenting, and
spiritual beliefs. Projections of satisfaction levels are
linked with degree of functionality of the relationship
(Goldberg, 2003). While marital satisfaction tends to
lessen over time, skills and insights can be gained to
prevent the erosion of satisfaction (Clements,
Cordova, Markman & Laurenceau, 1997; Dunn,
2002). Thus the need for ongoing skilling of couples
is critical.
Communication training is perhaps the one skill
found universally in marital therapies. Components
that couples endorse in their communication include:
identifying factors that improve communication over
the course of the marriage; recognising that the timing
of communication is important; a cooling off
separation period before resolving arguments when
they escalate. Difficulties that have been identified
include lack of understanding of what is being
communicated, lack of time together and time spent
on improving communication. (Lerner, 2003).
Conflict is often avoided in relationships, yet conflict
resolution is associated with romantic satisfaction
(Cantor, Acker & Cook-Flannagan, 1992). Some
Relationship transitions
theorists base their assessments of marriages on the
style of conflict employed (Gottman, 1998). Thus,
ongoing research is required to identify the factors
that couples find improves their communication.
Individual differences have been linked to
relationship outcomes. These are factors such as
personality differences: some people tend to be happy
across relationships while others are not (Robins,
Caspi & Moffitt, 2002); and birth-order, with firstborns showing the most irrational beliefs about
relationships and last-borns the least (Sullivan &
Schwebel, 1996). Social influences include network
approval (Sprecher & Felmlee, 2000). Helpful
attitudes include seeing relationships as a process
(Weigel & Murray, 2000), and viewing individual
performance differences from a ‘team’ point of view
(Beach, Whitaker, Jones & Tesser, 2001). Resilience
appears in recent literature as determining relationship
happiness and longevity. Skills can be learnt to
develop and maintain resilience (Reivich & Shatte,
2002).
Empathy has been found to be critical to the
harmony in relationships, promoting positive social
interactions and inhibiting antisocial behaviour
(Sezov, 2002). Many marital intervention programs
have been based on building empathy-based skills
(Rogge, Cobb, Johnson, Lawrence & Bradbury,
2002). Increased empathy has been positively related
to relationship satisfaction (Long, Angera, Carter,
Nakamoto & Kalso, 1999). Existential dilemmas must
importantly be faced and addressed in couple
transformation (Dunn, 2002). Mortality salience, or
death reminders, of one’s partner has been found to
lead to more willingness to work on the relationship
(Miller, 2003; Taubman, Findler & Mikulincer, 2002).
Therapeutic Implications and Future
Research
The implications of these findings for therapeutic
interventions are significant. The experience of
marital dissolution is a common reason for individuals
to seek therapy. Marital life is more complex today
and represents a significant challenge for couples and
therapists. Couples need ongoing education and
therapists must provide this dual role (Goldberg,
1985). The findings discussed here show how
important it is that individuals interpret their divorce
in terms of themselves as well as post-divorce
relationships, where the promotion of a more
objective assessment of the relationship rather than
blaming individuals can promote optimism about
future relationships. Many divorced individuals have
difficulties in redefining themselves and their close
relationships after divorce. Therapists can formulate
better intervention processes the more they understand
the complexities of the process that divorced people
undergo in making the transition from marital
dissolution to other relationships. Educational systems
can also be informed by these findings, and in
particular, Schneller’s (2003) constructive approach to
post-divorce experiences as developmental processes.
In this way, eventually the increase in major
transitions might attract less stigmatism.
43
Further research is needed to support the usefulness
of the theoretical frameworks posed here and to more
fully examine the interpretive processes that
individuals undergo as part of transitions between
significant relationships. Future research would be
well directed as to the more flexible and inventive
approaches people take to relationships after divorce
(Weiss, 1975) and to the pros and cons of this trend
for the breakdown of traditional family units, and the
psychological and social effects on individuals,
couples and children. Finally, ongoing research is
required to develop the understanding of how to help
adults traverse relationships and to develop better
social resources and competencies for adult
relationships.
References
Arnett, J.J., (2000). High hopes in a grim world:
Emerging adults’ view of their futures and
‘Generation X’. Youth & Society, 31(3), 267-687.
Backermans-Kranenburg, M. J., & van IJzendorn, M.
H. (1997). Adult attachment and the break-up of
romantic relationships. Journal of Divorce and
Remarriage, 27(3/4), 121-139.
Baumeister, R. F., & Bratslavsky, E. . L. (1999).
Passion, intimacy, and time: Passionate love as a
function of change in intimacy. Personality &
Social Psychology Review, Vol 3(1), 49-67.
Beach, S. R. H., Whitaker, D. J., Jones, D. J., &
Tesser, A. (2001). When does performance
feedback prompt complimentarity in romantic
relationships? Personal Relationships, 8(3).
Birtchnell, J. (1986). The imperfect attainment of
intimacy: A key concept in marital therapy. Journal of
Family Therapy, 8(2), 153-172.
Bumpass, L., Sweet, J. A., & Martin, T. C. (1990).
Changing patterns of remarriage. Journal of
Marriage and the Family, 52, 747-756.
Buunk, B. (1987). Conditions that promote breakups
as
a consequence of extradyadic involvements. Journal
of Social and Clinical Psychology,5(3), 271-284.
Cantor, N., Acker, M., & Cook-Flannagan, C. (1992).
Conflict and preoccupation in the intimacy life task.
Journal of Personality and Social Psychology,
63(4), 644-655.
Clements, M. L., Cordova, A. D., Markman, H. J. &
Laurenceau, J. (1997). The erosion of marital
satisfaction over time and how to prevent it. In
Sternberg, R. J. (Ed), & Hojjat, M. (Ed),
Satisfaction in close relationships. (pp. 335-355).
New York, NY, US: Guilford Press. xiii, 429 pp.
Drigotas, S. M., & Rusbult, C. E. (1992). Should I
stay or should I go? A dependence model of
breakups. Journal of Personality & Social
Psychology 62(1), 62-87.
Dunn, A. B., (2002). Existential dilemmas and skill
building in couple therapy. In Kaslow, F. W. (Ed),
Comprehensive handbook of psycotherapy:
Interpersonal/humanistic/exisitential John Wiley &
Sons, Inc, New York, NY, US.
Felmlee, D. H. (1998). Fatal Attraction. In Spitzberg,
B. H. (Ed), & Cupach, W. R. (Ed), The dark side of
close relationships.
44
FERGUSON
Felmlee, D., Sprecher, S., & Bassin, E. (1990). The
dissolution of intimate relationships: A hazard
model. Social Psychology Quarterly, 53(1), 13-30.
Furstenberg, Jr., F. F.,, & Spanier, G. B. (1987).
Recycling the Family: Remarriage after divorce.
Beverly Hills, CA: Sage.
Glick, P. C. (1989). Remarried families, stepfamilies
and stepchildren: A brief demographic profile.
Family Relations, 38, 24-27.
Goldberg, D. C. (Ed). (1985). Contemporary
marriage: Special issues in couples therapy.
Homewood, IL, US: Dorsey Press. (1985). xxi, 565
pp.
Goldberg, N. A., (2003). Changes in marriage over
time. Dissertation Abstracts International: Section
B: the Sciences & Engineering, 63(11-B), 5515,
US: Univ Microfilms International.
Gottman, J. (1998). Why marriages succeed or fail:
And how you can make yours last. Bloomsbury
Publishing, London.
Harvey, J. H., & Weber, A. L. (2001). Odessey of the
heart: Close relationships in the 21st century. (2nd
ed).
Hill, C. T., Rubin, Z., & Peplau, L. A. (1976).
Breakups before marriage: The end of 103 affairs.
Journal of Social Issues, 32(1), 147-168.
Hortacsu, N. & Karanci, A. N. (1987). Premarital
breakups in Turkish sample: Perceived reasons,
attributional dimensions and affective reactions.
International Journal of Psychology, 22(1), 57-74.
Kirkpatrick, L. E., & Hazan, C. (1994). Attachment
styles and close relationships: A four year
prospective study. Personal Relationships, 1, 123142.
Larson, R. W., Wilson, S., Brown, B. B., Furstenberg,
F. F., & Verma, S., (2002). Changes in adolescents’
interpersonal experiences: Are they being prepared
for adult relationships in the twenty-first century?.
Journal of Research on Adolescence, 12(1), 31-68.
Laursen, B. (1997). Close relationships across the
lifespan: A symposium. International Journal of
Behavioral Development, 21(4), 641-646.
Laursen, B., & Bukowski, W. M. (1997). A
developmental guide to the organisation of close
relationships. International Journal of Behavioural
Development, 21 (4), 747-770.
Lerner, L. N. (2003). A qualitative examination of
what couples feel are important components in
communication
Dissertation
Abstracts
International: Section B: the Sciences &
Engineering, 64(1-B), 465, US: Univ Microfilms
International.
Long, E. C. J., Angera, J. J., Carter, S. J., Nakamoto,
M., & Kalso, M. (1999). Understanding the one you
love: A longitudinal assessment of an empathy
training program for couples in romantic
relationships. Family Relations, 48, 235-242.
Kirkpatrick, L. E., & Hazan, C. (1994). Attachment
styles and close relationships: A four year
prospective study. Personal Relationships, 1, 123142.
Martin, T.c., & Bumpass, L.L. (1989). Recent trends
in marital disruption. Demography, 26, 37-51.
Miller, E. D., (2003). Imagining partner loss and
mortality salience: Consequences for romantic –
relationship satisfaction. Social Behaviour &
Personality, 31(2), 167-180.
Reivich, K., & Shatte, A. (2002). The resilience
factor: 7 essential skills for overcoming life’s
inevitable obstacles. Broadway Books, New York,
NY. 342pp.
Robins, R. W., Caspi, A., & Moffitt, T. E. (2002) It’s
not who you’re with, it’s who you are: Personality
and relationship experiences across multiple
relationships. Journal of Personality, 70(6), 925964.
Rogge, R. D.; Cobb, R. M.; Johnson, M.; Lawrence,
E. & Bradbury, T. N. (2002). The CARE program:
A prevention approach to marital intervention. In
Gurman, A. S. (Ed), & Jacobson, N. S. (Ed),
Clinical handbook of couple therapy (3rd ed).
Guilford Press, New York, NY.
Ross, C. E. (1995). Reconceptualizing the marital
status as a continuum of social attachment. Journal
of Marriage and the Family, 57, 129-140.
Schneller, D. P. (2003). After the breakup: Adult
perceptions and expectations of post-divorce
intimate relationships. Dissertation Abstracts
International, 63(8-A). 3017, US: Univ Microfilms
International Abstract, March 2003.
Sezov, D. D. (2002). The contribution of empathy to
harmony in interpersonal relationships. Dissertation
Abstracts International: Section B: the Sciences &
Engineering, 3046, US: Univ Microfilms
International, 63(6-B).
Sprecher, S., & Felmlee, D. (2000). Romantic
Partners’ perceptions of social network attributes
with the passage of time and relationship
transitions. Personal Relationships, 7(4), 325-340.
Sprecher, S. & Metts, S. (1999). Romantic Belief:
Their influence on relationships and patterns of
change over time. Journal of Social & Personal
Relationships. 16(6)-834-851.
Sullivan, B. F., & Schwebel, A. I. (1996). Birth-order
position, gender and irrational relationship beliefs.
Individual Psychology, 52(1), 54-64.
Taubman-Ben-Ari, O., Findler, L., & Mikulincer, M.,
(2002). The effects of mortality salience on
relationship strivings and beliefs: The moderating
role of attachment style. British Journal of Social
Psychology, 41, 419-441.
Tashiro, T. & Frazier, P. (2003). ‘I’ll never be in a
relationship like that again’: Personal growth
following romantic relationship breakups. Personal
Relationships, 10(1), 113-128.
Thompson, M. M. (1995). 'Should I stay or should I
go?' Ambivalence and the dissolution of dating
relationships. Dissertation Abstracts International
55(8-A), 2591, US: University Microfilms
International.
Wang, H., & Amato, P. R., (2000). Predictors of
divorce adjustment: Stressors, resources, and
definitions. Journal of Marriage and the Family,
62,
655-669.
Weigel & Murray, 2000. The paradox of stability and
change in relationships: What does chaos theory
Relationship transitions
offer for the study of romantic relationships?
Journal of Social & Personal Relationships, 17(3),
425-449).
Weiss, R. S., (1975). Marital separation. New York:
Basic Books.
Correspondence to:
Amanda Ferguson
PO Box 201
Neutral Bay NSW 2089, Australia
amanda@lifethatworks.com
45
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 46-51
The relationship between love styles and relationship dependent and
independent outcome variables: A within couple analysis
Barry J. Fallon (b.fallon@patrick.edu.au)
School of Psychology
Australian Catholic University, Melbourne, 3065, Australia
Terry Bowles (t.bowles@patrick.edu.au)
School of Psychology
Australian Catholic University, Melbourne, 3065, Australia
Of particular interest in the present research is the
agape love style. (Lee, 1998) described this love style
as a combination of eros and storge, which makes this
style of love intense and friendly with the added
quality of altruism whereby the loved one’s needs are
placed before their own. Consistent significant
correlations between agape and eros and storge have
been reported (e.g., Hendrick & Hendrick, 1986).
Previous research investigated the association
between love styles and various relationship variables.
In what is reported here only those findings
concerning agape, eros, and storge will be reported.
Frazier, Byer, Fischer, Wright, & De Bord (1996)
reported that relationship satisfaction was higher for
those who were high in eros and those high in agape.
Contreras, Hendrick, & Hendrick (1996) reported
that eros was the strongest predictor of relationship
satisfaction and agape was also a positive predictor of
relationship satisfaction for women. Worobey (2001)
in research into the relationship between love styles
and temperament reported that fearfulness and distress
were most predictive of agape.
It is likely that feelings about a close relationship
permeate every day functioning. Consequently
outcome measures which are not necessarily
relationship dependent will be explored in the present
research. After a review of literature on general well
being and the relationship literature, it became
apparent that a number of related constructs have been
investigated. At least to some degree, it would seem
that well being and psychological distress have been
confused, although Heady & Wearing (1993) argue
that they represent separate dimensions. In a
substantial study by Barnett and colleagues (see
Barnett, & Marshall, 1993; Barnett, 1994; Barnett,
Marshall,
Raudenbush,
&
Brennan,
1993)
investigating psychological distress in dual earner
couples, the authors measured psychological distress
via a frequency of symptoms scale. Derived from the
SCL-90R (Derogatis, Yevzeroff, & Wittelsberger,
1975) the symptoms scale measures anxiety and
depression - a key to mental health (Barnett, Marshall,
& Pleck, 1992; Barnett, 1994).
Measures of psychological distress concentrate on
the negative aspects of functioning, and how this
distress is exhibited. Well being, however, seeks to
tap into how happy individuals are in their life. Strong
links have been drawn between overall life
satisfaction and satisfying interpersonal relationships
(Traupmann, Hatfield & Wexler 1983). Further, an
absence of close relationships is argued to have an
Abstract
The present research investigates feelings about close
relationships and the ways these feelings are related to
outcome measures that are not necessarily relationship
dependent - overall life satisfaction, subjective wellbeing and psychological distress as well as a
relationship
outcome
variable–
relationship
satisfaction. In this research feelings about
relationships is operationalised by a measure of love
styles. The focus of the data for the present research is
the couple in a romantic heterosexual relationship. Data
from 118 heterosexual couples ranging in age from 18
years through 86 years formed the basis for analysis of
the associations between agape, eros, and storge love
styles, affective relationship variables, and the more
general variables of psychological distress and general
well-being. Questions addressed include: Is there a
consistent association between the agape, eros, and
storge love styles of the partners? What is the
relationship between the agape scores of one partner
and their partner’s perceptions of that relationship –
both relationship dependent outcomes as well as nonrelationship dependent outcomes? While many of the
associations are consistent with previous research, there
are different patterns of results for males and females.
The research highlights the importance of analysing
couple data appropriately.
The varied attitudes and beliefs about love which
individuals bring to a relationship impact
differentially on the perceptions which those in that
relationship have about their partner and the
relationship itself. While the concept of love has been
examined and written about in a wide variety of
disciplines including philosophy, psychology,
sociology, and theology, the term remains elusive.
Many definitions allow for a variety of forms of
other-centered love that promotes behaviors designed
to support the quality of life of loved ones. Other
terms, which can be used to describe other-centered
love, include unconditional love, agape, compassion,
and selfless love.
Lee (1973) proposed a typology of love and he
maintained that different attitudes to love stem in part
from previous experiences especially within the
family of origin. The attitudes influence emotions and
behaviors towards the partner and the relationship.
Lee (1973) proposed that the emotions and behaviors
could be classified into six different types. There were
three primary love styles: eros, ludus, and storge and
three secondary types: mania, pragma, and agape.
46
Love styles
association with poor subjective well being (SalmelaAro & Nurmi, 1996). A well being scale devised by
Barnett (1992) which adopts a positive assessment of
well being, represented by symptoms of life
satisfaction will be used in this investigation.
In the present investigation, the notion that
psychological distress and well being are separate
dimensions will be adopted. The definition of
psychological distress in terms of the symptoms of
anxiety and depression will also be adopted. Well
being will be contrasted with psychological distress in
tapping the positive aspects of life experience for the
individual.
The focus of the data for present research is the
dyad – the couple in a romantic heterosexual
relationship. By far the majority of research reporting
on love styles has been at the individual level and
examining associations between variables within
individuals. Of interest in the present study is the
association between the love style of one partner and
the relationship dependent outcomes for their partner
as well as some variables of the partner which are not
relationship dependent e.g., well being. This leads to
questions such as: What is the relationship between
the agape scores of one partner and their partner’s
perceptions of that relationship – both relationship
dependent outcomes as well as non-relationship
dependent outcomes? If the agape love style is as
described by Dwyer (2000) as “An unconditional,
caring, giving, and forgiving type of love. There is no
expectation of reciprocity, love is self-sacrificing”
(p.22) then it may be reasonable to expect that there
will not be any consistent association between one
partner’s agape score and the other partner’s
outcomes. Hendrick & Hendrick (2000) claim that
one of the most consistent themes in love style
research has been the exploration partner similarity.
They report that for the most part, a person’s own
love style was correlated with their preferences for
partners with the same love style and that the most
appealing love styles were agape and storge. The
present research will enable an examination of the
correlations of love styles between partners.
In essence, the love style of particular interest in the
present research is agape. The outcome variables of
interest are positive and negative affects, relationship
satisfaction, general well-being, and psychological
distress.
Method
Participants
Data was collected using a snowball technique from
couples in a heterosexual relationship. Two hundred
and sixty two individuals aged between 17 and 86
years of age completed the questionnaires.
Respondents were primarily in exclusive dating
relationships (55.3%) with the remainder being in
married (34.7%), defacto (6.1%) or engaged (3.8%)
relationships. Due to the small numbers of engaged
(N = 10) and defacto (N = 16) respondents they were
excluded from further analysis. There were equal
numbers of men and women in the sample, and the
47
mean age of the remaining 236 respondents was 31.0
years (sd. = 14.2, min = 17, max = 86).
In the present study 44.5% of respondents were
employed in either full-time (n = 84) or in parttime/casual work (n = 21), 49.1% of respondents
were students in full-time (n = 107) or part-time (n =
9) study and 6.4% of respondents were unemployed.
Measures and Materials
Respondents were presented with a questionnaire
booklet that included a brief demographic section
followed by the scale measures
Love Styles Respondents completed the Love
Attitudes Scale (LAS) described by Hendrick &
Hendrick (1986). While the scale taps six love styles:
eros, ludus, storge, pragma, mania, and agape, each
measured by seven items, the present research only
uses the scales of agape, storge, and eros. Items are
responded to on a 5 point Likert rating scale ranging
from 1 = strongly disagree to 5 = strongly agree.
Relationship Satisfaction The Kansas Marital
Satisfaction scale (KMS; Schumm et al. (1986)) was
adapted in the measurement of relationship
satisfaction. There are three items in this scale each of
which requires a response on a 7 point Likert scale.
The higher the rating the greater the satisfaction.
Relationship Affects Relationship affects were
determined from six affect items; anger, guilt,
contentment, and happiness (Austin & Walster
(1974), and depression and loving (Sprecher (1992).
The respondents were asked to rate each of these
affects on a 5 point Likert scale where 1 = strongly
disagree to 5 = strongly agree. The affects of
contentment, happiness and loving formed the
positive affect variables and guilt depression, and
anger formed the negative affect variable
.
General Well-Being The 19 item scale (Barnett,
Marshall, & Pleck, 1992) required respondents to
indicate how much of the time during the past month
they had felt each of the items. The 5 point rating
scale for the items in this scale ranged from 1 = not at
all to 5 = always.
Psychological Distress A twenty four item scale
examined the anxiety and depression aspects of
psychological distress Barnett and Marshall (1993). A
5 point rating scale was provided for each item. The
scale was anchored at 1 = strongly disagree to 5 =
strongly agree.
Procedure
Instructions were given for respondents to complete
the questionnaire booklet away from their partner,
without consultation before placing the completed
questionnaire in an envelope, and then placing both
envelopes in a larger one for returning to the
researcher.
FALLON & BOWLES
48
Results
Before conducting the major statistical analyses to
address the research questions, analysis of the factor
structure and reliability of the scales was undertaken.
For relationship satisfaction the Cronbach’s alpha
found in this study was 0.91. Using the GLM
procedure in SPSS a repeated measures analysis of
variance explored differences in responses on the
three items. A significant effect (F(2,464) = 14.386, p
< .001) indicated that individuals endorsed the items
differently from each other, suggesting that the items
tap non-redundant aspects of satisfaction.
There were six affects. were examined in AMOS
(Smallwaters, 1997), testing for the superiority of a
two factor solution over a single factor. Results
indicated that the two factor model of correlated
positive and negative affect provided a much better fit
of the data (Δχ2(1) = 71.38, p < .001). The goodness
of fit indices also supported the superiority of this
solution (GFI = 0.981, AFI = 0.949, PFI = 0.374, CFI
= 0.990) and the two factors correlated -0.580. A
measure of positive affect was derived by summing
the three positive affect items (Cronbach’s alpha =
0.89), responses (mean = 13.01, SD = 2.16) indicated
a generally high level of positive affect (scale range =
3 to 15). A negative affect measure was derived by
summing the negative affect items (Cronbach’s alpha
= 0.67) and responses on this measure (mean = 4.41,
SD = 1.73) indicated a very low level of negative
affect (scale range = 3 to 15).
A single measure of well being was generated by
taking the mean of the nineteen items (Cronbach’s
alpha = 0.94). Responses indicated a moderate level
of well being (mean = 4.41, SD = 0.74; scale range 1
to 6).
The anxiety and depression sub-scales described by
Barnett et al. (1993) were highly reliable (αanxiety =
0.85, αdepression = 0.88). Single measures were
generated by taking the mean of the 10 items for
anxiety (mean = 1.60, SD = 0.61) and 14 items for
depression (mean = 1.84, SD = 0.59). Responses
indicated low levels of both in the sample. The two
sub-scale measures were highly correlated (r = 0.82, p
< .001). It seems clear that the sub-scales measure
different aspects of a single construct. Consequently a
single measure of psychological distress was
developed by taking the mean of all 24 items (mean =
1.74, SD = 0.59) for use in further analyses.
Reliability analyses of the three love style subscales revealed moderate to high reliabilities. The
Cronbach alphas for the three scales were as follows:
agape = .83, eros, .74, and mania = .70
Descriptive statistics are presented in Table 1.
Preliminary analyses indicated that there were no
significant differences between those reporting that
they were married and those who reported that they
were dating. Due to the dyadic nature of the data,
partner’s scores are likely to be non-independent
(Kashy, 2000; Kenny, 1996). As such, cross-couple
correlations on the measures obtained were examined
(see Table 2) before proceeding with further analyses.
Pearson’s correlations were used to investigate the
associations between the variables.
Table 1
Means for males and Females
Variable
Agape
Eros
Storge
Males
3.93
3.99
3.69
Females
3.55***
4.00
3.80
Relationship Satisfaction
Positive Affect
Negative Affect
Psychological Distress
Wellbeing
6.10
4.28
1.48
1.69
4.40
6.15
4.28
1.49
1.80
4.40
In order to test for possible sex differences on the
research variables dependent t-tests were conducted.
The means (see Table 1) for both the females and the
males on relationship satisfaction, positive affect, and
the love styles were relatively high which is in
contrast to the means for the more negative aspects
such as negative affect and psychological distress.
There was a significant difference (t = 5.67, df = 121,
p < .001, η2 = .21) on only one of the variables agape. The male mean was significantly higher (3.93)
than the female mean (3.55).
The correlations between the variables of interest
within couples are set out in Table 2. For the three
love styles there was a significant correlation between
partners with the eros love style providing the highest
correlation (r = .50) and storge the lowest (r = .39).
For the outcome variables with the exception of
psychological distress there was also a significant
correlation between partners. Relationship satisfaction
resulted in the highest correlation (r = .41).
The significant correlations and the non-significant
dependent t tests provide substantial support for the
proposition put forward by Hendrick & Hendrick
(2000) of partner similarity. The exception to this
general proposition was for the agape love style where
the males were significantly higher than the females.
Table 2
Within Couple Correlations
Variable
Agape
Eros
Storge
Correlation
.42 ***
.50 ***
.39 ***
Positive Affect
Negative Affect
Relationship Satisfaction
Wellbeing
Psychological Distress
.27 **
.23 *
.41 ***
.36 ***
.16 NS
* p < ,05; ** p < .01; ***p < .001
An examination of the intraindividual correlations
between the love styles indicated that for the females
there were significant correlations between agape and
eros (r = .47, p < .001) but not between agape and
storge (r = .02, p > .05). For the males the correlations
of eros and storge with agape were both significant ((r
= .51, p < .001 and r = .29, p < .001 respectively).
Love styles
With respect to the correlations between the variables
within the individuals, agape correlates significantly
with all of the relationship dependent variables for
both the males and the females. For the nonrelationship dependent variables the correlations with
agape were non-significant for the females and for the
males the correlation between agape and well being
was positive and statistically significant where as it
was not for agape and psychological distress. For the
females eros correlated significantly with both the
relationship dependent variables as well as the nonrelationship dependent variables. It was similar for the
males with the exception of the correlation between
eros and distress, which was not statistically
significant. Storge was not correlated significantly
with any of the outcome variables for the females,
while for the males it only correlated significantly
with one variable – relationship satisfaction.
To examine the relationship between agape and the
outcome variables more closely a high, medium, and
low group of males and a high, medium, and low
group of females were formed based on their scores
on their agape love style scores. Each group had as
close to equal numbers as possible. A chi-square test
of association between agape groups and relationship
type indicated that there was no association between
these two variables for either the males or the females.
ANOVAs were computed with agape group as the
independent variable and the partner’s relationship
dependent outcomes (positive affect, negative affect,
relationship satisfaction) and non-relationship
dependent outcomes (psychological distress and wellbeing) as the dependent variables. Given that there
were a large number of statistical tests α was set at
.01.
There were no significant differences on any of the
female partner variables between the high, medium,
and low male agape groups. For the females while
there was no difference on the partner’s negative
variables of negative affect and psychological
distress, there were significant differences between
the three female agape groups and their partner’s
positive outcome variables. In each case there was
only a minor non-significant difference between the
low and medium agape groups while the mean for the
high agape group was significantly higher than the
means for both the medium and low agape groups.
The results of the ANOVAs (see Table 3) were as
follows: for positive affect F = 9.27, df = 2, 118, p <
.001, η2 = .14; for relationship satisfaction F = 9.53,
df = 2, 118, p < .001, η2 = .14, and for well-being F =
6.35, df = 2, 117, p = .002 η2 = .09.
In addition to conducting the analyses with the
partner’s variables as the dependent variables, the
same analyses were conducted with the individual’s
own variables as the dependent variables. For the
males there was a significant difference between the
three agape groups on only two of the dependent
variables – positive affect and relationship
satisfaction. The results of these ANOVAs were as
follows: for positive affect F = 17.00, df = 2, 118, p <
.001, η2 = .22 (means of 3.79, 4.21, and 4.64 for the
low, medium and high groups respectively) and for
relationship satisfaction F = 13.38, df = 2, 118, p <
49
.001, η2 = .18 (means of 5.59, 6.00, and 6.49 for the
low, medium, and high groups respectively). In both
instances the mean for the high agape group was the
highest and there was only a minor non-significant
difference between the means for the low and medium
agape groups.
Table 3
Means for male outcome variables for each of the
three female agape groups
Variable
Relat Satis
Positive Affect
Negative Affect
Psych Distress
Wellbeing
Low
Agape
5.83
4.09
1.53
1.64
4.26
Med
Agape
5.91
4.09
1.59
1.81
4.21
High
Agape
6.54 ***
4.66 ***
1.28
1.57
4.75 **
** p < .01, *** p < .001
For the females, the results of these ANOVAs were
the same as for the males – significant differences on
positive affect F = 8.33, df = 2, 123, p < .001, η2 =
.11 (means of 4.19, 4.24, and 4.76 for the low,
medium and high groups respectively) and for
relationship satisfaction F = 8.67, df = 2, 122, p <
.001, η2 = .11 (means of 5.96, 5.92, and 6.58 for the
low, medium, and high groups respectively). In both
instances the mean for the high agape group was the
highest and there was only a minor non-significant
difference between the means for the low and medium
agape groups.
In summary, for both the three male and the three
female agape groups, positive affect and relationship
satisfaction were significantly different when it was
their own scores which were the dependent variables.
When it was the partner’s outcome variables as the
dependent variables, the parallel situation pertained
for the female agape groups, i.e., there were
significant differences for the males on relationship
satisfaction, positive affect but also with the addition
of well being. The female outcome variables did not
differ between the three male agape groups..
Discussion
The data analyses indicated that there were more
similarities than differences between partners on their
love styles and the outcome variables, especially those
which can be seen to be relationship dependent.
Following the line of argument proposed by Dwyer
(2000) it would have been expected that there would
not have been any particular relationships between the
scores of the partners. However, as Hendrick and
Hendrick (2000) argued one of the most consistent
themes in love research has been that of partner
similarity. There was support for the Hendrick and
Hedrick (2000) proposition. Of all the variables
involved in the present research there was only one on
which there was a significant difference within
partners. That was for agape. The proposition that we
seek out those with a similar love style is statistically
supported not only by the one significant dependent t
50
FALLON & BOWLES
result but also by the correlations between partners.
With only one exception, the correlations between
partner’s scores were statistically significant. The
coefficients ranged from .39 to .50. While these
correlations are statistically significant, they share
only from 15% (for storge love style) to 18% (for
agape love style), and 25% (for eros love style) of the
variance. Is this sufficient to claim that we seek out a
partner who is similar in love style?
With previous research often reporting either no
difference between partners or, on occasion females
scoring higher on agape than males the present result
is in slight contrast to the earlier research. It may well
be that the present results are as much part of a within
couple analysis in contrast with much of the previous
research where a between males and females analysis
was used. It may also be that we are at the early stage
where we may find that the males are changing more
in their reporting of relationship characteristics than
are the females. In terms of societal expectations
about males and females in relationships, the females
have been exposed to pressures and opportunities for
change for more than thirty years. It is far more
recently that males are now expecting to, as well as
being expected to be more attuned to relationship
characteristics.
When the participants were grouped into high,
medium, or low based on their agape scores and their
own outcomes variables were the dependent variables,
the same pattern occurred for both the males and the
females. The positive relationship outcome variables
of positive affect and relationship satisfaction varied
between the three agape groups with those in the
highest agape group reporting the highest satisfaction
and positive affects. There were no differences for the
negative relationship dependent outcome variable or
for either of the positive or negative non-relationship
dependent outcome variables.
When the partner’s outcome variables were the
dependent variables, a similar pattern of results
occurred for both the males and the females. Once
again for both the males and the females, the positive
relationship dependent outcome variables of
relationship satisfaction and positive affect were
significantly different between the three groups with
those highest in agape scoring highest on these
variables. In addition for the males, the nonrelationship dependent variable of well being
followed the same pattern with significant between
the three female agape groups. This suggests that for
the males the way they feel about their partner
especially in terms of intensity and friendliness with
the added quality of altruism and the way their partner
feel about them does permeate their every day
functioning. Not only do the differences between the
agape groups result in differences in relationship
dependent outcomes but also the more general nonrelationship dependent variable of well being.
Conversely, for the females the data suggests that
there is no carry over into everyday functioning.
There were no differences between their partner’s
agape groups and the outcome variables, and there
was only a difference on the positive relationship
dependent variables for the groups based on the
female agape scores. Neither their own, nor their
partner’s agape scores resulted in differences on the
non-relationship dependent variables.
It is interesting to note that no significant
differences between the agape groups, based on one’s
partner’s scores or on one’s own scores, for the
negative outcome variables whether they be
relationship dependent or non-dependent. That there
were differences for some of the positive outcome
variables but not at all for any of the negative
outcome variables gives some weight to the notion
that there is a relative independency between the
positive and the negative aspects of psychological
functioning. That is, the factors which contribute to
the positive outcomes do not by definition “reduce”
the negative aspects of a relationship and conversely,
those factors which result in negative outcomes do not
necessarily result in a reduction of the positive
outcomes.
This research highlights the importance of treating
dyadic in an appropriate way. If the analyses had been
left at simply a between groups (males and females)
design the results and implications would have been
basically that there were no differences between the
males and the females. However, by acknowledging
and appropriately utilising the dependency in couple
data, the results indicated that there are differences
between the males and the females. In this research it
indicates that females are less dependent on their
partner for their relationship outcomes than are the
males.
References
Austin, W., & Walster, E. (1974). Reactions to
confirmations and disconfirmations of expectancies
of equity and inequity. Journal fo Personality and
Social Psychology, 30, 208-216.
Barnett, R. C. (1994). Home-to-work spillover
revisited: A study of full-time employed women in
dual-earner couples. Journal of Marriage and the
Family, 56, 647-656Barnett, R. C., & Marshall, N. L. (1993), Men,
family-role quality, job-role quality, and physical
health. Health Psychology, 12, 38-55.
Barnett, R. C., Marshall, N.L., & Pleck, J.H. (1992).
Men's multiple roles and their relationship to men's
psychological distress. Journal of Marriage and the
Family, 54, 358-367.
Barnett, R. C., Marshall, N. L., Raudenbush, S. W., &
Brennan, R. T. (1993). Gender and the relationship
between job experiences and psychological distress:
A study of dual-earner couples. Journal of
Personality and Social Psychology, 64, 794-806.
Contreras, R., Hendrick, S.S., & Hendrick, C. (1996).
Perspectives on marital love and satisfaction in
Mexican American and Anglo couples. Journal of
Counseling and Development, 74, 408-415.
Derogatis, L.R., Yevzeroff, H., & Wittelsberger, B.
(1975). Social class, psychological disorder, and the
nature of the psychopathologic disorder. Journal of
Consulting and Clinical Psychology, 43, 183-191.
Dwyer, D. (2000). Interpersonal relationships.
London: Routledge.
Love styles
Frazier, P.A., Byer, A.L., Fischer, A.R., & Wright,
D.M. (1996). Adult attachment style and partner
choice: Correlational and experimental findings.
Personal Relationships, 3, 117-136..
Heady, B. W., & Wearing, A.J. (1993) Dimensions of
mental health: Life satisfaction, positive affect,
anxiety and depression. Social Indicators Research,
29, 63-82.)
Hendrick, C., & Hendrick, S. (1986). A theory and
method of love. Journal of Personality and Social
Psychology, 50, 392-402.
Hendrick, S. S., & Hendrick, C. (2000). Romantic
love. In C. Hendrick & S. S. Hendrick (Eds.), Close
relationships: A sourcebook. Thousand Oaks: CA:
Sage.
Lee, J. A. (1973). The colors of love: An exploration
of the ways of loving. Don Mills, Ontario: New
Press.
Lee, J.A. (1998). Love styles. In R.J. Sternberg &
M.L. Barnes (Eds.). The psychology of love. (pp.
38-67). New Haven, CT: Yale University Press
Salmela-Aro, K., & Nurmi, J.E. (1996). Uncertainty
and confidence in interpersonal projects:
Consequences for social relationships and wellbeing. Journal of Social and Personal
Relationships, 13, 109-122.
Schumm, W. R., Paff-Bergen, L. A., Obiorah, F. C.,
Copeland, J. M., Meens, L. D., & Bugaighis, M. A.
(1986). Concurrent and disciminant validity of the
Kansas Marital Satisfaction Scale. Journal of
Marriage and the Family, 48, 381-387.
Smallwaters Corporation (1997). AMOS (Version
3.6). Chicago: Smallwaters Corporation.
Sprecher, S. (1992). How men and women expect to
feel and behave in response to inequity in close
relationships. Social Psychology Quarterly, 55, 5769.
Traupmann, J., Hatfield, E., & Wexler, P. (1983).
Equiity and sexual satisfaction in dating couples.
British Journal of Social Psychology, 22, 33-40.
Worobey, J. (2001). Sex differences in associations of
temperament with love-styles. Psychological
Reports, 89, 25-26.
Correspondence to:
Barry J Fallon
School of Psychology
Australian Catholic University
115 Victoria Parade
Fitzroy, VIC 3165
51
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 52-57
Is The Extent Of My Lying And Deception With The One I Love Related To
My Attachment Style?
Barry J Fallon (b.fallon@patrick.acu.edu.au)
Christopher Holt
School of Psychology
Australian Catholic University Melbourne, Victoria, 3065, Australia
anxious style have a model of themselves as being
misunderstood, unconfident, and under appreciated
and they view others as being unreliable and
unwilling to commit themselves. Those with an
avoidant style possess models of themselves as
suspicious, aloof, and skeptical and they possess
mental models of others as being unreliable or too
eager to commit to relationships. It is these beliefs and
expectations (mental models) which are the blue print
for future relationships later in life (Collins & Read,
1990). Research conducted by Collins and Read
(1990), Feeney and Noller (1990), and Hazan and
Shaver (1987) have all found evidence to support the
claim that the mental models formed in early infant
years are also found in adult relationships.
Bartholomew and Horowitz (1991) proposed a four
category model of attachment styles among young
adults. It is based on an extension of Hazan and
Shaver’s (1987) three category model, as it has two
types of avoidant attachment; dismissing and fearful
styles. Each attachment style is defined in terms of the
person’s model of self and the person’s model of
others (Simpson & Rholes, 1998).
Each attachment style can be defined using a two
dimensional approach: positivity of a person’s model
of self and positivity of a person’s model of others.
The self-model is associated with levels of anxiety,
due to their perceptions of self worth and whether or
not they will gain their partner’s approval. The model
of others is associated with whether or not they should
seek or avoid closeness, due to their perceptions of
others as being available and supportive Simpson &
Rholes, 1998).
Those with a secure attachment style are defined as
a person with a positive self model and positive model
of others, they generally have a sense of self-worth
and are not anxious as they do not fear abandonment.
They also are comfortable with being intimate and
therefore are not avoidant in their behaviour, they
perceive others as being available and supportive
(Simpson & Rholes, 1998). Those with a preoccupied
attachment style have a negative self-model and a
positive model of others, they anxiously seek to gain
acceptance from others, whilst being non-avoidant in
behaviour (Simpson & Rholes, 1998). This
attachment style is similar to Hazan and Shaver’s
(1987) anxious-ambivalent style (Keelan, Dion, &
Dion, 1998). Those with a fearful attachment style
have both a negative model of themselves and others.
They are similar to that of the preoccupied in the
sense that they are dependent on others. They are
anxious as they fear abandonment, and avoid intimacy
for reasons such as fear of rejection (Simpson &
Rholes, 1998). Those with a dismissing attachment
Abstract
Thirty-five married couples and 14 dating/engaged
couples provided the data for the present study.
Participants were asked about the deception they use in
their relationship as well as their perception of their
partner’s deception. They also were asked about the
frequency of lying to their partner and their perceptions
of their partner’s lying to them. The participants also
completed an attachment style measure. As there was
only a small number of participants in each of the
dismissive, preoccupied, and fearful attachment style
categories these were combined into a not-secure
category. ANOVAs with sex and attachment style as
the independent variables and own and partner’s lies
and deceptions as the dependent variables resulted in
no main effects or interaction effects involving sex.
However, there were significant effects for attachment
style on each of the four dependent variables with those
reporting a secure attachment style also reporting less
deception of their own and their partner and lower use
of lying by themselves and by their partner. The results
are discussed in terms of the use of deception and lying
as mechanisms for maintaining one’s own attachment
style.
Intimate or close relationships with others can be the
source of the deepest satisfaction and of the blackest
misery. Many people are lonely and unhappy, some
are mentally ill, simply because they are unable to
establish and sustain an intimate relationship in
today’s world (Argyle, 1983).
Attachment theory is one way in which
interpersonal relationships can be understood.
Bowlby, considered to be the founder of attachment
theory, believed that an attachment bond is developed
between the primary caregiver and infant to provide
children with a sense of security (Collins & Read,
1990). It is believed that the events and experiences
which occur during the child/caretaker relationship
strongly influence the nature and quality of close
relationships in adulthood (Collins & Read; 1990,
Fraley & Davis, 1997). Attachment theory is most
suited as a framework for understanding interpersonal
relationships, as it explains how love can develop and
how it can be shaped by social experience to produce
different relationship styles. It explains how both
healthy and unhealthy forms of a relationship
originate (Hazan & Shaver, 1987).
Three attachment styles were originally identified
(Hazen & Shaver, 1987). Individuals with a secure
attachment style have mental models of themselves as
being friendly, good natured, and likable and they
have mental models of others as being well
intentioned, reliable, and trustworthy. Those with an
52
Lying, deception and attachment style
style have a positive model of self and a negative
model of others, they fear intimacy for reasons such
as negative expectations, and lack anxiety as they
maintain an ideology of self worth by criticizing the
values of close relationships (Simpson & Rholes,
1998).
Although trust contributes to feelings of closeness
and intimacy, occasionally the costs of telling the
truth can do more harm than good. People are more
likely to use deception if the cost of telling the truth is
too high. Deception is considered to be a fact of daily
life (De Paulo & Kashy, 1998). However society
frowns upon deception and those caught deceiving are
looked down upon by other people (Boon & McLeod,
2001). Deception can be defined as “intentionally
trying to mislead someone (DePaulo, Kashy,
Kirkendol, Wyer & Epstein, 1996, p981). DePaulo, et
al, (1996) examined the nature of lying in community
members and college students. It was found that both
groups were more likely to tell lies which benefited
themselves than tell lies which benefited others. Their
research also suggested that the majority of the self
centered lies told were told to enhance the
individual’s self esteem. There are several
explanations for the use of deception in close
relationships. They are all to some extent linked to the
social exchange perspective in that people are less
likely to the tell the truth when the costs of doing so
outweigh the benefits (Cole, 2001). One explanation
for deceptive behavior is the norm of reciprocity,
meaning that an individual will tend to match their
partner’s contributions in a ‘tit for tat’ transaction.
Deceiving a partner leads to a decrease in
commitment and satisfaction (Cole, 2001). The
perception that a partner is dishonest results in a
decline in commitment and satisfaction in the
relationship. Therefore in the present research it is
expected that own use of deception is associated with
the perception that their partner engages in deception.
Another explanation for deceptive behavior in
relationships is the costs associated with telling the
truth. Although telling the truth contributes to feelings
of closeness and intimacy, people however are more
likely to use deception if the costs of telling the truth
are too high (Cole, 2001).
A final explanation for deceptive behavior is based
on individual’s differing needs for intimacy and
closeness. In examining Bartholomew and
Hororwitz’s (1991) four category model of attachment
styles, it can be seen that there are differing levels of
intimacy and closeness in this model. The secure
attachment style is considered to be comfortable with
being intimate, in contrast the dismissing attachment
style, fear being intimate. Deceit may be used by
individuals uncomfortable with intimacy, to keep
others at a safe distance. Cole (2001) examined the
prevalence of deception in romantic relationships.
One of his assumptions was that fear of intimacy is
positively related to the use of deception. Fear of
intimacy is a characteristic of people categorized into
the avoidant attachment style. His findings found that
a positive relationship does exist between fear of
intimacy and use of deception. It is expected in the
present research that deception is higher for those
53
individuals with the dismissing attachment style
compared to those with a secure attachment style.
Also, deception may be used to increase partner’s
interest in closeness and devotion. Individuals with
the preoccupied attachment style, where a person
anxiously tries to seek acceptance he/she may lie to
their partner by saying, ‘I love you’ simply to increase
partner’s interest in them (Cole, 2001). In the present
research it can be expected that deception is also
higher in those with a preoccupied attachment style.
The aim of this research is to investigate the
differing levels of deception among attachment styles.
It can be hypothesised that own use of deception is
associated with the perception that partner engages in
deception. It can be hypothesised that deception is
high in those with a dismissing/fearful attachment
style. It can also be by hypothesised that deception is
higher in those with a pre-occupied attachment than
those with a secure attachment style.
Method
Participants
Participants were 49 couples who had been in a
heterosexual relationship for six months or more.
Couples were recruited using a snowball technique.
The mean age of males was 38.8 years (range 18 to
67, SD= 13.50), the mean age of females was 37.29
years (range 18 to 64, SD = 13.00) the average length
of the relationship was 14.52 years (range .8 to 41.40
years, SD = 12.27). Out of the 49 couples 13.3% were
exclusively dating, 10.8% were in a de-facto
relationship and 70.4% were married.
Measures
Demographics This section incorporated questions
regarding participants’ age, sex, level of education,
current employment status, type, and duration of
relationship.
Deception The deception scale developed by Cole
consists of nine items pertaining to one’s own
deception and four items relating to perceived partner
deception. The nine-item scale developed by Cole
(2001) measures how often one uses deception using a
seven point Likert scale (1 = strongly disagree, 7 =
strongly agree) e.g.: “I sometimes find myself lying to
my partner about things I have done”. The inter-item
reliability of this nine item scale was acceptable with
a Cronbach alpha of .84. The four item scale
developed also by Cole (2001) measures how often
people perceive that their partner deceives them using
a seven point Likert scale (1 = strongly disagree, 7 =
strongly agree). The inter-item reliability was
acceptable with a Cronbach alpha of .80. This current
study obtained a Cronbach alpha of .64 for own
deception and perceived partner deception .47.
Attachment Bartholomew and Horowitz (1991) 35item attachment scale consists of two sub scales:
avoidance (e.g. “ I prefer not to show a partner how I
feel deep down”) and anxiety ( e.g. “ I worry about
being abandoned”). It is rated using a seven point
Likert scale (1 = strongly disagree, 7 = strongly
FALLON & HOLT
54
agree). Items 3, 15, 19, 22, 25, 27, 29, 31, 33, 35 are
reversed scored. The standardized reliabilities
(coefficient alpha) for the sub scales are: avoidance .
94 and anxiety .91. This current study obtained a
Cronbach alpha for the sub scales: avoidance .92;
anxiety .88. Using scoring instructions found in
Simpson and Rholes (1998) participants are classified
into either a secure, fearful, preoccupied or dismissing
attachment style.
Procedure
Questionnaires and information letters were
distributed to 75 couples who were eighteen years or
older and who were currently in a heterosexual
relationship lasting 6 months or more. Out of the 75
couples complete data from both partners on all
variables was obtained from 49 couples indicating a
65% return rate. Both partners from each couple were
asked to complete questionnaires in a comfortable,
isolated environment away from their partner.
Questionnaires for each couple were returned in a
sealed envelope so that data for each couple was not
separated. Participants were informed that they could
not be identified by name and the completion of the
questionnaire indicated that they had given voluntary
consent and could withdraw at any time.
Results
Results were analysed using correlations to test for
significant relationships among dimensions of
deception, avoidance, and anxiety for both males and
females. A two by three multivariate analysis of
variance was used to test for significant main and
interaction effects on deception, as the dependant
variable, gender and the different attachment styles as
the independent variables. Gender yielded no
significant main or interaction effects and was not
included in further analyses. Correlations were also
used to examine if partner scores are dependant on
each other.
To examine if associations exist between anxiety,
avoidance, and deception, mean scores, standard
deviations, alpha levels and correlations were
calculated. These are presented in Table 1.
According to the social exchange theory norm of
reciprocity, one’s own use of deception may be
associated with their partner’s use of deception (Cole,
2001). To examine if partner’s scores on deception
are related, correlations were calculated. These are
presented in Table 2.
Table 1
Correlations between anxiety, avoidance, and deception
Avoidance
Anxiety
Own
deception
Partner
deception
Mean
SD
1.92
2.87
2.83
1.00
1.97
1.27
.92
.89
.64
.47**
.44**
.70**
.28
2.90
1.04
.47
.59**
.37**
Alpha
Avoidance
Anxiety
Own
Deception
.55**
.49**
.63**
Partner’s
Deception
.29*
.37**
.45**
-
Note. Lower diagonal half represents female correlations, upper diagonal half represents male correlations.
* p < .05; ** p < .01
Table 2
Correlations between partner’s scores on deception and perceived partner deception
1. Female deception
2. Female perceived partner deception
3. Male deception
4. Male perceived partner deception
1.
2.
3
.63**
.43**
.44**
.25
.29
.45**
* p<.05, ** p < .01
As shown in table 2 males’ perceived partner
deception has a significant positive relationship with
their own partner’s use of deception (r = .44, p < .05),
and their own deception (r = .45, p < .05), indicating
that if a female partner engages in deception, the male
partner believes that their partner is engaging in
deception, and they too engage in deception
themselves. Females perceived partner deception was
not related to their partner’s use of deception but a
significant positive relationship with their own
deception (r = .63, p < .05), indicating that there is no
association between female’s perceived partner
deception and their partner’s actual use of deception,
however if the female partner perceives that their
partner is engaging in deception they too engage in
deception. In relation to the second hypothesis
analysis reveals that own use of deception was
associated with the perception that their partner
engages in deception.
Correlations and analysis of variance were used to
test the association between deception and avoidance.
As shown in table 1 for males, the attachment
Lying, deception and attachment style
avoidance scale was found to have a significant
positive relationship with own use of deception (r =
.55, p < .05), partner deception (r = .29, p < .05. A
similar pattern was found for the females. The
attachment avoidance scale was found to have a
significant positive relationship with, own use of
deception (r = .44, p < .05), partner deception (r = .59,
p < .05. Indicating that having a negative view of
others is associated with a high level of deception, and
perceived partner deception for both males and
females.
The anxious attachment scale for males was found
to have a significant positive relationship with own
use of deception (r = .49, p < .05), partner deception
(r = .37, p < .05). Indicating that having a negative
view of self is associated with an increase in
deception and partner deception for males. The
anxious attachment scale for females was found to
have a significant positive relationship with partner
deception (r=.37, p <.05). Indicating that females
having a negative view of self, is associated with high
partner deception for females.
In examining gender differences it was only found
that females differed in comparison to males as there
was no association between the anxious attachment
scale and own use of deception in females. All other
correlations were similar among male and female.
The degree to which an individual perceives that
their partner is trustworthy can be associated with the
attachment style they possess (Guerrero, 1996). As
trust is associated with deception. Deception may also
vary among the attachment styles. In analyzing the
55
four attachment styles, it was found that there were
only three participants who were categorized into the
dismissive attachment style. The dismissing and
fearful attachment styles were combined as they both
represent characteristics of Hazan and Shaver’s
(1987) avoidant attachment style. Calculated means
and standard deviations of own deception and
perceived partner deception were examined to observe
differences among attachment styles. These are
presented in table 3.
It was found that there was a significant difference
on own use of deception between the attachment
styles (F (2,95) = 9.344, p < .05). Using Tukey posthoc comparisons it can be said that the difference lies
between those with a secure and preoccupied and
those with a secure and dismissive attachment styles.
In examining the means among the three different
attachment styles it can be seen that people are who
posses a dismissing/fearful attachment style (m =
3.91) are more likely to engage in their own use of
deception than those with a preoccupied (m= 3.41)
and those with a secure (m = 2.52) attachment styles.
Using Tukey post-hoc comparisons it can be said
that there was a significant difference also between
those with a secure and those with a dismissive
attachment style. In examining the means among the
three different attachment style groups it can be seen
that people are who posses a pre-occupied attachment
style (m = 3.41) are significantly more likely to
perceive that their partner engages in deception than
those with a secure (m = 2.52) attachment style.
Table 3
Means and standard deviations of own deception and partner deception for each attachment style
Secure
Preoccupied
Dismissive/Fearful
n
70
17
11
Own deception
M
SD
2.51
.98
3.41
1.76
3.91
1.25
Discussion
The aim of this research was to investigate the levels
of deception within attachment styles. Use of
deception was associated with the perception that
their partner engages in deception. Deception was
high in the dismissing/fearful attachment style. Also,
deception was higher for those with a pre-occupied
attachment than those with a secure attachment style.
Cole (2001) suggested that that if a person
perceives that their partner is engaging in deception,
they too will engage in deception. Results from this
research support this finding as a significant
association between own use of deception and
perceived partner deception was obtained. This also
supports the norm of reciprocity (social exchange
theory) in that, if someone believes that their partner
is deceiving them, they too will retaliate by engaging
in deception themselves in order to match their
partner’s deceit. Whilst from the current research
males were generally more successful in predicting if
n
70
17
11
Partner deception
M
SD
2.76
.83
3.23
1.39
3.70
1.23
their partner engaged in deception, females own use
of deception was associated more to the belief or
perception that their partner engages in deception.
Possible reasons for this stronger association in
females compared to males, can be due to the fact that
females are considered to be more concerned with
interpersonal behaviour and have a more integrated
view of trust, they often are more dependant on their
partner (Rempel, et al, 1985). They tend to place more
value in the relationship and therefore if they suspect
deceit, they will retaliate more than their male
counterpart.
Simpson and Rholes (1998) suggested using
Bartholomew’s four category attachment style
measure. Those who have a negative view of others
avoid intimacy for reasons such as fear of rejection
and or have negative expectations about their
relationship. These are characteristics of the fearful
and dissimive attachment styles. Ainsworth similarly
states that the avoidant attachment style (fearful
/dismissive) lack confidence in themselves and they
56
FALLON & HOLT
view others as being unreliable and unwilling to
commit themselves (Collins & Read, 1990). Results
from this research found that those who possess a
fearful/dimissive attachment style, are associated with
an increase in own use of deception, the belief that
their partner engages in deception, relationship
concern and a decrease in faith, dependability, being
responsive and the ability to deal with conflict.
Simpson and Rholes (1998) also suggested that
those who have a negative model of themselves, again
anxiously gain to seek acceptance from others which
is a characteristics of the pre-occupied attachment
style and are also dependant on others which is a
characteristic of the fearful attachment style. Results
from this research found that those who have a
negative view of themselves (pre-occupied attachment
style) is associated with an increase in own use of
deception, belief that their partner engages in
deception. This is also associated with those who have
a negative view of others (fearful/ dismissive
attachment style).
Whilst there is a lack of research examining the
association between deception and attachment styles
Cole (2001) found an association between deception
and those who possess a negative model of
themselves (fearful and preoccupied attachment
styles) and those who possess a negative model of
others (fearful/dismissive attachment styles). The
present findings support Cole’s (2001) association, as
deception was significantly correlated with avoidance
(negative model of others) and with anxiety (negative
model of self) measures. Differences in deception
were also found among attachment styles as those
securely attached scored significantly lower on own
use of deception than the pre-occupied and the
fearful/dismissive attachment styles. The present
findings support Cole’s (2001) assumptions that
people who possess a negative model of others (fear
of intimacy) engage in deception possibly to keep
others at a distance and those who possess a negative
model of themselves (anxiously seek acceptance)
engage in deception to increase a partner’s interest in
them.
Although the results obtained relate to previous
research and theories, there are however some
methodological limitations which need to be
considered. The use of self report measures may not
be a true representation of individuals’ actual
characteristics, as people are less likely to report their
use of deception as it is socially frowned upon. They
may fear that their partner might find out that they are
being lied too, as in conversation with participants
most where curious about their partner’s responses to
questions regarding deception. Also people may have
different views in what they define as deception,
omissions or white lies may not be viewed as deceit
by some participants (West, 2001)
Future research could include an ‘observer report’
for example a close friend(s) of the couple, who is in a
position to report each partner’s or a partner’s level of
trust and deception. Another threat to validity of the
results is due to the comparison of unequal and small
group sizes, as the majority of participants in this
study categorized into the securely attached style, and
few participants represented the fearful/dismissive
and pre occupied attachment styles.
The use of dyadic data is another methodological
concern, as partner’s scores are interdependent of
each other. Previous research has found significant
low to moderate correlations between partner scores,
however interdependence does not account for all
variation among scores (West, 2001).
In accordance with the social exchange theory, own
use of deception is associated with the belief that their
partner engages in deception. In order to protect one’s
self-esteem, own use of deception is associated with a
decrease in trust as previous research suggests that
individuals who engages in deception automatically
create false impressions that their partner engages in
deception to avoid feelings of guilt. This investigation
also demonstrated that the infant-care taker bond
formed, being the blue-print for future relationships,
is associated with trust and deception. Differences
were found between attachment styles. Those with
pre-occupied and fearful/dismissive attachment styles
are associated with deception, for possible reasons of
fear of abandonment (pre-occupied attachment style)
and fear of intimacy (fearful/dismissive attachment
style). Although results support previous literature,
caution must be made with current findings as selfreport measures were used as the basis of analysis,
interdependence of partner scores were not taken into
consideration and some group sizes were relatively
small. Further research could investigate the causal
links between attachment styles and deception to
further our understanding in why some find it easy to
fall in love and others don’t.
References
Argyle, M. (1983). The psychology of interpersonal
behavior 4th ed. Hamandsworth: Pengiun.
Bartholomew, K., & Horowitz, L.M. (1991).
Attachment styles among young adults: A test of a
four-category model. Journal of Personality and
Social Psychology, 61, 226-244.
Boon, S.D., & McLeod, B.A. (2001). Deception in
romantic relationships: Subjective estimates of
success at deceiving and attitudes towards
deception. Journal of Social and Personal
Relationship, 18, 463-476.
Cole, T. (2001). Lying to the one you love: The use of
deception in romantic relationships. Journal of
Social and Personal Relationships, 18, 107-129.
Collins, N.L., & Read, S.J. (1990). Adult attachment,
working models, and relationship quality in dating
couples. Journal of Personality and Social
Psychology, 58, 644-663.
DePaulo, B.M., & Kashy, D.A. (1998). Everyday lies
in close and casual relationships. Journal of
Personality and Social Psychology, 74, 63-79.
DePaulo, B.M., Kashy, D.A., Kirkendol, S.E., Wyer,
M.M., & Epstein, J.A. (1996). Lying in everyday
life. Journal of Personality and Social Psychology,
70, 979-995.
Feeney, J.A., & Noller, P. (1990). Attachment style as
a predictor of adult romantic relationships. Journal
of Personality and Social Psychology, 58, 281-291.
Lying, deception and attachment style
Fraley, R.C., & Davis, K.E. (1997). Attachment
formation and transfer in young adults’ close
friendships and romantic relationships. Personal
Relationships, 4, 131-144
Guerrero, L.K. (1996). Attachment-style differences
in intimacy and involvement: A test of the fourcategory model. Communication Monographs, 63,
269-292.
Hazan, C., & Shaver, P. (1987). Romantic love
conceptualized as an attachment process. Journal of
Personality and Social Psychology 52, 511-524.
Keelan, J.P., Dion, K.K. & Dion, K.L. (1998).
Attachment style and relationship satisfaction: Test
of a self-disclosure explanation. Canadian Journal
of Behavioral Science, 30, 24-35.
Rempel, J.K., Holmes, J.G., & Zanna, M.P. (1985).
Trust in close relationships. Journal of Personality
and Social Psychology, 4, 95-112.
Simpson, J.A., & Rholes, W.S.(Eds.). (1998).
Attachment theory and close relationships. New
York: The Guilford press.
West, A.E. (2001). The Effect of deception on trust,
satisfaction and commitment in intimate
relationships.
Unpublished
postgraduate
dissertation, Australian Catholic University, St
Patrick’s Campus, Melbourne, Victoria, Australia..
Correspondence to:
Barry J Fallon
School of Psychology
Australian Catholic University
115 Victoria Parade
Fitzroy, VIC 3165
57
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 58-62
Adoption, Attachment, and Relationship Experiences
Judith A. Feeney (judy@psy.uq.edu.au)
School of Psychology
University of Queensland, Brisbane, QLD 4072 Australia
Nola L. Passmore (nolapass@usq.edu.au)
School of Psychology
University of Southern Queensland, Toowoomba, QLD 4350 Australia
Candida C. Peterson (candi@psy.uq.edu.au)
School of Psychology
University of Queensland, Brisbane, QLD 4072 Australia
Abstract
Adoption, Family Experiences and
Adjustment
There is evidence that adopted persons are overrepresented in clinical populations, and fare worse than
non-adoptees on various psychosocial variables,
including relational adjustment. However, some studies
have yielded null results, and others have highlighted
the greater variability in adjustment among adopted
persons. Despite considerable discussion about loss,
rejection, and trust in adoptees, no research has
comprehensively explored the impact of adoption on
dimensions of attachment security and relationship
outcomes in adulthood, or the moderating role of
family and search/reunion experiences. This paper
presents the first phase of a longitudinal study of adults
who were adopted as infants, and a comparison sample
of adults who grew up with both biological parents.
The samples were generally comparable in terms of
relevant background variables. Overall, attachment
profiles indicated less security in the adopted sample
than the comparison sample. However, variability was
somewhat greater among adoptees, and those who had
not searched for birth relatives did not differ from the
comparison group. For adopted persons, attachment
security was related to perceptions of childhood
relationships with adoptive parents and, to a lesser
extent, relationships with birth mothers. Ongoing work
will focus on stability and change in relationship
variables, and in-depth exploration of adopted persons’
experiences.
Many researchers have investigated the extent to
which adoption is a risk factor for general adjustment
difficulties. There is evidence that adoptees are overrepresented in clinical populations (Wierzbicki,
1993), and fare worse than non-adoptees on a range of
psychosocial variables including self-esteem and
depression (Borders, Penny, & Portnoy, 2000; Cubito
& Obremski-Brandon, 2000; Levy-Shiff, 2001).
However, Collishaw, Maughan, and Pickles (1998)
found that adoptees were no different to the general
population in terms of psychological distress, and
Borders et al. (2000) found no difference between
adoptees and their friends in terms of life satisfaction.
Further, Sharma, McGue, and Benson (1996) reported
higher levels of prosocial behaviour among adopted
than nonadopted adolescents. While methodological
differences may explain some of these mixed results,
an alternative explanation is that factors such as
search status and family experiences moderate the
association between adoption status and psychosocial
outcomes. Research shows that adoptees who have
searched for birth parents (‘searchers’) have lower
self-esteem than non-searchers (Aumend & Barrett,
1984; Borders et al., 2000; Sobol & Cardiff, 1983).
Further, Sobol and Cardiff (1983) found a trend
among non-searchers such that those who had less
favourable relationships with adoptive parents were
more likely to indicate a desire to search in the future.
Hence, a well-functioning adoptive family may act as
a buffer against unfavourable psychosocial outcomes.
Indeed, there is evidence that when the adoptive
family is open and supportive, adoptees are more
likely to develop healthy self-esteem (Kelly, TownerThyrum, Rigby, & Martin, 1998; Levy-Shiff, 2001).
If adoption is a risk factor for psychosocial
difficulties, at least for some adoptees, then many
adoptees may also experience difficulties in
interpersonal relationships. Issues concerning loss and
betrayal (which are inherently relational) are central
to the adoption experience. Not only have adoptees
lost their birth parents; they have also lost other birth
relatives, knowledge of their genetic heritage, and the
knowledge of being biologically tied to significant
others (Jones, 1997; Schechter & Bertocci, 1990). In
cases where an unsuccessful reunion has taken place,
adoptees may feel rejected again, and thus experience
Personal relationships play a crucial role in meeting
needs for intimacy and security, and in promoting
physical and psychological well-being (Baumeister &
Leary, 1995; Weiss, 1991). For these reasons, issues
regarding the development and deterioration of
personal relationships are of prime importance. More
specifically, there are crucial reasons for investigating
the relationship experiences and relationship concerns
of adult adoptees. Adoptees have lost the major
person(s) with whom attachments are normally
formed (i.e., biological parents). Further, with recent
changes in legislation, many adult adoptees are now
searching for birth relatives, and search and reunion
experiences sometimes entail further loss and
rejection. Hence, adoption may be a risk factor for
negative relational attitudes and relationship
difficulties in adult life.
58
Adoption, attachment & relationship experiences
a double loss. Parental loss through death or divorce
has already been linked to insecurity and interpersonal
difficulties (Kobak, 1999), but the losses associated
with adoption have unique features that may
predispose individuals to relationship problems.
Specifically, these losses are covert and often
unacknowledged or downplayed, and may entail a
sense of betrayal and abandonment (Brodzinsky,
1990; Jones, 1997; Nickman, 1985). Although several
studies have investigated the family relationships of
adoptees, few have assessed the impact of adoption on
the peer relationships of adult adoptees. As argued
below, attachment theory is uniquely suited to
addressing these issues.
The Adult Attachment Perspective
Hazan and Shaver (1987) argued that adults’ close
relationships (especially those between spouses and
dating partners) share important emotional and
behavioural similarities with infant-caregiver bonds,
and that the concept of ‘attachment style’ is relevant
to both types of relationships. These researchers
proposed that differences in adult attachment security
predict key relationship outcomes and reflect, in part,
childhood experiences with attachment figures.
Subsequent research has supported these propositions
(Feeney, 1999).
Recently, Edens and Cavell (1999) argued for the
utility of attachment theory in the study of adoption,
noting that current conceptualisations of adult
attachment are directly relevant to relationship
phenomena that are unique to adoptees, including loss
of biological ties, and the potential for search and
reunion. To our knowledge, however, Borders et al.
(2000) are the only researchers to have systematically
investigated attachment security in adult adoptees.
These researchers studied adoptees and their nonadopted friends. While these two groups did not differ
in marital satisfaction or sensitivity to rejection,
differences did emerge with regard to adult
attachment and social support. In terms of attachment
categories, non-adoptees were over-represented in the
secure group, and adoptees (regardless of search
status) were over-represented in the preoccupied and
fearful groups. Adoptees also reported lower social
support than their non-adopted friends, though this
association was moderated by search status: Searchers
reported lower support than non-searchers and nonadopted respondents. This study provided a first step
in exploring the link between adoption and adult
attachment, but was limited by its cross-sectional
nature, its reliance on a categorical measure of
attachment, and its failure to fully consider the role of
early parenting and ongoing relationship experiences.
In summary, despite the considerable literature on
issues of infant attachment, loss, rejection,
abandonment and trust in adoptees, no study to date
has comprehensively explored the impact of adoption
on dimensions of attachment security and relationship
outcomes in adulthood, or the moderating role of
family and search/reunion experiences on these
outcomes. This paper reports on the first phase of a
study addressing these issues. We expected that adults
who were adopted as infants would report higher
59
levels of insecurity than adults who grew up with both
biological parents (Hypothesis 1a). However, the
adopted sample was also expected to show greater
variability on attachment measures (Hypothesis 1b).
Within the adopted sample, insecurity was expected to
be higher for those who had searched for birth
relatives (Hypothesis 2), and those who perceived
relationships with adoptive parents and birth mothers
in a more negative light (Hypothesis 3).
Method
Participants
Participants were 131 adults who were adopted as
infants, and a comparison sample of 124 adults who
grew up with both biological parents. The
demographic characteristics of the two samples are
shown in Table 1. Analysis of variance indicated that
the samples were similar in terms of age, and
frequency comparisons generally revealed similar
patterns of gender, relationship status, parental status,
education level, and employment status.
Table 1
Demographic characteristics of comparison and
adopted samples
Variable
Age
Comparison
M = 37.63 years
Adopted
M = 37.67 years
Gender
74.19% females
77.86% females
Relationship
status
33.62% single
11.21% de facto
44.00% married
11.21% sep/ divorced
26.32% single
12.28% de facto
42.98% married
18.42% sep/ divorced
Parental
status
47.15% with children
57.25% with children
Education
level
7.26% some high school
6.45% Year 12
24.19% some further study
62.10% university degree
9.92% some high
school
9.92% Year 12
35.11% some further
study
45.04% university
degree
Employment
status
43.90% full-time
32.52% part-time
23.58% not employed
49.62% full-time
34.35% part-time
16.03% not employed
Measures
As part of a larger study, participants completed
measures of parental bonding and attachment security.
In addition, adoptees reported on their search and
reunion experiences.
Parental Bonding The Parental Bonding Instrument
(PBI; Parker, Tupling, & Brown, 1979) contains 25
items describing various parental attitudes and
behaviours. This measure requires participants to
think back over the first 16 years of life, and to rate
each item (separately for mother and father) from 0
60
FEENEY, PASSMORE, & PETERSON
(very unlike this parent) to 3 (very like this parent).
Adoptees answered these questions with respect to
their adoptive parents. (Those in the comparison
sample answered with respect to their biological
parents, but these data are not considered in this
paper.) The PBI yields scores on the dimensions of
care (12 items, e.g., ‘spoke to me with a warm and
friendly voice’), and overprotection (13 items, e.g.,
‘tried to control everything I did’). Both scales were
highly reliable, with alpha coefficients exceeding .90
for each parent.
Attachment Security Current attachment security
was measured in two ways. First, attachment style
was assessed by asking participants to endorse one of
the four attachment descriptions (secure, preoccupied,
dismissing, fearful), developed by Bartholomew and
Horowitz (1991). Second, participants completed the
40-item Attachment Style Questionnaire (ASQ),
developed by Feeney, Noller, and Hanrahan (1994).
The ASQ measures five dimensions of adult
attachment: confidence in self and others (8 items;
e.g., ‘I feel confident about relating to others’),
discomfort with closeness (10 items; e.g., ‘I prefer to
keep to myself’), need for approval (7 items; e.g., ‘It’s
important to me that others like me’), preoccupation
with relationships (8 items; e.g., ‘I worry a lot about
my relationships’), and relationships as secondary to
achievement (7 items; e.g., ‘Achieving things is more
important than building relationships’). Each item is
rated on a 6-point scale, from 1 (totally disagree) to 6
(totally agree). All five scales were reliable, with
alpha coefficients ranging from .74 to .88.
Search and Reunion Experiences Adoptees
answered a series of questions about their search and
reunion experiences. This paper considers only three
questions dealing with relationships with birth
mothers; these questions tapped satisfaction with the
initial reunion and satisfaction with the current
relationship (1 = extremely dissatisfying to 6 =
extremely satisfying), and emotional closeness of the
current relationship (1 = extremely distant to 6 =
extremely close).
Procedure
Participants for both samples were recruited through
the first-year Psychology pools at the University of
Queensland and University of Southern Queensland,
through brochures and flyers placed on university
campuses and in various community centres, and
through advertisements in local media. Those
interested in participating were informed of the
purpose and confidential nature of the study, and were
mailed a questionnaire package with a pre-paid
envelope for return of the materials. The major
sections of the questionnaire (search and reunion
experiences, attachment security, relationships with
adoptive parents) were presented in counterbalanced
order.
Results
Adoption and Attachment Characteristics
The association between adoption and attachment
security was assessed in two ways. First, a frequency
comparison was conducted, relating sample
(comparison versus adopted) to the four-group
(forced-choice) measure of attachment style. Results
revealed a significant association between sample and
attachment category, χ2 (3) = 19.41, p < .001.
Adopted persons represented only 37% of the secure
group; in contrast, they represented 55% of the
dismissing group, 63% of the preoccupied group, and
72% of the fearful group.
Second, a MANOVA was conducted to assess
differences between the samples on the five scales of
the ASQ. This analysis revealed a significant overall
difference, multivariate F (5, 249) = 3.49, p < .005;
further, univariate tests showed that this difference
applied to all five scales. Adopted persons obtained
lower scores than comparison persons on confidence,
and higher scores on all remaining attachment scales
(see Table 2, top rows). The multivariate test of
homogeneity variance was also marginally significant,
indicating greater variability in attachment scores
within the adopted sample than the comparison
sample. However, this effect applied only to the
confidence scale.
Search Status
To assess the role of search status, a more finegrained MANOVA was conducted in which searchers
(n = 97), nonsearchers (n = 32) and comparison
participants (n = 124) were compared on the
attachment scales. Significant differences emerged on
all scales except for relationships as secondary. Post
hoc (Tukey) tests showed a consistent pattern,
involving significant differences between the reports
of searchers and comparison participants (see Table
2): Searchers reported lower levels of confidence, and
higher levels of discomfort, preoccupation, and need
for approval. Interestingly, non-searchers and
comparison participants did not differ on any of the
five scales.
Table 2
Mean scores and standard deviations on the attachment
scales according to group
Comparison
Adopted
Searchers
Nonsearchers
Conf.
Disc.
4.50
(0.81)
4.10
(1.01)
4.05
(0.99)
4.24
(1.10)
2.95
(0.92)
3.25
(1.00)
3.29
(1.04)
3.18
(0.96)
Relate
second.
2.06
(0.75)
2.27
(0.80)
2.24
(0.84)
2.37
(0.73)
Need
approv.
2.97
(0.92)
3.34
(1.03)
3.40
(1.04)
3.18
(0.99)
Note. Conf. = Confidence, Disc. = Discomfort, Relate second. =
Relationship as secondary, Need approv. = Need for approval, Preocc.
= Preoccupation
Preocc.
3.14
(0.95)
3.55
(0.96)
3.61
(0.93)
3.31
(1.08)
Adoption, attachment & relationship experiences
Relationships with Adoptive Parents and with
Birth Mothers
To assess the role of family relationships (within the
adopted sample), the five attachment scales were
correlated with reports of childhood relationships with
adoptive parents, and, for those who had met their
birth mother (n = 75), with ratings of that relationship.
These results are shown in Table 3. The strongest
associations with reports of relationships with
adoptive parents were for the confidence, discomfort,
and need for approval dimensions of attachment.
61
Confidence was related positively to parental care
(from both adoptive mother and adoptive father), and
negatively to parental overprotection. Conversely,
discomfort and need for approval were related
negatively to parental care, and positively to parental
overprotection.
Associations with items assessing relationships with
birth mothers were more scattered. However,
perceptions of a satisfying reunion were associated
with more confidence and less discomfort.
Table 3
Correlations between attachment security and adoptees’ reports of parental relationships
Relationship variable
Adoptive parents
Maternal care
Paternal care
Maternal overprot.
Paternal overprot.
Birth mother
satisfaction (reunion)
satisfaction (current)
closeness
Conf.
Disc.
Relate
second.
Need
approv.
Preocc.
.42***
.32***
-.38***
-.29***
-.41***
-.27**
.37***
.24**
-.18*
-.09
.17+
-.02
-.23**
-.30***
.28***
.25**
-.22*
-.23**
.22*
.10
.23+
.10
.11
-.25*
-.10
-.10
-.22
-.15
-.17
-.07
-.07
-.02
-.09
-.01
-.02
Note. Conf. = Confidence, Disc. = Discomfort, Relate second. = Relationship as secondary, Need approv. = Need for approval, Preocc. =
Preoccupation;
+ p < .06, * p < .05, ** p < .01, *** p < .001
Discussion
Before discussing the results in more detail, it is
important to consider the limitations of the study.
Sample size was relatively small for some categories
(e.g., those who had met their birth mothers). Further,
because all these data were drawn from Phase 1 of the
study, they are cross-sectional in nature. This issue
complicates, in particular, interpretations of the
associations between measures of attachment security
and relationships with birth mothers. It is possible, for
example, that avoidant attachment (which is reflected
in high levels of discomfort with closeness) colours
reports of the reunion experience, rather than the
reverse.
Overall, the results support the suggestion that
insecure attachment is more widespread among
adoptees than among the general population.
Adoptees scored lower than comparison participants
on confidence in self and others, and higher on all five
scales tapping dimensions of insecurity. Adoptees
were also over-represented in the insecure attachment
categories.
This
over-representation
applied
particularly to the fearful style; attachment theory
suggests that fearful attachment represents the most
negative pattern of working models (Bartholomew &
Horowitz, 1991), and empirical research on the
characteristics of the four styles supports this
assertion (e.g., Feeney, Noller, & Hanrahan, 1994).
These findings on the attachment characteristics of the
samples support Hypothesis 1a; moreover,
supplementary analyses indicated that the results were
robust across gender. Hence, these findings fit with
the suggestion that attachment theory provides a
useful perspective on the relationship issues that arise
for adoptees, including loss of biological ties and the
potential for search and reunion (Edens & Cavell,
1999).
However, consistent with Hypothesis 1b, the results
also pointed to the wide variability in adoptees’
responses to measures of attachment security. Further,
on four of the five attachment scales, it was only those
who had searched for birth relatives who reported
higher levels of insecurity than the comparison
sample. These results fit with previous studies linking
search status to indices of psychological adjustment
(e.g., Sobol & Cardiff, 1983), and support Hypothesis
2.
As expected (Hypothesis 3), our results supported
the association between positive childhood
relationships with adoptive parents (high levels of
care, low levels of overprotection) and current
attachment security. This finding fits with the broader
literature on the link between responsive parenting
and offspring’s attachment security (e.g., Rothbard &
Shaver, 1994). To a lesser extent, the results also
point to the relevance of relationships with birth
mothers, at least in terms of the reunion experience.
The weaker nature of the findings pertaining to birth
mothers probably reflects not only the smaller sample
size, but also the less formative nature of these
relationships, compared to those with primary
caregivers.
Finally, it is worth noting some of the future
directions in this research project. We are interested in
the stability of adult attachment over time, given
previous suggestions that working models may be
more tentative and unstable in insecure individuals.
Further, we have collected a large body of qualitative
FEENEY, PASSMORE, & PETERSON
62
data tapping relationships with adoptive parents and
experiences of search and reunion, and these data will
be supplemented by in-depth interviews with the
adoptees. One of the complex issues to be addressed
in the interviews concerns adoptees’ experiences of
negotiating roles, relationships, and boundaries
between birth and adoptive families. These issues
need to be addressed in order to develop best practice
for adult adoptees who may be at risk of relationship
problems.
References
Aumend, S. A., & Barrett, M. C. (1984). Self-concept
and attitudes toward adoption: A comparison of
searching and nonsearching adult adoptees. Child
Welfare, 63, 251-259.
Bartholomew, K., & Horowitz, L. M. (1991).
Attachment styles among young adults: A test of a
four-category model. Journal of Personality and
Social Psychology, 61, 226-244.
Baumeister, R. F., & Leary, M. R. (1995). The need to
belong: Desire for interpersonal attachments as a
fundamental human motivation. Psychological
Bulletin, 117, 497-529.
Borders, L. D., Penny, J. M., & Portnoy, F. (2000).
Adult adoptees and their friends: Current
functioning and psychosocial well-being. Family
Relations, 49, 407-418.
Brodzinsky, D. M. (1990). A stress and coping model
of adoption adjustment. In D. M. Brodzinsky & M.
D. Schechter (Eds.), The psychology of adoption
(pp. 3-24). Oxford: Oxford University Press.
Collishaw, S., Maughan, B., & Pickles, A. (1998).
Infant adoption: Psychosocial outcomes in
adulthood. Social Psychiatry and Psychiatric
Epidemiology, 33, 57-65.
Cubito, D. S., & Obremski-Brandon, K. (2000).
Psychological adjustment in adult adoptees:
Assessment of distress, depression, and anger.
American Journal of Orthopsychiatry, 70, 408-413.
Edens, J. F., & Cavell, T. A. (1999). A review and
reformulation of adoptive relationships from an
attachment perspective. Adoption Quarterly, 3, 4370.
Feeney, J. A. (1999). Adult romantic attachment and
couple relationships. In J. Cassidy & P. R. Shaver
(Eds.), The handbook of attachment: Theory,
research, and clinical applications (pp. 355-377).
New York: Guilford.
Feeney, J. A., Noller, P., & Hanrahan, M. (1994).
Assessing adult attachment: Developments in the
conceptualization of security and insecurity. In M.
B. Sperling & W. H. Berman (Eds.), Attachment in
adults: Theory, assessment, and treatment (pp.128152). New York: Guilford.
Hazan, C., & Shaver, P. R. (1987). Romantic love
conceptualized as an attachment process. Journal of
Personality and Social Psychology, 52, 511-524.
Jones, A. (1997). Issues relevant to therapy with
adoptees. Psychotherapy, 34, 64-68.
Kelly, M. M., Towner-Thyrum, E., Rigby, A., &
Martin, B. (1998). Adjustment and identity
formation in adopted and nonadopted young adults:
Contributions of family environment. American
Journal of Orthopsychiatry, 68, 497-500.
Kobak, R. (1999). The emotional dynamics of
disruptions in attchment relationships: Implications
for theory, research, and clinical intervention. In J.
Cassidy & P. R. Shaver (Eds.), The handbook of
attachment: Theory, research, and clinical
applications (pp. 355-377). New York: Guilford.
Levy-Shiff, R. (2001). Psychological adjustment of
adoptees in adulthood: Family environment and
adoption-related correlates. International Journal of
Behavioral Development, 25, 97-104.
Nickman, S. L. (1985). Losses in adoption: The need
for dialogue. Psychoanalytic Study of the Child, 40,
365-398.
Parker, G., Tupling, H., & Brown, L. B. (1979). A
Parental Bonding Instrument. British Journal of
Medical Psychology, 52, 1-10.
Rothbard, J. C., & Shaver, P. R. (1994). Continuity of
attachment across the life span. In M. B. Sperling &
W. H. Berman (Eds.), Attachment in adults: Theory,
assessment, and treatment (pp. 31-71). New York:
Guilford.
Schechter, M. D., & Bertocci, D. (1990). The
meaning of the search. In D. M. Brodzinsky & M.
D. Schechter (Eds.), The psychology of adoption
(pp. 62-90). Oxford: Oxford University Press.
Sharma, A. R., McGue, M. K., & Benson, P. L.
(1996). The emotional and behavioral adjustment of
United States adopted adolescents: Part 1. An
overview. Children and Youth Services Review, 18,
83-100.
Sobol, M. P., & Cardiff, J. (1983). A
sociopsychological investigation of adult adoptees’
search for birth parents. Family Relations, 32, 477483.
Weiss, R. S. (1991). The attachment bond in
childhood and adulthood. In C. M. Parkes, J.
Stevenson-Hinde, & P. Marris (Eds.), Attachment
across the life cycle (pp. 66-76). London:
Tavistock/Routledge.
Wierzbicki, M. (1993). Psychological adjustment of
adoptees: A meta-analysis. Journal of Clinical
Child Psychology, 22, 447-454.
Acknowledgements
This study has been supported by a grant from the
Australian Research Council.
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 63-67
Favouritism and Rejection in Families: Black sheep and Golden-haired
Children
Julie Fitness (jfitness@psy.mq.edu.au)
Psychology Department
Macquarie University, Sydney 2109
“as though they are irretrievably bad and blamed for
all the tension and strife in the family” (p. 31).
Therapists argue that scapegoating means families can
maintain levels of solidarity and cohesiveness that
they could not otherwise maintain (Vogel & Bell,
1960). However, such treatment is extremely painful
for the family member who is object of scapegoating.
Furthermore, a scapegoated child is frequently
threatened with ‘eviction’ from the family, and may
even be sent away, e.g., to boarding schools, reform
schools, or foster families.
An interesting question, then, is why one child in a
family might be scapegoated, while another child
might be particularly highly regarded. Rohde et al.
(2003) have argued that, from an evolutionary
perspective, some offspring are simply better
investments than others, depending on factors such as
their sex, age, and health; the mother’s age and health
(and opportunities for future reproduction); and
environmental factors such as the availability of
resources and support. Offspring who, for one reason
or another, are ‘riskier’ investments (or who perhaps
require less investment than more vulnerable siblings)
may quite accurately perceive that they are,
comparatively speaking, disfavoured. Another
possibility is that offspring who jeopardize their own
and/or their siblings’ reproductive success (e.g., via
deviance, criminality, etc.) may be relatively
disfavoured, whereas offspring who enhance their
own and their siblings’ reproductive success (e.g.,
through talent, attractiveness etc.), may be relatively
favoured.
A second and related factor that might contribute to
differential parental treatment derives from the welldocumented human tendency to invest more in
genetically-related others than in non-genetically
related others (e.g., see Daly & Wilson, 1998). This
implies that adopted and foster children may well feel
less included and more rejected than biological
children in a family. Indeed, Daly and Wilson have
marshaled an impressive body of evidence
demonstrating that stepchildren are at relatively
greater risk for assault and murder by step-fathers
than are fathers’ biological offspring. Step-children
are also more likely than biologically-related children
to be killed by more violent and malicious means, as
opposed to being killed in their sleep. Clinical data
suggest that step-children, adopted children and foster
children are all more vulnerable to being scapegoated,
as are children with physical or intellectual disabilities
and signs of ‘difference’ from other children in the
family (Brody et al., 1998).
A more proximal factor that may elicit differential
treatment of children involves birth rank. Sulloway
(1996) proposed that offspring can enhance their
Abstract
Despite the assumption that parents aim to treat their
children equally, children frequently perceive and
complain about parental favouritism and its mirror
image, disfavouritism. Psychologists and evolutionary
theorists have proposed several factors that might
account for familial favouritism and rejection,
including offspring viability, genetic relatedness,
similarity to parents, sex, and birth rank. Little is
known, however, about laypeople’s theories of the
causes of familial rejection. In the current study, 70
respondents reported on the incidence of, and perceived
reasons for, family favouritism and rejection in their
immediate and extended families. Nearly 69% of
respondents reported the existence of a family
favourite, with perceived reasons for favouritism
involving birth order, sex, ‘goodness’ and similarity to
parents. Some 80% of respondents reported that there
was a black sheep in the family, with perceived reasons
for rejection involving a sense of difference, genetic
unrelatedness,
trouble-making,
and
deviance.
Interestingly, respondents’ perceptions of familial
favoritism and rejection were negatively associated
with reported closeness to extended family while
growing up. Implications for family-related theory and
research are discussed.
There is an implicit assumption (in Western society,
at least) that parents treat their children equally
(Klagsbrun, 1992). However, children frequently
perceive and complain about parental favoritism
(Harris & Howard, 1988; Kowal, Kramer, Krull, &
Crick, 2002; Zervas & Sherman, 1994). For example,
in one study of 127 males and females aged 17-30 and
62 of their siblings (aged 18-32 years), Brody,
Copeland, Sutton, Richardson, and Guyer (1998)
found that 65% of participants reported favouritism in
their families and 24% reported disfavouritism (i.e.,
not receiving the same benefits as other siblings).
There was modest sibling agreement both on the
existence of favouritism and on the identity of the
favoured/disfavoured child. Crucially, the presence of
disfavouritism was positively related to lower family
cohesion, higher family disengagement and higher
family conflict. Moreover, participants who perceived
themselves to be disfavoured experienced more
frequent and intense shame and fear.
Given that most parents at least try to treat their
children equally, how is it that perceptions of
favouritism and disfavouritism arise in families? A
number of potentially important contributing factors
have been suggested. For example, some children
may experience disfavouritism as a function of family
dynamics. Clinicians with interests in family therapy
have described so-called family ‘scapegoats’,
described by Dare (1993) as children who are treated
63
64
FITNESS
chances of survival by occupying different family
‘niches’ and so reduce inter-sibling competition for
resources. Based on his investigations, Sulloway
argued that middle born children are more likely to
rebel against traditional family values than firstborns,
who tend to be family-conscious and relatively
conservative. Middleborns may also be less
‘indulged’ than lastborns, who are frequently in a
position to enjoy more exclusive use of parental
resources (including time and accumulated financial
assets). Middleborns, then, may be correspondingly
more likely to perceive disfavouritism than firstborns
or lastborns.
There is some empirical evidence supporting the
idea that middleborns differ from firstborns and
lastborns. For example, Salmon (2003) found in a
study of 245 Canadian students that middleborns
expressed more positive views towards friends, and
less positive opinions of family, than firstborns or
lastborns. They were also less inclined to help the
family in need than firstborns or lastborns.
Along with birth-order, another proximal factor
contributing to parental favouritism is a child’s sex.
As Hrdy (2000) has pointed out, there is a widespread historical preference around the world for
male, rather than female, children. Within any one
family, however, a preference for one sex or the other
may depend on factors such as how many children of
a particular sex have already been born.
Clearly, there are a number of factors that play
potentially important roles in the differential treatment
of children in families. In particular, theories derived
from evolutionary social psychology suggest that the
most important factors may derive from offspring
qualities (such as genetic relatedness) and/or
behaviours (such as criminality) that potentially
impact on a person’s reproductive success, and/or that
of his/her siblings. However, these theories have not
yet been empirically examined. Moreover, the
question of how laypeople think about and explain
their favoured or disfavoured status within the family
to themselves and others is also unexplored. For
example, to what extent do disfavoured offspring
blame themselves for their relatively rejected status?
And how do favoured offspring justify their relatively
enhanced familial status?
Overview of Current Study
The aims of this preliminary study were first, to
explore laypeople’s accounts of their experiences and
perceived causes of familial favouritism and
disfavouritism; and second, to examine the extent to
which these accounts would be in line with the kinds
of causal factors suggested by evolutionary theorists.
A semi-structured questionnaire was administered to
individuals participating in a larger-scale study on
family rules, rule violations, and forgiveness (see
Fitness, in press; Fitness & Parker, 2003).
Participants’ accounts were coded for details
pertaining to the reported targets and causes of
familial favouritism and disfavouritism.
Method
Participants
Seventy respondents were recruited from the 1st year
Psychology pool at Macquarie University (26 males
and 44 females, M age = 27.5 years). Of these, 34%
were firstborns, 36% were middleborns, and 30%
were lastborns. The majority (94%) were of European
background. Participants received one hour’s course
credit for participating in the study.
Materials and Procedure
Along with a battery of questionnaires unrelated to the
current study, participants completed a questionnaire
containing the following, open-ended questions: a)
was there a favorite in your family? And if so, how
did you know s/he was the favourite? and b) was there
a ‘black sheep’ (defined for participants as “someone
who was not approved of, or liked, or included as
much as the others”) in your family? And if so, how
did you know s/he was the black sheep? They were
also asked demographic questions about their sex,
age, birth order, and number of siblings, and were
asked to rate their perceived closeness to extended
and nuclear family members when they were growing
up on 5 pt. Likert scales (end points, not close at all to
extremely close). Responses to the open-ended
questions were coded by two, independent raters into
a category system based on theoretical accounts of the
potential causes of familial favouritism and
disfavouritism
(e.g.,
birth
order;
genetic
unrelatedness). Responses that did not fall into one of
these categories (e.g., similarity to parent) were coded
initially as ‘other’, then placed in discrete subcategories. Inter-rater reliability was high, with an
overall Cohen’s k of .88.
Results
Family Favourites
In line with the findings from other, larger-scale
studies, nearly 69% of respondents reported there was
a family favourite, with 48% reporting it was
themselves, 35% reporting it was a brother, and 17%
reporting it was a sister. Moreover, and in accord with
Sulloway’s (1996) birth-order theory, 58% of
firstborns and 62% of lastborns believed they were the
family favourites; however, only 31% of middleborns
reported favoured status.
As shown in Table 1, the most frequently reported
reasons for favouritism involved birth order
(specifically, being first or last born, but never
middleborn); sex (in particular, being the only boy or
girl in the family); so-called ‘goodness’ (having
talent, being attractive, likable, ‘fitting in’); similarity
to a parent, and illness (specifically, having a sibling
who required special attention and who was treated in
a way that could be construed as ‘favoured’).
Respondents were just as likely to cite birth order,
sex, goodness, and similarity to a parent as reasons for
their own favoured status as for the favoured status of
siblings.
Favouritism & rejection in families
Table 1
Most Frequently Reported Reasons for Favouritism
(%)
___________________________________
Birth order
33
Sex
23
Goodness
21
Similar to parent 19
Sibling illness
4
___________________________________
Black Sheep
Remarkably, 80% of respondents reported that there
was a black sheep in the family, with 21% claiming it
was themselves, 13% naming a brother, 16% a sister,
25% an uncle, and 25% naming a cousin, in-law or
distant relative. Some 38% of middleborns believed
they were the ‘black sheep’ of the family, compared
to only 7% of firstborns and 18% of lastborns.
The reasons given for black sheep status differed,
depending on who the target was (see Table 2).
Respondents made the kindest attributions to
themselves for having black sheep status, with 100%
citing ‘difference’ to the rest of the family (including
genetic unrelatedness). Perceived difference was also
frequently mentioned to explain a sister’s black sheep
status, along with her perceived rejection of the
family by marrying someone undesirable, and/or
severing contact with family members. Very few
respondents mentioned ‘badness’ (drugs, criminality)
in relation to a sister’s black sheep status. Nearly
three-quarters of black sheep brothers, on the other
hand, were believed to have attained their status via
trouble-making, including drug and alcohol
involvement, rebelliousness, and criminality. The
remaining 29% were perceived as ‘different’ from the
rest of the family.
Finally, the majority of uncles and other extended
family members (99% of whom were male) were
considered to have earned their disfavoured status
through behaving badly. This category included
perceived moral deviance (e.g., homosexuality;
adultery; family desertion; gambling) as well as drugrelated and criminal behaviour (e.g., fraud, theft).
There was also a category best described as
‘strangeness’ that included mental illness and
behaviour regarded by participants as eccentric (e.g.,
nudism). Perceived difference accounted for 10% of
extended family black sheep, and most of these
involved religious differences.
Treatment of Black Sheep
The majority of participants (42%) reported exclusion
or ostracism of the black sheep. Indeed, some 10% of
these participants reported that a whole ‘side’ of a
family had been rejected or ostracized for religious
reasons or because of long-standing feuds and hatreds
(the origins of which were not aalways known or
understood by respondents). Another 16% of
participants described cold, distant, ‘chilly polite’
behaviour toward the black sheep. The remaining
33% reported active rejection of the black sheep,
65
including criticism, nagging, meanness, and ridicule
(including public humiliation).
Table 2
Most Frequently Reported Reasons for
Disfavouritism According to Target (%)
____________________________________
Self
Perceived difference
100
Sister
Perceived difference
72
Rejection of family
25
‘Badness’
3
Brother ‘Badness’
71
Perceived difference
29
Extended
Family
‘Badness’
70
‘Strangeness’
20
Perceived difference
10
____________________________________
Extended Family: A Protective Factor?
There was an interesting and unexpected finding that
emerged in the section of the study designed to
measure perceptions of family closeness. Specifically,
the more contact respondents reported having had
with their extended family as they grew up, the less
likely they were to report there had been a black sheep
(r = -.47) or a favourite (r = -.29) anywhere in the
family.
Discussion
Overall, these findings suggest that laypeople’s
accounts of the incidence and causes of familial
favouritism and disfavouritism are remarkably
consistent with evolutionary and social psychological
accounts. In line with Sulloway’s (1996) theory, birth
order (first or last) was a frequently mentioned reason
for perceived favoured status; however, although
considerably more middleborns than first or lastborns
considered they were the black sheep of the family,
none of them cited birth order as the reason for their
own disfavoured status. Rather, perceived difference
was the most frequently cited reason for a
participant’s own disfavoured status. Indeed, many
participants discussed at length their difference to the
rest of the family – they looked different (frequently
reported), had different personalities, talents, or
interests, and just didn’t ‘fit into’ or feel they
belonged to the family, whether genetically related or
not. This failure to ‘fit in’ was also a prominent
reason for black sheep sisters.
It is interesting to speculate whether, from an
evolutionary perspective, such perceptions of
‘difference’ may indeed serve as proximal cues to the
possibility of genetic unrelatedness (signaling a risky
investment, with respect to parents’ and siblings’
reproductive interests). In relation to this point, it is
worth noting that 19% of favourites were considered
to have attained their status through similarity to a
parent (a cue for genetic relatedness). This possibility
is supported by the findings of a study involving a
hypothetical adoption task in which Volk and Quinsey
(2002) found that men responded most favorably to
66
FITNESS
infants whose facial features most resembled their
own.
There was also an interesting contrast between socalled ‘goodness’ and ‘badness’ in relation to
favouritism and disfavouritism. The majority of
disfavoured males, in particular, were considered to
have attained their status through behaviours that
effectively dishonoured or shamed the family
(criminality, deviance). Such behaviours may
diminish a male’s reproductive potential and
relatedly, through reputational damage, the
reproductive success of his siblings. Indeed, a
prominent aspect of many participants’ accounts
involved the notion of the family as an entity to which
members owe allegiance. For example, a sizable
proportion of disfavoured sisters were considered to
have effectively betrayed their families by rejecting
family values, marrying ‘out’ and/or severing contact.
No doubt such behaviours involve the risk that a
daughter may ‘waste’ her limited opportunities to
produce high quality offspring. However, comments
were also made about a daughter’s duty to remain
close and loyal to family, and to facilitate contact
between grandchildren and grandparents.
Family favourites and scapegoats: The
products of self-fulfilling prophecies?
One important point to consider with respect to both
familial favouritism and disfavouritism concerns the
direction of causality. Brody et al. (1998) noted that if
a child is disfavoured, he or she receives relatively
few parental rewards and so exhibits fewer and fewer
positive behaviours; in effect becoming less ‘likable’.
Similarly, children who are perceived by parents,
teachers, and peers as attractive and likable tend to
receive a disproportionate number of rewards which
may, in turn, increase their popularity. As noted by
Baumeister and DeWall (in press), people who feel
rejected by society frequently exhibit the kinds of
behaviours observed in laboratory studies of rejection,
including self-defeating behaviours and impaired selfregulation. This begs the question re so-called ‘black
sheep’: To what extent were they rejected by their
families because they were different (or deviant), and
to what extent did they become different (or deviant)
as a function of having been (or felt) rejected by their
families? No doubt, there is a bi-directional
relationship here. However, and as discussed earlier,
there are also factors, such as birth rank and degree of
genetic relatedness, that may push children toward the
edge of their families. In this regard, it is worth noting
the associations found between apparent lack of
favouritism or disfavouritism and extended family
involvement. It may be the case that having a large,
available, and involved network of kin provides a
buffer against parental scapegoating of particular
children.
Conclusions
There is an enormous amount of work to be done in
this fascinating field. The current study has a number
of limitations, including the small sample size and the
fact that the data were drawn from a culturally and
socio-economically homogenous population. Largescale cross-cultural studies are needed to more
systematically investigate the evolutionary arguments
proposed here, along with detailed analyses of
relevant familial demographics (e.g., parental age,
health, birth spacing). Importantly, however, research
in this field has applied, as well as theoretical
significance. The need to belong is fundamental to
human beings, and the pain of feeling unacceptably
‘different’ may be severe. Indeed, strong links have
been found between parental rejection and depression
in adolescents and adults (Robertson & Simons, 1998;
Nolan, Flynn, & Garber, 2003). A better
understanding of the causes, beliefs, and assumptions
underlying such rejection may enhance the health and
happiness of humans in their first, and arguably most
crucial, context of belonging: The family.
References
Baumeister, R., & DeWall, C.N. (in press). The inner
dimension of social exclusion: Intelligent thought
and self-regulation among rejected persons. In K.
D. Williams, J. P. Forgas & W. von Hippel (Eds.),
The social outcast: Ostracism, social exclusion,
rejection, and bullying. New York: Psychology
Press.
Brody, L., Copeland, A., Sutton, L., Richardson, D.,
& Guyer, M. (1998). Mommy and daddy like you
best: Perceived family favoritism in relation to
affect, adjustment, and family process. Journal of
Family Therapy, 20, 269-291.
Daly, M., & Wilson, M. (1998). The truth about
Cinderella: A Darwinian view of parental love.
London: Orion Publishing.
Dare, C. (1993). The family scapegoat: An origin for
hating. In V. Varma (Ed.), How and why children
hate (pp. 31-45). London: Jessica Kingsley.
Fitness, J. (in press). Bye bye, black sheep: The
causes and consequences of rejection in family
relationships. In K. D. Williams, J. P. Forgas & W.
von Hippel (Eds.), The social outcast: Ostracism,
social exclusion, rejection, and bullying. New York:
Psychology Press.
Fitness, J., & Parker, V. (2003). Breaking the rules:
Exploring the causes and consequences of
perceived rule violations by family members. Paper
presented at the 32nd Annual Meeting of the
Society of Australasian Social Psychologists,
Sydney.
Harris, I. D., & Howard, K. I. (1988). Correlates of
perceived parental favouritism. Journal of Genetic
Psychology, 146, 45-56.
Hrdy, S. B. (2000). Mother nature. London: Vintage.
Klagsbrun, F. (1992). Mixed feelings: Love, hate,
rivalry, and reconciliation among brothers and
sisters. New York: Bantam Books.
Kowal, A., Kramer, L., Krull, J., & Crick, N. (2002).
Children’s perceptions of the fairness of parental
preferential treatment and their socioemotional
well-being. Journal of Family Psychology, 16, 297306.
Nolan, S., Flynn, C., & Garber, J. (2003). Prospective
relations between rejection and depression in young
Favouritism & rejection in families
adolescents. Journal of Personality and Social
Psychology, 85, 745-755.
Robertson, J., & Simons, R. (1998). Family factors,
self-esteem, and adolescent depression. Journal of
Marriage and the Family, 51, 125-138.
Rohde, P.A., Atzwanger, K., Butovskayad, M.,
Lampert, A., Mysterud, I., Sanchez-Andres, A., &
Sulloway, F. J. (2003). Perceived parental
favouritism, closeness to kin, and the rebel of the
family: The effects of birth order and sex. Evolution
& Human Behavior, 24, 261-276.
Salmon, C. (2003). Birth order and relationships:
Family, friends, and sexual partners. Human
Nature, 14, 73-88.
Sulloway, F. (1996). Born to rebel: Birth order,
family dynamics, and creative lives. New York:
Pantheon.
Vogel, E. F., & Bell, N. W. (1960). The emotionally
disturbed child as the family scapegoat. In: N.W.
Bell and E. F. Vogel (Eds.), A modern introduction
to the family (pp. 382-397). NY: Free Press.
Volk, A., & Quinsey, V. L. (2002). The influence of
infant facial cues on adoption preferences. Human
Nature, 13, 437-455.
Zervas, L. J., & Sherman, M. (1994). The relationship
between perceived parental favoritism and selfesteem. Journal of Genetic Psychology, 155, 25-33.
Correspondence to:
Julie Fitness
Psychology Dept.,
Macquarie University, Sydney 2109
jfitness@psy.mq.edu.au
67
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 68-72
Excessive Internet Use as a Behavioural Addiction
Carina J. Henry (cjhenr001@student.patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
Cecelia Winkelman (c.winkelman@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
these users, time online is not excessive. They tend
not to have personal problems from which Internet
use is an escape, nor does Internet use usually result
in any social or emotional disturbance (Davis, 2001).
For those whose time spent online is dissociated
from real life, their use of the Internet may be
unhealthy. The Internet may be used as an escape
(Suler, 1996). These individuals may portray a
personality to others online different to that of their
offline personality. They may keep their cyber world
separate from their real world, and may not discuss
their time spent online with those outside the Internet.
Rather than viewing these as clear-cut categories,
Internet use may be viewed along a continuum
ranging from healthy to unhealthy use (Davis, 2001).
When individuals engage in use that is skewed
towards the unhealthy end of this continuum, they
may be regarded as addicted. Since the phenomenon
of Internet addiction is relatively new, it can be
compared with aspects of other addictions in order to
validate it as a syndrome.
Addiction is difficult to define, but the consensus is
that it is repetitive abuse of a substance or an activity
that is problematic, either to the individual or society,
or both (Walker & Lidz, 1983). Substance addiction
involves the ingestion of a substance, including food,
nicotine, heroin and alcohol (Greenberg, Lewis &
Dodd, 1999). Addictions to an activity include
addictions to exercise, gambling, television and video
games and are often referred to as behavioural
addictions. Usually those addicted lose control over
use of the substance or activity with the inability to
regain control. This loss of control makes it
unacceptable, as one of the core societal values of
Western culture is to retain self-control.
Many abused substances and activities are legal and
often enjoyed in the company of others, such as
gambling and drinking alcohol (Walker & Lidz,
1983). They are acceptable in moderation, but become
intolerable when they are performed repeatedly.
Internet addiction is defined as problematic use of the
Internet that results in dysfunction in occupational,
academic, social and/or psychological aspects of life
(Beard & Wolf, 2001).
A number of models exist to explain addiction,
though not all are relevant to Internet addiction. For
example, the medical model is the most prevalent
model explaining addiction, though it does not
account for behavioural addictions such as Internet
addiction (Marlatt, Baer, Donovan & Kivlahan, 1988).
This model views addiction as a disease, and assumes
Abstract
The Internet offers many activities, such as chat rooms
and multi-user dungeons. The Internet may eventually
have as profound an impact on people's lives as the
telephone and television. This may be problematic for
high Internet users, who, compared to low users, have
reported higher levels of depression and less social
support. Low self-esteem has also been related to
excessive Internet use. Excessive use of the Internet
may be seen as an impulse control disorder akin to
pathological gambling. This paper discusses excessive
internet use as a behavioural addiction.
The Internet has rapidly changed from simply a
communication tool, similar to email, to include an
expanded range of information and activities
(Wallace, 1999). World use of the Internet went from
0.4% in 1995 to 9.5% in 2002. In 2001, 42% of
Australians reported having used a personal computer
at home in the week preceding the census (ABS,
2004). Almost seven million people in Australia had
used the Internet in that same week. Australian
Internet usage has grown from 6.7% in 1997 to 54%
in 2002 (NUA, 2004). This is an explosion in use,
considering the Internet is a relatively contemporary
technology.
The Internet started to take off in 1993, when user
friendly features were added to the world wide web
(WWW). These features, such as email, discussion
forums, chat rooms and multi-user dungeons, made it
more interesting to use (Jonscher, 2000). People from
all walks of life created home web pages and joined
discussion forums (Wallace, 1999). The social and
psychological impact of the Internet may be of
concern in the same way that other communication
technologies, such as television, radio, and
newspapers, have had a significant impact on people's
lives (Howard, Rainie & Jones, 2002). That the
Internet is rather abruptly becoming a routine part of
life makes its potential social and psychological
impact pervasive.
The Internet, if moderately used, can be beneficial
(Young and Rodgers, 1997). A distinction can be
made between healthy and unhealthy use of the
Internet. Healthy Internet use involves integrating
time spent online and time spent in the real world
(Suler, 1996). Individuals can bring their interests and
personalities with them online. They can use the
Internet to keep in touch with those with whom they
already have firm relationships. When face-to-face
with family and friends, they do not feel as if they
must hide how they have spent their time online. For
68
Excessive internet use as behavioural addiction
the person with the addiction is suffering from a
disease (Thombs, 1994).
The maladaptive behaviour model accounts for both
substance and behavioural addictions (Marlatt et al.,
1988). Addiction is viewed as the result of behaviour
learned within the context of personal and
environmental factors. It is based on social learning
theory, and explains the addictive substance or
activity as being reinforcing for the individual
(Donegan, Rodin, O'Brien & Solomon, 1983).
The neurobiological model offers an explanation
based on the activity of neurotransmitters
(Panksepp,1998). The dopamine hypothesis states that
addictive urges are initiated by the activity of the
neurotransmitter dopamine. Interaction on the Internet
may trigger the release of neurotransmitters that make
the experience pleasurable, acting as a conditioned
reinforcer.
A psychoanalytic model proposes that addiction is a
response to an inner emptiness, and that the addict's
control over the substance creates an illusion that this
emptiness is not there (Good, 1995). The Internet
addict may have a sense of control over the Internet
that masks this inner void.
Regardless of which model is used to explain
addiction, there exists several common properties of
both substance and behavioural addictions that
predispose them to become addictive and make them
difficult to give up (Donegan et al., 1983). Two such
components of addictions include tolerance and
withdrawal.
Tolerance occurs when the initial levels of a
substance that produced satisfaction no longer
produce the same satisfying effects. This results in a
greater perceived need, which consequently leads to
an obsession with seeking the high associated with the
addiction (Van Den Bergh, 1991). Compulsive
behaviour thus follows in order to obtain the
substance or engage in the activity. This compulsive
behaviour interferes with the ability to function in
various facets of life. Attempts to abstain result in
withdrawal effects.
Withdrawal is a consequence of physiological and
psychological dependence and involves an unpleasant
consequence of abstaining from the addiction (Elster
& Skog, 1999). Physiological dependence is apparent
when the substance is no longer available and the
body suffers physiological effects, such as irritability,
sweating and shaking (Donegan et al., 1983). In rare
cases, these somatic effects also take place in
behavioural addictions (Elster, 1999). Psychological
dependence is manifested when the substance or
activity is no longer available, resulting in withdrawal
effects of emotional distress and cravings.
These have been documented in cases of addictions.
In a study of substance and behavioural addictions,
which examined cravings, withdrawal, tolerance and
lack of control, the highest level of addiction was
found in exercise, caffeine, television and alcohol
(Greenberg et al., 1999). The study also found
evidence for Internet addiction using these
components. Brenner (1997) also reported evidence
that withdrawal, tolerance and cravings occur in
Internet addiction.
69
As with other behavioural addictions, Internet
addiction does not involve some of the symptoms of
substance abuse, such as physiological dependence
and withdrawal (Beard & Wolf, 2001). It may be
better described as an impulse control disorder.
Young (1996) modified the DSM-IV criteria for the
impulse control disorder of pathological gambling to
suit Internet addiction, yielding eight criteria: 1)
preoccupation, 2) increasing use, 3) inability to
control use, 4) irritability when not using, 5) staying
online longer than intended, 6) jeopardizing
relationships or career, 7) lying to conceal use, and 8)
use as an escape.
Internet addiction is assumed if people meet five of
the eight criteria. Young (1996) found that those
addicted increased their time spent on the Internet
substantially. This is similar to developing tolerance.
These addicted individuals would make drastic
attempts to cut back time spent online, such as
disconnecting their Internet services. This resulted in
becoming preoccupied with thoughts of being online.
This is comparable to cravings. In this way, Internet
addiction can be viewed as an impulse control
disorder.
With addiction, short-term gratification is generally
followed by long-term consequences (Lang, 1983).
These consequences involve difficulties in various
facets of life, including financial, occupational and
personal. This may occur as a result of the impulsive
nature of the individual. Impulsivity, a feature of
impulse control disorder, is a personality trait defined
as reacting hastily to impulses or urges without
stopping to think before acting (Lorr & Wunderlich,
1985). This trait has been implicated in a variety of
addictions.
For example, cocaine-dependent individuals have
greater levels of self-reported impulsivity than
matched controls (Coffey, Gudleski, Saladin, &
Brady, 2003). Those with high levels of impulsivity
are also less able to delay gratification in order to
receive a better reward. Also, those with pathological
gambling problems have greater impulsivity than
social gamblers, who have greater impulsivity than
non-gamblers (Nower, Derevensky & Gupta, 2004).
This implies that the greater the problem, the greater
the impulsivity. Similarly, those with co-morbid
addictions of gambling and substance abuse have
greater impulsivity than those with only a gambling or
only a substance abuse problem (Vitaro, Ferland,
Jacques & Ladouceur, 1998).
Sensation seeking is a feature of impulsivity, where
those who are sensation seekers tend to prefer novelty
and spontaneity (Zuckerman & Link, 1968). Lin and
Tsai (2002) found that sensation seeking is related to
Internet addiction. As explained by the maladaptive
behavioural model (Marlatt et al., 1988), sensation
seekers may find the Internet to be a stimulating,
adventurous activity, reinforcing to their impulsive
personality (Lin & Tsai, 2002).
People continue their addictive behaviour despite
the problems when the consequences outweigh the
benefits (Falk, Dews, & Schuster, 1983). Internet
addicts tend to spend so much time online that their
personal and professional lives suffer (King, 1996).
70
HENRY & WINKELMAN
They also spend less time engaging in what were
formerly pleasurable activities (Davis, 2001). As they
engage less in other activities and engage more in the
primarily solitary activity of the Internet, they may
become socially isolated.
Currently, the Internet is still growing, and may be
viewed as a continuation of the shift towards spending
more time with technology and less time with families
and friends (Ajayi, 1995, cited in Griffiths, 1997).
Social history has witnessed the introduction of radio
in the 1930s and television in the 1950s (Kraut,
Patterson, Lundmark, Kiesler, Mukopadhyay, &
Scherlis, 1998). As a means of privatising
entertainment, both have been held responsible,
especially television, for reduced social participation.
Even when radio or television are enjoyed in the
company of others, there is little interaction between
viewers.
Unlike television, the Internet includes two-way
interaction, allowing people to have their say to the
rest of the world and to receive a response
(Gackenbach & Ellerman, 1998). Rather than people
using the Internet as a means to further privatise their
entertainment, they may actually become more social
in their entertainment, using the Internet to break
geographical barriers and to communicate with the
rest of the world (Kraut et al., 1998).
The Internet is a way of connecting socially, yet,
paradoxically, at the same time isolates users
(DeAngelis, 2000). Those who spend more time on
the Internet reported lower quality relationships with
their families and friends than those spending less
time online (Sanders, Field, Diego & Kaplan, 2000).
Another study found similar results, with participants
reporting greater use of the Internet also reporting
greater levels of loneliness and less social support
than those who used it less (Kraut et al., 1998). This
occurred despite the fact that the Internet was used
primarily for social and communicative purposes.
Although relationships are often formed online, in
chat rooms and multi-user dungeons, these tend to be
weak rather than strong relationship bonds (Kraut et
al., 1998). Weak bonds are superficial and easily
broken, whereas strong bonds are long-lasting with
deep feelings of affection and obligation. What
separates relationships maintained through other
communication devices, such as telephone and fax,
from the Internet is that the Internet encourages the
development of relationships with complete strangers
(King, 1996).
People may be more inclined to form a relationship
with a stranger on the Internet than in real life due to
the anonymity that the Internet provides. This means
that Internet users do not have to worry about the
impression their appearance will make to the other
person (Wallace, 1999). It also allows Internet users
to form an idealistic image of the person with whom
they are forming a relationship. In this way,
connections are formed rather quickly (Young, 1997).
Even intimate details about their lives may be
disclosed upon chatting for the first time. These may
be things that they do not feel comfortable revealing
to people already in their lives, such as partners or
friends.
Preferring online to face-to-face interaction may
have deleterious consequences. The more one prefers
online interaction, the less psychosocial wellbeing
they possess (Caplan, 2002). One such psychosocial
factor, depression, is associated with greater levels of
Internet use (Kraut et al., 1998). Depression involves
a low mood, lack of motivation and the inability to
gain pleasure from one's usual activities (Bennett,
2003). Depressives often have a negative outlook
towards themselves and the future.
Self-esteem is a psychosocial factor that often
accompanies depression (Rosenberg, 1965). Selfesteem refers to how people evaluate themselves
(Baumeister, 1993). Those with low self-esteem tend
to lack positive self-views. Depression and selfesteem have been implicated as factors in other
addictions, including alcoholism (Corrigan, 1991) and
gambling (Lesieur & Blume, 1991).
In regards to Internet addiction, people with lower
self-esteem spend more time online and have higher
rates of addiction to the Internet (Armstrong, Phillips
& Saling, 2000). Internet addiction is also associated
with higher levels of depression (Young & Rodgers,
1997). Higher levels of depression have also been
identified in those who spend a greater amount of
time online (Kraut et al., 1998).
People with Internet addiction may have negative
cognitions that tend to involve either themselves or
their world (Davis, 2001). Those Internet addicts who
have a negative view of themselves can use the
Internet to achieve a more positive view through
affirmations from online relationships. Possessing a
negative view of the world may lead them to believe
that the Internet is a place, if not the only one, where
they can feel good about themselves.
The Internet may serve as an escape for people with
low self-esteem and depression (Armstrong et al.,
2000). They can create an online persona that is
different to that which they portray in the real world
(Young, 1997). People with low self-esteem tend to
be cautious in their self-presentation when interacting
with others (Baumeister, 1993). This may be because
they fear losing any of what little self-esteem they
possess. When people are online, they can create an
identity that represents an ideal rather than an actual
view of themselves (Young, 1997). This is what they
portray over the Internet, making them more confident
in their interactions and temporarily boosting their
self-esteem.
A possible reason individuals feel more confident
interacting on the Internet is that they can meet people
in a less intimidating manner than when meeting
people face-to-face (Young & Rodgers, 1997). They
also have more control over what they communicate
and can avoid the embarrassment of saying something
they may regret.
Individuals who are addicted to the Internet tend to
use excessively those features that allow
communication with other users (Young, 1996).
These include chat rooms, multi-user dungeons and
email. Those who are not addicted are more inclined
to use information protocols, newsgroups and the
WWW. This suggests that people may not be addicted
Excessive internet use as behavioural addiction
to the Internet as a whole, but only to these particular
features.
The attraction to these features may be for a variety
of reasons. In multi-user dungeons, individuals create
a character to interact with other online characters
(Turkle, 1995). These online personas contain aspects
of the self that may not be expressed in ordinary real
life. As it is easy to find groups with the same point of
view, individuals may enjoy interacting in
newsgroups and chat rooms (Wallace, 1999). This
may enable them to have in-depth discussions with
like-minded others, which is unavailable to them
elsewhere. This shared common interest may
reinforce these individuals, allowing them feel good
in themselves about having a point of view that others
agree with.
In conclusion, Internet addiction is a new
phenomenon and appears to have components in
common with other addictions, such as impulsivity,
low self-esteem, depression, and problems with social
support. Some features of Internet use are more
addictive than others. Internet addiction needs to be
recognised as a disorder, so that individuals can be
offered help, rather than viewed as wasting their time
on the Internet.
References
Armstrong, L., Phillips, J.G., & Saling, L.L. (2000).
Potential determinants of heavier internet
usage.
International Journal of Human-Computer Studies,
53, 537-550.
Australian Bureau of Statistics. (2004). 2001 census
data:
Computers
and
the
internet.
http://www.abs.gov.au/Ausstats.
Baumeister, R.F. (1993). Understanding the inner
nature of low self-esteem: Uncertain, fragile,
protective, and conflicted. In R.F. Baumeister (Ed.).
Self-esteem: The puzzle of low self-regard. New
York: Plenum Press.
Beard, K.W., & Wolf, E.M. (2001). Modification in
the proposed diagnostic criteria for internet
addiction. Cyberpsychology and Behavior, 4(3),
377-383.
Bennett, P. (2003). Abnormal and clinical
psychology: An introductory textbook. Maidenhead:
Open University Press.
Brenner, V. (1997). Psychology of computer use:
XLVII. Parameters of internet use, abuse and
addiction: The first 90 days of the internet usage
survey. Psychological Reports, 80, 879-882.
Caplan, S.E. (2002). Problematic internet use and
psychosocial well-being: Development of a theorybased
cognitive-behavioral
measurement
instrument. Computers in Human Behavior, 18,
553-575.
Coffey, S.F., Gudleski, G.D., Saladin, M.E., & Brady,
K.T. (2003). Impulsivity and rapid discounting of
delayed hypothetical rewards in cocaine-dependent
individuals.
Experimental
and
Clinical
Psychopharmacology, 11(1), 18-25.
Corrigan, E.M. (1991). Psychosocial factors in
woman’s alcoholism. In N. Van Den Bergh (Ed.).
Feminist perspectives on addictions. USA: Springer
Publishing Company.
71
Davis, R.A. (2001). A cognitive-behavioral model of
pathological internet use. Computers in Human
Behaviour, 17, 187-195.
DeAngelis, T. (2000). Is internet addiction real?
Monitor on Psychology, 31(4) [online version]
Available
at
www.apa.org/monitor/apr00/addiction.html.
Donegan, N.H., Rodin, J., O’Brien, C.P., & Solomon,
R.L. (1983). A learning-theory approach to
commonalities. In P.K. Levison, D.R. Gerstein &
D.R. Maloff (Eds.). Commonalities in substance
abuse and habitual behavior. Canada: Lexington
Books.
Elster, J. (1999). Gambling and addiction. In J. Elster
and O. Skog (Eds.). Getting hooked: Rationality
and addiction. UK: Cambridge University Press.
Elster, J. & Skog, O. (1999). Introduction. In J. Elster
and O. Skog (Eds.). Getting hooked: Rationality
and addiction. UK: Cambridge University Press.
Falk, J.L., Dews, P.B., & Schuster, C.R. (1983).
Commonalities in the environmental control of
behavior. In P.K. Levison, D.R. Gerstein & D.R.
Maloff (Eds.). Commonalities in substance abuse
and habitual behavior. Canada: Lexington Books.
Gackenbach, J., & Ellerman, E. (1998). Introduction
to psychological aspects of internet use. In J.
Gackenbach (Ed.). Psychology and the internet. San
Diego: Academic Press.
Good, L. (1995). Addiction as a narcissistic defence:
The importance of control over the object. In J.
Cooper and N. Maxwell (Eds.). Narcissistic
wounds: Clinical perspectives. London: Whurr
Publishers.
Greenberg, J.L., Lewis, S.E., & Dodd, D.K. (1999).
Overlapping addictions and self-esteem among
college men and women. Addictive Behaviors,
24(4), 565-571.
Griffiths, M. (1997). Psychology of computer use:
XLIII. Some comments on ‘addictiveuse of the
internet’ by Young. Psychological Reports, 80, 8182.
Howard, P.E.N., Rainie, L. & Jones, S. (2002). Days
and nights on the internet. In B. Wellman and C.
Haythornthwaite (Eds.). The internet in everyday
life. USA: Blackwell Publishing.
Jonscher, C. (2000). Wired life: Who are we in the
digital age? London: Anchor.
King, S. A. (1996). Is the internet addictive, or are
addicts using the internet? Retrieved [15/07/04]
from
the
World
Wide
Web:
http://www.concentric.net/~Astorm/iad.html.
Kraut, R., Patterson, M., Lundmark, V., Kiesler, S.,
Mukopadhyay, T., & Scherlis, W. (1998). Internet
paradox: A social technology that reduces social
involvement
and
psychological
wellbeing?
American Psychologist, 53(9), 1017-1031.
Lang, A.R. (1983). Addictive personality: A viable
construct? In P.K. Levison, D.R. Gerstein & D.R.
Maloff (Eds.). Commonalities in substance abuse
and habitual behavior. Canada: Lexington Books.
Lesieur, H.R., & Blume, S.B. (1991). When lady luck
loses: Women and compulsive gambling. In N. Van
Den Bergh (Ed.). Feminist perspectives on
addictions. USA: Springer Publishing Company.
72
HENRY & WINKELMAN
Lin, S.S.J., & Tsai, C. (2002). Sensation seeking and
internet dependence of Taiwanese high school
adolescents. Computers in Human Behavior, 18(4),
411-426.
Lorr, M., & Wunderlich, R.A. (1985). A measure of
impulsiveness and its relations to extraversion.
Educational and Psychological Measurement, 45,
251-256.
Marlatt, G.A., Baer, J.S., Donovan, D.M., &
Kivlahan, D.R. (1988). Addictive behaviours:
Etiology and treatment. Annual Review of
Psychology, 39, 223-252.
Nower, L., Derevensky, J.L., & Gupta, R. (2004). The
relationship of impulsivity, sensation seeking,
coping and substance use in youth gamblers.
Psychology of Addictive Behaviors, 18(1), 49-55.
NUA Internet Surveys. (2004). Nua Internet how
many online. New York: Nua Ltd. Available:
http://www.nua.ie/surveys/how_many_online/index
.html.
Panksepp, J. (1998). Affective neuroscience: The
foundations of human and animal emotions. New
York: Oxford University Press.
Rosenberg, M. (1965). Society and the adolescent
self-image. Princeton, NJ: Princeton University
Press.
Sanders, C.E., Field, T.M., Diego, M., & Kaplan, M.
(2000). The relationship of internet use to
depression and social isolation among adolescents.
Adolescence, 35(138), 237-242.
Suler, J. (1996). Computer and cyberspace addiction.
[online
version]
Available
at
www.rider.edu/suler/psycyber/psycyber.html.
Thombs, D.L. (1994). Introduction to addictive
behaviors. New York: The Guilford Press.
Turkle, S. (1995). Life on the screen: Identity in the
age of the internet. New York: Simon and Schuster.
Van Den Bergh, N. (1991). Having bitten the apple:
Feminist perspective on addiction. In N. Van Den
Bergh (Ed.). Feminist perspectives on addictions.
USA: Springer Publishing Company.
Vitaro, F., Ferland, F., Jacques, C., & Ladouceur, R.
(1998). Gambling, substance use, and impulsivity
during adolescence. Psychology of addictive
behaviors, 12(3), 185-194.
Wallace, P. (1999). Psychology and the internet.
USA: Cambridge University Press.
Walker, A.L., & Lidz, C.W. (1983). Common features
of troublesome habitual behaviours: A cultural
approach. In P.K. Levison, D.R. Gerstein & D.R.
Maloff (Eds.). Commonalities in substance abuse
and habitual behavior. Canada: Lexington Books.
Young, K.S. (1996). Internet addiction: The
emergence of a new clinical disorder.
Cyberpsychology and Behavior, 1(3), 237-244.
Young, K.S. (1997). What makes the internet
addictive: Potential explanations for pathological
internet use. [online version]. Available at
www.netaddiction.com/articles/habitforming.htm.
Young, K.S., & Rodgers, R. (1997). The relationship
between depression and internet addiction.
Cyberpsychology and Behavior,1(1), 25-28.
Zuckerman, M., & Link, K. (1968). Construct validity
for the sensation-seeking scale. Journal of
Consulting and Clinical Psychology, 32(4), 420426.
Correspondence to:
Cecelia Winkelman
Australian Catholic University
Locked Bag 4115, Fitzroy, Vic 3065
c.winkelman@patrick.acu.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 73-78
Bullying Relationships in High Schools
Dorothy Lenthall (dotl@iprimus.com.au)
Seton Catholic College, Western Australia 6156
situation worse (Slee, 2002). The study explored the
reasons for students’ resistance to intervening in
bullying incidents, even though they claimed they
would.
Abstract
In response to evidence that bullying in schools persists
in the presence of bystanders, this study sought to add
to the existing knowledge about its reinforcing effects.
The objectives of this research project were to
investigate non-intervention in bullying incidents by
students. This was a multi-dimensional investigation of
the emotional, cognitive and behavioural factors from
the bystander’s perspective, within the context of a coeducational Catholic high school, in a sample of
eighteen Year 8 students. In-depth and group
interviews, participant observation, case studies and the
input of a focus group of teachers formed the data
collection. Previous findings that fear prevents
bystanders from taking action, were extended by this
study which revealed was that there are several sources
of this fear. The study found reasons for students’ fear
included embarrassment at making a mistake, the
importance of the teachers’ responses, the need to
assimilate into the new school culture, to be “cool” and
to avoid a negative, conforming self-image.
Importantly, the existing focus on fear does not explain
why students do not anonymously report bullying. This
study found that students resisted taking responsibility
for intervening, and unexpected findings included that
students categorised victims, only caring enough to
report bullying if the victim were a friend or sibling.
The thrill of watching bullying was also a strong
deterrent to bystander intervention. The study suggests,
therefore, information based anti-bullying policies will
be ineffective unless students are motivated to
intervene. It is crucial that programmes now address
the emotional deterrents of fear, excitement and apathy
before considering educational approaches, and that
future policies need to examine the culture of the
school, including teachers’ responses to bullying,
which dictates the behavioural code for incoming Year
8 students.
The Need to Address Bullying
The people most obviously suffering from bullying
are the victims, who experience learning difficulties
(Smith and Sharp, 1994) post-traumatic stress
(Mynard et al, 2000), low self-esteem, anxiety,
depression (Olweus, 1992, Gilligan et al, 1990) and
even suicide (Marr and Field, 2001). But it is not just
the victims who suffer. Bullies are more likely than
non-bullies to suffer depression, guilt, anger and
shame (Dietz, 1994). These negative effects are longlasting. One in four male bullies will have a criminal
record by the age of thirty and is likely to have
convictions for violent crime, be abusive to their
wives and children and have children who bully.
(Eron et al, 1987). School is an efficient training
ground for adult bullying behaviour. Children will
take their experiences with them into adulthood and
perpetuate them.
Until recently, the focus in schools has been on
victims and bullies, while bystanders have received
little attention. It is now known that bullying
adversely affects them too (Lampert, 1998).
Bystanders at school are also likely to suffer longterm psychological and physiological distress (Janson,
2000).
Clearly, the culture of the school is established by
its interpersonal relationships. The commitment of the
school community is needed to address aggressive
and bullying cultures. Close examination of the school
climate and the employment of effective strategies
can create a culture of peace, respect and dignity.
Method
Research on bullying to date has mainly been
quantitative, providing extensive evidence of the
degree of bullying in schools. This qualitative study
sought to answer some questions about why bullying
is so prevalent.
Year 8 participants were chosen for this
investigation because bullying increases during early
adolescence (Pellegrini et al, 1999). In order to
investigate the bystanders’ point of view, the study
explored the various contributors to non-intervention
by bystanders, or ‘bystander apathy’, as Latane and
Darley (1970) named it. It is well known that bullying
behaviours are reinforced by the presence of
bystanders (Salmivalli et al, 1996; Simmons, 2002)
and Year 8 students are quite well educated in
primary school about this aspect of bullying.
Bystanders and victims do not report bullying because
they fear being bullied themselves (Olweus, 1993;
Rigby, 1996) and students do not ask for help from
teachers because they fear that doing so will make the
Participants
Participants were eighteen Year 8 students, eleven
girls and seven boys, who volunteered to take part in
the study. Their ages ranged from twelve to thirteen
years. Recruitment was from within the coeducational Catholic high school where the study was
conducted. Participants included a volunteer focus
group of two male and two female teachers at the
same school. Two were full time teachers, one was in
middle management and one in upper management.
Procedure
Definitions of bullying were explained to the
participants and possible actions to counter bullying
were clarified. Intervention was explained as being
any action to counter bullying, such as reporting the
bullying, either verbally or anonymously or taking
more direct action, such as telling the bully to stop.
73
LENTHALL
74
Two group interviews were conducted with the
students, to allow interaction within the group and to
facilitate a stimulating environment (Fontana & Frey,
2000). Individual, in-depth interviews were conducted
with the student participants. The interviews were
unstructured, with the aim of allowing students to
speak for themselves, with as little input from the
interviewer as possible (Stringer, 1996). The software
package QSR N6 was used for coding data from
interviews to discern strategic or consequential
relationships.
The researcher’s role was also diminished during
discussions with the focus group of teachers, allowing
a different, adult perspective to be introduced
(Madriz, 2000).
Being the school psychologist facilitated the
collection of data from participant observation
(Denzin & Lincoln, 2000). Observations of Year 8
students on a daily basis corroborated and
strengthened the trustworthiness of data from the
interviews. Endeavouring to be unobtrusive, field
notes and reflective notes were written immediately
after events, rather than at the time they took place.
Bullying behaviours and particularly, the collective
behaviour of bystanders were noted.
To maximize understanding of bystanders’
behaviour, the sampling of case studies was
purposeful (Patton, 1990; Yin, 2003). Cases were
selected that would produce the most in-depth
information.
Results
Students felt that adults in the school community
should counter bullying and were resentful of the
assumption that it was their responsibility. They did
not intervene in bullying incidents, despite claiming
that they would. They responded to the prevailing
school culture of inaction by students and teachers
and ignored primary school education, which taught
them to take action against bullying. The reason they
said they would intervene was to agree with the
perceived social justice ethos.
It was found that the fear of being bullied
themselves was only one of the fears preventing
students from taking action against bullying. It was
reasoned that if this was the only fear, then why did
students not anonymously report bullying?
There was a marked lack of empathy for victims,
who were categorized as being worthy or not of
support according to the relationship they had with the
bystanders. If they were a friend or a sibling,
bystanders were more likely to help. If they were not
known to the bystander, especially if they were
deemed to “deserve” the bullying, then bystanders
would not intervene.
Students
feared
being
embarrassed
at
misunderstanding aggressive behaviours. Bullies
capitalized on these fears, maintaining that their
aggression was “just mucking around” and that their
victims were really their friends.
Students strived for a “cool” image, and wished to
be seen as “tough”, rejecting a self-image of
conforming to school rules. Not reporting bullying
was a way of asserting that they were part of the
popular group.
The thrill of watching violence created a dependent
relationship between bystanders and bullies and
enforced the code of silence that supports bullying
behaviour. Victims also supported their antagonists in
an effort to gain favour with them.
Bystander-Victim Relationship
Bystander intervention relied heavily on the
relationship between the bystander and the victim.
Participants were quite clear about the strength of the
relationship being the only acceptable criterion in the
school culture for confronting bullies. Protecting a
sibling was considered an automatic response. They
thought it honourable to help a best friend and an
obligation to help a sibling.
When the victim was only an acquaintance, there
was a noticeable shift in attitude towards helping.
Participants did not intervene because they feared
being told to mind their own business. “Sticking
your nose in” was viewed as a cardinal offence –
evidence that one does not understand the rules of the
culture. The school’s cultural expectation that
bystanders mind their own business, or else risk
ridicule and rejection from the dominant group, acted
as a controlling factor for those students who would
have liked to stop the bullying. The relationship
bystanders had with victims was weighed carefully
against their own place in the group. Placing victims
on a “relationship scale” thus made bullying easier,
with the bully enjoying the protection afforded by a
culture that accepts aggression as normal.
The Code of Silence
One of the main difficulties for schools dealing with
bullying is the code of silence that students maintain
to protect bullies. The refusal of all players in bullying
incidents to report them is a theme which repeats
itself throughout this paper.
One of the bystanders’ biggest fears was being
embarrassed at making a mistake by misinterpreting
an aggressive incident. Two situations were cited as
being open to misinterpretation. The first was when
two people were friends and indulging in physical or
verbal sparring – “just mucking around”. The other
situation was when a victim “deserved” to be subject
of aggression.
A favourite excuse of bullies for their aggression,
silently verified by bystanders, was to say that their
actions were a form of friendly play. This explanation
effectively blocked any action, even by teachers,
against bullying
Teachers adopted a “What can you do?” attitude to
aggressive behaviour. In one instance, a teacher on
duty stood and watched as three boys pushed,
punched and kicked another boy. Later, she said she
did not approach them because “Boys are always
doing that sort of thing. I thought they were just
friends mucking around.” In fact, the victim was
terrified of his antagonists but, because no-one, not
even the teacher, was stopping them, he thought their
aggression must be normal behaviour.
Bullying relationships in high schools
The Bystander-Bully Relationship
The Excitement Element As previously stated,
bystanders provide protection for bullies by not
reporting their actions. The thrill of watching
aggressive behaviour is one of the more subtle and
complicated
relationship
factors
involved.
Participants likened it to watching wrestling and
boxing in the wider community. By gathering around
and attracting a crowd, bystanders can encourage
violence and accelerate it so that minor incidents
become serious. One interviewee, Aaron, said this:
People know what’s right in their head, but it’s like
they’ve forgotten when it comes to one of those
situations. They want to see the result.
When the bully is surrounded by an “audience” of
enthusiastic onlookers, they experience the pressure
of the audience’s demand for a performance. The
lines between victim, bully and bystander thus
become less defined. Some participants who admitted
to bullying, said that they were subjected to extreme
pressure by classmates and peers to continue bullying.
One girl said that, feeling remorseful for hurting
others, she had resolved to stop bullying. Her
determination to mend her ways was short-lived,
however, when her classmates ostracized her at
lunchtime, ridiculing her for becoming a “goody-twoshoes”. They had been deprived of the entertainment
they had come to expect, and they were disappointed.
Under threat of being left out of the dominant group,
she resumed bullying and was welcomed back. She
said she felt better, even though she was in constant
trouble for her behaviour, but felt ashamed of the hurt
she was causing.
The Power of Bullies Participants were aware that
countering bullying was difficult because of the
power they gave to the bullies. The bullies were also
aware of this, and responded to the positive
reinforcement by bullying more. Victims usually face
not only the bully, but also the crowd of supporters,
who join in, laughing, cheering and adding to the
insults. Even when bystanders do nothing, their
presence gives implied approval and confirmation of
the behaviour. The bully is safe in the knowledge that
none of the bystanders will report the bullying and
their power renders victims helpless. The whole group
galvanizes towards him or her, providing almost
impregnable protection from authority.
Being tough was repeatedly cited as a contributor
to the bullying culture. Students felt they had to look
tough, so they bullied others and victims did not ask
for help for the same reason. Bystanders wanted
vicarious toughness, so they supported the bully,
completing the cycle of reinforcement.
The Bystander-Victim Relationship
Participants realized that some victims were
provocative. They thought it would be a mistake to
intervene if the victim “deserved” to be bullied. They
would not act if they did not know the background to
the incident.
They considered some bullying to be justified. If the
victim was not well liked, they could rely on no
support from bystanders. Satisfaction was expressed
75
by some participants at this situation, with one boy
saying, “If you don’t like them, you might want them
beaten up”. A girl said, “If the victim deserves it for
being a pest…You’re punching him for me as well,
because I feel like doing it.”
The behaviour of provocative victims is familiar in
all schools, and more research is needed on the
behaviour of such victims.
The Victim-Bully Relationship
There was an inexplicable relationship between
some victims and the people who bullied them. Group
affiliation is so strong that it is very difficult to get
victims to report being bullied or leave the dominant
group (Mynard et al, 2000), even when there are other
groups willing to befriend them. They maintained a
fierce loyalty to their antagonists. The pain of being
apart from their preferred group was even greater than
the often considerable pain inflicted on them by that
same group. The victims protected the bully, denying
to teachers that they were being bullied, claiming to
be “just mucking around”, often adding, “We’re
friends”.
Participants in the study recognized this as a way to
gain positive reward from the bully and his/her
supporters. The self-esteem of these victims was
extremely poor and they simply grasped any chance
they could to curry favour with the leaders.
One Year 11 boy at the school was called obscene
names, was ridiculed, locked in the toilets and had
food smeared on him by his peers. They instructed the
Year 8 boys to bully him every day at lunch time. A
large crowd of Year 8 boys laughing and jeering
outside the toilets, where they had barricaded him in
with a table, drew the duty teacher’s attention. The
boy’s peers were outraged when reprimanded by the
Principal for bullying, denying any involvement.
Their denials were upheld by the victim, who
concurred with their story and said that the bullying
was just friendly banter. He was emphatic that he did
not want his “friends” to get into trouble on his
behalf.
On another occasion, a girl was repeatedly rejected
by her group and was never sure if she was in or out
of the friendship circle. When another group of girls
invited her to join them, her original group threatened
her with permanent rejection and so she stayed with
them. She wanted to be part of the dominant group,
even at great cost to her wellbeing.
Why Asking for Help is so Difficult Even when
victims want bullying to stop, they sometimes endure
an enormous amount of misery before asking for help,
if they ever do (Marr & Field, 2001). Although fear of
being bullied even more is part of the reason, Besag,
(1989) found that they felt ashamed for being
unpopular. Their relationship with the bullies was so
degrading that they often believed what the bullies
said to them, sending their self-esteem plummeting.
When bystanders did not intervene on their behalf,
victims’ suffering intensified. They felt betrayed and
concluded that bystanders supported the bullies.
The new Year 8 students felt that now they were in
high school, they should deal with their own
76
LENTHALL
problems. In the masculine culture of this school, it
was especially important to maintain this tough
image. Victims therefore did not show how upset they
were, or tell teachers they were being bullied because
they wanted to appear tough, even to teachers. They
expected to be on their own and not to expect
assistance from staff. This expectation arose from
their self-image as being capable of taking care of
themselves and being independent.
Staff Responses as a Deterrent to Asking for Help
In view of the central importance of a caring ethos in
the school, it was surprising that the students did not
rely on it to protect themselves from this most feared
aspect of school life. In this school’s culture,
however, both staff and students normalized bullying
as if it were innate behaviour. Some of the staff
members held particularly disturbing views, giving
their opinion that countering bullying was futile
because it had always been part of everyday school
life and no-one could change it. Many favoured the
toughening up of victims – an eye for an eye
approach. They were strongly resistant to the
introduction of anti-bullying strategies, even though
they had attended professional development sessions
on ways to deal with bullying. Proven strategies,
recommended by world experts, were presented and
still a minority of staff rejected any challenge to their
preferred way of dealing with bullying.
The Year 8 students were new to high school, and
in their heightened state of awareness and tension,
were particularly sensitive to teachers’ actions. If a
teacher’s response was not completely supportive –
even raised eyebrows, a frown, a question about the
bullying, or a slight pause before answering – they
assumed that they had done something wrong to
report bullying. Inadvertently, teachers were
indicating that bullying was part of the culture and
that these students should not be bothering them by
telling them about it. They felt these messages as
being criticism of them for not fitting into their new
school culture, and resolved never to reported
bullying again.
Participants said they felt angry when teachers did
nothing to stop bullying. They cited examples of
students being called obscene names in class, being
punched, having their books thrown on the floor and
their property stolen, and claimed that teachers simply
told the bullies to stop it. The excuse of “just mucking
around” was too readily accepted by teachers, they
said. They concluded that staff did not care about it,
did not know what to do, or had simply given up
trying to deal with it.
Teachers’ incompetence in dealing with bullying
was a source of exasperation. If they dealt with the
situation inadequately, the victim suffered renewed
and more forceful attacks. As Slee, (2002) noted,
“Girls are very clear about the fact that adults make it
worse”.
Relationships with Teachers
The student-teacher relationship is important in
determining the students’ overall view of bullying
behaviours. Teachers provided poor role models when
they were confused and unsure of what action to take
against bullying.
Students defined themselves against what they
viewed as authoritarian teachers. Students were
disdainful of the cultural ideals of this Catholic
school, believing that they were not being upheld,
particularly by teachers.
The focus group of teachers thought that power
structures within the school were masculine, with an
emphasis on male sport, management positions being
male-dominated and the entire office staff being
female. They criticized the school culture as being
oppressive, punitive, hierarchical and regimented.
Power was viewed as being distributed in masculine
ways, like shouting, modeling aggressive behaviour
and supporting a way of being negative, preventing
mutual respect between teachers and students.
That bullying is a behaviour rejected by Catholic
ideals is largely detached from and has little practical
impact on students’ perspectives. The term “antibullying” is not linked in any convincing manner with
the school’s Catholic ideals or anti-bullying policy.
The institutional culture of bullying appeared when
teachers bullied students, raising their voices, using
aggressive gestures and punitive measures for
behaviour management. This caused resentment
among the students, who then viewed the school’s
efforts to address bullying as hypocritical. They
reported feeling embarrassed and angry when teachers
bullied them. One girl described some of her teachers’
behaviour:
Jaimie: They yell at you, then, when you try to
explain something, like, why you didn’t do your
homework, or why you’re late, or something, they,
like, yell more, or just go (raised her eyes and
groaned).
DL: And what does that mean?
Jaimie: It means that you’re not worth listening to.
You’re nothing.
There is likely to be little connection or
communication in such situations, which, in turn,
makes it difficult for students to ask such teachers for
help if they are being bullied.
Discussion
The results of this study revealed that Year 8 students
arrive at high school in a certain state of tension,
which they try to relieve by adapting to the prevailing
culture as quickly as possible. Therefore, to
effectively counter bullying, the school culture that
needs to be changed to a non-bullying one. Critical to
an anti-bullying programme’s success is a wholeschool approach to countering bullying where there is
a change in the school culture to one where all
members of the community are treated with respect
(Rigby, 1996).
Purely educational anti-bullying programmes will
not be effective unless the students’ motivation to
counter bullying is first addressed. It is essential that
students take responsibility for countering bullying, as
the impact of adult intervention on the incidence of
bullying is practically negligible (Rigby & Johnson,
2004).
Bullying relationships in high schools
The results of this study show that unless the
students take responsibility for countering bullying,
there will be no action taken by them to do so. Raising
the expectations of student behaviour in bullying
situations is therefore fundamental to creating a nonbullying culture, instilling the notion that every
individual in the school community has the potential
to prevent bullying. Programmes are needed that teach
appropriate methods of intervention, particularly in
situations such as violence, where there is enjoyment.
One barrier to bystander intervention the study
found was the lack of empathy for the victim. This
can be increased with the employment of strategies
such as Salmivalli et al’s (1996) role-play. This
highlights the role of the bystander, allowing
participants to experience the plight of the victim.
This is a clear illustration of how bystanders usually
give power to the bully, when they could give it to the
victim. People who play the part of the bully report
feeling abandoned and powerless when the bystanders
defend the victim, which is a realistic reflection of
what happens among students.
As part of the whole-school approach, the
administration has to support staff to be proactive,
rather than reactive to bullying situations. The staff
needs to be informed of study’s findings that the
students view them as incompetent, unreliable and
unsure about how to deal with bullying. Teachers
underestimated the importance of their responses to
bullying. A professional development programme to
raise their awareness of the meanings students attach
to their behaviour is urgently required. The teachers in
this school were shocked at the students’
interpretation of their reactions to bullying. The No
Blame Approach (Maines and Robinson, 1992), for
example, should be adopted by all staff members. It is
a most effective method of dealing with bullying
(Rigby, 1996; Cross, 2003).
The study found that bullying is protected by a code
of silence. Without such protection, bullying would be
exposed and much easier to deal with. Some of the
participants thought that making bullying public
would reduce it.
Enlisting the aid of peers is the most effective
strategy within a school because students respond
more readily to the opinions of their peers than they
do to adults’ (Olweus, 1992, Rigby & Johnson, 2004,
Slee, 1997). A culture where it is the expectation that
all students take responsibility for countering bullying
is essential, so that a future of schools with a culture
where students tap each other on the shoulder and say,
“We don’t bully in this school” would not be beyond
the realms of possibility.
The study’s findings clearly show that the Year 8
students find themselves in a predicament because the
behaviour of other students and teachers repudiates
the anti-bullying education they have received. The
students’ confusion is much worse than was realized.
They feel under constant pressure to adapt to the
school culture and daily school life for them is a
minefield of uncertainty. It is not surprising that in
this state of anxiety, they conform.
In conclusion, the school leadership needs to
address the confusion of staff in dealing with cullying.
77
They, in turn, can then assist students, as a body, to
reclaim their independence so that it will become the
victims, not the bullies who receive their protection.
With the leadership promoting the new culture,
bullying can be confronted with students quoting from
the school Anti-Bullying Policy, “We don’t tolerate
bullying in this school”.
References
Besag, V. (1989), Bullies and Victims in Schools,
Open University Press, Bristol
Besag, V. (2002), Keynote Speaker, School
Psychologists Association of Western Australia
(Inc.) Conference, Perth, Western Australia, July,
2002
Cross, D. (2003), ‘Bullying and Harassment in
Schools’, paper presented by Associate Professor
Cross at Bullying in Schools and Communities
Seminar, April 2003, Perth, W.A.
Denzin, N.K. & Lincoln, Y. (Eds.) (2000), Handbook
of Qualitative Research, (2nd Edn.) Sage, Thousand
Oaks, California
Dietz, B (1994) ‘Effects on subsequent heterosexual
shyness and depression of peer victimization at
school’, Paper presented at the International
Conference on Children’s Peer Relations, Institute
of Social Research, University of South Australia,
Adelaide.
Eron L.D., Huesmann, L.R., Dubow, E., Romanoff,
R. and Yarmel, P.W. (1987), ‘Aggression and its
correlates over 22 years’, in Crowell, D., Evans, I.
& O’Donnell, C. (Eds) (1987), Childhood
Aggression and Violence, Plenum Press, New York
Fontana, A. & Frey, J.H. (2000), ‘The Interview:
From Structured Questions to Negotiated Text’, in
Denzin, N.K. & Lincoln, Y. (Eds.) (2000),
Handbook of Qualitative Research, (2nd Edn.) Sage,
Thousand Oaks, California
Gilligan, C., Lyons, N.P. & Hanmer, T. (1990),
Making Connections: The Relational Worlds of
Adolescent Girls at Emma Willard School, Harvard
University Press, Cambridge, Massachusetts
Janson, G.R. (2000), ‘Exposure to repetitive abuse:
Psychological distress and physiological reactivity
in bystanders as compared to victims’, Dissertation
Abstracts International Section A: Humanities &
Social Sciences, 61, 6-A, P2196 US: University
Microfilms International
Lampert, J. ‘Voices and Visions: Adolescent girls’
experiences as bullies, targets, and bystanders’,
Dissertation Abstracts International Section A:
Human & Social Sciences: 58, 8-A, p2986 U S:
University Microfilms International
Latane, B. & Darley, J.M. (1970), The unresponsive
bystander: Why doesn’t he help? AppletonCentury-Crofts, New York
Madriz, E. (2000), ‘Focus Groups in Feminist
Research’, in Denzin, N.K. & Lincoln, Y. (Eds.)
(2000), Handbook of Qualitative Research, 2nd
Edn.), Sage, Thousand Oaks, California
Maines, B. & Robinson, G. (1992), The No Blame
Approach, Lame Duck Publishing, Bristol
Marr, N. & Field, T. (2001), Bullycide: Death at
Playtime, Wessex Press, Oxford
78
LENTHALL
Mynard, H., Joseph, S. & Alexander, J. (2000), ‘Peervictimisation and posttraumatic stress in
adolescents’,
Personality
and
Individual
Differences, 29, 815-821
Olweus, D. (1992), ‘Victimisation by peers:
antecedents and long term outcomes’, in Rubin,
K.H. & Asendorf, J.B. (Eds) (1992), Social
withdrawal, inhibition and shyness in children,
Erlbaum, Hillsdale, New Jersey.
Olweus, D. (1993), Bullying at School: What we know
and what we can do, Blackwell, Oxford
Patton, M.Q. (1990), Qualitative Evaluation and
Methods, (2nd Edn), Sage, London
Rigby, K. (1996), Bullying in schools and what to do
about it, The Australian Council for Educational
Research Ltd., Melbourne
Rigby, K.& Johnson, B. (2004), Bystander
Intervention: Its role in countering school bullying,
(in press)
Salmivalli, C., Lagerspetz, K., Bjorkvist, K.,
Osterman, K., Kaukiainen, A. (1996), ‘Bullying as a
group process: participant roles and their relations
to social status within the group’, Aggressive
Behaviour, 22,1,1-15
Simmons, R. (2002), Odd Girl Out: The Hidden
Aggression in Girls, Harcourt, California
Slee, P. (1997) The P.E.A.C.E. Pack. A Programme
For Reducing Bullying In Our Schools (2nd Edn),
School of Education, Flinders University, S.A.
Slee, P.T. (2002), ‘Bully Tactics’, The Sunday Age,
June 9, 2002
Stringer, E.T. (1996) Action Research: A Handbook
for Practitioners, Sage, Thousand Oaks, California
Yin, R. (2003), Applications of Case Study Research
(2nd Edn), Sage, Thousand Oaks, California
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 79-84
The healthy balance among work, family, and personal relationships: Fact
or fiction?
Kathleen A. Moore (kmoore@deakin.edu.au)
School of Psychology
Deakin University, Burwood Vic 3125 Australia
work demands on employees’ relationships, in
particular work-family conflict, their health and
intention to quit using a range of mediating factors,
such as control, social and organizational support. The
focus will be on outcomes and a discussion rather
than on a review of the pre-existing literature. This
detail can be found in each of the published studies.
Working weeks in excess of 50 hours are becoming
endemic among workers typically employed for a
37½ hour week (Australian Bureau of Statistics,
2001) leaving many of us on a treadmill which seems
to be going faster and faster. These long hours have
been shown to increase work–family conflict and to
negatively affect health (Cooper & Cartwright, 1994;
Maruyama & Morimoto, 1996; Sparks, Cooper, Fried
& Shirom, 1997). Accordingly, the aim of the first
paper to be discussed (see Tinker & Moore, 2003)
was to investigate the interrelationships among long
working hours, work–family conflict and health, and
to test employees’ perceived control over working
hours and organisational support as mediators of the
impact of long working hours on work–family
conflict and health. We analysed useable data from a
sample of 91 teachers and academics who had been in
their profession for an average of 12 years.
The factors in the structural model (Fig. 1)
explained a substantial amount of variance (69%) in
health which was represented more strongly by
physical (β=.90) (e.g., not been feeling well and in
good health) than psychological (β=.67) health (e.g.,
been feeling nervous and strung up). Significant levels
of variance in control over the number of hours
worked (12%) and work-family conflict (20%) were
explained by the relevant paths in the models. It is
interesting to note that work-family conflict was
reduced in circumstances where participants
perceived the organisation to be supportive (β = -.22)
(e.g., The organization strongly considers my goals
and values) which can be interpreted to suggest that
this relationship is important not only in the
workplace itself but the effect transfers to family life
as well.
Abstract
In this paper, I present a series of studies which look at
the role of work demands on family relationships, in
particular work-family conflict, and their several and
joint impact on health. The role of neuroticism on these
factors will also be considered. In addition, a series of
studies which investigate the impact of perceived
control as a mediator of work stress as well as the role
that organizational support, another form of
interpersonal relationship, has on work-family conflict
and health outcomes is presented, as are studies which
focus on people’s intention to quit their job because of
the workplace stressors experienced. Common
measures of work-family conflict and health have been
used throughout these studies thus making the results
directly comparable.
The industrial revolution resulted in a major shift in
how most of us human beings earn our living. Many
folk moved from rural to urban living and from
outdoor manual work to indoor mechanized activities.
Much literature, and indeed literature it was (e.g,
Trollope, 1873:1967), focused on the plight of people
working in less than desirable conditions where heat;
cold; crowding, poor light, sanitation and long hours,
in addition to work demands for poor pay and
conditions, were the norm.
Industrialization did not stop in the 18th century or,
it seems, the demands and stressors within the
workplace. It has in many ways been superceded by
the age of computerization and internationalism,
where the former feeds into the latter and time of day
world-wide has, to some extent, become obsolete as
someone, somewhere is always contactable or
demanding via the internet. What was and dare I say,
should be, a tool, has become one of our masters.
Other masters include our own sense of
acquisitiveness, our desire for more of life’s
opportunities and goodies which, for those of us in the
western world are of a higher standard than ever
before, and the competition to succeed or exceed
within our chosen career paths. But what costs have
these changes and ambitions brought to our lives, to
our relationships with others? In this series of papers,
I would like to present the results of an ongoing
research program which has as its focus the role of
79
MOORE
80
6%
Organisational
Support
Physical
Psychological
-.26
-.22
-.25
.90
.67
.24
69%
20%
Long Work Hours
(Demand)
.21
Work-Family
Conflict
Health
.61
-.19
-.19
-.24
Control Over
Work Hours
12%
Figure 1
Impact of long work hours on work-family conflict and health mediated by organizational support and
perceived control
Work-family conflict was also reduced where
participants perceived themselves to have some
control over their work hours (β = -.19) (e.g., I could
cut back on the hours I work if I wanted). Indeed, in
some instances control might not be related directly to
the number of hours worked but simply to when the
hours were worked, that is, where some degree of
flexibility was involved.
Interestingly, there was no direct effect for long
worked hours on health in the current model. This
finding supports our earlier findings (e.g.,
Greeenglass, Burke & Moore, 2003; Moore, 2001)
where, whilst we found bivariate correlations between
stress and health, we failed to demonstrate a structural
relationship between work stress and health. Those
results suggested to us that the impact of mediating
variables is paramount in determining outcomes.
In summary, these findings confirm the nexus
between major facets of our lives: our work, our
relationships – here with the organization and with
our families, and our health outcomes. In the next
study, King and Moore (1999) extended the types of
stressors to examine simultaneously the impact of
work stress and family stress on work-family conflict
and health using data from 78 medical scientists.
Interestingly, we found that 40% of the variance in
work-family conflict was explained primarily by
family stress (β = .62), and it was also family stress in
combination with work-family conflict which
negatively predicted health status (6%) (Fig. 2).
Interestingly, there was no significant relationship
between work stress and family stress, or between
work stress and work-family conflict. Rather it was
family stress that impacted upon work-family conflict
and on health (β = -.12) suggesting that it is our
family life that is more influential in terms of domain
conflict and health. We decided then to extend this
model to include satisfaction with these two life
domains (see Fig. 3).
Work
Stress
-04
09
40%
Work-family
conflict
6%
-17
Health
08
62
Family
Stress
-12
Figure 2
Impact of work stress, family stress on work-family
conflict and health (χ2 4.3, p = .503) (Decimal points
are omitted for clarity)
The relationships among work and family stress and
work-family conflict as would be expected remained
stable (see left hand side of model), work stress
negatively predicted job satisfaction (β = -.14) and
family stress also had a negative but stronger impact
on family satisfaction (β = -.43). Interestingly, workfamily conflict contributed negatively to both job
satisfaction (β = -.26) and health (β = -.11) but had no
impact on family satisfaction: family satisfaction (β =
.29) and job satisfaction (β = .12) both contributed to
better health.
It seems, from the overall pattern of results in these
data, that family stress and family satisfaction are
more significantly related and play a more significant
role in health outcomes than do workplace stress and
job satisfaction. While not to downplay the role of
work stress and satisfaction on health, the finding that
relationships have a stronger influence should not be
surprising as our relationships tend to be more long
lasting than our job roles (even within the one
organization), occupy more of our time and typically
are more multi-faceted than work.
Healthy balance between work, family and personal relationships
We then decided to investigate the role of
personality factors, in particular neuroticism, on
stress, work-family conflict and health (see Fig 4).
Neuroticism had a significant impact on some factors
in the model, in particular, on family but not work
stress, on health and on the amount of variance
explained in health, which increased in this model as a
function of Neuroticism from 15 to 42%. Neuroticism
was very strongly related to health (β = .61) and to
family stress (β = .63) but not to work stress. Clearly,
neuroticism typified by worry, is an important factor
in family stress and health outcomes. It might be that
these worries are concerned with family strain,
children and even the dyadic relationship itself which
might subsume any work related worries.
We then used these same data to determine if there
were discernible clusters in the data using the same
variables as in the models (see Table 1) as well as an
indicator of family-work conflict.
Table 1
Comparison of Clusters
1
Family Stress
Family-work conflict
Work stress
Work-family conflict
Health
Neuroticism
10%
Job
satisfaction
-14
Work
stress
09
40%
Cluster
2
30.54
20.33
14.38
49.21
94.08
1.68
F
15%
-11
Health
62
19%
Family
stress
29
Family
satisfaction
-43
Figure 3
Extending Figure 2 to include job satisfaction and family satisfaction
-63
Work
stress
02
10
Neuroticism
10%
31
40%
42%
Workfamily
conflict
63
Family
stress
Figure 4
Influence on neuroticism on work stress, family stress and health
-17
09
p
76.89 109.44 .000
23.54
4.67 .034
10.96
2.99 .088
70.89 54.54 .000
88.68
4.45 .038
2.44
3.91 .052
12
-26
Workfamily
conflict
08
81
Health
MOORE
82
The results of the ANOVAs on the clusters revealed
that those participants who demonstrated more family
stress also exhibited more family-work conflict,
greater work-family-conflict and poorer health. These
findings are interesting in that they suggest that the
domains of health and work are, to some extent,
discreet and this separateness was also suggested in
the models presented in Figures 2 to 4.
The dominant role of family stress on health was
largely independent of health differences (p = .052)
and surprisingly, of work place stress as, for the group
high on family stress, work stress tended to be lower
(p = .088).
We then included job satisfaction and family
satisfaction in the clusters (see Table 2 for additions).
The groups differed only on level of family
satisfaction where those higher on family stress and
family-work conflict were also less satisfied with their
family life.
Table 2
Cluster Differences on Satisfaction
Cluster
1
2
F
p
Family satisfaction
Job satisfaction
9.53
4.67
.003
.427
6.00
83.83
5.00
81.05
Overall these results paint a picture where aspects
of the work domain (i.e., work stress, job satisfaction)
are interrelated and where the family domains (i.e.,
family stress and family satisfaction) as well as workfamily conflict and to a lesser extent family-work
conflict are also interrelated. These findings, while
also intuitively correct, contradict the earlier work of
Parasuraman, Greenhaus and Granrose (1992) who
found that only work role stressors were negatively
related to job satisfaction and that family role
stressors and work family conflict were unrelated. If
these findings are robust in further samples, then they
might be indicative of the beginnings of a sea-change
where people are beginning to place more importance
and value of the role of life outside the workplace.
However, these results, while clearly indicative of
discrete domains and effects, are limited in terms of
attributing causal directions. The data are crosssectional and provide at best, an indication of the
interrelationships among the factors at the time of
testing. It is therefore important that future studies are
longitudinal in design. It may also be relevant to
investigate for possible gender differences as some
past research (e.g., Lewis & Cooper, 1988) has
indicated that there may be differential spillover
effects in that work role demands are more likely to
be related to family life for men while family role
stress is more likely to affect the work life of women.
Although this premise sounds rather stereotypical, it
does warrant investigation, to either confirm or reject,
in samples of sufficient size to differentiate between
genders.
In two further studies (Firth, Mellor, Moore &
Loquet, 2003; Moore 2003) we looked at the role of
work place relationships, more specifically social
support in the work place, as a mediator of intention
to quit, that is the degree these peer relationships were
deemed to buffer stressors and their impact on
employees intention to leave their job Although
intention is not always indicative of actual behaviour,
it does indicate a negative attitude towards the job or
organization which may, in future, be actualized as
quitting behaviour.
In the first of these studies, using data from 173
employees of a major department store, we found that
social support was directly predictive of greater levels
of job satisfaction (β = .28) and job commitment (β =
.14) and lower levels of job stress (β = -.19); and
through these variables, it was indirectly predictive of
a reduced intention to quit (see Fig 5).
In the second study using data from 201 nurses
whose hospital had recently undergone a restructuring
process, we found that social support from colleagues
was indicative of lower levels of burnout (β = -.25)
and an increased sense of challenge (β = .21) to deal
with the restructuring process and its impact (see Fig
6). However, it must be countered that the greater the
level of impact of restructuring (e.g.., less time for
patient care, poorer working conditions) the less these
nurses felt themselves to be supported by their peers
(β = -.37). It might be that nurses were too busy
dealing with increased demands to offer the required
levels of support to each other although where this
was perceived to be present it was implicated in
reduced burnout and an increased sense of challenge
as well as indirectly reducing nurses’ intention to quit.
It seems from both these studies that the
relationship one has with peers at work is important in
terms of one’s health especially what might be termed
workplace health (e.g., burnout – cynical about the
job; job satisfaction and commitment to seek or work
out job-related problems) and one’s attitude, in this
case, seeing the impact of organizational change as a
challenge.
Another factor which increased nurses’ perceptions
of the impact of restructuring was the level of
communication by the organization. It would be
expected that the greater the communication by
management the less the impact of restructuring
however, this was not the case in the current data.
Rather the obverse was true (β = .60). This finding
seems incongruent until one looks at the actual type of
communications engaged in by management:
‘Management held meetings and told us what would
happen’; Management issued newsletters advising us
of what would happen’. Clearly, this was not
consultation at all: it was simply top-down directives.
Healthy balance between work, family and personal relationships
83
12%
21
20
Stress
-18
-19
Esteem
22
30%
Stressors
52%
-48
Job
commitment
-15
Intention to
quit
14
Social
support
-28
28
37
31%
-25
Job
satisfaction
-23
Figure 5
Department store data including social support on intention to quit
20%
Level of
Consultation
.46
.60
.38
14%
-37
70%
.36
Impact of
Restructuring
25%
-.25
Social
Support
Burnout
.41
Restructuring
Index
-.08
.09
30%
13%
-.37
Challenge
Intention to
Quit
.15
.21
33%
.42
Profession
Efficacy
.56
-.23
-.25
.26
Figure 6
Effects of hospital restructuring, peer social support and consultation by management on nurses’
burnout, efficacy and intention to quit
In summary, these studies indicate that work place
relationships, whether of an organizational or peer
nature, are vitally important. They are necessary to
help staff to deal with demands, to promote health and
to maintain a sense of job satisfaction which, in these
two studies was operationalised as an intention to
quit.
Although by no means conclusive, it does seem that
work stress and work support and work-health (e.g.
burnout, intention to quit, job satisfaction) are
relatively independent domains to family stress and
family satisfaction although the latter tends to
spillover over to create work-family conflict more
than vice versa.
In presenting a synopsis of these studies, I have
attempted to broaden the concept of relationship
beyond the dyadic relationship we most often think of
when using the term relationship. Rather, life is
comprised of many different types of relationships
and they all make important contributions to our
health and overall sense of wellbeing. As we spend
such a large part of our lives in the workplace, it is
vitally important that we have a sense of support from
both peers and the organization in order that this
system functions well for the individual and, indeed,
the organization. It seems from these data that it is
also important for our health that work-family conflict
is kept to a minimum and one mechanism for this
might be the maintaining of good interpersonal
relationships within the home.
In brief, good relationships wherever they might be
are vital for health and wellbeing. This may be
especially so where long hours and high workplace
demands prevail.
MOORE
84
References
Australian Bureau of Statistics (ABS). (2001).
Economic and social data. Retrieved 18th June,
2004, from http://www.abs.gov.au/ausstats
Cooper, C.L., & Cartwright, S. (1994). Healthy mind;
healthy organization: A proactive approach to
occupational stress. Human Relations, 47, 455-471.
Firth, L., Mellor, D.J., Moore, K.A., & Loquet, A. (in
press). How managers handle intention to quit.
Journal of Managerial Psychology. 19(2),
Greenglass, E.R., Burke, R.J., & Moore, K.A. (2003).
Reactions to increased hospital workload: Effects
on professional efficacy of nurses. Applied
Psychology: An International Review. 52(4), 580597.
Lewis, S.C., & Cooper, C.L. (1988). Stress in dual
earner families. In B.A. Gutek, A.H. Stromberg &
L.Larwood (Eds.). Women and work: An annual
review (pp. 57-79). Beverley Hills: Sage
Publications.
Maruyama, S., & Morimoto, K. (1996). Effects of
long workhours on life style, stress and quality of
life among intermediate Japanese managers.
Scandinavian Journal of Work, Environment and
Health, 22, 353-359.
Moore, K.A. (2001). Hospital restructuring: Impact on
nurses mediated by social support and a perception
of challenge. Journal of Health and Human
Services Administration, 23, 490-517.
Parasuraman, S., Greenhaus, J.S., & Granrose, C.S.
(1992). Role stressors, social support, and well
being among two career couples. Journal of
Organisational Behavior. 13, 339-356.
Sparks, K., Cooper, C.L., Fried, Y., & Shirom, A.
(1997). The effects of hours of work on health: A
meta analytic review. Journal of Occupational and
Organizational Psychology, 70, 391-408.
Tinker, S., & Moore, K.A. (2003). (2003). Work
hours, work-family conflict and health: the
mediating effects of perceived control and
organisational support. Korean Journal of Health
Psychology, 8(3), 663-677.
Trollope, A. (1967). Australia and New Zealand. St
Lucia, Brisbane: Uni Qld (First published London
1873).
Correspondence to:
Kate Moore, PhD, MAPS
School of Psychology
Deakin University
221 Burwood Hwy
Burwood 3125
kmoore@deakin.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 85-90
Melancholic Mothers and their Mourning Infants
Clare O’Bree (clare_obree@hotmail.com)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
Lisa N. Eisen (l.eisen@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
Zita Marks (z.marks@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
Jeannette Milgrom (Barbara.Frazer@austin.org.au)
Department of Clinical and Health Psychology
Heidelberg Repatriation Hospital, University of Melbourne, Melbourne VIC 3081 Australia
Kate Neilson (Kate.Neilson@austin.org.au)
Infant Clinic Research Project
Heidelberg Repatriation Hospital, Melbourne VIC 3081 Australia
potential to undermine intuitive mothering capacities.
It is well established that the mother-infant interaction
is the primary context in which infants develop the
ability to regulate their affect (Brazelton, Koslowski,
& Main, 1974; Papousek & Papousek, 1997; Stern,
1985; Trevarthen, 1993; Tronick & Gianino, 1986).
Depression may interfere with a mother’s capacity to
respond to an infant with the necessary empathic
attunement, sensitivity, and responsivity. Interaction
with a depressed mother, who is herself experiencing
difficulties in managing her affect, is unlikely to
provide the infant with essential assistance to
modulate distress, or to increase positive emotions.
Infants of depressed mothers may spend considerable
time in negative affective states as a result of
empathically resonating or attuning with the affect of
their mothers. Furthermore, the difficulties of
depressed mothers in interpreting and responding to
the cues of their infants may result in infants
experiencing a sense of helplessness and lack of
control at not being able to influence maternal
behaviour (Tronick & Gianino, 1986). Infants whose
affective cues are not responded to consistently,
would be expected to experience greater levels of
negative affect, and may develop a predominant style
of withdrawal and reliance on their own self-soothing
abilities, rather than drawing on external support for
affect
regulation.
The
intricate
emotional
communication between mother and infant is
vulnerable to disruption associated with maternal
depression. Such disruption has implications for the
effectiveness of the infant’s current and future
regulation of attention and affect. There is a
substantial body of research documenting difficulties
within the mother-infant interaction associated with
postnatal depression. It has been noted that the
affective quality of maternal behaviour, and the
temporal relationship of maternal behaviour to infant
behaviour are important aspects of the mother-infant
Abstract
This study focuses on the mother-infant relationship.
The impact on this developing relationship of a
Depressive Episode in the postnatal period was
investigated. Dyadic interactions between depressed
mothers and their 6-9 month old infants were observed
at the conclusion of a 12-week cognitive-behavioural
(CBT) group program designed to treat Postnatal
Depression. The quality of the dyadic interactions was
assessed using the Infant-Caregiver Behavioural
Measures and the Parent-Caregiver Involvement Scale,
and compared with the interactions of non-depressed
mothers and their infants. The interactive qualities
assessed included: verbal involvement; positive and
negative regard for child; responsiveness; enjoyment;
maternal affect; and infant affect and attention.
Depressed mothers treated with CBT had significantly
poorer quality of interaction with their infants,
compared with non-depressed mothers. This difference
was most pronounced for those whose depression had
not remitted, but was also significant for those who had
recovered. These findings have important implications
for the development of treatment strategies for
postnatally depressed mothers and their infants.
Depression occurring in the postnatal period has been,
and continues to be, the focus of considerable
research and clinical interest. Postnatal Depression
(PND) represents a significant mental health problem
with serious implications for maternal well-being,
couple and family functioning, and infant
psychological, social, emotional and cognitive
development (Murray & Cooper 1997).
Depressed mothers may be inhibited in their ability
to interact spontaneously, sensitively, responsively or
joyfully with their infants. The very nature of
depression, characterised in part by low mood and
resulting displays of flat or negative affect, loss of
interest or motivation, fatigue or low levels of energy,
and psychomotor agitation or retardation, has the
85
O’BREE, EISEN, MARKS, MILGROM, & NEILSON
86
interaction which are compromised when mothers are
experiencing depressive symptoms (Cohn, 1993).
This study aimed to evaluate whether difficulties in
the mother-infant interaction persist after postnatally
depressed mothers have received cognitivebehavioural treatment for their depression. It was
assumed, based on the theoretical literature and
empirical research, that interactional problems
accompany the experience of postnatal depression for
many mother-infant dyads. However, the research
literature is currently unclear as to whether motherinfant interactions continue to be impaired after initial
periods of postnatal depression have resolved.
This study assessed differences in the quality of
mother-infant interaction between two groups of
mothers and their infants: one group in which the
mothers had been postnatally depressed and
subsequently received a cognitive-behavioural group
treatment for depression (the “treated” group); and
another group in which the mothers had not been
depressed during the postnatal period (the “control”
group.
Method
Participants
The participants were 58 mother-infant dyads
recruited through a postnatal depression screening
program conducted at Maternal and Child Health
Centres (MCHCs). Thirty-two of the mothers had
experienced a Depressive Episode with an onset in the
postnatal period and had subsequently been treated
with 12 weeks of group cognitive-behavioural
therapy. The depressed mothers averaged 31 and the
control group 33 years of age, with their infants
averaging 31 and 32 weeks respectively. All infants
were born between 37 and 42 weeks gestation, and all
were single births. Of the depressed mothers, 41%
were primiparous, and 59% had either two or three
children. Within the control group, 58% of the
mothers were primiparous, and 42% had either two or
three children. The majority of mothers in both groups
were married or in defacto relationships, with the
exception of one depressed mother who was single,
and one control group mother who was separated.
The Infant Clinic Research Project This study was
designed to extend research in progress at the Infant
Clinic of the Parent-Infant Research Institute based at
Heidelberg Repatriation Hospital, under the direction
of Professor Jeannette Milgrom. The Infant Clinic
research project aimed to evaluate the efficacy of
treatment for Postnatal Depression and its effects on
the mother-infant relationship disturbance. The 12week cognitive-behavioral program was developed
specifically for PND and has been found superior to
routine care (Milgrom, Martin, & Negri, 1999;
Milgrom et al 2004).
Measures
Edinburgh Postnatal Depression Scale (EPDS)
Recruitment of participants involved mothers in the
North-East region of Melbourne being routinely
screened for depressive symptoms using the EPDS.
The EPDS was administered by Maternal and Child
Health Nurses at, or as close as possible to, the 12week postnatal visit. The EPDS is a 10-item selfreport scale specifically designed to detect symptoms
of depression experienced by women during the
postnatal period. It was administered and scored using
the methods described by Cox, Holden & Sagovsky
(1987). The EPDS is simple to administer and score
by persons not training in psychometric testing,
making it an acceptable tool to be administered by
Maternal and Child Health Nurses.
Hamilton Rating Scale for Depression (HAM-D)
Mothers who scored above 12 on the EPDS, a score
indicating
the
possibility
of
depressive
symptomatology, were offered an assessment session
with a psychologist at the Infant Clinic. The
assessment involved administration of the HAM-D
clinical interview to determine whether the mothers
met full diagnostic criteria for a Major Depressive
Episode according to the Diagnostic and Statistical
Manual for Mental Disorders, Fourth Edition (DSMIV: APA, 1994). The HAM-D (Hamilton, 1960) was
used within this research as a diagnostic interview,
rather than in its original rating scale form. A version
of the HAM-D revised to facilitate diagnostic
interviewing was used. The HAM-D contains 14
items, 9 of which correspond with the DSM-IV
diagnostic criteria for a Major Depressive Episode.
The interview requires systematic consideration of
each of the nine diagnostic criteria, for the purpose of
arriving at an accurate diagnosis.
Beck Depression Inventory, Second Edition (BDIII) Mothers who were diagnosed with a Major
Depressive Episode (MDE), and whose depression
had commenced during the postnatal period, were
invited to participate in the Infant Clinic’s research
program, which involved participation in a 12-week
cognitive-behavioural group treatment program. Prior
to and at the conclusion of the 12-week treatment
program participants completed the BDI-II (Beck,
Steer, & Brown, 1996). The control group was
recruited via letters (inviting them to participate in a
study of the effects of Postnatal Depression on
mother-infant interaction) sent to mothers who scored
below 12, on the EPDS when screened at MCHC, and
whose infants were aged between six and eight
months of age to match for age of those of the
depressed mothers in the main study. The control
group also completed the BDI-II. The BDI-II was
administered and scored according to the methods of
Beck et al. (1996).
Infant/Caregiver Behavioural Measures (ICBM)
At the end of the 12-week group treatment for the
depressed mothers, the quality of mother-infant
interaction was assessed. The control group were
scheduled to attend the Infant Clinic to assess motherinfant interaction. Each mother-infant dyad engaged
in a 15-minute play session when the infant was
awake and alert, and this interaction was videotaped.
A standard set of toys was used for each of the play
sessions. The mothers were asked to ignore the
Melancholic mothers
camera as much as possible and instructed as follows:
“Play with your baby as you would normally”. After
ten minutes, mothers were further instructed as
follows: “Try to get your baby to talk to you”. Efforts
were made to focus on the faces of mothers and
babies during the interaction, while keeping the play
activity in frame. At times when the mother and baby
were not in close proximity, the camera was “zoomed
out” to decrease the magnification of the image, so
that both mother and baby were continually in the
frame being filmed. Sound was also recorded using
the video camera’s inbuilt microphone. The ICBM
(Milgrom & Burn, 1988) assesses interactions
between caregivers and infants, according to qualities
of caregiver behaviour, observed aspects of infant
behaviour, and the qualities of the interaction between
infant and caregiver. There are 12 Caregiver
Measures: Respond; Stimulate/Arouse – object;
Stimulate /Arouse – event; Stimulate/Arouse – using
self; Vocalising; Positive Affect – physical; Positive
Affect – non-physical; Negative Affect – anxiety;
Negative Affect – hostility; Flat Affect; Body
Language Stillness; and Attending. The six Infant
Measures are: Infant Clarity of Cues; Explore;
Smile/Excite; Fuss/Cry; Avert; and Attending to the
Caregiver. The joint Caregiver/Infant Measures, of
which there are four, are: Mutual Attention;
Reciprocity/Synchrony; Joy; and Connectedness.
Three further measures, labeled Situational Measures,
are included to determine whether an individual
profile is valid, and these include: Caregiving; Infant
attending to other children; and Infant Alertness. High
scores on either of the first two Situational Measures,
or a low score on the third, indicate the need for
caution in interpreting the result as it may reflect an
atypical interaction for the dyad. Each of the abovelisted measures are rated, for each three minutes of
interaction, on a four-point scale: a score of zero
indicating the absence of the behaviour; a score of one
indicating the behaviour occurred from one to 30
percent of the time, a score of two indicating the
behaviour occurred 31 to 60 percent of the time, and a
score of three indicating that the behaviour was
present for more than 60 percent of the time. Interrater reliability for this measure has been established
at 84%.1
The Parent/Caregiver Involvement Scale (PCIS)
used in this research was a revision of several
previously developed scales. The current version,
revised by Seifer and colleagues (Seifer, Schiller,
Sameroff, Resnick, & Riordan, 1996), was based on
the Parent/Caregiver Involvement Scale developed by
Farren and colleagues (Farren, Kasari, Comfort, &
Jay, 1986). The scale was designed to focus on the
behaviour of the adult caregiver in interaction with an
infant or child (aged up to 2.5 years), and does not
specifically evaluate the infant or child’s contribution
to the interaction. The PCIS was designed to assess
caregiver interactions during home visits based on a
10-minute play session, however, its items are
1
Inter-rater reliability for the ICBM was established using 10
mother-infant interaction videos, each of which was rated by Clare
O’Bree and Kate Neilson prior to the commencement of this study.
87
applicable for play sessions in other contexts. The
PCIS, revised version, encompasses 10 sub-scales.
The initial six subscales focus on different caregiver
behaviours, for which two ratings are made: the
amount of the specified behaviour, and the quality of
that behaviour: Verbal Involvement, caregiver’s
verbalisations and vocalisations; Responsiveness,
caregiver’s reactions to infant signals; Play
Interaction, joint playful or game-like activities;
Directives and Control Over Child’s Activities,
organisation and management of the infant’s activities
by the caregiver; Positive and Negative Regard for the
Child, types of physical or verbal overtures or
responses made towards the infant, including praise,
criticism, affection or rough handling. The final four
subscales each require a rating to be made regarding
general impression of caregiver involvement with the
child: Availability of caregiver to child; General
acceptance and approval manifested by caregiver;
General atmosphere of caregiver involvement with
child; and Enjoyment (as experienced by the
caregiver). Each of the first six subscales require two
five-point Likert ratings, and each of the final four
subscales require one five-point Likert rating. Each of
the odd numbers (1, 3 and 5) is behaviourally
anchored, meaning that a description is provided to
enhance accuracy and consistency of coding. The
revised version of the PCIS has been found to be
significantly correlated with marital and family
functioning (Hayden et al., 1998) and with measures
of attachment security at six- and nine-months of age
(Seifer et al., 1996). Inter-rater reliability data was not
available for this measure.
Materials
All mother-infant interactions were recorded using a
portable video camera. The toys provided to mothers
and infants included a soft toy, a set of wooden
blocks, a plastic ball, plastic beakers, a rattle, an
activity frame, and a toy car.
Results
For the depressed mothers who received cognitivebehavioural treatment (n=32), BDI scores pretreatment were significantly higher than BDI scores
post-treatment (Wilcoxon Z = 4.01, p < .001). Despite
this significant reduction in depression following
treatment, the BDI scores of only 13 mothers who had
received the treatment fell in the range reflecting no
or minimal symptoms (Beck et al., 1996). Twelve of
the treatment group mothers received scores in the
mild range; four in the moderate range; and three
received scores reflecting severe depression. As a
result of this range of BDI scores following treatment,
the treatment group was divided into a
“treated/depressed” group (n = 19) whose BDI scores
were 14 or above, and a “treated/recovered” group (n
= 13) whose BDI scores were in the range of zero to
13. All mothers in the control group (n = 26) had BDI
scores of between zero and 13. There was a
significant difference in BDI scores between the three
groups (χ2(2) = 38.01, p < .001). Treated/depressed
mothers received higher BDI scores than either the
O’BREE, EISEN, MARKS, MILGROM, & NEILSON
88
control group (U = 0, p < .001), or the treated/
recovered group (U = 0, p < .001), who did not differ
significantly (U = 148, p = .55).
In terms of demographic characteristics, the three
groups (control; treated/recovered; treated/depressed)
did not differ in the mean age of mothers (F(2,55) =
2.58, p = .09), or infants (F(2,55) = .91, p = .41), or in
whether or not the infant was first born (χ2(2) = 3.2, p
= .20). Mothers’ level of education was not associated
with group membership (χ2(6) =7.93, p = .24).
However, a significant difference was found in
family income across the three groups (χ2(2) = 12.52,
p < .01). Mann-Whitney tests indicated that the
control group reported a significantly higher family
income than either the treated/depressed group (U =
108.5, p < .01) or the treated/recovered group (U =
88, p < .02).
A series of Kruskal-Wallis tests were conducted to
evaluate differences between the three groups of
mothers on items of the ICBM and the PCIS. The
relevant items of each of these measures were
analysed individually and the results are presented in
Table 1. Significant differences across the three
groups were found for a number of ICBM and PCIS
items, as outlined in Table 1.
Table 1
Comparisons of Mother-Infant Interaction Variables between Groups
Mother-Infant
Interaction Variables
• Maternal affect
Positive regard - qualitya
Positive regard - amounta
Negative affect - hostileb
Negative affect - anxiousb
Flat affectb
Stillnessb
Positive affect - physicalb
Positive affect - non-physicalb
• Maternal vocalising
Verbal involvement - qualitya
Verbal involvement - amounta
Vocalisingb
• Maternal play
Play interaction - qualitya
Play interaction - amounta
• Maternal responsiveness
Responsiveness - qualitya
Responsiveness - amounta
Respondb
• Infant affect
Smile/exciteb
Fuss/cryb
• Infant attention
Avertb
Attending to caregiverb
• Interaction/Enjoyment
Atmospherea
Enjoymenta
Joyb
Connectednessb
Treated/
depressed
(n = 19)
Mean Rank
Treated/
recovered
(n = 13)
Control
df
χ2
p
(n = 26)
23.47
25.82
30.53
27.29
35.76
33.53
32.45
24.45
28.96
32.08
29.00
35.73
31.50
34.08
32.08
28.00
33.06
30.90
29.00
28.00
23.92
24.27
26.06
33.94
2
2
2
2
2
2
2
2
4.57
1.55
2.05
5.32
6.23
5.82
2.11
4.27
.10
.46
.36
.07
.04*
.05
.35
.12
24.00
22.84
21.74
22.19
27.31
29.81
35.87
35.46
35.02
2
2
2
9.30
7.77
8.06
.01*
.02*
.02*
25.03
24.82
24.12
32.50
35.46
31.42
2
2
6.69
7.92
.04*
.02*
24.05
22.11
24.21
22.92
28.73
25.19
36.77
35.29
35.52
2
2
2
10.24
8.40
7.73
.01**
.02*
.02*
28.42
25.82
31.85
30.00
29.12
31.94
2
2
0.36
1.85
.84
.40
35.39
24.00
33.12
29.65
23.38
33.44
2
2
11.23
4.15
.004**
.13
24.11
24.55
30.03
26.87
24.50
24.35
28.88
28.04
35.94
35.69
29.42
32.15
2
2
2
2
7.71
7.07
0.43
1.30
.02*
.03*
.98
.52
Note. The PCIS items of Negative regard (amount & quality) were omitted as this behaviour was observed in only one mother.
Item from Parent/Caregiver Involvement Scale (PCIS), b Item from Infant/Caregiver Behavioural Measures (ICBM)
*p < .05, **p < .01
a
Melancholic mothers
Mann-Whitney tests were employed, for each of the
items showing significant differences, as a method of
making
planned
comparisons
between
the
treated/recovered group and the control group. Out of
the twelve items with group differences, six items had
significant differences between the treated/recovered
and control groups.
The control group had a significantly higher score
on the PCIS measure of quality of verbal
involvement, compared with the treated/recovered
group (U = 90.5, p < .02), however there was no
difference between those two groups on the amount of
verbal involvement, as measured by the PCIS (U =
121.5, p = .10) or the ICBM (U = 138, p = .28).
The control group had a significantly higher
quality of responsiveness as measured by the PCIS,
compared with the treated/recovered group (U = 89, p
< .01), and they also had a significantly higher score
on the “respond” item of the ICBM (U = 99.5, p
<.02). However, there was no significant difference
found between the control group and the
treated/recovered group on the PCIS amount of
responsiveness item (U = 132, p = .18).
The control group demonstrated a higher score on
PCIS quality of play interaction, compared with the
treated/recovered group, the difference approaching
statistical significance (U = 108.5, p = .05). However,
PCIS amount of play interaction, did not differ
significantly between the control and treated
/recovered groups (U = 162.5, p = .48).
The control group had a significantly lower score
on the ICBM item of “avert” (U = 108.5, p < .01) and
a significantly higher score on the PCIS atmosphere
item (U = 105.5, p < .05) and on the PCIS enjoyment
item (U = 104, p < .05) compared to the treated
/recovered group.
Discussion
The aim of this study was to evaluate whether
difficulties in the mother-infant interaction persist
after postnatally depressed mothers have received
cognitive-behavioural treatment for their depression.
The behaviour of mothers and infants was examined
through observation of a fifteen-minute videotaped
interaction.
Only 13 of the 32 depressed mothers in this study
had recovered from depression following the
cognitive-behavioural group treatment. These
treated/recovered mothers and their infants differed
significantly from the control group in important
ways. Although the treated/recovered mothers in this
study resembled the control mothers in terms of
amount of responsiveness, vocalising and play, they
showed significant disturbance when compared with
the control mothers, on the quality of each of these
dimensions. Furthermore, treated/recovered mothers
and their infants experienced less interactive
enjoyment than the control mothers and their infants.
The infants of treated/recovered mothers tended to
gaze avert more frequently, when compared with the
control group. While treated/recovered mothers may
have experienced improvement in their energy and
mood coinciding with remission from depression,
which may have resulted in some improvements in
89
their interaction with their infants, such as increased
time playing or talking with infants, mother-infant
relational disturbance appeared to persist in more
subtle forms.
The findings of this study are consistent with the
theoretical literature, which suggests that early
mother-infant interactional disturbance is likely to
lead to ongoing relational difficulties, in the absence
of appropriate intervention. Papousek and Papousek
(1997) assert that difficulties in mother-infant
communication can lead to serious interactional
disturbances, and possibly to maternal rejection of the
infant. These authors suggest that vicious circles can
form, whereby early problems elicit chains of
secondary deterioration. After postnatal depression
has remitted, disturbance established within the
mother-infant interaction may continue and gain
momentum through such a process. Even brief and
remitted episodes of postnatal depression, from this
perspective, have serious implications for infant affect
regulation and subsequent development.
It is acknowledged that considering mother-infant
interaction in isolation of the social context in which
the dyad is embedded is somewhat limited. This study
did not undertake examination of factors such as
social and partner support, which are known to be
significant to the experience of postnatal depression
and mother-infant relationships (Coyne et al., 1992).
Furthermore, the control group in this study reported a
significantly higher family income than either the
treated/recovered or treated/depressed groups. Socioeconomic status, reflected in family income level, is
likely to be related to other psycho-social variables
that may exert influence on the quality of motherinfant interaction independently of maternal
depression. Further research examining the
differential influences on mother-infant interaction of
maternal postnatal depression and the social context is
required, and would be highly useful in informing
targeted treatment strategies.
The data within this study suggests that there may
be differences between the treated/recovered and
treated/depressed mothers on several dimensions of
the mother-infant interaction, however, the modest
sample size prohibited analysis of this difference.
This study makes a unique contribution to the body
of research examining the impact of postnatal
depression on mother-infant interaction. The results
indicate that following recovery from postnatal
depression, mothers and infants continue to display
disturbances within their interaction. These findings
focus attention on the importance of considering the
mother-infant interaction when planning treatment for
mothers with postnatal depression. The cognitivebehavioural group treatment utilised in this research
does not appear to have effectively addressed
disturbance within the developing mother-infant
relationship. Effective treatment to address such
disturbances may need to focus on the mother-infant
relationship, and involve both members of the dyad.
Acknowledgments
This research was conducted under the auspices of the
Parent Infant Research Institute, Austin Health.
O’BREE, EISEN, MARKS, MILGROM, & NEILSON
90
References
American Psychiatric Association. (1994). Diagnostic
and statistical manual of mental disorders (4th ed.).
Washington, DC: Author.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996).
Manual for the Beck Depression Inventory (2nd
ed.). San Antonio, TX: The Psychological
Corporation.
Brazelton, T. B., Koslowski, B., & Main, M. (1974).
The origins of reciprocity: The early mother-infant
interaction. In M. Lewis & L. A. Rosenblum (Eds.),
The effect of the infant on its caregiver. (pp. 49-76).
New York: Wiley.
Cohn, J. F. (1993). Mother-infant play and maternal
depression. In K. MacDonald (Ed.), Parent-child
play: Descriptions and implications (pp. 239-256).
Albany: State University of New York Press.
Cox, J.L., Holden, J.M., & Sagovsky, R. (1987).
Detection of postnatal depression: Development of
the 10-item Edinburgh Postnatal Depression Scale.
British Journal of Psychiatry, 150, 782-786.
Coyne, J. C., Downey, G., & Boergers, J. (1992).
Depression in families: A systems perspective. In
D. Cicchetti & S. L. Toth (Eds.), Rochester
symposium on developmental psychopathology:
Vol. 4. Developmental perspectives on depression
(pp. 211-249). NY: University of Rochester Press.
Farren, D. C., Kasari, C., Comfort, M., & Jay, S.
(1986).
Parent/Caregiver
Involvement
Scale.Unpublished manuscript, Department of
Child Development and Family Relations,
University of North Carolina, Chapel Hill.
Hamilton, M. (1960). A rating scale for depression.
Journal of Neurology, Neurosurgery and
Psychiatry, 23, 56-62.
Hayden, L. C., Schiller, M., Dickstein, S., Seifer, R.,
Sameroff, A. J., Miller, I., et al. (1998). Levels of
family assessment: I. Family, marital, and parentchild interaction. Journal of Family Psychology,
12(1), 7-22.
Milgrom, J., & Burn, J. (1988). Synchrony coding
scale in maternal perceptions of infant
temperament, maternal competency, mothering
style and later mother-infant synchrony.
Unpublished manuscript, Melbourne College of
Advanced Education, Melbourne.
Milgrom, J., Martin, P., & Negri, L. (1999). Treating
postnatal depression: A psychological approach for
health care practitioners. Chichester: Wiley.
Milgrom, J., Negri, L.M., McNeil, M. & Martin P.R.
(2004). A Randomised Controlled Trial of
Psychological
Interventions
for
Postnatal
Depression. (In preparation).
Murray, L. & Cooper, P.J. (Eds.) (1997). Postpartum
depression and child development. New York:
Guilford Press
Papousek, H., & Papousek, M. (1997). Fragile aspects
of early social integration. In L. Murray & P.
Cooper (Eds.), Postpartum depression and child
development (pp. 35-53). New York: The Guilford
Press.
Seifer, R., Schiller, M., Sameroff, A. J., Resnick, S.,
& Riordan, K. (1996). Attachment, maternal
sensitivity, and infant temperament during the first
year of life. Developmental Psychology, 32(1), 1225.
Stern, D. N. (1985). The interpersonal world of the
infant: A view from psychoanalysis and
developmental psychology. New York: Basic
Books, Inc.
Trevarthen, C. (1993). The self born in
intersubjectivity: The psychology of an infant
communicating. In U. Neisser (Ed.), The perceived
self: Ecological and interpersonal sources of selfknowledge (pp. 121-173). Cambridge: Cambridge
University Press.
Tronick, E. Z., & Gianino, A. F. (1986). The
transmission of maternal disturbance to the infant.
In W. Damon, E. Z. Tronick & T. Field (Eds.), New
directions for child development: No. 34. Maternal
depression and infant disturbance (pp. 5-11). San
Francisco: Jossey-Bass.
Correspondence to:
Clare O’Bree
School of Psychology
Australian Catholic University
115 Victoria Pde. Fitzroy, VIC 3065
clare_obree@hotmail.com
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 91-95
The disparate olfactory determination of sex and age in humans, reflecting
the reproductive and relationship characteristics of the species
Robert N. Paddle (r.paddle@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne Vic 3065 Australia
Jacqui Robert (jrobert@ysas.org.au)
School of Psychology
Australian Catholic University, Melbourne Vic 3065 Australia
than women (Powers, 1971). The theoretical
framework of all these early studies has, however,
been considered less than compelling, and more
recent studies have used an evolutionary perspective
on human mating behaviour (Feingold, 1992).
Sexual selection was originally proposed by Darwin
(1871) and refers to the evolution of characteristics
that give organisms reproductive advantage, as
opposed to daily survival advantage. Triver’s (1972)
parental investment model, borrowed directly from
Darwin with a “new-synthesis” spin, states that
females hold a larger investment in their offspring
than do males; women can realistically only bear a
limited amount of children whereas men can
theoretically impregnate large numbers of women.
Parental investment theory suggests that men are
mainly attracted to the opposite sex by visual cues
that signify reproductive ability, such as youth and
physical attractiveness, as women can only reproduce
up until a certain age. Women, on the other hand, are
said to seek features in a mate that indicate resource
attainment, as it maximises the survival prospects of
their offspring, rather than appearance-related factors.
Parental investment has been defined as “any
investment by the parent in an individual offspring
that increases the offspring’s chance of surviving (and
hence reproductive success) at the cost of the parent’s
ability to invest in other offspring” (Trivers, 1972,
p139), and extensive support for parental investment
theory and greater female selectivity has been found
throughout many vertebrate and invertebrate species
(Trivers, 1985). Such female acuity has been used to
explain the extremely high incidence of monogamy
found amongst primate species (Paddle, 1989, 1991).
In sex-role reversed species, such the Mormon cricket
(Anabrus simplex) and Panamanian poison-arrow frog
(Dendrobates auratus), in which males, through
parental care and protection, invest more in their
offspring than females; females are often larger than
males, and they compete more aggressively with each
other for the more heavily investing males (Summers,
1989; Trivers, 1985). These results draw attention to
the fact that relative parental investment, not purely
biological sex, drives the process of sexual selection
(Buss & Schmitt, 1993).
Humans are typical mammals, in which females
invest more heavily in reproduction. This occurs
partly because offspring develop internally within
women, and after the birth of a child, women carry the
further
parental
investment
of
lactation;
Abstract
Olfactory information is critical to mammalian sexual
behaviour. Parental investment theory suggests that, in
constructing relationships, the relative importance of
olfaction compared with vision, touch, taste and
hearing should be different for human males and
females. A convenience sample of 36 subjects, 30
females and six males, with a mean age of 24.7 years,
completed the Sensory Stimuli and Sexuality Survey
(Herz & Cahill, 1997) to ascertain whether olfactory
stimuli would be a more important determinant of
sexual choice and arousal for females than it would be
for males, and this was supported. Subjects were then
asked to consider the odours of four selected well-worn
and unwashed t-shirts. As expected, females were more
accurate than males in determining the sex of an
individual by the odour of a worn article of clothing,
but females were not more accurate than males in
determining the age of an individual by the odour of a
worn article of clothing. Combined-sex data showed
that age estimates for the wearers of the female t-shirts
were more accurate than age estimates for the t-shirts
worn by males. In the light of the more limited span of
reproductive fertility amongst human females, this
result was not surprising. It was concluded that the
human olfactory identification abilities demonstrated in
this study reflect basic mammalian relationship
proclivities within our species.
Freud and Jung considered that humans seek in mates,
characteristics that resemble their opposite sex parent
(Eckland, 1968). Winch (1958) believed that people
seek characteristics that they themselves lack. Others
have examined the popular perception that people
seek similarity in mates, that like attracts like (Buss &
Schmitt, 1993). Exchange and equity theories propose
that people search for those with whom an equal
exchange of valuable resources can be obtained
(Berscheid & Walster, 1974; Clark & Reis, 1988). All
these theories share the view that human mating is
strategic and that selections are made, either
consciously or unconsciously, to capitalise on a
particular characteristic or equipoise. However, they
also seem to share a lack of specificity, broadly
seeking fairness, similarity and complementarity
between the sexes, but not identifying the specific
areas in which individuals make their choice. Interest
in mate selection criteria has been widespread in
research on the sociology of the family, and early
findings supported the widely held sex-linked belief
that physical attractiveness is more important to men
91
92
PADDLE & ROBERT
accompanying this, however, is a certainty of
maternity. Due to these forms of investments, the
number of children a women can successfully produce
is limited. Men do not have these obligatory forms of
heavy parental investment and paternity is never
certain. The minimum investment by the male is in
contributing easily-replaced sperm. As a result of this,
men have a higher potential capacity to produce
offspring (Buss & Schmitt, 1993). According to
sexual selection theory (Darwin, 1871), when
choosing a mate, the sex differences in parental
investment cause women to be the more selective and
discriminating, and men to be less discriminating and
more competitive. Expressed more simply, males
(certainly in the absence of expressed monogamous
social structures) should mate with as many fertile
females as possible, while females should select mates
who are most likely to assure the survival of their
offspring (Herz & Cahill, 1997).
Buss and Schmitt, (1993) have proposed that the
assessment of human female fertility is determined by
the physical indications of youth and health, for
example, full lips, clear and smooth skin, clear eyes,
and high activity level, and thus male mate-search
strategies are expected to be based mainly on the
visual evaluation of physiological cues to a female’s
maternal fitness (Herz & Cahill, 1997). Conversely,
females have been characterised as predominantly
concerned with evaluating males’ behaviour, loyalty
and contribution of resources towards herself and her
prospective offspring, a mate-search strategy
conventionally described as based mainly on the
evaluation of behavioural cues to a male’s possible
paternal attitude (Buss & Schmitt, 1993; Herz &
Cahill, 1997). On reflection, however, the
physiological aptitude of the male must also play a
part in female assessment, albeit at a more complex
level than the simple age-based criterion posited for
males, and parental investment theory predicts that
females should be more concerned than males with
physiologically adaptive mating, because females
invest more in their offspring than do males (Trivers,
2002).
Physiologically adaptive mating involves genetic
compatibility between a mating couple, including
allele combinations in offspring that maximise disease
protection. Therefore females should be the more
sensitive sex to any physiological cues indicative of
immunological genotype.
The major histocompatibility complex (MHC) was
first studied because of its importance in tissue
transplantation and the immune system in humans
(Hendrick, 1993). Products of the MHC have an
extremely important role in immune recognition
(Wedekind, Seebeck, Bettens, & Paepke, 1995). Mice
can recognise each other by individually characteristic
phenotypic body odours that reflect their genetic
constitution of the MHC (Beauchamp, Curran &
Yamazaki, 2000), and female mice base their matechoice on odours, seeking out or avoiding certain
alleles depending on their own genotype (preferring
mates who are genetically dissimilar) (Herz & Cahill,
1997; Wedekind et. al., 1995).
This preference for genetically dissimilar mates
could have evolved either through a strong selective
advantage in offspring possessing a wider range of
immunological response, presumably provided by
variability in MHC genes, or as a means for
inbreeding avoidance (Jordan & Bruford, 1998).
Certainly, the greater the MHC similarities existing
between breeding humans, the greater the chance of
infertility and spontaneous abortion (Ho, Gill, Nsieh,
Hsieh & Lee, 1990). Recent research has shown that
human MHC type is an important variable in human
female mate choice, and that, as with rodents, it is
demonstrated in response to body odour.
Wedekind et. al. (1995) typed female and male
students for their human leukocyte antigens which
correspond to the mouse MHC. Each male subject
then wore a t-shirt for two consecutive nights, after
which the t-shirts were collected and placed in
identical cardboard boxes for the female subjects to
sniff and assess. For each female, half of the boxes
contained t-shirts from men who were similar to her
in MHC-type and half contained t-shirts from men
who were dissimilar. Their results found that females
preferred the smell of males who were most dissimilar
from them in MHC-type, thus indicating that female
preferences for male body odour inversely correlate
with MHC complementarity. (The odour assessment
preference was reversed, however, when the women
rating the t-shirts were taking oral contraceptives.)
It is well known that animals use olfactory signals
to communicate their sexual status and individuation.
Such olfactory information is a critical component of
sexual and social behaviours expressed in all
mammalian types (Jackson, 2003). Odour-based
discrimination probably plays a major role in
organising group behaviours in most social-living
species with well-developed olfactory abilities. That
personal odour plays a significant part in human nonverbal communication has been studied by Russell
(1976), Schleidt (1980) and Schleidt, Hold, and Attili
(1981).
Humans are highly scented mammals, with scentproducing glands associated with each hair follicle.
Given that the species is almost as hairless as aquatic
and marine mammals, their only function would seem
to be to produce various scents (Stoddart, 1990).
Apocrine glands are most dense around the armpits,
pubic region, the area around the anus, the face, the
scalp and the umbilical region. These glands are
responsible for most of a healthy person’s body
odour. In addition to genetic differences,
environmental factors, such as diet, also contribute to
a person’s distinctive body odour. Environmental
factors are also seen in the organization of human
social-group behaviour, with menstrual synchrony
among women living together achieved through
pheromones (McClintock, 1971).
The investigation of female mate-search and
relationship strategies based on olfactory cues has
been somewhat overlooked in psychology. The
primary purpose of this research was to investigate the
function of the olfactory system with regard to sexual
attraction and mate choice. It was hypothesised that
olfactory stimuli would be a more important
Disparate olfactory determination
determinant of sexual choice and arousal for females.
It was also hypothesised that females would be more
accurate than males in determining the sex and age of
an individual by the odour of an article of clothing.
Method
Subjects
The convenience sample used consisted of 36
subjects, 30 females and six males, who were
studying second year Psychology at an Australian
university and participated as part of their course
requirements. The subject’s age ranged from 19.3
years to 46.2 years, with the mean age being 24.7
years. No attempt was made to establish whether or
not female subjects were currently taking a
contraceptive pill.
Materials
To evaluate how different types of sensory
information are used by males and females to assess
sexual partners and evoke sexual arousal, the Sensory
Stimuli and Sexuality Survey was used (Herz and
Cahill, 1997). It consists of 18 Likert scaled questions
(with 1 = strongly disagree, and 7 = strongly agree)
grouped under three topics: lover/potential lover
choice, sexual arousal during sexual activity and
sexual arousal during nonsexual activity. The
questions under each topic measured the subjective
importance of sight, hearing, touch and smell to
subjects’ behaviour.
The second part of the research involved
determining whether subjects could identify
characteristics of an individual by the odour of a worn
t-shirt. Three males (between 19 and 29 years) and
three females (between 24 and 28 years) were
recruited to assist the researchers. They were each
given a new t-shirt and asked to wear it to bed for
three consecutive nights. Before going to bed, the tshirt wearers were asked to shower, but not to use any
soap, shampoo, perfume or deodorant. Upon waking,
they were asked to remove their t-shirt and place it in
a resealable plastic bag. The six t-shirts were then
collected and bilaterally scissored, to produce two
identical odour test samples, thus enabling the use of a
replicate t-shirt in the study
A set of six questions was asked about each t-shirt
concerning the level of attraction felt by a subject,
familiarity of odour, probable sex and age of t-shirt
wearer, point of reproductive cycle (if applicable),
and stimulated romantic interest of subject.
Procedure
Each subject was tested individually in a small room
in the presence of one of the researchers, and given a
brief explanation of the questionnaires and activity to
be completed. They each sniffed and commented
upon four different t-shirts and one replicate t-shirt.
The choice and order in which the four different tshirts were to be sniffed was randomised for each
subject, with the addition into fourth position of the
replicate half of the t-shirt first sniffed. After they had
answered the six questions on the third t-shirt, the
instructions suggested that they may like to return to
93
the first two t-shirts and readjust their ranking scales
as they now had some comparisons. When each
subject had completed the exercise, they placed their
anonymous response in a designated collection box.
The responses of all 36 subjects were obtained in
six hours of testing completed in one day – ensuring
an equivalent state of “freshness” for the t-shirts under
consideration.
Results
Responses for each subject were scored and
descriptive statistics, z tests and χ2 analysis, where
appropriate, were conducted. For frequency
comparisons involving sex, given the small numbers
of male subjects and their resulting limited and
unequal range of responses across the available tshirts, χ2 analysis, even with Yates correction, proved
problematical (Siegel, 1956). Frequency comparisons
in these cases were undertaken by a z Test for
significance of difference between two proportions on
a logically dichotomous variable (Bruning and Kintz,
1977)
To examine the consistency and reliability of
subject’s answers, correlative responses to
attractiveness for the first t-shirt sniffed and it’s later
replication, in position four of the sequence, were
examined. It was found that there was a significant
positive correlation between the first t-shirt and the
replicate t-shirt, r=.737, p<0.001, representing a
perfectly credible level of reliability to the test
situation. For comparison, correlative responses to the
first t-shirt and the third t-shirt sniffed (sourced to
different bodies) found no significant correlation,
r=.068, p>.05.
One way independent sample t-tests were calculated
for each of the 18 items on the Sensory Stimuli and
Sexuality Survey to discover whether sex differences
existed in responses. Females rated item four (“how
someone smells can make a big difference to me”)
significantly higher than did males, t(32)=2.223,
p<.02. No significant sex differences were found for
any of the other items, including that of appearance
(“how someone looks”).
The second hypothesis suggested that females
would be more accurate than males in determining the
sex of an individual by the odour of an article of
clothing. Females correctly identified the sex in 72%
of cases. At 52% males were no better than chance.
The data available for sex comparison in this study
unfortunately do not originate from equal source and
choice bases, and thus fail to meet essential criteria
for χ2 analysis. Hence, a one-tailed z Test for
significance of difference between two proportions on
a logically dichotomous variable was conducted on
sex estimation, and this proved significant for the
hypothesised female superiority (z=1.83, p<.03).
The third hypothesis suggested that females would
be more accurate than males in estimating the age of
an individual by the odour of an article of clothing.
The age range for a correct estimate of age was
chosen to be plus or minus five years, within which,
rather than showing evidence of greater female
ability, males were correct 63% of the time, females
PADDLE & ROBERT
94
only 48% of the time, but this difference was not
significant (z=1.27, p>.05).
Discussion
The results of the current study supported the
hypothesis that olfactory stimuli would be a more
important determinant of sexual choice and arousal
for females than it would be for males and that
females would be more accurate than males in
determining the sex of an individual by the odour of
an article of clothing. The results did not support the
hypothesis that females would be more accurate than
males in determining the age of an individual by the
odour of an article of clothing.
The fourth item on the Herz and Cahill (1997)
Sensory Stimuli and Sexuality Survey, “how someone
smells can make a big difference to me”, displayed
the only significant difference between male and
female respondents. The survey result showed that
females believed their sexual interest was affected
more by body odour than any other sensory stimuli.
Herz and Cahill (1997) found that, for males, when
selecting a mate, visual and olfactory cues were
equally important, while for females, olfactory cues
were the single most important variable in mate
choice and were found to be significantly more
important than visual, tactile or auditory cues. The
current study supported this typical mammalian
finding.
Recent research has indicated the importance of
human MHC type as a variable in female mate choice
and, just as in rodents, it is mediated through body
odour (Herz & Cahill, 1997). Ober, Weitkamp, Cox,
Dytch, Kostyu and Elias (1997) found that human
mate choice in isolated areas of northern United States
and western Canada was influenced by MHC
haplotypes, with an avoidance of partners with similar
haplotypes to one’s own. Wedekind et. al (1995)
found that women scored male body odours as more
pleasant when they differed from the men in their
MHC, than when they were similar. The present
findings, that females were significantly more
accurate than males in the olfactory determination of
sex, support previous research that has demonstrated
the importance of olfactory discrimination in female
mammal mate-selection strategies (Herz & Cahill,
1997).
The third hypothesis, that females, because of their
greater focus on the olfactory in relationships, would
also be superior at estimating the age of the t-shirt
wearer was not supported.
Interestingly, a post-hoc analysis of this data,
combining the sexes, revealed a significant effect in
terms of the sex of the t-shirt sniffed (see Table 1). By
combining the sexes, the existing problem of disparity
in source and choice bases of responses was obviated,
and a χ2 test became the most appropriate measure for
statistical analysis. Subjects as an whole, were
markedly better at predicting the age of a female tshirt wearer than they were at predicting the age of a
male t-shirt wearer, χ2=8.450, p<.005. One possible
explanation for this may be found in a different
orientation towards age in social mammalian species.
The male’s orientation towards mating is based on the
physiological, females live as long as, or longer than
males, but are reproductively more restricted in the
length of time of fertility in their life-cycle (and less
reproductively successful with age), than males.
Consequently, males should be biologically
programmed to decipher age-related cues from female
odours. In contrast, within a social mammalian
species, the female orientation towards mating
involves both behavioural and physiological
assessment, but the latter, at least in terms of the
major histocompatibility complex, post-puberty, is not
age-related. With males remaining reproductively
fertile throughout the bulk of their life-cycle, age is
not such a significant factor in female mate-choice,
but it may well be important for a female social
mammal to recognise females of similar reproductive
potential in the vicinity, who may, depending on the
species’ period of fertility, represent competitors for
the most desirable males, but probably more
importantly, they represent potential co-operators in
the communal care of young.
Table 1
T-shirt age estimates
Male t-shirt
Female t-shirt
Correct
Estimate
25
44
Incorrect
Estimate
41
26
A number of methodological issues and limitations
associated with this research should be noted. While a
larger number of subjects, including a greater number
of males and broader age range, would be beneficial,
potential changes in worn t-shirt odours over time
could confound data collection spread over two or
more days.
This study was a direct examination of the role of a
rarely considered form of sensory information in
human sexual behaviour. It found that olfactory
stimuli were a more important determinant of sexual
attraction and arousal for females than for males. It
also found that females were more accurate than
males in determining the sex of an individual by the
odour of a worn article of clothing. Finally, it was
found that estimates for the age of the wearer of a tshirt were more accurate for t-shirts worn by females,
than t-shirts worn by males. Sexual selection was
originally proposed by Darwin (1871) and it clearly
paved the way for Trivers’ (1972) theory of parental
investment. The human olfactory identification
abilities demonstrated in this study support parental
investment theory and appear to reflect basic
mammalian relationship proclivities within our
species.
References
Beauchamp, G. K., Curran, M., & Yamazaki, K.
(2000). MHC-mediated fœtal odour types expressed
by pregnant females influence male associative
behaviour. Animal Behaviour, 60, 289-295.
Berscheid, E., & Walster, E. (1974). Physical
attractiveness. In L. Berkowitz (Ed.). Advances in
experimental social psychology, (pp. 157-215). San
Diego: Academic Press.
Disparate olfactory determination
Bruning, J. L., & Kintz, B. L. (1977). Computational
handbook of statistics (2nd ed.). Glenview, Illinois:
Scott Foresman.
Buss, D. M., & Schmitt, D. P. (1993). Sexual
strategies theory: An evolutionary perspective on
human mating. Psychological Review, 100, 204232.
Clark, M. S., & Reis, H. T. (1988). Interpersonal
processes in close relationships. Annual Review of
Psychology, 39, 609-672.
Darwin, C. (1871). The decent of man and selection in
relation to sex. London: John Murray.
Eckland, B. (1968). Theories of mate selection. Social
Biology, 15, 71-84.
Feingold, A. (1992). Gender differences in mate
selection preferences: A test of the parental
investment model. Psychological Bulletin, 112,
125-139.
Hendrick, P. W. (1993). Evolutionary genetics of the
major histocompatibility complex. American
Naturalist, 143, 945-964.
Herz, R. S., & Cahill, E. D. (1997). Differential use of
sensory information in sexual behaviour as a
function of gender. Human Nature, 8, 275-286.
Ho, H. N., Gill, T. J., Nsieh, R. P., Hsieh, H. J., &
Lee, T. Y. (1990). Sharing of human leukocyte
antigens in primary and secondary recurrent
spontaneous abortions. American Journal of
Obstetrics and Gynaecology, 163, 178-188.
Jackson, S (ed.). (2003). Australian Mammals.
Biology and captive management. Melbourne:
CSIRO Publishing.
Jordan, W. C., & Bruford, M. W. (1998). New
perspectives on mate choice and the MHC.
Heredity, 81, 127-133.
McClintock, M. K. (1971). Menstrual synchrony and
suppression. Nature, 229, 244-245.
Ober, C., Weitkamp, L. R., Cox, N., Dytch, H.,
Kostyu, D., & Elias, S. (1997). HLA and mate
choice in humans. American Journal of Human
Genetics, 61, 497-504.
Paddle, R. N. (1989). The pedantic monopoly on
primate monogamy. Australian Primatology, 4 (3),
2-3.
Paddle, R. N. (1991). Deconstructing primate
monogamy: Increasing its incidence, and requiring
global models for its phylogenetic origins. In E.
Akiyoshi, T. Kimura, O. Takenaka and M. Iwamoto
(eds) Primatology Today. Amsterdam: Elsevier.
Powers, E. A. (1971). Thirty years of research on
ideal mate characteristics: What do we know?
International Journal of Sociology of the Family, 1,
207-215.
Russell, M. J. (1976). Human olfactory
communication. Nature, 260(5551), 520-522.
Schleidt, M. (1980). Personal odour and nonverbal
communication. Ethology and Sociobiology, 1, 225231.
Schleidt, M., Hold, B., & Attili, G. (1981). A crosscultural study on the attitude towards personal
odour. Journal of Chemical Ecology, 7, 19-31.
Siegel, S. (1956). Nonparametric statistics for the
behavioural sciences. Tokyo: McGraw-Hill
Kogakusha.
95
Stoddart, D. M. (1990). The scented ape: The biology
and culture of human odour. Cambridge:
Cambridge University Press.
Summers, K. (1989). Sexual selection and intrafemale competition in the green poison-dart frog,
Dendrobates auratus. Animal Behaviour, 37, 797805.
Trivers, R. (1972). Parental investment and sexual
selection. In B. Campbell (Ed.). Sexual
selection
and the decent of man (pp. 136-179). Chicago:
Aldine-Atherton.
Trivers, R. (1985). Social evolution. Menlo Park, CA.:
Benjamin/Cummings.
Trivers, R. (2002). Natural selection and social
theory. New York: Oxford University Press.
Wedekind, C., Seebeck, T., Bettens, F., & Paepke, A.
J. (1995). MHC-dependant mate choice in humans.
Proceedings of the Royal Society of London: Series
B, 260, 245-249.
Winch, R. (1958). Mate selection. New York: Harper
and Row.
Correspondence to:
Robert Paddle
School of Psychology,
Australian Catholic University, St. Patrick’s Campus
115 Victoria Parade
Fitzroy. Victoria 3065
r.paddle@patrick.acu.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 96-101
Parental Decision-Making surrounding Complicated Births: The Impact on
Parent-Infant and Infant-Marital Relationships
Sarah J. Phillips (sjph@deakin.edu.au)
School of Psychology
Deakin University, Melbourne VIC 3125 Australia
Gregory A. Tooley (tooley@deakin.edu.au)
School of Psychology
Deakin University, Melbourne VIC 3125 Australia
increase the strain placed on the infant-parent
relationship, especially during lengthy NICU stays.
Additionally, the delay in the much-awaited event of
taking the baby home may increase feelings of anxiety
and ambivalence between the parent and infant, also
impacting the marital relationship (Harrison, 1993).
Although attention has been alerted to the adverse
consequences of poor decision-making processes, and
infant-parent separation on relationships, parents
currently receive little guidance as to how to make
difficult treatment decisions under extraordinary
physical and emotional stress. This paper provides a
brief overview of the different types of medical and
general care decisions that new parents often face, and
examines the consequences of poor decision-making
for the infant, parent, and family unit. The benefits of
kangaroo care (skin to skin contact) and maternal
sensitivity and their impact on family functioning will
be discussed. Finally, the future research area of
education and preparation of couples for serious
decision-making surrounding complicated births, and
the induction of strategies to facilitate strong maternal
and paternal attachment will be highlighted.
Abstract
This paper reviews current research regarding the
impact of birth complications on parental decisionmaking, and the resulting effects on parent-infant and
infant-marital relationships. It discusses the importance
of informed decision-making on parental satisfaction of
the birthing experience, and the benefits of certain
strategies, such as kangaroo care, in the facilitation of
greater levels of attachment and improved
relationships. The paper concludes with suggestions for
future research areas to focus on finding better ways to
prepare and support parents in these situations, thus
improving the quality of relationships between parents
and with their child.
The technical advances that have characterized
newborn care in the last forty years have been
accompanied by an increase in the survival rate of
preterm (birth before 37 completed weeks), low birth
weight (<2,500g) and critically ill infants. Twins and
higher order multiples, which have increased in birth
rate due to assisted reproductive technology (Ventura,
Martin, Curtin, Menacker, & Hamilton, 2001), are
also more likely to be born prematurely and have low
birth weights (Gardner et al., 1995). Such infants are
often at a higher risk of long-term neurological,
cognitive and behavioural problems (Breslau, 1995).
Accompanying the technological advancements,
however, is an increase in the need for parents to
participate in very serious treatment decisions
regarding resuscitation, mode of delivery, and other
neonatal intensive care unit (NICU) decisions. Studies
investigating the experience of such tasks have lead to
an increased appreciation of the psychological strain
and emotional stresses encountered by the family of
the sick infant (Siegel, Gardner, & Merenstein, 2002).
Amidst trying to cognitively and emotionally assess
competing treatment options in decision-making for
their infant’s care, parents have reported an impaired
ability to retain and to understand important
information in stressful periods such as labour, time
constraints, anxiety, and even doubts in the
willingness and confidence to be involved in the
decision-making process (Harrison, 1993; Zupancic et
al., 2002). Qualitative studies investigating parents’
experiences of birth complications have highlighted
the need for the parents to be able to see, hold, and
touch their newborn in order to facilitate early
attachment and bonding. The initial separation of the
infant from the mother and father has been found to
Decisions surrounding Complicated Births
There are many reasons that infants require NICU
care. The most common reason is preterm birth,
which occurs in almost 7 percent of all births in the
state of Victoria (Ridley & Halliday, 2001). Birth
complications other that prematurity are also
common, and in the year 2000, documented
difficulties included emergency caesareans (11
percent), unplanned vacuum extraction (6 percent),
and stillbirths (0.7 percent). As a consequence of
some of these complications, a striking 43 percent of
all births in the year 2000 required some form of
resuscitation, ranging from external cardiac massage
and ventilation, to suction and O2 therapy (Riley &
Halliday, 2001).
As can be inferred from these statistics,
complications during birth are not infrequent events
and often require serious decision-making in
determining the appropriate level of treatment (or
non-treatment) of sick infants. Society has, for many
reasons, allocated the parents primary responsibility,
albeit in collaboration with health care providers, in
making decisions about their infant’s care (Harrison,
1993; Swaney, English, & Carter, 2002). Of all
people, parents know best the values of the family
culture or environment in which the infant will be
96
Parental decision-making and complicated births
raised (Swaney et al., 2002). They often desire to be
active participants in the many decision tasks that are
necessary during the NICU treatment course, and
report daily decision-making tasks (such as whether to
contribute to medical care plans, to raise issues with
medical staff, or to seek second opinions) as being
just as important as serious treatment decisions
(Sudia-Robinson & Freeman, 2000).
However, it has been consistently reported that
parents find it difficult to cope with such decisionmaking tasks due to the overwhelming emotional
response elicited by unexpected birth complications
(Harrison, 1993; Lau & Morse, 2003). Qualitative
interviews of parents who have experienced birth
complications have revealed that parents are
frequently physically and emotionally overwhelmed,
and confused and/or intimidated by the high-tech
environment of the NICU and the complexities of
their infant’s care (Lau & Morse, 2003). Amid
feelings of grief, fear, anxiety and wonderment over
their preterm or critically ill infant, they may be
uncertain of their proper roles and responsibilities as
parents (Swaney et al., 2002). Before examining the
potential consequences of poor decision processes, it
is important to place in context the influence of the
NICU environment itself.
Impact of NICU Environment
Not only has the number of infants requiring NICU
care increased, but the length of time required for an
infant to stay in the NICU has also dramatically
increased due to lower gestational ages and smaller
birth weights (Ventura et al., 2001). The impact of
NICU care on mothers and fathers has been well
documented. Findings generally concur that mothers,
who frequently do not get the chance to see or touch
their baby before they are whisked away for medical
attention, often feel distressed about being separated
from their newborn (Tully, Arseneault, Caspi, Moffitt,
& Morgan, 2004). During the pregnancy, parents
often build up a rather idealized picture of the baby
they are expecting to be born. However, many parents
of even healthy babies later volunteer that they were
shocked and appalled when they first saw their baby
(Kitchen, Ryan, Rickards, & Doyle, 1998). Among
the negative sequelae of complicated births, such as
prematurity, is the disruption in the attachment
process, resulting in part from maternal-infant
separation caused by standard incubator care.
Maternal separation during the post-birth period
exerts a negative impact not only on the infant’s
physiology and behaviour but also on the mother’s tie
to her infant (Hofer, 1995). The frequency of maternal
affiliative behaviour, affectionate touch, and care
taking decrease in the immediate post-birth period
following premature birth, partly due to disruptions in
the psychobiological process of maternal bonding
(Feldman, Weller, Sirota, & Eidelman, 2003). Studies
comparing the mother’s global relationship style
among preterm and full-term infants have found lower
levels of maternal sensitivity and adaptation in the
preterm group (Feldman et al., 2003). Mothers of
premature infants often exhibit an intrusive interactive
style (Brachfield, Goldberg, & Sloman, 1980), which
97
places the infant at a higher risk for deficient socialemotional growth (Minde, 2000). Such a style may
result from the premature infant’s lower selfregulatory capacities, as well as from the decreased
opportunities for mother-infant touch and contact in
the first post-birth period (Field, 1996). Qualitative
studies involving interviews with mothers have also
revealed that their main, if only, purpose during this
time is to express milk. Mothers reported believing
that in the high tech environment of the NICU, this is
all they can do for their infant, leaving them feeling
quite helpless and unneeded (Harrison, 1993). In
addition to coping with their own feelings during this
time, parents are often required to make, or consent
to, serious treatment decisions, which may affect the
outcome of their newborns life.
Consequences of Poor Decisions
Poor decision-making processes may result in
unfavourable outcomes for the infant, parents, and
family as a whole. By examining the potential
consequences of poor decision processes for each of
the parties involved, a need for improvement in this
area will be highlighted.
Infant
For the infant, a poor decision-making process may
result in a worse outcome than necessary, an inability
to achieve his/her full potential, or a more distressing
death (Harrison, 1993). Preterm and low birth weight
infants are often at a high risk of long-term
neurological, cognitive, and behavioural problems
(Breslau, 1995). By the very nature of the situation,
infants are placed at the mercy of their parents in their
willingness and ability to respond, touch, and interact
(Swaney et al., 2002). Studies have found that mothers
of premature infants, unlike those of blind or deaf
infants, do not increase the level of touch to
compensate for their infant’s difficulties in
maintaining gaze and affective synchrony; to the
contrary, the early separation in the NICU decreases
proximity and touch (Weizman et al., 1999).
Disruptions in other components of the co-regulatory
system have also been observed, for example, the
emotional expressions of premature infants are often
unclear. Infants’ reduced self-regulation combined
with lower maternal contact may lead to the lower
level of synchrony observed between mothers and
premature infants, leading to long-term relationship
issues (Lester, Hoffman, & Brazelton, 1985). The
effects of maternal depression on infant cognitive and
neurological development and emotional selfregulation have also been the topic of thorough
investigation. The influence of maternal depression on
child outcomes has been thought to be due, in part, to
marital conflict. Children exposed to inter-parental
conflict have been shown to be vulnerable to
externalising and internalising problems, with infants
and toddlers expressing more insecure attachments
and older children displaying academic and social
difficulties (Tulley et al., 2004). In addition, the
irritability and agitation associated with depression
also may be reflected in mothers’ overt expressions of
negative affect, which are relatively rare phenomena
98
PHILLIPS & TOOLEY
in routine mother-baby interactions in nondepressed
dyads (Campbell, 1995).
Parents
For the parents, poor decision-making processes may
result in poor parental attachments, guilt, failure, and
regret that they had made a poor decision (Siegel et
al., 2002).
Parents of a newborn with a disability normally
experience lowered self-esteem and view this event as
an affront to their reproductive capabilities (Swaney
et al., 2002). More specifically, the mother views it as
a failure of her feminine role. Parents commonly
experience anticipatory grief when serious treatment
decisions are required, as the necessity to make lifeand-death decisions indicates to the parents that their
infant’s chances for survival are diminished.
Researchers have shown that the decision to transfer
an infant to a NICU alone is likely to initiate an
anticipatory grief reaction (Benfield, Leib, & Reutor,
1976; Siegel et al., 2002). Parents may also be
experiencing feelings of sadness over the loss of the
expected, idealized child that they had wished for
during the pregnancy. For some parents, attaching to a
critically ill infant may be too overwhelming; parents
may withdraw from the infant in an attempt to protect
themselves from their feelings of hurt, disappointment
and guilt (Bialoskurski, Cox, & Hayes, 1999; Miles,
Holditch-Davis, & Burchinal, 1999). Some parents
may feel ambivalent about the infant; they may feel
they could not love or cope with an infant who might
die or who would have significant physical or mental
problems. Feeling uncertain about whether they want
their infant to survive can cause overwhelming
feelings of guilt, shame and responsibility (Siegel et
al., 2002).
Not surprisingly, parents have reported that these
feelings have affected the development of their
relationship with their newborn. For example,
qualitative interviews with parents, who had decided
to do everything possible for the survival of their
handicapped infant without having all of the relevant
knowledge, have revealed that some mothers in
particular, being the main caregivers for their child,
were in strong favour of euthanasia. One mother said
“If I knew as I know now, I’d have euthanasiaed her.
It’s cruel for me and it’s cruel for her. There’s no life
for me while she’s here and none for her” (Kuhse &
Singer, 1985, p149).
Family
Clearly, in deciding the lengths to which resuscitation
efforts for their preterm/critically ill infant should go,
parents need to consider the potential impact a
handicapped child will have on their family and
themselves. Very important in any neonatal illness
and subsequent hospitalization is the disruption and
stress that is frequently created in the nuclear family
system. It has been demonstrated that the family’s
functioning and its adaptation to stress have important
effects on the family’s relationship with the infant and
the infant’s later development (Siegel et al., 2002).
Potentially, the survival of a handicapped premature
infant may require constant care and a great amount
of physical, cognitive, emotional and financial
resources (Kuhse & Singer, 1985). Empirical studies
of objectively measurable features of the lives of
families with, as compared to families without,
handicapped children, have found differences between
these groups including a higher rate of marital breakup and a decrease in the likelihood for further children
in the former. Siblings are also affected, in part
because their parents have to give so much of their
time and attention to the disabled member of the
family, and in part because they – especially if they
are sisters, are likely to be asked to do a good deal of
babysitting and behave in ways not appropriate to
their ages (Kew, 1984, as cited in Kuhse & Singer,
1985).
These findings have important implications for
furthering our understanding of how families function
during this time, as it is predominantly viewed that
families function as unitary systems and are best
understood and treated within a systemic perspective
(Feldman et al., 2003). In infancy, studies of family
interactions have shown important differences
between triadic mother-father-infant interactions and
parent-infant relatedness, emphasizing the need for a
fuller exploration of family-level processes. One
difficulty encountered in studying family-level
processes is the multiple sources of influences
impacting on the family system. Therefore, several
levels of direct and indirect influences need to be
considered. These include (a) the influence of each
individual on the behaviour of other individuals (e.g.
infant-to-parent) or dyads (infant-to-marital) in the
family; (b) the influence of each dyad (e.g. marital) on
the other dyads (e.g. parenting); and (c) the effects of
higher order process on individuals, dyads, and their
interrelationships in the family (Feldman et al., 2003).
Such influences may emerge through direct
interactions, imitation and modeling, or the effects of
other relationships. The quality of marital
relationships, be they satisfying or distressed, also
impacts on parenting behaviour, particularly the
father-child relationship (Parke & Beitel, 1988).
Within the context of triadic interactions, moments of
shared marital pleasures increase the quality of
parenting (Belsky, Crnic, & Gable, 1995), whereas
distressed marital relationships reduce the level of
sensitive parenting, particularly maternal sensitivity
(Feldman et al., 2003).
The systems perspective can also provide several
testable hypotheses on change in family systems. For
example, the systemic nature of the family dictates
that intervention targeting one relationship would also
affect other individuals, relationships, and the higher
order process (Feldman et al., 2003). To illustrate,
Cowan and Cowan (1992) found that intervention that
promotes marital closeness positively impacts the two
parenting systems. Furthermore, on the basis of the
dyadic-to-triadic influences, gains in one dyadic
system are expected to persist when these two
individuals interact within the triad. Finally, it is
likely that the specific gains for the target system
following intervention would be expressed in the
same domains in the other systems (Feldman et al.,
2003). For example, if intervention affects the degree
Parental decision-making and complicated births
of sensitivity or positive affect in marriage, those
same properties would also be enhanced in the two
parenting subsystems. Testing such hypotheses is not
only important for the evaluation of intervention
outcomes in infancy, but moreover, for a fuller
understanding of how family systems evolve and
function (Feldman et al., 2003).
Strategies to Facilitate Better Relationships
With the potential consequences of poor decisionmaking resulting in poorer outcomes for all involved,
it is paramount that clear and effective strategies be
employed to facilitate bonds between parents and
infants, and parents themselves.
Kangaroo care, or skin to skin contact, may
attenuate the negative effects of maternal separation
on the mother-child relationship, both in terms of
global sensitivity and in relation to micro-regulatory
patterns of gaze, affect, and touch (Feldman et al.,
2003). Research in human and animal models has
shown that touch and handling during periods of early
maternal separation had a positive impact on maternal
and infant behaviour (Weizman et al., 1999). The
provision of maternal-infant body contact during a
period of maternal separation has also been examined
for its effect on parent-infant and triadic interactions
(Feldman et al., 2003). In a study of three month old
preterm infants and their parents, these researchers
found that following kangaroo care, mothers and
fathers were more sensitive and less intrusive, infants
showed less negative affect, and family style was
more cohesive. In addition, maternal and paternal
affectionate touch of infant and spouse was more
frequent, spouses remained in closer proximity, and
infant proximity position was conducive to mutual
gaze and touch during triadic play (Feldman et al.,
2003).
In regards to the father-child relationship and the
family process, studies on co-parenting have found
associations between mothers’ and fathers’ behaviour
toward their children, which is thought to lay the
foundation for the family process (Belsky et al.,
1995). The researchers thus hypothesized that
improvements
in
maternal
sensitivity
and
intrusiveness following kangaroo care may also be
observed between father and child in the same
domains.
Maternal warmth has also been the topic of
investigation in studies of low birth weight infants
and their developmental outcomes. For example, in a
study investigating the moderating effect of maternal
warmth on the association between low birth weight
and children’s attention-deficit/hyperactivity disorder
(ADHD) symptoms and low IQ, Tully and colleagues
(2004) found a significant interaction between
children’s birth weight and maternal warmth in
predicting mothers’ and teachers’ ratings of ADHD,
but no significant interaction for IQ. These findings
suggest that the effect of children’s birth weight on
their ADHD symptoms can be moderated by maternal
warmth and that enhancing maternal warmth may
prevent behaviour problems among the increasing
population of low birth weight children (Tully et al.,
2004). These results support the findings of earlier
99
longitudinal research showing a moderating effect for
maternal emotional responsivity to their low birth
weight children (Laucht, Esser, & Schmidt, 2001).
The findings of the aforementioned studies suggest
that parents’ emotional attitudes may also be an
important factor in the development of child
psychopathology. Researchers have found that
mothers of children with behavioural and emotional
disorders express more critical comments, fewer
positive comments, and less warmth to their children
than control parents (Scott & Campbell, 2001). As
maternal warmth is important for children’s
competence,
behaviour,
and
development
(MacDonald, 1992), high levels of maternal warmth
may be critical for low birth weight children and may
moderate the risk of long-term cognitive and
behavioural difficulties (Tully et al., 2004). In
addition, some researchers have found that the effects
of low birth weight may be greater in twins than in
singletons, due to the evidence that parents of twins
may experience greater stress, depression, and marital
difficulties, owing to the burden of having two
children
simultaneously
(Thorpe,
Golding,
MacGillivray, & Greenwood, 1991). Thus, such
strategies to improve infant and family relationships
could be applicable to twins and singletons alike.
Future Research Directions
Given the current state of the literature on the impact
of birth complications on parent-infant and infantmarital relationships, coupled with the frequency of
such complications, it is clear that advances in this
area are required. As previous attempts to aid parents
during or immediately before labour have been
limited at best, a key area that remains
underdeveloped at present is the education and
preparation of couples for serious decision-making.
Couples could, if willing, participate in an educative
intervention program that would focus on helping
parents clarify their involvement in the decision
process, and familiarize them with common
procedures and events following birth complications.
If complications during birth do occur, couples may
then be aware of how they feel about different
decision outcomes, and may be less affected by the
time constraints and emotional responses commonly
associated with such decision tasks.
Alternatively, research in this area may be useful in
helping to guide clinical staff in the way they assist
parents to negotiate not only the decision processes,
but also care involvement if complications do occur.
Perhaps by providing parents with the knowledge of
previous research findings about kangaroo care and
maternal sensitivity, parents would be more aware of
the options they have in their infants care and the
benefits, for all parties involved, of touch, gaze, and
interactions. The induction of strategies to facilitate
strong maternal and paternal attachment would result
in the improvement of relationships between family
members, and would most likely minimize the
negative effects of the NICU environment on
relationships and later infant development.
PHILLIPS & TOOLEY
100
References
Belsky, J., Crnic, K., & Gable, S. (1995). The
determinants of coparenting in families with toddler
boys: Spousal differences and daily hassles. Child
Development, 66, 629-642.
Benfield, D. G., Leib, S., & Reutor, J. (1976). Grief
response to parents after referral of the critically ill
newborn a regional centre. New England Journal of
Medicine, 294, 975-986.
Bialoskurski, M., Cox, C., & Hayes, J. (1999). The
nature of attachment in a neonatal intensive care
unit. Journal of Perinatal and Neonatal Nursery,
13, 66-72.
Brachfeld, S., Goldberg, S., & Sloman, J. (1980).
Parent-infant interaction in free play at 8 and 12
months: Effects of prematurity and immaturity.
Infant Behavior and Development, 3, 289-305.
Breslau, N. (1995). Psychiatric sequelae of low birth
weight. Epidemiologic Reviews, 17, 96-106.
Cowan, P., & Cowan, C. (1992). When partners
become parents: The big life change for couples.
New York: Basic Books.
Feldman, R., Weller, A., Sirota, L., & Eidelman, A. I.
(2003). Testing a family intervention hypothesis:
The contribution of mother-infant skin-to-skin
contact (kangaroo care) to family interaction,
proximity, and touch. Journal of Family
Psychology, 17, 94-107.
Field, T.M. (1996). Attachment and separation in
young children. Annual Review of Psychology, 47,
541-561.
Gardner, M. O., Goldberg, R. L., Cliver, S. P.,
Tucker, J. M., Nelson, K. G., & Copper, R. L.
(1995). The origin and outcome of preterm twin
pregnancies. Obstetrics and Gynecology, 85, 553557.
Harrison, H. (1993). The principles for family-centred
neonatal care. American Academy of Pediatrics, 92,
643-650.
Hofer, M. A. (1995). Hidden regulators: Implications
for a new understanding of attachment, separation,
and loss. In S. Golberg, R. Muir, & J. Kerr (Eds.),
Attachment theory: Social, developmental, and
clinical perspectives (pp. 203-230). Hillsdale, NJ:
Analytic Press.
Kitchen, W. H., Ryan, M. M., Rickards, A. L., &
Doyle, L. W. (1998). Premature babies: A guide for
parents. Melbourne: Hill of Content Publishing.
Kuhse, H., & Singer, P. (1985). Should the Baby
Live? The Problem of Handicapped Infants. New
York: Oxford University Press.
Lau, R., & Morse, C. A. (2003). Stress experiences of
parents with premature infants in a special care
nursery. Stress and Health, 19, 69-78.
Laucht, M., Esser, G., Schmidt, M. (2001).
Developmental outcome of infants born with
biological and psychosocial risks. Journal of Child
Psychiatry, 38, 843-853.
Lester, B. M., Hoffman, J., & Brazelton, T. B. (1985).
The rhythmic structure of mother-infant interaction
in term and preterm infants. Child Development, 56,
15-27.
MacDonald, K. (1992). Warmth as a developmental
construct: An evolutionary analysis. Child
Development, 63, 753-773.
Miles, M., Holditch-Davis, D., & Burchinal, P.
(1999). Distress and growth in mothers of medically
fragile infants. Nursing Research, 48, 129-138.
Minde, K. (2000). Prematurity and serious medical
conditions
in
infancy:
Implications
for
development, behavior, and intervention. In C. H.
Zeanah (Ed.), Handbook of infant mental health
(pp. 176-194). New York: Guilford Press.
Parke, R. D., & Beitel, A. (1988). Disappointment:
When things go wrong in the transition to
parenthood. Marriage and Family Review, 12, 221265.
Riley, M., & Halliday, J. (2001). Births in Victoria
1999-2000. Melbourne: Department of Human
Services.
Scott, S., & Campbell, C. (2001). Expressed emotion
about children: Reliability and validity of a
Camberwell Family Interview for Childhood (CFIC). International Journal of Methods in Psychiatric
Research, 9, 3-10.
Siegel, R., Gardner, S. L., & Merenstein, G. B.
(2002). Families in crisis: theoretic and practical
considerations. In G. B. Merenstein, & S. L.
Gardner (Eds.), Handbook of neonatal intensive
care (pp. 725-753). Missouri: Mosby.
Sudia-Robinson, T. M., & Freeman, S. B. (2000).
Communication patterns and decision making
among parents and health care providers in the
neonatal intensive care unit: A case study. The
Journal of Acute and Critical Care, 29,143-148.
Swaney, J., English, N., & Carter, B. S. (2002). Ethics
in neonatal intensive care. In G. B. Merenstein, &
S. L. Gardner (Eds.), Handbook of neonatal
intensive care (pp. 801-821). Missouri: Mosby.
Thorpe, K., Golding, J., MacGillivray, I., &
Greenwood, R. (1991). Comparison of prevalence
depression in mothers of twins and mothers of
singletons. British Medical Journal, 302, 875-878.
Tully, L. S., Arseneault, L, Capsi, A., Moffitt, T. E.,
& Morgan, J. (2004). Does maternal warmth
moderate the effects of birth weight on twins’
attention-deficit hyperactivity disorder (ADHD)
symptoms and low IQ? Journal of Consulting and
Clinical Psychology, 72, 218-226.
Ventura, S. J., Martin, J. A., Curtin, S. C., Menacker,
P. H., & Hamilton, B. E. (2001). Births: Final data
for 1999. National Vital Statistics Reports, 49, 1100.
Weizman, R., Lehmann, J., Leschiner, S., Allman, I.,
Stoehr, T., Heidbreder, C., Domeney, A., Feldon, J.,
& Gavish, M. (1999). Long-lasting effects of early
handling on the peripheral benzodiazepine receptor.
Pharmacology, Biochemistry, and Behavior, 64,
725-729.
Parental decision-making and complicated births
Zupancic, J. A., Kirpalani, H., Barrett, J., Stewart, S.,
Gafni, A., Streiner, D., et al. (2002). Characterising
doctor-patient communication in counseling for
impending preterm delivery. Archives of Disease in
Childhood, 87, 113-117.
Correspondence to:
Sarah Phillips
School of Psychology, Deakin University
221 Burwood Hwy, Burwood VIC 3125
sjph@deakin.edu.au
101
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 102-108
Attachment Security and Intimate Relationship Satisfaction: The
Mediational Role of Maladaptive Attributions
Zoë J. Pearce (z.pearce@griffith.edu.au)
School of Applied Psychology
Griffith University – Mt Gravatt Campus, Brisbane, QLD 4111 Australia
W.K. Halford (k.halford@griffith.edu.au)
School of Applied Psychology
Griffith University – Mt Gravatt Campus, Brisbane, QLD 4111 Australia
Hutchinson, 1993). Current relationship functioning
has been strongly related to attributions (about
negative behaviour) towards the partner causing and
considered blameworthy for the behaviour (Fincham
& Bradbury, 1987), with attributions of negative
intent more likely to occur when relationships are
unstable, distressed, or close to separation (Bradbury
& Fincham, 1990; Fletcher, Fincham, Cramer, &
Heron, 1987). Gallo and Smith (2001) found anxious
attachment was a strong predictor of perceptions of
marital support and conflict, and in some cases this
relationship was mediated by attribution style. This
raises the possibility that attributions might mediate
the association between attachment and relationship
satisfaction.
Of further interest is the question of whether
attributions for negative partner behaviour can
mediate attachment and relationship satisfaction
across different relationships, representing a
consistent pattern of cognitive processing. The
earliest studies in social perception support the
presence of consistency using real and imagined
targets (Burron, Carlson, Getty, & Jackson, 1971).
Further support comes from Baldwin’s (1992) theory
of relational schemas, cognitive structures for patterns
of interpersonal interactions that are cognitions about
relationships, and not merely self and other in
isolation. The presence of globality—the extent to
which the cause for behaviour is perceived to be
typical across a number of situations or particular to
just that instigating event—in conjunction with stable
attachment insecurity would predict the presence of a
consistent cognitive bias. In this case, a global
attribution style would consistently mediate
attachment regardless of the relationship in question,
thus the behaviour of current and potential intimate
partners would be interpreted in the same manner.
The present study investigated whether attributions
mediate the relationship between attachment and
relationship satisfaction. Consistent with early
attachment research it was hypothesised that avoidant
and anxious attachment would be associated with
relationship dissatisfaction, and that this relationship
would be mediated by maladaptive attributions for
negative partner behaviour. Based on the widely made
assumption that attachment is a stable personal
characteristic (Collins, 1996; Collins & Read, 1990;
Feeney & Noller, 1990; Gallo & Smith, 2001;
Simpson, 1990) it was also predicted that attributions
about the behaviour of a hypothetical future partner’s
Abstract
Attachment insecurity is a well-established predictor of
low couple relationship satisfaction. In this study we
tested the hypothesis that negative attributions mediate
this association. Both partners in 59 heterosexual
couples completed self-report measures of attachment,
attributions in their current relationship and
relationship satisfaction, and watched video vignettes
of a potential partner and rated their attributions for
that potential partner’s behaviour. As predicted,
negative attributions mediated between insecure
attachment and relationship satisfaction in the current
relationship, but not with the potential partner.
Attributions seem to be a relationship- specific process
rather than a general interpersonal attribute.
Hazan and Shaver’s (1987) influential studies showed
that the formation and maintenance of adult romantic
relationships can be conceptualised as an attachment
process analogous to the development of infant–
caregiver relationships (Bowlby, 1977). Later
research demonstrated a robust association between
attachment and relationship satisfaction (Collins &
Read, 1990; Feeney & Noller, 1990; Simpson, 1990).
However, there is a dearth of research investigating
the processes through which attachment and
satisfaction are related to one another.
Adult attachment is argued to reflect the
representation that individuals have of themselves and
others as either positive or negative. Recent
approaches (e.g. Brennan, Clark, & Shaver, 1998)
classify attachment insecurity across two dimensions
– avoidance, and anxiety. While securely attached
individuals are comfortable depending on others and
having others depend upon them, anxious styles are
characterised by the desire to be closer than others
would like and worry that their partners do not love
them or will leave them; and avoidant styles desire to
avoid close involvement and have a negative
expectation of others. Attachment style can change in
response to new relationship experiences, but is
generally assumed to be formed by early experiences
and be relatively stable across time (Collins & Read,
1994).
While psychodynamic in origin, attachment can be
explained in terms similar to the cognitive
components of attribution theory (Fletcher &
Fincham, 1991). Attribution theorists hypothesise that
couples have biases in the explanations they make for
partner
behaviour
(Holtzworth-Munroe
&
102
Attachment & attributions
negative behaviour would mediate between
attachment and current relationship satisfaction.
Method
Participants
Fifty-nine couples were recruited via an
advertisement placed in a metropolitan newspaper for
a study of relationship communication. To be
included in the study participants had to be in a
heterosexual relationship for a minimum of six
months, and had to spend a minimum of four nights
per week together. The mean duration of relationship
was 11.7 years (SD = 10.3). The mean age for males
was 41.4 years (SD = 10.9), and for females was 38.1
years (SD = 10.8). Approximately half of the couples
(n = 33) were married, and the remaining couples
were defacto (n = 26). The majority of males (n = 47)
were in full-time employment with 24 females in fulltime employment.
Measures
Attachment The Experiences in Close Relationships
(ECR) is a 36-item self-report measure of adult
attachment (Brennan et al., 1998) consisting of two
18-item scales, Avoidance and Anxiety. Participants
rate the extent to which they agree with each item on
a 7-point likert scale with higher scores reflecting
more avoidant or anxious attachment. There is high
internal consistency for both Avoidance (α = .94) and
Anxiety (α = .91), and the scales of the ECR correlate
significantly with both Collins and Read (1990) and
Griffin and Bartholomew’s (1994) measures.
Attributions The Relationship Attribution Measure
(RAM) is a 60-item measure of attributions for ten
hypothetical spousal behaviours (Fincham &
Bradbury, 1992). Respondents rate on a 6-point likert
scale the extent to which they agree with six
statements for each of the eight negative behaviours
and two positive behaviours (included as filler items).
Derived scores are the extent to which the partner is
seen as causing the negative behaviour, and is
responsible or blameworthy for the behaviour. Higher
scores represent a greater tendency to make negative
attributions of this nature. The RAM displays good
internal consistency for both scales.
Relationship Satisfaction The Dyadic Adjustment
Scale (DAS; Spanier, 1976) is a 32-item assessment
of couple relationship satisfaction. The DAS is a
widely used measure in relationship research, with
high scores indicating greater levels of relationship
satisfaction. The scoring results in the calculation of a
total adjustment score that is highly reliable (α = .96).
Furthermore, the DAS distinguishes between married
and divorced couples, and correlates highly (r = .86)
with other measures of dyadic adjustment (Spanier,
1976).
Potential Partner Partner attributions were assessed
in response to behaviours of a hypothetical potential
partner. A Potential Partner video was created using
actors and consisting of headshots of the actor making
103
12 different statements, chosen to prompt attribution
processes. The statements spanned three emotional
valences (four positive, four negative, and four
ambiguous) and were designed to be potentially
relevant to a cross-section of relationships.
Preliminary testing of the video showed it to be a
valid and reliable measure. Attributions made in
response to video statements were assessed using a
modified version of the RAM that specifically
targeted each video statement.
Procedure
This study formed part of a larger research project
conducted as partial requirement for the first author’s
Doctor of Philosophy degree. Couples responding to
the advertisement were sent an ethical consent form
and information sheet outlining the broad goals of the
study and explaining what to expect in the interviews.
All interviews took place at the Psychology Clinic,
Griffith University, Brisbane, Australia. Couples
participated in two 2-hour interviews during which
time they completed the questionnaires and Potential
Partner protocol. An information sheet listing local
counselling providers was made available should
couples wish to seek further assistance. Each couple
received $50.00 payment for participation.
Results
The within-gender correlations between the key
variables are shown in Table 1. While significant,
many of these correlations were moderate at best.
Table 1 shows that, for males, there was a small but
significant association between attachment style and
relationship satisfaction, where greater levels of
insecurity were related to reports of lower
satisfaction. A similar small but significant
association was found between attachment insecurity
and causal attributions but this result was not
replicated with responsibility-blame attributions.
Finally, attributions of both dimensions were
moderately associated with relationship satisfaction
where making more negative attributions was
associated with reports of lower relationship
satisfaction. For females, attachment insecurity was
associated with relationship dissatisfaction, but the
correlations were low to moderate at best. A low
strength of association was also evident between
attachment and attributions but on the whole greater
attachment insecurity was significantly associated
with making more negative attributions. Finally, there
was a moderate association between attributions and
relationship satisfaction, where making more negative
attributions was associated with reports of lower
relationship satisfaction. Additional between-gender
correlations were conducted separately and significant
associations were found between female avoidant
attachment and male causal attributions (r = .29, p <
.05), and female anxious attachment and male
responsibility-blame attributions (r = .27, p < .05).
PEARCE & HALFORD
104
Table 1
Correlations for Questionnaire Measures
Measure
1. Avoidance
2. Anxiety
3. Cause
4. Resp/Blame
5. DAS
1.
.24
.20
.22
.33*
-.48**
2.
.43**
-.05
.32*
.37**
-.31**
3.
.28*
.37**
.31*
.70**
-.55**
4.
.12
.14
.68**
.33**
-.64**
5.
-.31*
-.38**
-.59**
-.52**
.57**
Note. N = 59. Correlations for male participants are presented above the diagonal; correlations for female participants are presented below
the diagonal. Between-gender correlations are presented on the diagonal in bold text. Avoidance = Experiences in Close Relationships
Avoidance scale; Anxiety = Experiences in Close Relationships Anxiety scale; Cause = Relationship Attribution Measure Cause scale;
Resp/Blame = Relationship Attribution Measure Responsibility-Blame scale; DAS = Dyadic Adjustment Scale.
*p < .05, **p < .01
Tests of mediation were conducted using
hierarchical regressions (Baron & Kenny, 1986;
Holmbeck, 2002; MacKinnon, Lockwood, Hoffman,
West, & Sheets, 2002; Shrout & Bolger, 2002) on
variable pathways that showed significant IV-MV and
MV-DV correlations. The significance of these
analyses was tested via the Freedman-Schatzkin test,
as suggested by MacKinnon et al. (2002). Initial
regression analyses with relationship satisfaction as
the criterion and attachment as predictor are presented
in Table 2. Male attachment accounted for 17% of the
variance in male relationship satisfaction and when
female attachment was entered at Step 2 it accounted
for 20% additional unique variance. Female
attachment accounted for 27% of the variance in
female relationship satisfaction, and male attachment
accounted for 13% additional unique variance. Table
2 also shows that avoidant attachment in males was
not a significant predictor for their own or their
partner’s relationship satisfaction. Conversely, female
avoidance significantly predicted male relationship
satisfaction. Anxious attachment in males was
predictive of female satisfaction but had no impact on
their own relationship satisfaction, whereas female
anxiety was the best predictor of their own and their
partner’s relationship satisfaction.
Table 2
Regression Analysis Summary – Relationship Satisfaction as Criterion
Predictor
B
SEB
β
DV = Male Relationship Satisfaction
Step 1
Avoidance
Anxiety
-3.62
-5.65
2.77
2.46
-.18
-.31*
Avoidance
Anxiety
Avoidance (P)
Anxiety (P)
-1.88
-3.78
-6.55
-4.89
2.51
2.44
2.42
2.02
-.09
-.21
-.34**
-.27*
Step 2
ΔR2
F
.17
5.84**
.20
8.09***
.27
10.56***
.13
8.95***
DV = Female Relationship Satisfaction
Step 1
Avoidance
Anxiety
-8.44
-3.90
2.27
2.08
-.43***
-.22
Avoidance
Anxiety
Avoidance (P)
Anxiety (P)
-4.76
-4.63
-1.97
-6.42
2.39
2.00
2.48
2.41
-.24
-.26*
-.10
-.35**
Step 2
Note. N = 59. Relationship satisfaction = Dyadic Adjustment Scale Total Score; Avoidance = Experiences in Close Relationships Avoidance
scale; Anxiety = Experiences in Close Relationships Anxiety scale; (P) = partner’s measure of this scale.
*p < .05. **p < .01. ***p < .001.
Subsequent steps in the mediation analysis were
conducted on the data for attachment style, attribution
style, and relationship satisfaction following
standardisation using z-score transformations. The
results are presented in Table 3. For males, more
causal attributions mediated the association of
avoidant attachment behaviour in their partner and
lower relationship satisfaction, and greater use of
responsibility-blame attributions mediated between
anxious female attachment and lower relationship
satisfaction. For females, Table 3 suggests the
association of females’ own anxious attachment on
their relationship satisfaction was mediated by both
causal and responsibility-blame attributions. The
Attachment & attributions
association of male anxious attachment on female
relationship satisfaction was mediated by female
responsibility-blame attributions. In all but a single
case full mediation was supported, and between 28
and 75 percent of the effect of attachment on
relationship satisfaction was mediated by attribution
style. Overall, the results show that attributions do
mediate between attachment and relationship
satisfaction.
Within-gender correlations for the Potential
Partner attachment style, attributions and relationship
satisfaction appear in Table 4. As the significant
correlations between attachment and relationship
105
satisfaction were already illustrated in Table 1, the
key relationship of interest in Table 4 is how
attachment relates to Potential Partner attributions and
how they, in turn, relate to relationship satisfaction.
Male attachment style was not significantly correlated
with any of the male attribution scales, nor was
female attachment security significantly correlated
with any of the female attribution scales. Similarly
none of the between-gender correlations of interest
were significant. Further analyses were not
conducted, as the key variables were not related to
one another.
Table 3
Tests of Mediation on Significant Standardised Regression Paths
α
β
τ
τ'
DAS
.29*
DAS
.27*
.51***
-.35**
-.37**
Resp/Blame
.49***
.46***
Females
Anxiety
Cause
DAS
.32*
Anxiety
Resp/Blame
DAS
Anxiety (P)
Resp/Blame
DAS
.37*
*
.28*
.50***
.61***
.59***
IV
Males
Avoidance
(P)
Anxiety (P)
MV
DV
Cause
Mediation Results
Type
Test
%
Partial
-4.14***
27.89
-.23
Full
-3.64***
35.32
-.31*
-.15
Full
-3.99***
51.83
-.31*
-.08
Full
-4.80***
74.55
-.35**
-.18
Full
-4.36***
47.54
Note. α = beta weight of IV to MV (MV regressed on to IV); β = beta weight of MV to DV controlling for IV, τ = IV to DV without MV, τ‘
= IV to DV with MV. Avoidance = Experiences in Close Relationships Avoidance scale; Anxiety = Experiences in Close Relationships
Anxiety scale; (P) = partner’s measure of this scale; Cause = Relationship Attribution Measure Cause scale; Resp/Blame = Relationship
Attribution Measure Responsibility-Blame scale; DAS = Dyadic Adjustment Scale.
* p < .05. ** p < 0.01. *** p < .001
Discussion
The first hypothesis was supported, where negative
attributions about partner behaviour mediated
attachment insecurity and relationship dissatisfaction.
Contrary to the prediction of hypothesis two, negative
attributions about a potential partner’s behaviour did
not mediate attachment insecurity and relationship
dissatisfaction.
The finding in the present study that attachment
security significantly predicts relationship satisfaction
largely replicates the well-documented association
between adult romantic attachment style and
relationship outcome (Collins, 1996; Collins &
Feeney, 2000; Collins & Read, 1990; Feeney, 1994;
Simpson, 1990). The fact that higher attachment
anxiety in women was inversely associated with
relationship satisfaction in male partners is supportive
of earlier research (Collins & Read, 1990; Kirkpatrick
& Davis, 1994; Kirkpatrick & Hazan, 1994; Simpson,
1990) and may represent the stereotypical
possessiveness (Davis & Oathout, 1987) and jealousy
(Sharpsteen & Kirkpatrick, 1997) often identified in
women. The finding that female avoidant attachment
was predictive of lower male relationship satisfaction
is somewhat contradictory to previous research, but
may be due again to gender stereotypes. Females are
stereotyped as being more nurturing of intimate
relationships and more comfortable with intimacy
(Eagly & Crowley, 1986) compared to males.
Avoidance in females, and anxiety in males, would
tend to go against social stereotypes, producing a
greater degree of concern in their partners, with both
gender effects potentially relevant to the current data.
As a whole, however, avoidant attachment was only
weakly associated with relationship satisfaction,
suggesting avoidant attachment may be a less
important predictor than anxious attachment (Gallo &
Smith, 2001).
The unique contribution of the current study is that
it provides fairly convincing evidence that
maladaptive attributions for negative partner
behaviour do mediate between attachment and
relationship satisfaction. Working models of
attachment have been found by Collins (1996) to bias
the attribution process, with insecure participants
providing more negative attributions for hypothetical
relationship events compared to secure participants.
Data from the present study suggest Collins’ findings
have been supported, albeit with modest correlations.
PEARCE & HALFORD
106
Table 4
Correlations for Potential Partner Protocol
Measure
1. Avoidance
2. Anxiety
3. Cause_PP neg
4. Resp/Blame_PP neg
5. Cause_PP ambig
6. Resp/Blame_PP
ambig
7.
DAS
1.
2.
3.
4.
5.
6.
.24
.20
.05
.12
-.04
.43**
-.05
-.09
.04
.01
.03
.20
.18
.77**
.67**
-.09
-.04
.47**
-.05
.64**
.05
.15
.62**
.39**
.26
.01
.03
.46**
.56**
.60**
.06
.22
.57**
.80**
.73**
.09
-.31**
-.08
-.28*
-.07
-.28
.48**
7.
-.31*
-.38**
-.03
.11
-.03
.15
.57**
Note. N = 58. Correlations for male participants are presented above the diagonal; correlations for female participants are presented below
the diagonal. Between-gender correlations are presented on the diagonal in bold text. Avoidance = Experiences in Close Relationships
Avoidance scale; Anxiety = Experiences in Close Relationships Anxiety scale; Cause_PP = Potential Partner using RAM Cause scale;
Resp/Blame_PP = Potential Partner using RAM Responsibility-Blame scale; DAS = Dyadic Adjustment Scale.
*p < .05, **p < .01
In perhaps the only empirical study that has
investigated both attachment style and attributions in
married couples, Gallo and Smith’s (2001) research
found attachment to be predictive of marital
functioning in a sample of 57 married psychology
undergraduate students. Furthermore, attributions of
negative intent, as assessed by the Relationship
Attribution Measure, were found to partially mediate
this association. Yet a number of distinct differences
between Gallo and Smith’s research and the current
study should be pointed out. By their own admission,
Gallo and Smith suspect their use of the Adult
Attachment Scale (Collins & Read, 1990) to measure
attachment was somewhat problematic and the
authors instead suggest utilising the ECR in future
research. Secondly, the use of a college sample,
despite the marital status, may also limit comparing
their results to those presented here, as differing
recruitment techniques and sample characteristics can
reduce generalisability (Amato & Keith, 1991;
Karney et al., 1995). Of course, the sample used in the
present study is also limited in terms of how
generalisable the results are. Our sample was largely
an older, long-term committed, middle-class sample,
and caution should be applied before generalising
these results too far. Nevertheless, it appears that
while earlier research has provided tentative support
for a link between attachment and attributions, the
mediational nature of attributions specific to
attachment (as measured by the ECR) and relationship
satisfaction, in a community sample of couples, is a
unique result.
The lack of results for the potential partner scenario
contradicts the global cognitive bias hypothesis. The
mediation of insecure attachment and relationship
satisfaction by attributions appears to be dependent on
the relationship, not on the individual making the
attributions. Is this result due to methodological
concerns or a genuine effect? The idea of asking
people to consider a complete stranger as a potential
intimate partner does require a fair degree of
imagination, even though early social perception
research supports this methodological technique
(Burron et al., 1971). It would be easy to assume that
the potential partner video was unable to elicit
attributions to the same successful degree as the
Relationship Attribution Measure, or asked
participants to engage in unimaginable situations. Yet
alpha coefficients for each measure were comparable
and both measures were modestly correlated with one
another suggesting convergent validity was not a
reason for the discrepant results. Additionally,
participants’ vocalisations during the potential partner
scenario suggested they had little difficulty
performing the required imagination task.
Previous attachment research that has used both real
and imagined partners has found an attachment style
difference (Collins, 1996) where avoidant attachment
differed across conditions, with interpretations for real
partner behaviour reported as less negative and
anxious attachment returning similarly negative
interpretations across both conditions. Converse to
Collins’ results, anxious attachment here was not
strongly related to the outcome variable in the
hypothetical partner condition. Thus while it’s
possible that the use of different stimuli can produce
different results, this may not be the only reason for
this result.
While contrary to predictions, this finding is
interesting in that it suggests attachment security and
attributions combine to produce a different pattern of
responding for different people. While attachment has
traditionally be considered a global individualdifference variable, there is growing support for reconsidering attachment as a within-person difference
variable (La Guardia, Ryan, Couchman, & Deci,
2000). The lack of support for the potential partner
scenario in the present research provides support for
some of the more recent attachment literature that
suggests people possess models of self and other that
are relationship-specific in addition to a global,
generalised attachment model (Pierce & Lydon,
2001). So rather than having a single global
attachment style that is generalisable to impact
similarly on all relationships, the lack of support for
the potential partner scenario lends support to the
notion that people possess different attachment styles
for different relationships, or different types of
relationships. A more recent interpretation of
Baldwin’s (1992) relational schema theory suggests
Attachment & attributions
multiple schemas exist for multiple relationships, a
fact that certainly appears to be supported in the
present study (Pierce & Lydon, 2001).
Finally, it is suggested here that future research
would benefit from adding an affective component to
the mediational model to account for these
differences. The potential partner scenario provided
participants with little opportunity to emotionally
engage in the task and evidence in favour of the role
that affect plays in both attachment (Feeney, Noller,
& Roberts, 1998; Mikulincer, 1998; Rowe &
Carnelley, 2003), and cognitive processing (Forgas,
1998; Teasdale, Lloyd, & Hutton, 1998) tends to
support including this variable in research of this
nature, the absence of which is arguably a limitation
of the present study. Thus while cognitive biases may
be an important mediator in current intimate
relationships, generalising this result to all potential
partnerships might prove difficult.
References
Amato, P. R., & Keith, B. (1991). Parental divorce
and adult well-being: A meta-analysis. Journal of
Marriage and the Family, 53, 43-58.
Baldwin, M. W. (1992). Relational schemas and the
processing of social information. Psychological
Bulletin, 112, 461-484.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable distinction in social psychological
research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social
Psychology, 51, 1173-1182.
Bowlby, J. (1977). The making and breaking of
affectional
bonds:
I.
Aetiology
and
psychopathology in the light of attachment theory.
British Journal of Psychiatry, 130, 201-210.
Bradbury, T. N., & Fincham, F. D. (1990).
Attributions in marriage: Review and critique.
Psychological Bulletin, 107, 3-33.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998).
Self-report measurement of adult attachment: An
integrative overview. In J. A. Simpson & W. S.
Rholes (Eds.), Attachment theory and close
relationships (pp. 44-76). New York: Guilford.
Burron, B. F., Carlson, K. A., Getty, G. R., &
Jackson, D. N. (1971). The effects of informational
characteristics on the perception of real and
hypothetical target persons. Psychonomic Science,
23, 145-147.
Collins, N. L. (1996). Working models of attachment:
Implications for explanation, emotion, and
behavior. Journal of Personality & Social
Psychology, 71, 810-832.
Collins, N. L., & Feeney, B. C. (2000). A safe haven:
An attachment theory perspective on support
seeking and caregiving in intimate relationships.
Journal of Personality and Social Psychology, 78,
1053-1073.
Collins, N. L., & Read, S. J. (1990). Adult
attachment, working models, and relationship
quality in dating couples. Journal of Personality
and Social Psychology, 58, 644-663.
Collins, N. L., & Read, S. J. (1994). Cognitive
representations of attachment: The structure and
107
function of working models. In K. Bartholomew &
D. Perlman (Eds.), Attachment processes in
adulthood (Vol. 5, pp. 53-90). London: Jessica
Kingsley.
Davis, M. H., & Oathout, A. (1987). Maintenance of
satisfaction in romantic relationships: Empathy and
relational competence. Journal of Personality &
Social Psychology, 53, 397-410.
Eagly, A. H., & Crowley, M. (1986). Gender and
helping behavior: A meta-analytic review of the
social psychological literature. Psychological
Bulletin, 100, 283-308.
Feeney, J. A. (1994). Attachment style,
communication patterns and satisfaction across the
life cycle of marriage. Personal Relationships, 1,
333-348.
Feeney, J. A., & Noller, P. (1990). Attachment style
as a predictor of adult romantic relationships.
Journal of Personality and Social Psychology, 58,
281-291.
Feeney, J. A., Noller, P., & Roberts, N. (1998).
Emotion, attachment, and satisfaction in close
relationships. In P. A. Andersen & L. K. Guerrero
(Eds.), Handbook of communication and emotion:
Research, theory, applications, and contexts (pp.
473-505). San Diego: Academic Press.
Fincham, F. D., & Bradbury, T. N. (1987). The
impact of attributions in marriage: A longitudinal
analysis. Journal of Personality and Social
Psychology, 53, 510-517.
Fincham, F. D., & Bradbury, T. N. (1992). Assessing
attributions in marriage: The Relationship
Attribution Measure. Journal of Personality and
Social Psychology, 62, 457-468.
Fletcher, G. J., Fincham, F. D., Cramer, L., & Heron,
N. (1987). The role of attributions in the
development of dating relationships. Journal of
Personality and Social Psychology, 53, 481-489.
Fletcher, G. J. O., & Fincham, F. D. (1991).
Attribution processes in close relationships. In G. J.
O. Fletcher & F. D. Fincham (Eds.), Cognition in
close relationships (pp. 7-35). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Forgas, J., P. (1998). On being happy and mistaken:
Mood effects on the fundamental attribution error.
Journal of Personality and Social Psychology, 75,
318-331.
Gallo, L. C., & Smith, T. W. (2001). Attachment style
in marriage: Adjustment and responses to
interaction. Journal of Social & Personal
Relationships, 18, 263-289.
Griffin, D. W., & Bartholomew, K. (1994). The
metaphysics of measurement: The case of adult
attachment. In K. Bartholomew & D. Perlman
(Eds.), Attachment processes in adulthood (Vol. 5,
pp. 17-52). London: Jessica Kingsley.
Hazan, C., & Shaver, P. (1987). Romantic love
conceptualized as an attachment process. Journal of
Personality and Social Psychology, 52, 511-524.
Holmbeck, G. N. (2002). Post-hoc probing of
significant moderational and mediational effects in
studies of pediatric populations. Journal of
Pediatric Psychology, 27, 87-96.
108
PEARCE & HALFORD
Holtzworth-Munroe, A., & Hutchinson, G. (1993).
Attributing negative intent to wife behavior: the
attributions of maritally violent versus nonviolent
men. Journal of Abnormal Psychology, 102, 206211.
Karney, B. R., Davila, J., Cohan, C. L., Sullivan, K.
T., Johnson, M. D., & Bradbury, T. N. (1995). An
empirical investigation of sampling strategies in
marital research. Journal of Marriage and the
Family, 57, 909-920.
Kirkpatrick, L. A., & Davis, K. E. (1994). Attachment
style, gender, and relationship stability: A
longitudinal analysis. Journal of Personality and
Social Psychology, 66, 502-512.
Kirkpatrick, L. A., & Hazan, C. (1994). Attachment
styles and close relationships: A four-year
prospective study. Personal Relationships, 1, 123142.
La Guardia, J. G., Ryan, R. M., Couchman, C. E., &
Deci, E. L. (2000). Within-person variation in
security of attachment: A self-determination theory
perspective on attachment, need fulfillment, and
well-being. Journal of Personality & Social
Psychology, 79, 367-384.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M.,
West, S. G., & Sheets, V. (2002). A comparison of
methods to test mediation and other intervening
variable effects. Psychological Methods, 7, 83-104.
Mikulincer, M. (1998). Adult attachment style and
individual differences in functional versus
dysfunctional experiences of anger. Journal of
Personality and Social Psychology, 74, 513-524.
Pierce, T., & Lydon, J. E. (2001). Global and specific
relational models in the experience of social
interactions. Journal of Personality & Social
Psychology, 80, 613-631.
Rowe, A., & Carnelley, K. B. (2003). Attachment
style differences in the processing of attachmentrelevant information: Primed-style effects on recall,
interpersonal expectations, and affect. Personal
Relationships, 10, 59-75.
Sharpsteen, D. J., & Kirkpatrick, L. A. (1997).
Romantic jealousy and adult romantic attachment.
Journal of Personality and Social Psychology, 72,
627-640.
Shrout, P. E., & Bolger, N. (2002). Mediation in
experimental and nonexperimental studies: New
procedures and recommendations. Psychological
Methods, 7, 422-445.
Simpson, J. A. (1990). Influence of attachment styles
on romantic relationships. Journal of Personality
and Social Psychology, 59, 971-980.
Spanier, G. B. (1976). Measuring dyadic adjustment:
New scales for assessing the quality of marriage
and similar dyads. Journal of Marriage and the
Family, 38, 15-28.
Teasdale, J. D., Lloyd, C.-A., & Hutton, J. M. (1998).
Depressive thinking and dysfunctional schematic
mental models. The British Journal of Clinical
Psychology, 37, 247-257.
Correspondence to:
Zoë Pearce
School of Applied Psychology
Griffith University
Brisbane QLD 4111
z.pearce@griffith.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 109-116
The Utility of Work and Family Stressors and Perceived Social Support in
the Prediction of Counselling Utilisation
Andrea Quah (andrea.quah@med.monash.edu.au)
Department of Psychology
Building 17, Monash University
Victoria 3800, Australia
Terry Bowles (t.bowles@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, VIC, Australia
however, is an exception as work and general life
stressors were measured amongst students attending
counselling. Results indicated that general anxiety
disorder and major depressive episodes were 5 times
more likely to be experienced by counselling users
experiencing work and general stress. The absence of
a control group in the study, however, prevents
conclusions to be made regarding counselling users’
vulnerability to stressors in comparison to nonusers.
Phillips and Murrell (1994), however, have
incorporated a control group in their study
investigating help-seeking behaviour amongst the
elderly. In particular, those seeking help, reported
higher levels of stressful life events compared to
nonseekers.
Considering the scarcity of research involving
stressors in the context of counselling use, this study
turns to investigations involving indices of health
amongst counselling users as a guide to their
experiences of stressors. Frequent counselling users
have reported high scores on the Minnesota
Multiphasic Personality Inventory-2 (MMPI-2)
College Maladjustment Scale, (Stewart, 1996). Other
studies involving control groups have found that
counselling users were more anxious than nonusers of
counselling (Dua & McNall, 1987) and had poorer
psychological health as indicated by scores on the
MMPI (Cooke & Kiesler, 1967). Taking into account
the stressor-strain link established by stress research
amongst the general population and findings of poor
well-being amongst counselling users, it is plausible
that stressors can be useful in discriminating
counselling users from nonusers.
More commonly, however, research investigating
the characteristics of counselling users has taken into
consideration the role of social support. A factor
characterised by a variety of dimensions (Cohen &
Wills, 1985; Sarason, Sarason, & Pierce, 1990), social
support entails the provision of “emotional concern,
instrumental aid, information, and/or appraisal…
between people” (House, 1981, p. 26). Results have
indicated that inadequate social support was a
common characteristic of university students who
attended counselling (Bosmajian & Mattson, 1980;
Snyder, Hill, & Derksen, 1972). In fact, having
alternate sources of social support was the most
important factor in discriminating those who accessed
counselling services from those who did not
(Bosmajian & Mattson, 1980). Amongst the elderly,
limited social support has been found to predict those
seeking professional help (Phillips & Murrell, 1994).
Abstract
This paper examines the utility of work and family
stressors (role conflict, role ambiguity, overload, and
underutilisation of skills), and perceived social support
in independently predicting counselling utilisation.
Support from partner, family members, supervisor, and
co-workers (in that order) significantly discriminated
adults seeking personal counselling from those seeking
career counselling and nonusers of counselling. There
was no evidence that stressors discriminated between
the three groups. Mean ages of participants within each
group were between 34 and 38, with a proportionate
number of men and women in each group. All 117
participants completed questionnaires measuring
perceived stressors and social support.
Most researchers view stress as a process (Pearlin,
Lieberman, Menaghan, & Mullan, 1981) which has
been conceptualised to involve external sources of
stress, strain, and outcomes or consequences of such
responses to stress (Elliott & Eisdorfer, 1982).
Work stressors are situations in the workplace that
have been associated with potential stress. Two
commonly researched work stressors based on role
theory (Kahn, Wolfe, Quinn, Snoek, & Rosenthal,
1964) are role conflict (incompatible and inconsistent
work demands) and role ambiguity (uncertainty
regarding aspects of work). Further potential sources
of workplace stress have been conceptualised by
Caplan and his colleagues (Caplan, Cobb, French,
Harrison, & Pinneau, 1975) as work overload
(excessive amount of work load) and work
underutilisation of skills (the poor usage of a worker’s
knowledge and abilities).
These four work stressors have been associated with
a number of strains - job dissatisfaction, depression
(Ganster, Fusilier, & Mayes, 1986; Terry, Nielson, &
Perchard, 1993), and irritation (Caplan, Cobb, French,
Harrison, & Pinneau, 1975) across a range of
occupations.
An important limitation that exists across studies of
stress is the use of participants from the general
population (Beehr, Jex, Stacy, & Murray, 2000;
Ganster et al., 1986; LaRocco, House, & French,
1980; Noblet, Rodwell, & McWilliams, 2001, Terry
et al., 1993). In the effort to control for possible
confounding variables such as negative affectivity,
researchers have neglected to analyse stress in varying
populations (Spector, Zapf, Chen, & Frese, 2000). A
study by Surtees, Pharoah, and Wainwright (1998),
109
QUAH & BOWLES
110
The role of social support in the stressor-strain
relationship within the general population can be
viewed as lending credence to its role in predicting
counselling utilisation. Lack of support from work
and family has been found to be associated with high
levels of depression, anxiety and irritation amongst
respondents from a range of occupations (LaRocco et
al., 1980).
The Present Study
The primary aim of this study is to examine the utility
of stressors and social support in discriminating
between users and nonusers of counselling. It is
expected that inadequate social support is most
characteristic of personal counselling users in
comparison to career counselling users and nonusers.
Whereas past research has indicated social support as
an effective discriminator, the experience of stressors
has not been researched extensively across such a
sample. This study, therefore, is exploratory in nature
as it applies the stresssor-strain framework developed
within the general population to counselling users.
Considering that past research has linked the
experience of stressors to poor mental and physical
health, and that counselling users have been
associated with poor well-being, it is plausible that
counselling users experience higher work and family
stressors than nonusers. Thus, the utility of work and
family stressors in distinguishing personal counselling
users from career counselling and nonusers will be
examined in this study.
Method
Participants
A total of 117 participants provided data for this study
through the completion of questionnaires. From this
total, 39 participants (49 % male) comprised the
personal counselling group. The mean age was 37.9
years (SD = 11.66). Participants in this group were
mostly married or partnered (80 %) and had made
appointments with a psychologist for problems of a
personal or social nature. A total of 39.5 % were
professionals, 31.6 % classified as tradespersons and
clerical workers, and 28.7 % blue-collar workers
(occupation classification based on Australian Bureau
of Statistics, McLennan, 1997).
Participants who sought counselling for work or
study related problems were categorised in the career
counselling group. This group consisted of 40
participants (50 % men) with an average age of 38
years (SD = 9.87). The majority of participants were
either married or partnered (83 %). In this group, 37.5
% were professionals, 52.5 % tradespersons and
clerical workers, and 10 % blue-collar workers.
The third group, noncounselling users, were those
who responded to advertisements seeking participants
interested in a study of stress. A total of 38
participants (52 % male) belonged to this group. On
average, participants were 33.7 years of age (SD =
12.99) and 76 % were married or partnered. A total of
52.6 % were professionals, 23.7 % tradespersons and
clerical workers, 2.6 % blue-collar workers and 13.2
% full-time students.
Measures
Work Stressors Role conflict Rizzo et al.’s (1970)
scale was used to measure role conflict. Participants
responded to a total of eight items regarding
conflicting work demands (e.g., “I receive
incompatible requests from two or more people at
work). Responses to each item were indicated on a 7point scale ranging from very false to very true. High
scores indicate high role conflict. Good reliability for
this scale has been demonstrated. Rizzo et al. (1970)
reported a Kuder-Richardson reliability coefficient of
.82.
Role ambiguity Role ambiguity was measured using
the scale developed by Rizzo et al. (1970). Consisting
six items in total, participants indicated on a 7-point
scale whether statements such as “I know exactly
what is expected of me at work” were very false or
very true. All items in this scale were reverse scored
so that high scores indicate high role ambiguity.
Adequate reliability has been established for this
measure, with Rizzo et al. reporting a KuderRichardson reliability coefficient of .78.
Quantitative work overload Caplan et al.’s (1975)
scale was employed to measure quantitative work
overload, the “amount of work the person is given to
do” (p. 43). Seven of the 11 items required
participants to indicate on a 5-point scale ranging
from hardly any to a great deal, the extent of overload
at work. An example of an item is, “How much time
do you have to do all your work?” For the remaining
four questions, participants responded to questions on
a five-point scale ranging from rarely to very often.
Upon reverse scoring four questions, high scores on
this scale were an indication of high work overload.
Good reliability for this scale has been established by
Terry et al. (1993), reporting a Cronbach’s alpha
coefficient of .91.
Underutilisation of skills Three items developed by
Caplan et al. (1975) were used to assess
underutilisation of skills. For example, “how often
can you use the skills from your previous experience
and training at work?” was a question presented to
participants. Responses were recorded on a 5-point
scale ranging from hardly/rarely to very often. Items
on this scale were reversed scored, thus, high scores
corresponded with high levels of this stressor. This
measure had an adequate reliability coefficient of .74
(Terry et al., 1993).
Perceived Availability of Work-related Support
Social support from five sources (supervisor, coworkers, partner, family, and friends) was measured
using a modified scale originally developed by Caplan
et al. (1975). Each social support source was
measured using six 5-point scale items. A total of five
measures of social support, therefore, were obtained
from these items. For the purposes of this study,
questions on his measure were modified to explicitly
name the domain for which social support was
proffered, in this case work-related problems (Terry et
al., 1993). An example of a question is “How much
can each of these people be relied on when things get
tough at work?”. Participants indicated whether the
perceived availability of social support ranged from
The prediction of counselling utilisation
very much to no such person. Reverse scoring was
undertaken so that high scores indicated the
perception of high levels of availability of social
support. Good reliability has been reported for each of
the five measures of social support, ranging from .90
to .95 (Terry et al., 1993).
Family Stressors Role conflict, role ambiguity,
underutilisation of skills, and home overload All
measures of family stressor variables (role conflict,
role ambiguity, underutilisation of skills, and home
overload) were based on scales employed for
measurement of work stressors described above. In
fact, except for minor word changes, the scales used
to measure dimensions of family stressors were
identical to work stressor measures. The statements “I
receive incompatible requests from two or more
people at home” and “I know exactly what is expected
of me at home” are examples of items on the family
role conflict and ambiguity scales, respectively. Thus,
questions and statements consisting of phrases “at
work”, were changed to “at home” in all four
measures of family stressors.
Perceived Availability of Family-related Support
The social support measure by Caplan et al. (1975)
was used as the basis to assess perceived availability
of family-related support. Questions in this measure
were modified so as to specifically ask whether social
support was available for family-related problems
rather than general problems. “How much can you
count on these people to listen to you when you need
to talk about family-related problems?” is an example
of one of the six items measuring support from five
sources (supervisor, co-workers, partner, family, and
friends). These measures, therefore, correspond with
the work-related support measures described above.
111
Procedure
Personal and Career Counselling Users Clients
who made an appointment with a registered
psychologist for personal or social problems (personal
counselling) and for work or study (career
counselling) were invited to complete the
questionnaires prior to the commencement of their
first counselling session.
Nonusers of Counselling Participants were recruited
through advertisements posted on notice boards in
Monash University. Additionally, the order in which
questionnaires were administered to all three groups
was counterbalanced.
Results
Data Screening
Inspection of correlations between the variables of
social support revealed multicollinearity. Correlations
approaching and exceeding .70 were obtained
between support for family problems and work
problems from identical sources, a pattern established
across all sources of support (see Table 3). In
accordance with the recommendation by Tabachnick
and Fidell (2001), composite variables were created
for inclusion in discriminant function analyses.
Therefore, the ten variables of social support were
combined to create five variables - social support
from supervisor, co-workers, partner, family, and
friends for both work and family problems.
Descriptive Statistics
Table 1 presents the means and standard deviations of
work and family stressor scores across the three
groups of counselling utilisation.
Table 1
Means and Standard Deviations for Work and Family Stressor Scores as a Function of Counselling Utilisation
___________________________________________________________________________
Personal
Career
counsellinga
counsellingb
Nonusersc
M
(SD)
M (SD)
M (SD)
__________________________________________________________________________________________
Work
Role conflict
4.25
(1.18)
3.71 (1.25)
3.66
(1.20)
Role ambiguity
2.48
(1.02)
2.45 (1.15)
2.29
(0.89)
Overload
3.82
(0.65)
3.69 (0.83)
3.63
(0.66)
Underutilisation
3.06
(0.81)
2.93 (1.10)
2.77
(1.13)
Family
Role conflict
3.56
(1.07)
3.05 (1.02)
3.26
(0.84)
Role ambiguity
2.48
(0.88)
2.37 (0.69)
2.39
(0.82)
Overload
3.18
(0.25)
3.20 (0.23)
3.17
(0.72)
Underutilisation
2.87
(1.13)
2.69 (1.07)
2.97
(1.05)
Stressor
a
N = 39. bN = 40. cN = 38
QUAH & BOWLES
112
In order to determine whether potential differences
between the groups might be dependent on sex of
participants, a 3 x 2 between-subjects MANOVA was
conducted on the eight stressor variables outlined in
Table 1. Independent variables were counselling
status (personal, career, and nonusers) and sex. Using
Wilks’ criterion, the interaction effect was not
significant, F (16, 208) = 0.62, p > .05.
Table 2 illustrates the means and standards
deviations of social support scores for all three
counselling utilisation groups. It can be observed that
personal counselling users consistently have the
lowest mean scores in comparison to career
counselling users and nonusers across all sources of
social support. The nonsignificant interaction between
counselling status and sex obtained from a 3 x 2
between-subjects MANOVA, confirmed that potential
group differences on the social support scores would
be independent of sex. Wilks’criterion F (10, 214) =
1.10, p > .05 was utilised.
Table 2
Means and Standard Deviations for Social Support Source as a Function of Counselling Utilisation
___________________________________________________________________________
Variable
Personal
counsellinga
M
( SD)
Career
counsellingb
M (SD)
Nonusersc
M
(SD)
Social Support Source
Supervisor
1.39
(1.15)
1.85 (1.26)
1.93
(1.09)
Co-workers
1.53
(0.99)
1.96 (1.00)
1.97
(1.20)
Partner
2.08
(1.36)
2.98 (1.41)
2.98
(1.26)
Family
2.26
(0.96)
2.56 (0.81)
2.88
(0.74)
Friends
2.37
(1.00)
2.60 (0.86)
2.62
(0.68)
__________________________________________________________________________________________
a
N = 39. bN = 40. cN = 38
Correlations Between the Measures
Table 3 presents the correlations between the stressors
and social support measures for the entire sample.
Generally, few significant correlations were obtained
between work and family stressors or within the
respective domains of stressors. An interesting pattern
of correlations emerges between social support
directed at work and family problems obtained from
the same source. Specifically, the correlations are
highly significant across all sources of social support,
ranging from .68 to .92. For example, a highly
significant positive correlation was obtained between
partner family support and partner work support.
Discriminant Function Analyses
The first discriminant function analysis was
conducted to determine whether eight stressor
variables, four variables of work stressors (role
ambiguity, role conflict, underutilisation of skills, and
work overload) and a corresponding four variables of
family stressors, could predict group membership.
The three groups of counselling utilisation were
personal counselling users, career counselling users
and nonusers. Neither of the two discriminant
functions calculated were significant, the overall
Wilks’Λ = .90, χ2 (16, N = 117) = 11.83, p > .05 and
the residual, Wilks’Λ = .97, χ2 (7, N = 117) = 3.73, p
> .05, indicating that the experience of stressors did
not differ between the groups.
The second discriminant function analysis involved
five composite variables of social support. Social
support from supervisor, co-workers, partner, family,
and friends for the domains of work and family were
used as predictors of membership in three groups of
counselling status. Whereas the overall Wilks’ lamda
was significant, Λ = .81, χ2 (10, N = 117) = 24.13, p <
.01, the second discriminant function was not, Λ =
.98, χ2 (4, N = 117) = 2.34, p > .05. Respectively the
two discriminant functions accounted for 91% and 9%
of the between group variability. Table 4 presents the
loading matrix of correlations between predictors and
discriminant functions.
The prediction of counselling utilisation
113
Table 3
Intercorrelations between Stressor and Social Support Subscales
Variables
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Work Stressors
1. Role ambiguity
-
2. Role conflict
.40**
3. Overload
.09
.35**
4. Underutilisation
.27**
.11
-.07
-
5. Role ambiguity
.34**
.12
.0
.15
6. Role conflict
.18
.39
.09
.16
.31**
7. Overload
-.05
-.17
.08
.12
-.06
-.03
-
8. Underutilisation
.23*
.14
-.14
.02
.18
.04
-.17
-
9. Supervisor
-.13
-.02
.22*
-.11
.08
-.07
.00
.01
10. Co-workers
-.08
-.04
.19*
-.08
-.01
-.01
.01
.03
.62**
11. Partner
-.05
-.10
.00
-.08
-.14
-.14
.01
-.12
-.08
.04
-
12. Family
.03
-.10
-.12
-.14
.11
-.20*
-.18
.08
.17
.08
.09
13. Friends
.09
-.02
-.05
-.09
.06
-.16
-.07
-.04
.13
.09
.21*
.54**
-
14. Supervisor
-.03
.07
.16
-.07
.05
-.09
.03
-.01
.81**
.45**
-.02
.21*
.16
-
15. Co-workers
-.06
.02
.17
-.10
-.02
-.12
.11
-.13
.49**
.72**
.10
.12
.15
.65**
-
16. Partner
-.09
-.08
-.01
-.02
-.14
-.11
.06
-.05
-.03
.09
.92**
.04
.16
-.01
.13
-
17. Family
-.02
-.21*
.00
-.20*
-.06
-.30
-.09
-.07
.36**
.27
.09
.72**
.38**
.27**
.23*
.11
-
18. Friends
.00
-.13
.03
-.09
-.01
-.26**
-.01
-.06
.32**
.28**
.03
.30**
.68**
.27**
.28**
.03
.45**
-
Family stressors
-
Social support (work)
-
-
Social support (fam.)
Note. *p < .05. **p < .01
QUAH & BOWLES
114
Table 4
Correlations of Social Support Variables with the
Two Discriminant Functions
___________________________________________
Variable
Function 1
Function 2
___________________________________________
Social Support Source
Supervisor
.45
.10
Co-workers
.45
.28
Partner
.67
.47
Family
.62
-.68
Friends
.28
.10
__________________________________________
Since only the first discriminant function and not
the second was found significant, interpretation is
limited to the former. Table 4 shows that the best
predictors for distinguishing between the three groups
of counselling utilisation were social support from
partner and family, followed by supervisor and coworkers (loadings less than .3 are not interpreted).
The means on the discriminant functions are
convergent with this interpretation. Nonusers (M =
.91) had the highest mean scores on the combination
of social support variables whilst personal counselling
users had the lowest mean social support scores (M =
-.67) followed by career counselling users (M = -.21).
Discussion
The primary aim of this study was to determine
whether stressors, and social support variables were
capable of predicting counselling utilisation. Firstly, it
was expected that personal counselling users in
comparison to career and nonusers of counselling
would be most closely linked with high reports of
stressors. Contrary to expectations, the four
dimensions of work and family stressors (role
ambiguity,
role
conflict,
overload,
and
underutilisation of skills) did not discriminate
between the groups who sought personal counselling,
career counselling, and nonusers. Consistent with
predictions, however, social support from various
sources was found to significantly differentiate the
three groups, specifically, support from partner,
family members, supervisor, and co-workers (in that
order). Nonusers perceived the highest levels of social
support, whereas career and personal counselling
users perceived low support.
Firstly, the finding that counselling users do not
differ from nonusers in terms of stressors experienced
is in disagreement with findings of past research.
Phillips and Murrell (1994) reported that in addition
to poor psychological well-being, stressful life events
were able to discriminate between those who sought
professional help and those who did not. However the
differing ages of participants in Phillips and Murrell’s
study in comparison to the present study may account
for the discrepancy in results. Older adults, above the
age of 55 characterize Phillips and Murrell’s study,
whereas the present findings have been obtained from
a younger adult sample. Thus, it is plausible that the
variables which prompt older people to seek help
cannot be generalised to a younger sample.
Secondly, it is essential to note the varying sources
of help approached by the respective samples. In this
study, participants with personal and career problems
made appointments with a psychologist, whereas a
majority of Phillips and Murrell’s (1994) older sample
sought mental health assistance from a medical
doctor. One could speculate that seeking help from a
medical doctor for mental health problems requires
people perceiving themselves in a more dire state of
need compared to those who do not, a state possibly
reflected by the high reports of stressful life events by
the former group. Seeing a psychologist on the other
hand, may not require higher perceptions of stressors
but rather a stronger desire to work through existing
stressful issues with a professional. These
explanations, however, can only be confirmed by
future research that compares characteristics of groups
who seek help from different health providers.
Further understanding of the obtained unexpected
result indicating the inability of reported stressors to
predict counselling utilisation, can be gained through
the collective examination of other past research. A
characteristic shared by numerous studies is that the
entire sample is derived from the university
population (Bosmajian & Mattson, 1980; Cooke &
Kiesler, 1967; Stewart, 1996; Surtees et al., 1998).
Thus, whilst Phillips and Murrell’s (1994) study
involve older adults, these studies involve younger
adults, hence obscuring outcomes of direct
comparisons between these studies and the present.
Due to the scarcity of past research involving
stressors as predictors of counselling use, theory
linking these variables in this study is based upon the
relationship of poor well-being and counselling use,
and the experience of stressors with poor well-being
in the general population. In light of the obtained
findings, however, it is possible that the relationship
between experienced stressors and well-being in the
general population cannot be extended to help-seeking
behaviour. Significantly poorer well-being may not
be associated with higher levels of stressors for
counselling users. Future research, therefore, needs to
explore the relationship of well-being and experienced
stressors amongst counselling users.
Additionally, it is important to distinguish between
the “magnitude and impact of stressers” (Schwartberg
& Dytell, 1988, p. 188). Past research has found that
although employed and nonemployed women did not
differ in terms of family stress levels, the latter were
more sensitive to its effects on self-esteem and
depression than the former group of women
(Schwartberg & Dytell, 1988). Thus, the fact that
counselling users in the current study did not report a
higher magnitude of stressors compared to nonusers
does not mean that stressors should be dismissed and
deemed an unimportant variable. A possibility that
has not been investigated by this study is that stressors
may still be able to discriminate counselling users
from nonusers based on its differential impact on each
group rather than its reported magnitude. Further
research with measures of well-being may be able to
The prediction of counselling utilisation
establish the respective sensitivity of each group to
the effects of stressors.
Results indicate that social support for family and
work problems from a variety of sources (partner,
family members, supervisor, and co-workers) is able
to predict counselling utilisation. In particular,
personal counselling users are most closely associated
with lower levels of perceived social support in
contrast to nonusers. These findings are congruent
with those obtained by other researchers despite the
varying age groups of samples in the respective
studies. Thus, social support seems to be an important
determinant of counselling use across all ages,
ranging from young college students (Bosmajian &
Mattson, 1980) to older adults (Phillips & Murrell,
1994), including middle-aged adults (as established
by the present study).
The nature of the obtained result involving social
support and counselling utilisation, can help explicate
the perceived function and role of counselling.
Specifically, the finding that those low in perceived
support are most likely to seek counselling services
from a psychologist, suggests that counselling is
perceived as a form of professional social support.
Thus, an interesting issue for future research to
undertake is the examination of the comparative
effectiveness of professional versus informal social
support.
Another deduction that can be made of counselling
use is that it is not the first choice for help.
Considering the finding that those high in perceived
support are least likely to initiate counselling use, it is
reasonable to conclude that the use of professional
mental health services replaces rather than
supplements the receipt of support from other sources.
Such a conclusion complements findings from other
studies that report that students with personal
problems prefer friends and close relatives rather than
counselling services as a source of help (Snyder et al.,
1972). In light of the perceived unavailability of
alternate sources of support, seeking professional help
may be an action taken as the last resort for
assistance. This finding, therefore, can contribute to
professional mental health providers’ understanding
of the circumstances under which their services are
being sought.
References
Beehr, T. A. (1995). Psychological stress in the
workplace. New York: Routledge.
Beehr, T. A., Jex, S. M., Stacy, B. A., & Murray, M.
A. (2000). Work stressors and coworker support as
predictors of individual strain and job performance.
Journal of Organizational Behavior, 21, 391-405.
Bosmajian, C. P. J., & Mattson, R. E. (1980). A
controlled study of variables related to counseling
center use. Journal of Counseling Psychology,
27(5), 510-519.
Caplan, R. D., Cobb, S., French, J. R. P., Jr., Harrison,
R. V., & Pinneau, S. R., Jr. (1975). Job demands
and worker health. Ann Arbor, MI: The University
of Michigan, Institute for Social Research.
Cohen, S., & Wills, T. A. (1985). Stress, social
support, and the buffering hypothesis.
115
Psychological Bulletin, 98(2), 310-357.
Cooke, M. K., & Kiesler, D. J. (1967). Prediction of
college students who later require personal
counseling. Journal of Counseling Psychology,
14(4), 346-349.
Dua, J. K., & McNall., H. M. (1987). Assertiveness in
men and women seeking counselling and not
seeking counselling. Behaviour Change, 4(1), 1419.
Elliott, G. R., & Eisdorfer, C. (Eds.). (1982). Stress
and human health: Analysis and implications of
research. New York: Springer-Verlag.
Ganster, D. C., Fusilier, M. R., & Mayes, B. T.
(1986). Role of social support in the experience of
stress at work. Journal of Applied Psychology,
71(1), 102-110.
Holmes, T. H., & Rahe, R. H. (1967). The social
readjustment rating scale. Journal of Psychosomatic
Research, 11, 213-218.
House, J. S. (1981). Work stress and social support.
Reading, MA: Addison-Wesley.
Kahn, R. L., Wolfe, D. M., Quinn R. P., Snoek, J. D.,
& Rosenthal, R. A. (1964). Organizational stress:
Studies in role conflict and ambiguity. New York:
John Wiley & Sons.
Kahn, R. L., & Byosiere, P. (1992). Stress in
organizations. In M. D. Dunnette, & L. M. Hough,
(Eds.), Handbook of industrial and organizational
psychology (2nd ed., pp. 571-650). Palo Alto, CA:
Consulting Psychological Press.
LaRocco, J. M., House, J. S., & French J. R. P. Jr.
(1980). Social support, occupational stress, and
health. Journal of Health and Social Behavior, 21,
202-218.
McLennan, W. (1997). Australian Bureau of
Statistics: Australian Standard Classification of
Occupations (2nd ed.). Commonwealth of
Australia.
Noblet, A., Rodwell, J., & McWilliams, J. (2001).
The job strain model is enough for managers.
Journal of Managerial Psychology, 16(8), 635-649.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., &
Mullan, J. T. (1981). The stress process. Journal of
Health and Social Behavior, 22, 337-356.
Phillips, M. A., & Murrell, S. A. (1994). Impact of
psychological and physical health, stressful events,
and social support on subsequent mental health help
seeking among older adults. Journal of Consulting
and Clinical Psychology, 62(2), 270-275.
Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970).
Role conflict and ambiguity in complex
organizations. Administrative Science Quarterly,
15, 150-163.
Sarason, I. G., Sarason, B. R., & Pierce, G. R. (1990).
Social support: The search for theory. Journal of
Social and Clinical Psychology, 9(1), 133-147.
Schwartzberg, N. S., & Dytell, R. S. (1988). Family
stress and psychological well-being among
employed and nonemployed mothers. In E. B.
Goldsmith (Ed.), Work and family: Theory,
research, and applications. Newbury Park, CA:
Sage.
Snyder, J. F., Hill, C. E., & Derksen, T. P. (1972).
Why some students do not use university
116
QUAH & BOWLES
counseling facilities. Journal of Counseling
Psychology, 19(4), 263-268.
Spector, P. E., Zapf, D., Chen, P., & Frese, M. (2000).
Why negative affectivity should not be controlled in
job stress research: Don't throw out the baby with
the bath water. Journal of Organizational
Psychology, 21, 79-95.
Stewart, D. W. (1996). Predicting counselling service
utilization patterns with the MMPI-2 college
maladjustment scale. Canadian Journal of
Counselling, 30(3), 211-216.
Surtees, P. G., Pharoah, P. D. P., & Wainwright, N.
W. J. (1998). A follow-up study of new users of a
university counselling service. British Journal of
Guidance and Counselling, 26(2), 255-272.
Tabachnick, B. G., & Fidell, L. S. (2001). Using
multivariate statistics (4th ed.). Needhan Heights,
MA: Allyn and Bacon.
Terry, D. J., Nielsen, M., & Perchard, L. (1993).
Effects of work stress on psychological well-being
and job satisfaction: The Stress-buffering role of
social support. Australian Journal of Psychology,
45(3), 168-175.
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 117-123
Improving family and peer relationships with children who engage in
problem sexual behaviour
Dr. Petra Staiger (pstaiger@deakin.edu.au)
School of Psychology, Deakin University
Dr. Nicolas Kambouropoulos (NKambourop@swin.edu.au)
School of Social and Behavioural Sciences, Swinburne University of Technology
Jari Evertsz
The University of Melbourne
Joe Tucci (jtucci@childhood.org.au)
Australian Childhood Foundation
Janise Mitchell (jmitchell@childhood.org.au)
Australian Childhood Foundation
knowledge base has been fraught with difficulties.
Firstly, community responses towards children who
engage in problem sexual behaviour are clearly
diverse and often dramatically so. Attitudes towards
these children vary from denial and minimisation to
outrage and condemnation (Burton, Nesmith &
Badten, 1997).. Beliefs about the innocence of
children are challenged by these problem behaviours
(Mitchell, 2001). Secondly, the responses by statutory
services, the police and child welfare organisations
have added to the confusion. At almost every level,
children who engage in problem sexual behaviour
have not been able to access specialist services. There
is little public policy which covers their needs and
there are few resources allocated to programs
specifically designed to support them and their
families.
Most of the research conducted in this area has
arisen from the United States (Araji, 1997). Yet, it is
clearly important to develop a knowledge base and
data on this topic within Australia. This paper
describes the first formal (albeit preliminary)
evaluation of a program for children who engage in
problem sexual behaviour in Australia.
There have been no attempts made to gauge the
prevalence rates of children who engage in problem
sexual behaviour either in Australia or internationally.
In contrast, it has been estimated that 2%-4% of
adolescent males commit sexual assaults (Ageton,
1983) and these offences account for between 30%
and 50% of all child molestation. A report by the
Children’s Protection Society in Victoria, estimated
that 20%-40% of child sexual assaults are the
responsibility of those aged under 18 (Mitchell,
2001).
Generally, studies have defined problem sexual
behaviour as that which “far exceeds the mutual
exploratory behaviour normally seen in young
children” (Friedrich & Luecke, 1988, p. 154) and is
“outside the normal developmental sexual activity
expected for children” (Johnson, 1988, p. 222). For
Abstract
The Transformers program uses cognitive behavioural
and family therapy strategies to assist children under
the age of 11 to stop engaging in problem sexual
behaviour. The program aims to improve family and
peer relationships by working on empathy, respect and
communication. This paper presents preliminary
findings from 17 children who participated in the
Transformers program. There were two groups: Group
A participated in assessment only and Group B
participated in both assessment and treatment. Data
was collected pre and post intervention and then two
years later. Questionnaires consisted of a range of
measures of child/parent functioning developed for the
purpose of evaluating this program. Wilcoxon signedrank tests compared the outcomes of the assessment
group (Group A) with those achieved by the
intervention group (Group B). Those in the treatment
group reported three times the frequency of the
problem sexual behaviour compared to the assessment
group prior to any intervention. Yet, following
intervention, the frequency of this behaviour in Group
B was lower with less than an average of one incident
per month whilst the assessment only group was still
engaging in at least one problem sexual behaviour
incident a month. Only the children in Group B
reported significant improvements in self-intervention
and awareness of personal risk factors. Only Group B
caregivers reported a significant increase in confidence
to manage the child-parent/caregiver relationship.
These findings are discussed in relation to the current
literature on children who engage in problem sexual
behaviour.
In a little over a decade, the literature in relation to
adolescent and adult sex offending has developed
rapidly (Araji, 1997). In that time, there has been an
increasing interest in building a framework for
intervening to stop children who engage in problem
sexual behaviour from continuing to develop more
harmful sexual behaviour into their adolescence and
adulthood. However, the development of this
117
118
STAIGER, KAMBOUROPOULOS, EVERTSZ, TUCCI, & MITCHELL
example, public masturbation, forceful penetrative
behaviour, excessive fondling and genital contact
characterise children presenting with sexual behaviour
problems (Johnson, 1988; Gray et al., 1999).
Research has shown that children presenting with
sexual behaviour problems often appear to come from
low income families, with many living below the
poverty level. In general males are over-represented in
samples of children who engage in problem sexual
behaviour (approximately 70% are boys). Studies
reviewed indicate that children who engage in
problem sexual behaviour are characterised by high
levels of behavioural and emotional problems, low
levels of empathy, restricted affect and depressive
symptoms (Araji, 1997; Gray et al., 1999).
Research has reported that children with sexually
abusive caregivers are more likely to develop problem
sexual behaviour (Johnson, 1988; Pithers, Gray,
Busconi & Houchens, 1998). In addition, high rates of
physical and emotional abuse among such children
have also been reported (e.g., Johnson, 1988; Ray &
English, 1995). Studies have found that approximately
two-thirds of mothers were victims of childhood
physical neglect and/or had been sexually abused
(Burton et al., 1997).
The majority of treatment programs for problem
sexual behaviour in children have been developed in
the United States. The various treatment programs
have a number of common components. Most
programs utilise cognitive and behavioural
approaches as their primary theoretical framework,
with treatment interventions specifically targeting the
problematic sexual behaviours. In all of the programs,
parents or caregivers are considered to be an
important part of the treatment process and are
required to participate in treatment. The primary mode
of treatment for children is group therapy, with
individual and family therapy also frequently utilised,
often as an adjunct to group treatment. It is important
to note however that the effectiveness of many of
these programs has not been empirically examined.
To date, there have been only three published
outcome evaluation studies, all of which were
conducted in the United States (Bonner, Walker &
Berliner, 1999; Pithers et al., 1998; Ray et al., 1995).
Each of these treatment outcome studies is limited
by methodological problems. Only one study had a
comparison group, this means that changes in test
scores from pre-treatment to post-treatment cannot be
attributed with certainty to the impact of the
treatment. The study by Ray et al. (1995) did not
include self report data from the children themselves.
Despite methodological limitations, these treatment
outcome studies provide some evidence to support the
efficacy of structured cognitive behavioural and
dynamic/expressive therapy for children who engage
in problem sexual behaviour. It also appears that a
number of important components are inherent in most
programs, including family involvement and group
therapy.
Over the last eight years the Australian Childhood
Foundation, based in Melbourne, has developed a
specialist treatment program for children who engage
in problem sexual behaviour. This program has a
particular focus on treating the whole family system
and not just the child. The program focuses on
improvements in psychosocial functioning and not
just reduction in problem sexual behaviours. The
evaluation measures utilised consist of a combination
of validated and newly developed measures. The latter
were piloted in a previous study (Mulhern, 1997) and
found to be sensitive to change and show adequate
face validity, yet require psychometric validation. The
new measures were developed in order to examine
clinically and theoretical constructs relevant to the
program. For example, taking responsibility for one’s
behaviour is a central theme of the clinical
intervention and hence a child friendly measure in the
form of a “blame cake” was developed to measure
this construct in a time efficient manner.
The current paper presents some of the data from
the preliminary evaluation of the Transformers
Program. Clients who participated in the program
were compared with those completed the assessment
phase only. The comparison group, although selfallocated, provides a unique contrast to those
participating in the intervention. In addition to the
problem sexual behaviour per se this paper focuses on
changes
in
specific
child
measures
of
psychoemotional functioning (depression, anger,
empathy, and program learning) and parent/caregiver
measures (understanding and confidence of child’s
problem behaviour) as a result of the intervention.
Method
Design
An evaluation of the Transformers Program was
conducted which included the collecting of
information via questionnaires and interviews. Two
groups were included in the evaluation. The first
group (Group A) participated in the assessment phase
of the intervention. The second group (Group B)
participated in the assessment and treatment phases of
the program. Due to ethical considerations, there was
no control group comprising children who did not
receive any treatment. The outcomes of the
assessment group (Group A) were compared with the
outcomes achieved by the intervention group (Group
B). Data was collected prior to children entering the
assessment phase for both groups and at the end of
their participation in the program.
Intervention
The Transformers Program is primarily based on the
principles of cognitive behavioural therapy and
systems theory. It emphasizes the importance of
understanding children as part of a family system
within the constraints of their developmental stage.
The program has three components: an assessment
stage; group program for the children and; a
concurrent parental group therapy program.
The assessment takes place over 4 – 6 sessions, the
assessment evaluates the systems issues and problems
for the child; the impact of parenting/family factors;
the onset, duration, triggers, and risk level for the
sexual behaviour.
Problem sexual behaviour
The group program for the children is conducted
weekly over a 20 week period and each session is
approximately 1.5 hours in duration. The sessions are
conducted by two therapists. The group intervention
program is based on a schedule incorporating
activities which:
• Develop personal responsibility for behaviour;
• Identify triggers to sexual behaviour;
• Provide alternative methods of dealing with
difficult feelings;
• Increase awareness of personal risk patterns;
• Promote empathy for other’s experiences and
feelings;
• Enhance self-intervention skills;
• Develop and maintain an appropriate support
network;
• Improve children’s self-esteem and self-confidence.
The parent/caregivers program occurs concurrently
with the children’s group and is also facilitated by two
therapists over a 20 week period. The parent’s group
has both a supportive and an educational function and
aims to:
• Understand, prioritise and respond to the needs of
the child;
• Identify and respond appropriately to the protective
needs of the child;
• Appropriately and consistently discipline;
• Understand the issues regarding sexually aggressive
behaviours in children;
• Cope with their own anger and denial;
• Appropriately respond to the child’s sexually
aggressive behaviours;
• Support better models of coping in the child;
• Identify and change familial maintaining factors to
the child’s behaviour;
Involvement in the parents/caregivers group is not a
prerequisite for the child to receive treatment but
participation is encouraged.
Participants and Procedure
During a 2-year period, all children who engaged in
problem sexual behaviour under the age of 12 referred
to the Transformers Program of the Australian
Childhood Foundation were asked to participate in the
study. Children were excluded if there were protection
issues pending, no parental consent and/or no stable
living environment, moderate/severe intellectual
impairment, global developmental delay. The mean
age of the children was 9.27 (SD = 1.74). For Group
A (assessment phase only) the mean age was 9.61 (SD
= 1.70) while for Group B (treatment phase) the mean
age was 8.83 (SD = 1.82). Group A was comprised of
seven boys and two girls, and Group B consisted of
six boys and one girl.
In total, 152 children were referred to the program
and 41 were eligible to participate. Twenty-two
families gave consent to take part and six of these
withdrew from the program due to clinical reasons,
protective and placement issues (see Results).
Children who were eligible and agreed to participate
were firstly assessed for sexual behaviour problems
and were administered the instruments described
below at the assessment phase. Those children
119
continuing on to treatment took part in the
intervention program. Measures obtained during the
assessment phase were again completed post
intervention. Relevant ethics approval was obtained
from the Standing Committee on Ethics in Research
with Humans.
Measures
Frequency of problem sexual behaviour Behaviour
report forms were completed by teachers, parents, and
statutory child protection workers to provide an
approximation of the frequency of problem sexual
behaviours. The questionnaires were used as
structured interview items. They elicited information
on the type, risk level, and frequency of sexualised
behaviours. The interviews with teachers and child
protection workers were conducted via the telephone
by the therapist. Where there was disagreement
between informants the average frequency was
recorded. For the purposes of this report, the number
of occurrences per month was taken as the index of
frequency of sexual behaviours.
Responsibility taking A “Blame Cake” was
presented to children with the explanation about the
cake representing the degree of responsibility for the
sexual behaviour. Children were asked to divide the
cake into proportions according to whom they believe
should accept responsibility for the problem sexual
behaviour. The proportion of blame which the child
allocated to themselves was used as the measure of
responsibility taking (0-100 expressed as a
proportion).
Empathy In order to provide a measure of empathy,
children were asked “How do you think (name of
target child) felt about the behaviours?”
The
expectation that the children would be able to
generate three possible feelings resulted in the score
range of 0 – 3 (0=no response, 3=3 feelings
generated).
Self-rated confidence by parents A single item was
used to assess parental confidence in managing their
child’s problem sexual behaviours. Parents were
asked to rate from 0 (not confident at all) to 4
(extremely confident), “How confident do you feel in
your ability to effectively manage these behaviours?”
For the purposes of this report, scores were computed
a percentage of scale maximum (e.g., 4 = 100%).
Parental understanding Parental knowledge and
management of their child’s problem sexual
behaviours was rated by therapists on a 39 item
questionnaire. This measure was divided into three
sections (13 items each). These were: actual
management responses; knowledge of risk issues; and
understanding of own impact upon the child. Every
question endorsed by the therapist was given a value
of 1, resulting in a possible score range of 0-39.
STAIGER, KAMBOUROPOULOS, EVERTSZ, TUCCI, & MITCHELL
Results
Sample Characteristics
In total, data from 16 children with sexual behaviour
problems were included in the analyses. Nine children
completed the clinical assessment phase (Group A)
and seven children completed assessment and
participated in treatment for problem sexual behaviour
(Group B). In Group A, two children lived at home
with both parents, the remainder were in foster care,
at home with a natural parent and a defacto, or in
residential units. Four of the seven children in Group
B lived at home with both parents and three were in
foster care or placed with extended family members.
Of the 16 children, seven were victims of physical
abuse, six were exposed to emotional abuse, and four
children had been sexually abused. Four of the 16
children had been exposed to family violence and four
had been victims of neglect. Over half of the sample
of caregivers were themselves victims of some form
of physical and/or sexual abuse.
Children
Sexual behaviour Wilcoxon analysis of pre-post data
indicated that children in Group A (p < 0.02) and
Group B (p < 0.02) demonstrated a significant
decrease in frequency of sexual behaviour from prepost intervention (See Figure 1). Scores on sexual
behaviour at pre-intervention ranged between 0.25 –
50. Post-intervention scores ranged between 0 – 8.
Sexual Behaviour
Group A
Group B
14
Freq. per month
120
12
10
8
6
4
2
0
pre
Intervention
post
Withdrawn children or children excluded
Initially, 22 children participated in the study however
six children were withdrawn for the following
reasons. One child was withdrawn for clinical reasons
and referred to a child and adolescent mental health
service. The remaining children were withdrawn due
to protective issues (2), placement issues (1) and
withdrawal by parents (2).
Figure 1
Sexual behaviour at pre and post intervention
Empathy Analysis of empathy ratings revealed no
significant changes for Group A children (p = 0.10).
Group B children displayed a significant increase in
empathy from pre-post intervention (p = 0.04).
Intervention Evaluation
Group A consisted of children who only participated
in assessment. Group B consisted of children who
participated in assessment and intervention. The
following results compare the two groups on the
dependent measures to examine whether the
intervention had an impact on the risk of the children
referred to the program continuing to engage in
problem sexual behaviour, through shaping their
capacity for empathy, and the degree to which they
assumed responsibility for the behaviour. The
analyses also addressed changes in parental
confidence in managing their child’s behaviours and
in their knowledge and understanding of managing
the behaviour.
Means and standard deviations for the dependent
measures are displayed in Table 1. Due to the small
sample size, individual Wilcoxon signed-ranks tests
were conducted to examine changes in these variables
from pre to post intervention.
Baseline Differences
An examination of Table 1 indicates that children in
Group B (treatment) were higher than those in Group
A on frequency of sexual behaviour problems. As is
often the case in intervention programs those with
more serious problems continue on to treatment. Even
though there are differences at baseline it should be
noted that as opposed to Group A, the degree of
improvement for the children who received treatment
was of a greater magnitude..
Responsibility Taking There were no statistically
significant changes in responsibility taking from prepost intervention for either Group A (p = 0.17) or
Group B children (p = 0.07; see Figure 3). The
increase in responsibility taking for Group B
approached statistical significance.
Parents
The changes in parental ratings of understanding
about problem sexual behaviour and their confidence
in responding to such behaviour is presented in Table
2.
Parental Understanding Wilcoxon analysis revealed
a significant increase in self-rated understanding for
Group A (p =0.04) and Group B parents (p = 0.03).
Self-rated Confidence by Parents The analysis of
confidence ratings indicated no significant changes
for parents in Group A (p = 0.07). For Group B,
parents reported significant increases in confidence
from pre-post intervention (p = 0.02).
Problem sexual behaviour
121
Table 1
Means and standard deviations for pre and post assessments for children
GROUP A (N = 9)
Pre
Post
M
SD
M
SD
Sexual Behaviour
4.25
4.61
1.61
2.57
Responsibility – taking
25.00
43.42
48.13
42.92
Empathy
0.03
0.08
0.18
0.22
Note. Resp. Taking = Responsibility taking expressed as percentages.
GROUP B (N = 7)
Post
Pre
M
SD
M
SD
0.64
1.49
12.43
16.95
59.29 24.56
20.00
36.51
0.35
0.33
0.03
0.09
Table 2
Means and standard deviations for pre and post parental assessments
Understanding
Confidence
GROUP A (N = 9)
Pre
Post
M
SD
M
SD
20.18
6.76
23.75
9.08
54.32
5.89
76.21
8.36
R e s po ns ibilit y T a k ing
Gr oup A
Gr oup B
70
60
50
40
30
20
10
0
Pr e
Post
Int e rv e nt io n
Figure 3
Responsibility taking at pre and post intervention
Discussion
Approximately 50% of the children in the current
sample had experienced some form of abuse. This is
consistent with the numbers reported in the recent
large scale study conducted by Bonner, Walker and
Berliner (1999) in the USA. From these results, it is
reasonable to conclude that previous experiences of
child maltreatment increases the probability of
children engaging in problem sexual behaviour.
However, as a contributing factor, a history of child
abuse is not sufficient to completely explain the
genesis of the behaviour. Almost two-thirds of the
children in the sample of this study had psychiatric
diagnoses. A similar proportion was found in previous
studies (Gray et al., 1997; Pithers et al., 1998). In
particular, children who engage in problem sexual
behaviour tend to display greater behavioral problems
and more symptoms of anxiety and trauma.
Pre
M
18.01
21.01
GROUP B (N = 7)
Post
SD
M
SD
6.21
25.28
10.19
10.17
70.87
7.72
The parents in the current sample showed high rates
of psychological difficulties, substance abuse and
histories of abuse. These findings are again consistent
with prior research (Burton et. al., 1997). Overall, it
suggests that the sample of children and their
caregivers who took part in the intervention program
are representative, or at least similar to, those reported
in the literature so far.
This study provides some tentative support for the
feasibility and possibly the effectiveness of the
Transformers program. More specifically, there was a
general decrease in frequency of problem sexual
behaviours for children in both groups. However,
those in the treatment group showed a greater
magnitude of change compared to those who only
participated in the assessment phase. Children who
participated in treatment showed an increase (a trend
approaching significance) in responsibility taking
whilst no change was observed for those in the
assessment only group.
There was a statistical difference between pre and
post empathy ratings in the intervention group only.
However, closer inspection of the means suggests that
this change is of minimal clinical significance. In
most cases, at post-treatment the children were not
able to identify one feeling that they thought the other
child might be experiencing. Whilst empathy is
considered an important prognostic indicator in the
adult offender field, it is unclear how predictive it is
with children who engage in problem sexual
behaviour (Araji, 1997). Once again the importance of
the developmental stage of the child is vital in
understanding the meaning of these findings. It is
highly likely that many children of this age would
have difficulties in identifying the feelings of another
child. It is unclear from these findings whether the
small changes in empathy ratings is due to the lack of
sensitivity in the measure or whether the program did
122
STAIGER, KAMBOUROPOULOS, EVERTSZ, TUCCI, & MITCHELL
not result in changes in the children’s capacity to
empathise with others.
For parents, understanding and management of their
child’s sexual behaviour problems improved
significantly for both groups. This indicates that
assessment may be just as effective as treatment in
addressing parental ‘knowledge’ issues pertaining to
sexual behaviour problems. However, parental
confidence in managing the child’s sexual behaviour
problems only improved significantly for those who
received the intervention. This reflects the additional
time and resources required to assist parents/carers to
integrate their new knowledge into new strategies for
understanding and responding to their child’s needs.
There was no statistically significant increase in
responsibility taking in the children who participated
in the assessment phase only. Those in Group B who
received the intervention showed a trend towards an
increase in responsibility taking which almost
approached significance. A closer inspection of the
means and standard deviations suggests that there was
considerable variability in the response to this
question. Once again further development of specific
measures to adequately capture the constructs of
interest is required before any clear conclusions can
be drawn.
Many problems are encountered when conducting
research in treatment settings. In this study, the
difficulties were magnified by the fact that the
subjects of the research were children who engaged in
a complex and sometimes confronting behavioural
problem. Due to the small numbers in the study, the
conclusions should only be considered preliminary.
Care was taken not to exceed the number of
participants by the number of statistical tests applied.
Even so, the ratio was somewhat high. There was no
formal control group in this study, that is there was no
group who did not receive assessment or treatment.
This was based on the ethical concerns for the
children of withholding or delaying treatment
particularly in the light of the behaviours of this
group. The assessment group who themselves chose
to not continue into treatment provided a unique
comparison group to allow an examination of how
effective the treatment was over and above the
comprehensive assessment phase. Previous studies
have not been able to provide this information as they
have had either only one treatment group or 2
different types of treatments. Finally, some of the
measures that were specifically developed to measure
relevant therapeutic change (i.e. responsibility taking)
are new and require validation before a larger clinical
trial can go ahead.
The current study suggests that treatment provided
through the Transformers Program can result in
behaviour change for children who engage in problem
sexual behaviour, however a larger clinical trial is
required.
Acknowledgements
This project was funded by the National Council
Against Child Abuse. The authors would like to thank
all the families for participating in this study. We
would also like to acknowledge the important
contribution made to this project by a number of
colleagues: Dr. Neerosh Mudaly; Dr. Stephen
Wallace; Professor Chris Goddard.
References
Ageton, S.S. (1983). Sexual assault among
adolescents. Lexington, MA: D.C. Heath.
Araji, S.K. (1997). Sexually aggressive children:
Coming to understand them. US: Sage Publications.
Bonner, B., Walker, E., & Berliner, L. (1999).
Children with Sexual Behavior Problems:
Assessment and Treatment. Report to the
national center on child abuse and neglect.
Washington: U.S. Department of Health and
Human Services.
Burton, D., Nesmith A., & Badten, L. (1997).
Clinician's views on sexually aggressive children
and their families: A theoretical exploration. Child
Abuse and Neglect, 21, 157-170.
Friedrich, W., & Luecke, W. (1988). Young SchoolAge Sexually Aggressive Children. Professional
Psychology: Research And Practice 19, 155-164.
Gil, E., & Johnson, T.C. (1993). Sexualised children:
Assessment and treatment of sexualised children
who molest children. Rockville: Launch Press.
Gray, A., Busconi, A., Houchens, P., & Pithers, W.
(1997). Children with sexual behavior problems and
their caregivers: Demographics, functioning, and
clinical patterns. Sexual Abuse: Journal of Research
and Treatment, 9, 267-290.
Gray, A., Pithers, W., Busconi, A., & Houchens, P.
(1999).
Developmental
and
etiological
characteristics of children with sexual behavior
problems: Treatment implications. Child Abuse and
Neglect, 23, 601-621.
Mitchell, J. (2001) Children who hurt other children: a
qualitative study. Unpublished manuscript.
Mulhern, P. (1997). The development of specific
measures to evaluate the Child Sexual Behaviour
Program. Unpublished Manuscript.
Johnson, T.C. (1988) Child perpetrators: Children
who molest other children. Child Abuse and
Neglect, 12, 219-229.
Pithers, W., & Gray, A. (1998). The other half of the
story: Children with sexual behaviour problems.
Psychology, Public Policy, and Law, 4, 200-217.
Pithers, W., Gray, A., Busconi, A., & Houchens, P.
(1998). Five empirically-derived subtypes of
children with sexual behaviour problems:
Characteristics potentially related to juvenile
delinquency and adult criminality. Irish Journal of
Psychology, 19, 49-67.
Ray, J., & English, D. (1995). Comparison of female
and male children with sexual behavior problems.
Journal of Youth and Adolescence, 24, 439-451
Ray, J., Smith, V., Peterson, T., Gray, J., Schaffner,
J., & Houff, M. (1995). A treatment program for
Problem sexual behaviour
children with sexual behaviour problems. Child and
Adolescent Social Work Journal, 12, 331-343
Staiger, P.K., Kambouropoulos,N., Everetsz, J.,
Mitchell, J. & Tucci, J. (2004). Children who hurt
other children: A preliminary investigation of the
Children’s Sexual Behaviour Program. Australian
Childhood Foundation: Melbourne.
Correspondence to:
Dr. Petra Staiger
School of Psychology, Deakin University,
Burwood, 3125. Australia
pstaiger@deakin.edu.au
123
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 124-131
The Role of Interactions between Parents and Children in Social Skills
Development of Prep Children
Gill Terrett (g.terrett@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, St Patrick’s Campus, Melbourne Vic 3065 Australia
Michelle Puttick (michelleputtick@hotmail.com)
School of Psychology
Australian Catholic University, St Patrick’s Campus, Melbourne Vic 3065 Australia
noted by Gresham and Elliott (1990), are: Assertion,
which includes initiating behaviours, such as asking
others for information, introducing oneself, and
responding to the actions of others; Cooperation,
which includes behaviours such as helping others,
sharing materials, and complying with rules and
directions; and Self-control, which includes
behaviours that emerge in conflict situations, such as
responding appropriately to teasing, and in
nonconflict situations such as taking turns and
compromising. The current study adopts Gresham and
Elliot’s (1990) conceptualisation of social skills and
considers predictors of these three specific social
skills behaviours.
Abstract
The aim of this study was to investigate predictors of
social skills in young children. Of particular interest
were interactions between child and environmental
characteristics. This study investigated the concept of
interactions in two ways. First, a ‘goodness of fit’
measure was used to tap aspects of the child’s fit with
the
environment,
specifically
with
parental
expectations. Second, a statistical approach was used to
assess the contribution of interaction terms created
using temperament and parenting characteristics. The
results showed that ‘goodness of fit’ and parental use of
inductive reasoning contributed to self-control,
however none of the hypothesised predictor variables
was
significantly
related
to
cooperation.
Temperamental sociability and ‘goodness of fit’ were
significantly related to assertion. An interaction
between temperamental sociability and parental
warmth in the development of assertion was also found.
Further investigation of this interaction effect revealed
that level of parental warmth did not impact upon the
development of assertion in children who were more
temperamentally
sociable.
For
children
temperamentally less sociable, however, lower levels
of parental warmth were associated with lower levels
of assertion, while higher parental warmth was
associated with higher levels of assertion. This study
highlights the value of focusing on specific social skills
outcomes, and of pursuing interaction effects in the
prediction of children’s social skill development.
Parenting Styles
Many investigators have examined the relation of
parenting behaviours to children’s prosocial behaviour or
social competence, although rarely to specific social skills
outcomes. Research has indicated, for example, that
parental warmth, as seen in displays of affection and
closeness to the child, is related to positive social
outcomes for children (Eisenberg & Fabes, 1998),
particularly when in combination with adequate control
(Putallaz & Heflin, 1990). Parents who show such
behaviours act as models for their children, and use
affection as effective positive reinforcement for
appropriate behaviour (Baumrind, 1973).
Parenting behaviour which emphasises reasoning
may also be an important contributor to children’s
social development. Reasoning that orients children
to the impact of their actions on others is associated
with prosocial behaviour (Krevans & Gibbs, 1996)
Furthermore, parental use of inductive reasoning
also models appropriate behaviour and may
ultimately lead to the internalisation of the inductive
process whereby children begin to consider for
themselves the pro’s and con’s of acting in certain
ways (Bandura, 1986).
This study investigates a range of within-child factors
and family characteristics which are believed to be
influential in determining socially skilled behaviours.
In addition, this study also addresses the question of
whether interactions between the child and his/her
environment are important in social skill
development. The need to pursue interactional
models, and also to continue to improve upon additive
models of influence on children’s behavioural
development, has been noted by a number of theorists
(Bates, 1989). The current study considers a range of
potential contributors to level of social skills.
Specifically these are parenting style, temperament,
marital functioning, and ‘goodness of fit’ between
child characteristics and parental expectations.
Developing socially skilled behaviours enables a
child to interact positively with peers, parents, and
teachers and represents one of the most important
developmental accomplishments (Gresham, 2002).
Three domains of socially skilled behaviour which are
Temperament and Social Skills
Temperament refers to relatively stable geneticallybased individual differences in behavioural style that
are visible from early childhood (Pedlow, Sanson,
Prior, & Oberklaid, 1993; Sanson, Hemphill, &
Smart, 2002). Some research has considered the role
of temperament in social skills development. For
example, temperamental approach has been linked to
prosocial behaviour (see Stanhope, Bell, & ParkerCohen, 1987). It could be suggested that a child who
124
Social skills development in children
is high in temperamental approach (high sociability)
is particularly likely to develop assertive behaviours.
An approaching (sociable) child is likely to evoke
positive feedback from the environment, which in turn
leads to greater confidence and opportunity for social
interaction, and the development of assertive
behaviours. In addition, some previous research has
implicated temperamental persistence and inflexibility
as contributors to social skills development
(Eisenberg, Fabes, Bernzweig, Karbon, Poulin, &
Hanish, 1993).
Marital Relationship and Social Skills
The marital relationship is an aspect of family
functioning that has also been viewed as an important
contributor to children’s social development, with
most of the previous work in this area focusing on the
impact of marital conflict. Marital conflict is the way
problems, disagreements, and conflicts in the marital
relationship are dealt with and resolved. Parke, et al.
(2002) however, argue that some degree of marital
conflict is inevitable in most parental relationships
and is not detrimental to children’s functioning under
all circumstances. Less work has addressed the effect
of the quality of the marital relationship on social
skills development. (Stocker & Youngblade, 1999).
Marital quality refers to satisfaction, adjustment, and
happiness (Parker, 2002). Research indicates that
children whose parents reported lower levels of
marital satisfaction played less with their best friends
than children from more satisfied marriages (Gottman
& Katz, 1989) and school-age children whose parents
were unhappily married showed low levels of
sociability in school (Long, Forehand, Fauber, &
Brody, 1987).
The link between marital satisfaction and conflict,
and children’s social skill development may be via a
process of modelling (Bandura, 1986).
Goodness of Fit and Social Skills
It has been suggested that social development of a
child is influenced by the ‘fit’ between the child and
his/her environment (Thomas, Chess, & Birch, 1968).
A good ‘fit’ occurs when the child’s characteristics
(capacities, motivations and temperament) are able to
cope with the demands and expectations of the
environment in which he/she functions. Such
consonance between the child and environment
promotes optimal positive behavioural adjustment
(Chess & Thomas, 1996).
It could be suggested that a good ‘fit’ would be
likely to influence the development of socially skilled
behaviours. The child with better ‘fit’ with their
family is likely to have more positive interactions, and
show better adjustment, than the child with poor ‘fit’.
As a result, this child would have more opportunities
to develop and display sociable behaviours. Evidence
for the influence of goodness of fit on social skills
development was provided by Paterson and Sanson
(1999) using a sample of 5-year old children. They
found that better ‘fit’ was associated with higher
scores on combined parent/teacher ratings of overall
social skills. The strength of the ‘goodness of fit’
125
approach is that it attempts to capture the complexity
of the developmental process.
Interaction Effects and Social Skills:
Statistical Approach
Another method of investigating interactional
processes utilises a statistical approach. Thus, apart
from considering child temperament and family
factors in terms of direct linear effects on social
development, of additional interest to this study is the
investigation of multiplicative effects. Although
findings of significant multiplicative effects are still
relatively scant, interaction effects continue to be
investigated in the current developmental literature
using this statistical approach (see Lengua, 2002).
While the majority of findings indicate direct linear
relationships between child temperament and family
factors, and social behaviour, evidence is
accumulating
that
temperament-by-parenting
interactions are important contributors to the
developmental process (Sanson et al., 2002). Sanson
and Rothbart (1995) suggest that more complex views
of temperament and parenting may prove to be a good
deal more useful than searching for simple direct
associations. However, the number of studies that
have explicitly examined interaction effects is still
quite small.
Nevertheless, some studies have found evidence of
the importance of interaction effects in the
development of externalising behaviour problems (
Paterson & Sanson 1999), and conscience
(Kochanska, 1997). The relative lack of previous
literature addressing the role of interaction effects in
social skill development means that the current
study’s investigation in this area is somewhat
exploratory. However, the literature points most
strongly to the individual roles of temperamental
approach (sociability) and warm parenting, and
temperamental inflexibility and parental use of
inductive reasoning, as possibly contributing to the
development of social skills.
This study therefore investigates the concept of
interactions in two ways. First, a ‘goodness of fit’
measure is used to tap aspects of the child’s fit with
the environment, specifically with parental
expectations. Second, a statistical approach is used to
assess interaction effects via hierarchical multiple
regression.
From this literature review the following
hypotheses are proposed:
Hypothesis 1: It was hypothesised that higher levels
of temperamental approach (sociability), parental
warmth, inductive reasoning, marital satisfaction, and
a good ‘fit’ would be linked to higher levels of
assertion.
Hypothesis 2: It was further anticipated that lower
levels of temperamental inflexibility, and higher
levels of persistence, warmth, inductive reasoning,
marital satisfaction, and a good ‘fit’ would be linked
to greater cooperation.
Hypothesis 3: It was hypothesised that low
inflexibility and marital conflict, and higher
persistence, warmth, inductive reasoning, and a good
‘fit’ would be linked to higher self-control.
TERRETT & PUTTICK
126
Hypothesis 4: In relation to interaction effects, it
was anticipated that the combination of
temperamental approach (sociability) and parental
warmth would account for additional variance in
assertion, over and above the main effects of these
two variables.
Hypothesis 5: The combination of temperamental
inflexibility and parental use of inductive reasoning
was expected to account for additional variance in
self-control and cooperation, respectively, over and
above the main effects of these two variables.
Method
Participants
Participants were 61 Prep children (33 boys and 28
girls) who were recruited from four lower middleclass Melbourne Catholic primary schools. The
children’s ages ranged from 5–6 years.
Predictor Variables Child Temperament Child
temperament was measured using the Short
Temperament Scale for Children (Prior, Sanson, &
Oberklaid, 1989). This questionnaire consists of 30
items rated on a 6-point frequency scale (1 = almost
never, 6 = almost always). Three dimensions were
relevant for this study; approach (sociability, reaction
to novelty), persistence (attention skills and on-task
behaviour), and inflexibility (child’s negative
emotionality and adaptability). Each subscale contains
seven items with the exception of inflexibility which
has nine items. High scores reflect withdrawal, nonpersistence, and inflexibility, while low scores reflect
the positive end of each temperament dimension;
approach (sociability), persistence and flexibility.
Adequate internal consistencies (alpha coefficients)
have been reported for each subscale in children aged
5-6 years; .75 for approach, .80 for persistence, and
.61 for inflexibility (Prior et al., 2000).
Parenting Style Parenting style was measured by a
questionnaire developed by Sanson (1995) which
contains factors of warmth, punishment, and inductive
reasoning. This questionnaire consists of 25 items
rated on a 5-point frequency scale (1 = never, 5 =
always). Although this questionnaire measures three
parenting styles, only two were relevant for this study.
These were levels of parental warmth (positive
emotional tone in parent-child interactions), and
inductive reasoning (parents’ tendency to discuss with
their children reasons for rules and limitations). The
warmth subscale contains 10 items, and the inductive
reasoning subscale contains 7 items. The alpha
coefficients for the scales were; .82 for warmth and
.66 for inductive reasoning reflecting adequate
internal reliability for each scale (Sanson, 1995).
Marital Relationship Marital relationship was
assessed by a questionnaire combining two scales.
Specifically these were The Hendrick Relationship
Assessment scale (1988) which assesses marital
quality, and a short form of the O’Leary and Porter
Hostility Scale (1980) which assesses marital conflict.
This questionnaire consists of 16 items rated on a 5point frequency scale (1 = never, 5 = always). The
scale for marital quality (level of satisfaction and
happiness in the marital relationship) contains 9 items.
The scale for marital conflict (the couple’s response to
problems and disagreements in the marital
relationship) contains 7 items. The alpha coefficients
for the scales were >.80, reflecting high internal
reliability.
Goodness of Fit (‘Bother’ Index) The
questionnaire used was based on the work of Feagans
et al. (1991). This instrument assesses the degree of
‘fit’ between parents’ desires and their child’s
characteristics. Parents are presented with a list of 19
potentially undesirable child behaviours compiled by
Feagans et al., covering areas such as temper
tantrums, and shyness. Parents are first requested to
choose five behaviours they would find most
undesirable in young children. Parents are then asked
to rate their own child on each of the five behaviours,
on a 4-point frequency scale (1 = never, 4 = often).
The remaining fourteen items are not rated. Each
child’s ‘fit’ score is the sum of rating on these five
items, with a minimum score of five and a maximum
score of 20, with higher scores reflecting poorer fit.
Outcome Variable Parent rated Social Skills Rating
System (Elementary Level) The Social Skills Rating
System (SSRS; Gresham & Elliot, 1990) was used to
assess three social skills. Although this questionnaire
measures four social skills, only three were relevant
for this study. The social skills scale consists of the
following subscales: assertion (showing social
confidence and initiative), cooperation (being helpful
and following rules), and self-control (ability to cope
in situations of conflict and compromise). Each social
skill subscale contains 10 items. Each social skill
subscale is rated on a 3-point frequency rating scale (0
=never, 1 = sometimes, 2 = very often). Scores range
from 0 to 20 on each subscale. High scores reflect
high levels of social skills. The SSRS is a well
validated questionnaire with adequate internal
consistencies (alpha coefficients); .74 for assertion;
.77 for cooperation; and .80 for self-control (Gresham
& Elliot, 1990).
Procedure
Parents of the Prep children were asked to complete
five questionnaires regarding dimensions of child
temperament, parenting style, marital relationship,
‘goodness of fit’ and social skills. All parental
questionnaires were distributed to Prep children to
take home to their parents, and were returned to the
school office. Approximately 210 envelopes were
distributed, with a return rate of 29%.
Results
Mean scores and standard deviations for each variable
are shown in Table 1. Mean ratings on the
temperament dimensions were similar to those found
for the 1366 5-6 year old children in the Australian
Temperament Project (Prior et al., 1989), and
indicated moderate levels of approach (sociability),
persistence and flexibility. Parents reported high
degrees of warmth, and relatively high levels of
inductive reasoning, and in both cases there was little
Social skills development in children
variation in parents’ responses. In relation to the
marital relationship, parents reported high levels of
marital quality and lower levels of marital conflict.
The mean score for the ‘goodness of fit’ variable
reflected moderate ‘fit’ between child characteristics
and parental desires for this sample. Mean scores for
social skills of self-control and cooperation indicated
the children, on average, displayed socially skilled
behaviour a little more than ‘sometimes’. However,
the mean score for assertion showed that assertive
behaviour occurred more often than self-controlled
and cooperative behaviours, and was reported as
occurring close to mid-way between ‘sometimes’ and
‘very often’.
Table 1
Mean ratings and Standard Deviations for Parents’
Ratings of Temperament, Parenting, Marital
Relationship, Goodness of Fit, and Social Skills
Domain
Subscale
Temperamenta
Parentingb
Approach
Persistence
Inflexibility
Warmth
Inductive
Reasoning
M
SD
2.81 0.79
2.98 0.81
2.54 0.67
4.21 0.52
3.43 0.50
Marital Relationshipc
Marital Quality 4.28 0.55
Marital Conflict 2.17 0.48
Goodness of Fitd
11.66 3.09
Social Skill Outcomese Self-Control
1.18 0.27
Cooperation
1.29 0.31
Assertion
1.58
0.26
___________________________________________
Note. N = 61.
a Scores can range from 1 to 6; high scores indicate withdrawal
(not approach), nonpersistence, and high inflexibility.
b Scores can range from 1 to 5; high scores indicate high warmth
and high explanation.
c Scores can range from 1 to 5; high scores indicate high marital
quality and high marital conflict.
d Scores can range from 5 to 20; high scores indicate poor fit.
e Scores can range from 0 to 2; high scores indicate high frequency
of each behaviour.
Intercorrelations between each predictor variable
and each social skills outcome were calculated. These
results are presented in Table 2.
127
Table 2
Intercorrelations between Parents’ Ratings of
Temperament, Parenting, Marital Relationship,
Goodness of Fit and Social Skills
Subscale
Self-Control
Cooperation
Assertion
___________________________________________
1. Approach
-.12
-.24
-.57**
2. Persistence
.06
-.05
-.29*
3. Inflexibility
-.41**
-.21
-.40**
4. Warmth
.20
.25
.30*
5. Ind. Reasoning .20
.21
.15
6. Marital Quality -.02
.03
.04
7. Marital Conflict -.06
-.01
.02
8. Goodness of Fit -.39**
-.27*
-.35**
____________________________________
Note. N = 61
Low scores are positive end of temperament dimensions
*p<.05. **p<.01.
Table 2 indicates that lower inflexibility (ie., high
flexibility) scores and lower ‘fit’ scores (ie., good
‘fit’) were significantly correlated with higher selfcontrol scores. Lower ‘fit’ scores (ie., good ‘fit) were
significantly correlated with higher cooperation
scores.
Lower approach scores (ie., high approach), higher
parental warmth scores and lower ‘fit’ scores (ie.,
good ‘fit’) were significantly correlated with higher
assertion scores.
A series of hierarchical regressions were
performed for each of the following outcome
variables; self-control, cooperation, and assertion.
Step 1 involved entry of the hypothesised predictor
variables, e.g., in relation to assertion: approach,
warmth, inductive reasoning, ‘fit’, and marital quality.
Step 2 involved entry of the interaction term, for
assertion, for example, the interaction term approach
x warmth. This strategy allowed for the assessment of
both additive and multiplicative effects. The summary
results are reported in Table 3, 4 and 5.
Table 3 indicates that in Step 2, 33% of the
variation in the self-control scores were explained by
the linear combination of the predictor variables:
inflexibility, persistence, warmth, inductive reasoning,
‘fit’, marital conflict, and the interaction term
inflexibility x inductive reasoning. The overall
relationship for Step 2 of the model was significant,
F(7, 53) = 3.72, p < .05. At Step 2 when the
interaction term, inflexibility x inductive reasoning
was included, inductive reasoning and ‘fit’
contributed significantly to the model.
128
TERRETT & PUTTICK
Table 3
Summary of Hierarchical Regression Analysis
Predicting Self-Control from Temperament,
Parenting, Goodness of Fit, and Marital Conflict (N
= 61)
Outcome
Predictor
B
SE B β
___________________________________________
Self-Control Step 1
Inflexibility
-0.12 0.06 -.29*
Persistence
0.04 0.04 .13
Warmth
0.03 0.06 .06
Ind. Reasoning 0.08 0.07 .14
Fit
-0.03 0.01 -.30*
Marital Conflict 0.02 0.07 .03
Step 2
Inflexibility
0.41 0.28 1.03
Persistence
0.06 0.04 .19
Warmth
0.02 0.06 .04
Ind. Reasoning
0.51 0.23 .93*
Fit
-0.03 0.01 -.30*
Marital Conflict 0.01 0.07 .01
Inflex x Ind. R. -0.17 0.09 -1.35
___________________________________________
Note.Adjusted R² = .20; R2 = .28 for Step 1; hR2 = .05 for Step 2
*p < .05.
Table 4
Summary of Hierarchical Regression Analysis
Predicting Cooperation from Temperament,
Parenting, Goodness of Fit, and Marital Quality (N
= 61)
Outcome
Predictor
B
SE B β
___________________________________________
Cooperation Step 1
Inflexibility
-0.01 0.07 -.03
Persistence
-0.00 0.05 -.01
Warmth
0.10 0.08 .17
Ind. Reasoning 0.12 0.09 .19
Fit
-0.03 0.01 -.26
Marital Quality -0.03 0.08 -.06
Step 2
Inflexibility
-0.01 0.36 -.01
Persistence
-0.00 0.05 -.01
Warmth
0.10 0.08
.17
Ind. Reasoning 0.13 0.30 .20
Fit
-0.03 0.01 -.26
Marital Quality -0.03 0.08 -.06
Inflex x Ind. R. -0.00 0.11 -.02
___________________________________________
Note. Adjusted R² = .07; R2 = .16 for Step 1; hR2 = .00 for Step 2
*p < .05.
Table 4 shows that in Step 2, 16% of the variation
in the cooperation scores were explained by the linear
combination of the predictor variables: inflexibility,
persistence, warmth, inductive reasoning, ‘fit’, marital
quality, and the interaction term inflexibility x
inductive reasoning. The overall relationship for Step
2 of the model was not significant. At Step 2 when the
interaction term, inflexibility x inductive reasoning
was included, none of the variables contributed
significantly to the model.
Table 5
Summary of Hierarchical Regression Analysis
Predicting Assertion from Temperament, Parenting,
Goodness of Fit, and Marital Relationship (N = 61)
Outcome
Predictor
B
SE B β
___________________________________________
Assertion Step 1
Approach
-0.17 0.03 -.52*
Warmth
0.10 0.05 .20
Ind. Reasoning
0.01 0.05 .01
Fit
-0.02 0.01 -.29*
Marital Quality
-0.03 0.05 -.06
Step 2
Approach
-0.77 0.28 -2.40*
Warmth
-0.33 0.20 -.68
Ind. Reasoning
0.03 0.05 .06
Fit
-0.03 0.01 -.33*
Marital Quality
-0.05 0.05 -.12
Approach x Warmth 0.15 0.07 1.98*
___________________________________________
Note. Adjusted R² = .41; R2 = .46 for Step 1; hR2 = .05 for Step 2
*p < .05.
Table 5 indicates that in Step 2, 51% of the
variation in the assertion scores were explained by the
linear combination of the predictor variables:
approach (sociability), warmth, inductive reasoning,
‘fit’, marital quality, and the interaction term
approach x warmth. The overall relationship for Step
2 of the model was significant, F(6, 54) = 9.05, p
<.05. At Step 2 when the interaction term, approach
(sociability) x warmth was included, approach
(sociability) temperament and ‘fit’,
and the interaction term contributed significantly to
the model. Introducing the interaction term, approach
(sociability) x warmth explained an additional 5% of
the variation in assertion scores.
Further investigation of the significant contribution
of the interaction term was conducted. Four groups
were created based on the median splits of the two
variables approach (Mdn = 2.71) and warmth (Mdn =
4.20). The four groups were: high approach (high
sociability)/high warmth; high approach (high
sociability)/low warmth; low approach (low
sociability) /high warmth; and low approach (low
sociability)/low warmth. Mean assertion scores were
calculated for each group (Ms = 1.68, n = 16; 1.64, n
= 11; 1.53, n = 12; and 1.39, n = 14, respectively).
Social skills development in children
The interaction between temperamental approach
(sociability) and parental warmth is illustrated in
Figure 1.
1.8
Assertion
1.6
High Warmth
1.4
Low Warmth
1.2
1
Low
Approach
High
Approach
Figure 1
Mean ratings for assertion among children above
and below the median on Approach (sociability) and
whose parents were above and below the median on
Parental Warmth.
It is apparent from Figure 1 that the combination of
high approach (high sociability)/high warmth, and
high approach (high sociability)/low warmth reported
similar levels of assertion, regardless of the levels of
parental warmth (M = 1.68, M = 1.64, respectively).
The combination of low approach (low sociability)
with high levels of parental warmth was associated
with lower assertion than the two high approach
groups, (M = 1.53) but this group had higher levels of
assertion than the low approach (low sociability)/low
parental warmth group, which had the lowest levels of
assertion (M = 1.39).
Discussion
In summary, the findings show that ‘fit’ and parental
use of inductive reasoning were contributing factors
to level of self-control. None of the hypothesised
predictor variables was significantly related to
cooperation. Temperamental sociability and ‘fit’ were
significantly related to assertion as was an interaction
between temperamental approach (sociability) and
parental warmth.
The emergence of goodness of fit as a predictor of
self-control suggests that the interrelationship
between child behaviour and parental expectations is
important for the development of this social skill. This
finding supports the views of Thomas and Chess
(1977) and Feagans et al. (1991). The significant
contribution of inductive reasoning to self-control
supports the idea that such a parenting style models
appropriate behaviour and highlights the impact of a
child’s action on others.
In relation to cooperation, the results indicated that
none of the hypothesised predictor variables was
significantly related to cooperation. Future research
may consider different variables to those focused on
in the current study in order to shed more light on the
development of cooperation, for example, aspects of
the home environment and other aspects of parents’
129
behaviour separate to the parenting style dimensions
considered here.
The finding that temperamental approach
(sociability) contributed to assertion is consistent with
a number of previous studies which also found an
association between this temperament dimension and
positive social behaviour (eg., Stanhope et al., 1987).
The results of the current study, however, expand
upon previous research by showing a link between
temperamental approach (sociability) and the specific
social skill of assertion.
As with self-control, goodness of fit was also a
contributor to assertion, emphasising again the
important role congruence between parental demands
and child characteristics play in children’s social
skills development.
In addition, results revealed a significant
interaction effect for the combination of
temperamental approach (sociability) and parental
warmth. Specifically, level of parental warmth did not
impact upon the development of assertion in children
who were more temperamentally sociable. For
children temperamentally less sociable, however,
lower levels of parental warmth were associated with
lower levels of assertion, while higher parental
warmth was associated with higher levels of assertion.
This finding provides support for an interactional
model of development. Parallels can be drawn
between the current study and previous research by
Paterson and Sanson (1999), using a similar age group
of Australian children. They found that a combination
of high temperamental inflexibility and high punitive
parenting was particularly problematic in the
development of externalising behaviour problems.
Both studies suggest that it is children who are
temperamentally ‘vulnerable’ who are most affected
by the parenting style they are exposed to.
It should be noted that the correlational nature of
the current data limits claims regarding causality.
Nevertheless, the relationships found in this study
were consistent with the causal processes believed to
be occurring.
Another limitation of this study was the reliance
upon questionnaire data. However, while additional
forms of data, for example, behavioural observations
and teacher ratings, would have been desirable, the
questionnaires used generally had sound psychometric
properties.
A further limitation in this current study, relates to
the reliance on parents as the only source of ratings.
Rothbart and Bates (1998), however, comment on the
benefits of using parental ratings. For instance,
parents can report on children over a greater range of
contexts than is possible for an observer, and parents
have a wider knowledge base on which to draw when
making judgements about a child’s general behaviour.
Despite these limitations, a number of implications
can be derived from the findings of the current study.
Klein (1992) suggests that environments need to be
made more flexible so that a wider range of
temperaments can receive positive input. This
recognises the importance of a good ‘fit’ between
parental demands and child characteristics, to increase
a child’s chances of developing skilled social
TERRETT & PUTTICK
130
behaviours. Second, the identification of individual
predictors of specific social skills outcomes may be
useful in the development of social skill intervention
programs in order to improve their efficacy. In
addition, the significant interaction between
temperamental approach (sociability) and parental
warmth could provide guidance for effective parent
training programs. That is, by taking child
temperament and parental style and their
interrelationships into account, for example, it may be
possible to promote the benefits of parental warmth
for parents of children low in temperamental approach
(low sociability).
In conclusion, the results of the present study
clearly demonstrate the value of identifying specific
predictors of specific dimensions of social skills. In
addition, the support found for an interactional model
of development as shown in the regression analysis
(using a statistical approach) and in the prominence of
the ‘goodness of fit’ measure as a predictor of social
skills, suggests that interactional processes are
important to models of social skill development, and
that future research should continue to pursue
evidence of interaction effects.
References
Bandura, A. (1986). Social foundations of thought and
action: A social cognitive theory. Englewood
Cliffs,
New Jersey: Prentice Hall.
Bates, J. E. (1989). Applications of temperament
concepts. In G.A. Kohnstamm, J.E. Bates, & M.K.
Rothbart (Eds.), Temperament in childhood (pp.
321355). Chichester, England: Wiley.
Baumrind, D. (1973). The development of
instrumental
competence through socialization. In A.D. Pick
(Ed.),
Minnesota Symposium on Child Psychology (Vol.
7,
pp. 3-46). Minneapolis: University of Minnesota
Press.
Chess, S., & Thomas, A. (1996). Temperament:
Theory
and practice. New York: Brunner/Mazel.
Eisenberg, N., Fabes, R. A., Bernzweig, J., Karbon,
M.,
Poulin, R., & Hanish, L. (1993). The relations of
emotionality and regulation to preschoolers’ social
skills and sociometric status. Child Development,
64,
1418-1438.
Eisenberg, N., & Fabes, R. A. (1998). Prosocial
development. In W. Damon (Series Ed.) & N.
Eisenberg (Vol. Ed.), Handbook of child
psychology,
Vol. 3: Social, emotional, and personality
development (5th ed., pp. 701-778). New York:
Wiley.
Feagans, L. V., Merriwether, A.M., & Haldane, D.
(1991). Goodness of fit in the home: its relationship
to school behaviour and achievement in children
with
learning disabilities. Journal of Learning
Disabilities,
24, 413-420.
Gottman, J. M., & Katz, L. F. (1989). Effect of
marital
discord on young children’s peer interaction and
health. Developmental Psychology, 25, 373-381.
Gresham, F. M., & Elliot, S. N. (1990). Social Skills
Rating System Manual. Circle Pines, Minnesota:
American Guidance Service.
Gresham, F. M. (2002). Social skills assessment and
instruction for students with emotional and
behavioural disorders. In K. L. Lane, F. M.
Gresham,
& T. E. O’Shaughnessy (Eds), Interventions for
children with or at risk for emotional and
behavioural
disorders (pp. 242-258). Boston: Allyn and Bacon.
Hendrick. (1988). Relationship satisfaction:
Relationship assessment scale.
Klein, H. A. (1992). Temperament and self-esteem in
late adolescence. Adolescence, 27, 589-693.
Kochanska, G. (1997). Multiple pathways to
conscience
for children with different temperaments: From
toddlerhood to age 5. Developmental Psychology,
33,
228-240.
Krevans, J., & Gibbs, J. C. (1996). Parents’ use of
inductive discipline: relations to children’s empathy
and prosocial behaviour. Child Development, 67,
3263-3277.
Lengua, L. J. (2002). The contribution of emotionality
and self-regulation to the understanding of
children’s
response to multiple risk. Child Development, 73,
144-161.
Long, N., Forehand, R., Fauber, R., & Brody, G.
(1987). Self perceived and independently observed
competence of young adolescents as a function of
parental marital conflict and recent divorce. Journal
of Abnormal Child Psychology, 15, 15-27.
O’Leary & Porter. (1980). Marital conflict: Overt
hostility scale.
Parke, R. D., Simpkins, S. D., McDowell, D. J., Kim,
M., Killian, C., Dennis, J., Flyr, M. L., Wild, M., &
Rah, Y. (2002). Relative contributions of families
and
peers to children’s social development. In P. K.
Smith
& C. H. Hart (Eds.), Blackwell handbook of
childhood
social development (pp. 156-177). Oxford:
Blackwell.
Parker, R. (2002). Why marriages last: A discussion
of
the literature. Research Paper No. 28, Melbourne,
Australia: Australian Institute of Family Studies.
Paterson, G., & Sanson, A. (1999). The association of
behavioural adjustment to temperament, parenting
and family characteristics among 5-year-old
children.
Social Development, 8, 293-309.
Pedlow, R., Sanson, A. V., Prior, M., & Oberklaid, F.
Social skills development in children
(1993).
Stability
of
maternally
reported
temperament
from infancy to 8 years. Developmental
Psychology,
29, 998-1007.
Prior, M., Sanson, A., & Oberklaid, F. (1989). The
Australian temperament project. In G. A.
Kohnstamm, J. E. Bates, & M. K. Rothbart (Eds.),
Temperament in childhood (pp. 537-554),
Chichester,
England: Wiley.
Prior, M., Sanson, A., Smart, D., & Oberklaid, F.
(2000). Pathways from infancy to adolescence: The
Australian Temperament project: 1983-2000.
Melbourne, Australia: Australian Institute of Family
Studies.
Putallaz, M., & Heflin, A. H. (1990). Parent-child
interaction. In S. R. Asher & J. D. Coie (Eds.), Peer
rejection in childhood (pp. 189-215). New York:
Cambridge University Press.
Rothbart, M. K., & Bates, J. E. (1998). Temperament.
In W. Damon (Series Ed.) & N. Eisneberg (Vol.
Ed.),
Handbook of child psychology: Vol. 3. Social,
emotional, and personality development (5th ed., pp.
105-176). New York: Wiley.
Sanson, A. V. (1995). Childrearing questionnaire.
Unpublished scale. University of Melbourne,
Australia.
Sanson, A. V., & Rothbart, M. K. (1995). Child
temperament and parenting. In M. Bornstein (ed.),
Handbook of parenting: Vol. 4. Applied and
practical
parenting (pp. 299-321). New York: Erlbaum.
Sanson, A., Hemphill, S. A., & Smart, D. (2002).
Temperament and social development. In P. K.
Smith
& C.H. Hart (Eds.), Blackwell handbook of
childhood social development (pp. 97-116). Oxford:
Blackwell.
Stanhope, L., Bell, R. Q., & Parker-Cohen, N. Y.
(1987). Temperament and helping behaviour in preschool children. Developmental Psychology, 23,
347353.
Stocker, C. M., & Youngblade, L. (1999). Marital
conflict and parental hostility: Links with children’s
sibling and peer relationships. Journal of Family
Psychology, 13, 598-609.
Thomas, A., & Chess, S. (1977). Temperament and
development. New York: Brunner/Mazel.
Thomas, A., Chess, S., & Birch, H. G. (1968).
Temperament and behaviour disorders in children.
New York: New York University Press.
Correspondence to: Dr. Gill Terrett,
School of Psychology
Australian Catholic University, St Patrick’s Campus,
Melbourne Vic 3065 Australia
g.terrett@patrick.acu.edu.au
131
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 132-136
Attraction for Out-Group Members: Is There a Role for Attitude
Importance and Attitude Similarity?
Gill Terrett (g.terrett@patrick.acu.edu.au )
School of Psychology
Australian Catholic University, Melbourne, VIC 3065 Australia
Cecelia Winkelman (c.winkelman@patrick.acu.edu.au )
School of Psychology
Australian Catholic University, Melbourne, VIC 3065 Australia
Barry Fallon (b.fallon@patrick.acu.edu.au )
School of Psychology
Australian Catholic University, Melbourne, VIC 3065 Australia
on liking or on disliking? As with many findings in
psychology, it depends. For one thing, it depends on
the perceived status of the target.
For example, it takes little to create the impression
that an individual is a member of an in-group. Once
an in-group is created, in-group members tend to
perceive those not part of the in-group as forming the
out-group. The theory of minimal groups explains
how easy this is to bring about (Tajfel, Billig, Bundy,
& Flament 1971). Two examples are 1) simply telling
participants they are in Group A or Group B or 2)
telling them they are over-estimators or underestimators of the number of dots in a picture. It is this
easy to create in-groups and out-groups.
Once an in-group is formed, social identity theory
proposes that certain processes may be activated to
help an individual protect and enhance self-esteem
(Tajfel & Turner, 1986; Turner, 1987). Such
processes include identifying with the achievements
of one's in-group and showing in-group favouritism,
which is the tendency to discriminate in favour of
one's in-group over the outgroup. Showing
favouritism to other members of one's in-group can
also be understood as an indication of attraction or
liking.
Research on similarity has demonstrated the
tendency for individuals to assume similarity where
none may exist. This is the false consensus effect, in
which people tend to overestimate the extent to which
others share their opinions, attributes, and behaviours
(Ross, Greene, & House, 1977). The fact that people
tend to assume others hold similar opinions implies
that when one finds the other does hold similar
opinions, it comes as no surprise. After all, it was
assumed to be so. In other words, expectancy theory
plays a part in this argument for assumed similarity
among in-group members.
Diehl (1988) found support for combining the
theory of minimal groups with that of assumed
similarity. In Diehl's study, participants assumed
similarity of attitude for the in-group target person
more so than for the out-group target.
Chen and Kenrick (2002) investigated the effect of
both group membership and assumed similarity on
attraction/repulsion to both an in-group and an out-
Abstract
Working within the theoretical frameworks of social
identity theory and interpersonal attraction this research
investigated the role of the importance of attitudes and
attitude similarity on attraction to out-group members.
Seventy-five undergraduate psychology students
participated in this research. Attitude similarity and the
importance of the attitudes for the out-group member
were manipulated. Initial attraction based only on the
knowledge of the target group membership was
compared with ratings of attraction after participants
learned about the target’s attitudes. The hypothesis that
attitude similarity on important issues would show the
largest increase in level of attraction was not supported.
The data indicated that those participants who
perceived that their attitudes were dissimilar to the outgroup member reported a higher degree of attraction.
The hypotheses that attitude similarity, regardless of
attitude importance, would increase the level of
attraction was not supported. The results are discussed
in terms of methodological issues and alternative
theory.
Two hypotheses offer explanations for the finding that
people tend to be attracted to similar others. One,
based on the 'birds of a feather flock together'
metaphor, is the similarity-attraction hypothesis. It
proposes that people tend to like others who are
similar or perceived to be similar in some way (Byrne,
1971, Neimeyer & Mitchell, 1988). For example,
Byrne (1971) found evidence for increased attraction
to others who are perceived to share similar attitudes.
The second is the dissimilarity-repulsion
hypothesis. It proposes that dissimilarity leads to
repulsion. In challenging the similarity-attraction
hypothesis, Rosenbaum (1986) argued that the
emphasis on similarity is misplaced. Rather, the
emphasis should be on dissimilarity, because it is
dissimilarity that leads to disliking and decreases
attraction.
The argument then is between similarity and
dissimilarity. Which has the stronger effect and offers
the better explanation for attraction? Is it a matter of
moving toward the similar other or moving away from
the dissimilar other? Is the better explanation based
132
Attraction for out-group members
group member. They based their manipulation of
assumed similarity on expectancy theory. That is,
because people assume that an in-group member is
more similar than an out-group one, finding the ingroup member to be similar is no more that what was
assumed and expected. Hence the effect should be
little or no increase in attraction to the in-group
member.
On the other hand, finding that an out-group
member is similar is not assumed nor expected. The
impact should led to an increase in attraction for the
out-group member who is discovered to be similar.
Chen and Kenrick (2002) predicted that an increase in
attraction to a similar out-group member would be
explained by the similarity-attraction hypothesis. In
their study they found that attraction was stronger for
the similar out-group member than for the similar ingroup member and that overall similarity was more
important than group membership in increasing
participants' attraction to the target individual.
The in-group target, who is perceived to hold
dissimilar attitudes, violates the assumed similarity;
this is not what expectancy theory predicts. One
would expect the out-group target to hold dissimilar
attitudes, but not the in-group one. The impact should
lead to a decrease in attraction for the in-group target,
i.e. repulsion. Chen and Kenrick (2002) predicted a
decrease in attraction, i.e. repulsion, to a dissimilar ingroup member and explained it with the dissimilarityrepulsion hypothesis.
The current study took up that part of Chen and
Kenrick's (2002) study that found an increase in
attraction for the out-group member with similar
attitudes. This study was a partial replication of Chen
and Kenrick. The aim of this study was to compare
the similarity-attraction hypothesis with that of
dissimilarity-repulsion toward an out-group target. In
addition to group membership, this study incorporated
a manipulation of attitude importance. Would it
matter whether the attitude was important to the
participant?
The first hypothesis predicted that the effect of
attitude similarity on attraction to an out-group
member would depend upon the importance of the
attitude to the participant. In other words, it was
expected that similarity of attitudes on important
issues would predict greater attraction for an outgroup member and dissimilarity would predict greater
repulsion. The second hypothesis predicted that
attitude similarity, regardless of attitude importance,
would increase attraction to an out-group member.
133
variable was attitude importance, with its two levels
of important versus unimportant. The dependent
variable was change in attraction after a one week
interval in which the participants were given
information about how similar/dissimilar the target's
attitudes were.
Materials
The materials were comprised of three scales.
The Attitude Scale has 15 items measuring pro and
con attitudes on a six point scale. This scale was
based on Byrne's (1971) Survey of Attitudes and
included three of his items: student needs, family
finances and money. The other 12 items were
generated by the researchers on the basis of two
criteria, that they be topical or relevant to students.
The issues for these 12 items were: bar on campus,
gay/lesbian adoption, detention of refugees, tertiary
fees, death penalty, smoking in restaurants, Australia:
a monarchy, stem-cell research, legalisation of
marijuana, entrance exams, final examinations, and
cars in the city centre. Instructions requested that the
participant circle the number that best represented
his/her attitude. For example, the six choices for
detention of refuges were "I am very much in favour
of detention of refugees", "I am in favour of detention
of refugees", "I am mildly in favour of detention of
refugees","I am mildly against detention of refugees",
"I am against detention of refugees", "I am very much
against detention of refugees". The scale formed the
basis of the manipulation for the first independent
variable.
The Importance of Issue Scale (IIS) assessed the
importance of each of the 15 issues on the Attitude
Scale on a 10-point Likert scale, from 1(Not at all) to
10 (Extremely). This scale was constructed
specifically for this study to operationalise the second
independent variable.
The participants were a convenience sample of 75
undergraduate psychology students. The 66 women
and 9 men ranged in age from 19 to 47 years (M =
22.3, SD = 6.01).
The Interpersonal Judgment Scale (IJS) (Byrne,
1971) has six items on which each participant rates
the 'other person': intelligence, knowledge of current
events, morality, adjustment, personal feelings, and
working together on a project. The response format
for each item was a 7-point scale, ranging from 1 (the
most negative, e.g., "I believe that I would very much
dislike working with this person on a project") to 7
(the most positive, e.g., "I believe that I would very
much enjoy working with this person on a project").
For the purpose of measuring attraction toward the
target, only the two items dealing with attraction were
included in the scoring: Personal feelings and working
together on a project. A mean attraction score based
on the combined score for these two items was
calculated. The dependent variable was the difference
in attraction score between weeks 1 and 2.
Design
Procedure
This study used a 2 X 2 design. One independent
variable was attitude similarity, with its two levels of
similar versus dissimilar; a second independent
Participants were told that the study was on students'
views about different issues, and that a number of
institutions were involved. They were asked to
Method
Participants
TERRETT, WINKELMAN, & FALLON
complete the Attitude Scale in relation to their own
views on the issues presented. When this was done,
they were told that they were each paired with a
student from another university in the same city (the
target). The purpose of this manipulation was to
create minimal groups (Tajfel, Billig, Bundy &
Flament, 1971), in order that the participants would
perceive themselves as members of the in-group and
the student from the other university, the target, as a
member of the out-group.
Next each participant completed the Interpersonal
Judgment Scale with regard to their judgments about
the target. Finally, each participant was given the
Attitude Scale again; but this time, instead of their
own attitudes, they were asked to complete it in the
manner in which they thought the target student
would do so. After writing an identification code on
their questionnaires, along with their age and sex, the
questionnaires were collected. This ended the first
part of the manipulation.
One week later, each participant was given an
Attitude Scale and told it was completed by the target
student with whom he/she had been paired. In fact,
each of these Attitude Scales had been completed by
the researchers specifically for each participant in one
of the four following ways:
Similar ratings to the participant on issues of
importance, random ratings on issues of lesser
importance.
Dissimilar ratings to the participant on issues of
importance, random ratings on issues of lesser
importance.
Random ratings on issues of importance, similar
ratings to the participant on issues of lesser
importance.
Random ratings on issues of importance, dissimilar
ratings to the participant on issues of lesser
importance.
After reading this Attitude Scale, each participant
again completed an Interpersonal Judgment Scale on
the target student. The mean rating on the two items,
personal feelings (i.e., degree of liking) and working
together (desire to work together), in week one minus
the mean rating on these two items in week two
yielded a measure of change in attraction.
The participants were debriefed as to the nature of
the manipulation.
Results
Table 1 shows the mean change in attraction to the
target for the four groups.
Table 1
Mean Change in Attraction to Target
Attitude Similarity:
Similar
Dissimilar
Attitude Importance:
Important
Unimportant
M
SD
M
SD
.05
.71 -.14 .74
.30
.92
.78 1.73
The means show that for similar attitudes,
attraction decreased slightly for unimportant (M= .14) compared to important (M=.05) attitudes. Change
in attraction toward the target was greater for
dissimilar attitudes, and increased more for
unimportant attitudes (M=.78) than for important ones
(M=.30).
A 2-way (between groups) ANOVA indicated no
interaction effect between the two independent
variables, attitude similarity and importance of
attitude, on the dependent variable, change in
attraction toward the target, F(1, 71) = 1.74, p > .19).
One independent variable, attitude similarity, had a
significant main effect on the dependent variable,
change in attraction toward the target, F(1,71) = 5.25,
p < .02. The second independent variable, importance
of attitude was not significant, F (1,71) = 0.32, p >
.58.
Figure 1 displays the means for the change in
attraction toward the target for the four groups. The
error bars in the figure represent one standard error.
1.0
Similar
Estimated Marginal Means
134
Disimilar
.8
.6
.4
.2
.0
-.2
Important
Not Important
Figure
Mean change in attraction toward target on
similarity and importance of attitude
Discussion
The first hypothesis that the effect of attitude
similarity on attraction to an out-group member would
depend upon the importance of the attitude to the
participant was not supported. The second hypothesis
that attitude similarity, regardless of importance of the
attitude, would increase attraction to an out-group
member was not supported. The results of this study
do not support that of previous research by Chen and
Kenrick (2002), Byrne (1971), Rosenbaum (1986),
nor Neimeyer and Mitchell (1988). Nor can the results
be explained by the similarity-attraction nor the
dissimilarity-repulsion hypotheses.
The only significant effect, increased attraction
toward the out-group member with dissimilar
attitudes, especially unimportant ones, goes against
both the similarity-attraction and dissimilarityrepulsion hypotheses as well as previous research.
How to explain this unusual finding? An explanation
may be in the methodology, in that the manipulation
to create minimal groups may have been ineffective.
Attraction for out-group members
The participants were university students in their
second year of study. The researchers manipulated
group status by describing the target as a student from
a different university. Might these participants have
perceived the target as a fellow in-group member? If
so, then expectancy theory would explain how the
participants may have assumed the target person
should share similar attitudes. As Chen and Kenrick
(2002) found, expecting the in-group target to share
similar attitudes, should lead the participants to
assume similarity of attitudes and produce little to no
change in attraction. As Figure 1 shows, it was not
the similar attitudes that led to the greater increase in
attraction to the target. This part of the results is
consistent with those of Chen and Kenrick (2002) and
is explained by
expectancy theory. Assuming
similarity of attitude yields no great change in
attraction.
But if the participants were expressing their change
in attraction toward a perceived in-group member,
then why would their attraction increase for dissimilar
attitudes?
The dissimilarity-repulsion hypothesis
would lead one to expect just the opposite of this
result. Assuming the participants perceived the target
as an in-group member, they should have been
repulsed by the dissimilar attitudes. They were not.
It may be that the way dissimilarity of attitudes was
operationalised in this study partially explains this
result. Unlike Chen and Kenrick (2002), who used a
method of constant discrepancy to manipulate all
items on their attitude survey so that the target was
perceived as entirely similar or entirely dissimilar, in
this study the manipulation of dissimilarity, as
explained above, was moderate. Perhaps this
methodological artifact may help to explain the result
that attitude dissimilarity elicited such a strong change
in attraction. When the manipulated dissimilarity is
constant as was the case for Chen and Kenrick (2002),
then dissimilarity is too dissonant and the target is
repulsed. In that case the repulsion-dissimilarity
hypothesis prevails.
Further explanation may be found in social identity
and cognitive dissonance theories. These two
explanations might help to explain this result of
greater change in attraction for the in-group target
with dissimilar attitudes. First, social identity theory
(Tajfel & Turner, 1986): Could it be that the
participants were doing something akin to in-group
favouritism in an attempt to retain the in-group target
as an in-group member?
Rather than showing repulsion and distancing
themselves from the perceived in-group target with
dissimilar attitudes, the participants focused on the
target's similarity in having in-group status and
expressed attraction as a way of retaining the target's
in-group status. The fact that the change in attraction
was greater for the unimportant attitudes suggests it
was easier to ignore dissimilarity on the unimportant
attitudes than on the important ones. That is, the level
of dissimilarity in attitude must not have been
sufficient to marginalize or to reject the in-group
target.
The second explanation may be found in cognitive
dissonance theory (Festinger, 1957). Cognitive
135
dissonance theory draws on Heider's (1958) balance
and consistency theories that explain the way
individuals prefer both balance and consistency in
their relationships. Having perceived the target as an
in-group member, the participants may have
experienced cognitive dissonance at finding this
individual appeared to hold dissimilar attitudes.
In an attempt to escape the dissonance and reestablish the balance and cognitive consistency that
comes from liking similar others, the participants may
have rated the target with dissimilar attitudes as more
attractive. That is, once the target is perceived as an
in-group member, expectancy theory proposes that the
in-group member is expected to hold similar attitudes.
Finding dissimilar attitudes, when similar ones are
expected, should challenge the assumptions of
consistency theory, resulting in cognitive dissonance.
Experiencing dissonance is uncomfortable. To
escape the dissonance, something has to change. The
participants cannot change the target's dissimilar
attitudes, but they can change their own. Cognitive
Dissonance theory predicts this would happen. Yet,
measuring change in participants' attitudes toward the
topics was not part of the design for this study, so we
cannot say whether such attitude change happened.
Cognitive dissonance theory may explain the change
in attitude toward the target, that is, the increase in
attraction toward the target. Increasing attraction
toward the target with the dissimilar attitudes would
compensate for and minimize the dissonance caused
by the target's dissimilar attitudes.
In conclusion, this study attempted to investigate
the role of attraction for an out-group member in
terms of attitude similarity and attitude importance.
While the results showed a main effect for the
similarity-dissimilarity variable, the change in
attraction was in the opposite direction to that
predicted - greater change in attraction for the target
with dissimilar attitudes, especially unimportant ones.
Part of the explanation appeared to be in the
methodology, in that the target may have been
perceived as a member of the in-group, not the outgroup, and in that the perceived dissimilarity in
attitude may have been perceived as moderate. These
two facts led to explaining how the participants may
have worked harder to retain the target as a member
of the in-group and one in good standing. In the end
this study contributes to those attempts to understand
the conditions under which attraction or repulsion for
another individual, whether a member of one's ingroup or of the out-group are maintained. Those with
divergent opinions, whether from within the in-group
or from without, have the potential to be constructive.
Whether their ideas get a hearing or are treated as
destructive, a priori, may be an implication of this
study
investigating
similarity-attraction
and
dissimilarity-repulsion.
References
Byrne, D. (1971). The attraction paradigm. New
York: Academic Press.
Chen, F.F., & Kenrick, D. T. (2002). Repulsion or
attraction: Group membership and assumed attitude
136
TERRETT, WINKELMAN, & FALLON
similarity. Journal of Personality and Social
Psychology, 83, 111-125.
Diehl, M. (1988). Social identity and minimal groups:
The effects of interpersonal and intergroup
attitudinal similarity on intergroup discrimination.
British Journal of Social Psychology, 27, 289-300.
Festinger, L. (1957). A theory of cognitive dissonance.
Stanford, CA: Stanford University Press.
Heider, F. (1958). The psychology of interpersonal
relations. New York: Wiley.
Neimeyer, R.A., & Mitchell, K.A. (1988). Similarity
and attraction: A longitudinal study. Journal of
Social and Personal Relationships, 5, 131-148.
Rosenbaum, M.E. (1986). The repulsion hypothesis:
On the nondevelopment of relationships. Journal of
Personality and Social Psychology, 51, 1156-1166.
Ross, L., Greene, D., & House, P. (1977). The false
consensus phenomenon: An attributional bias in
self-perception and social-perception processes.
Journal of Experimental Social Psychology, 13,
279-301.
Tajfel, H., Billig, M., Bundy, R., & Flament, C.
(1971). Social categorization and intergroup
behavior. European Journal of Social Psychology,
1, 149-178.
Tajfel, H., & Turner, J.D. (1986). An integrative
theory of intergroup relation. In S. Worchel & W.G.
Austtrin (Eds.), Psychology of intergroup relations.
Chicago; Nelson-Hall.
Turner, J.D. (1987). Rediscovering the social group:
A self-categorization theory. Oxford: Blackwell.
Correspondence to:
Dr Gill Terrett
School of Psychology
Australian Catholic University
Locked Bay 4115, Fitzroy 3065
g.terrett@patrick.acu.edu..au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 137-140
Lying in relationships: Protecting ourselves or protecting our partners?
Alexandra E.West (a.west@patrick.acu.edu.au)
Barry J. Fallon (b.fallon@patrick.acu.edu.au)
School of Psychology
Australian Catholic University, Melbourne VIC 3065 Australia
Millar and Tesser (1988) suggest that deceptive
behaviour is “a mechanism by which a threat to the
maintenance of a social relationship is temporarily
resolved" (p.263). Similarly, De Paulo et al. (1996)
argued that lying (and deception) can sometimes be
used to benefit others, by avoiding conflict, protecting
another individual’s self-esteem and minimising hurt
feelings.
While it has been established that the use of
deception is common, its use in intimate relationships
is still not clearly understood, particularly with regard
to the motivations for engaging in deception with an
intimate partner, and the strategies people employ to
deceive their partners as opposed to other individuals.
Furthermore, research has often divided deception
into types or strategies (such as falsification, omission
and distortion), not all of which involve blatantly
lying to another person (DePaulo et al., 1996;
Peterson, 1996). Roloff and Cloven (1990) further
noted that although individuals may not overtly
deceive their partners, many still withhold
information.
DePaulo et al. (1996) investigated the use of lying
in everyday interactions, and using diary-study data,
examined rates of lying, types of lies, and
characteristics of the lies. The researchers believed
that lies are more frequently told to benefit oneself
rather than to benefit others, but that these lies are told
for psychological rewards, such as esteem and
affection. Their results showed that lies were indeed
commonly used. By developing a taxonomy of lies,
De Paulo et al. (1996) found that individuals told
more self-centered lies than other-oriented lies, which
supported their initial contention. In addition, the
researchers found a gender difference in the reasons
for lying, with women tending to tell more otheroriented lies, and less self-centred lies than the men in
the sample.
DePaulo et al. (1996) further suggested that often,
lies are told to avoid conflict and tension, and to
minimize hurt feelings. This is a position supported
by research by Peterson (1996), who found that
individuals often use deception as a method of
conflict avoidance.
In later research, DePaulo and Kashy (1998)
suggested that because lies (and deception) may be
used to benefit others, in that individuals lie to
compliment others, pretend to agree, or in an attempt
to show understanding, the underlying messages of
these lies may be supportive instead of threatening.
Metts (1989) also noted that because of the emotional
and behavioural interdependence between intimate
partners, the main function of deception in the context
of close relationships might be to maintain relational
harmony and cohesion.
Abstract
The present research presents an exploratory study on
the reasons that individuals give for their use of
deception. The aim was to explore the different reasons
individuals provided for engaging in deceptive
behaviour within their relationships, and to investigate
whether these reasons are differentially related to
different types of deceptive strategies. A series of openended questions about the use of deception in
relationships was completed by 152 participants (73
male, 79 female), who were a combination of
undergraduate psychology students, and individuals
recruited by the students. The categories of reasons
formed from the participants’ responses were Conflict
Avoidance, Protecting Self, Protecting Relationship,
Protecting
Partner,
Maintaining
Impressions,
Provoking Partner, and Manipulation. Protecting one’s
partner was the most common reason for engaging in
deception, and was given as a reason by almost half the
participants. Avoiding conflict was the next most
common reason, with provoking one’s partner being
the least common reason reported. Results showed no
significant differences between males and females in
their preference for giving different reasons for
engaging in deception. Further findings, regarding the
relationship between reasons for deception and specific
deceptive strategies, and their implications for the use
of deception in relationships are discussed.
The use of deception has been variously viewed as a
threat to social mores (Bok, 1978), a predictor of
negative life circumstances (DePaulo, Kashy,
Kirkendol, Wyer & Epstein, 1996), and as a social
skill (DePaulo & Jordan, 1982), and has long been an
area of interest for relationship researchers. Since
early recognition that deception played an important
role in communication in close relationships (see
Metts, 1989), research has shown that deception is
commonly used in social interactions. Defined as
deliberately invoking a false belief in another
individual that the communicator knows to be untrue
(Miller, Mongeau & Sleight, 1986), deception is used
to achieve communication goals, to facilitate
interactions, and to maintain impressions (DePaulo &
Kashy, 1998).
While disclosure is thought to have positive
outcomes, through enhancing trust and intimacy,
deception is generally regarded as damaging
(Finkenauer & Hazam, 2000), as it produces
suspicion, which in turn may undermine the values of
the relationship (Fehr, 2001). In a culture where
individuals are expected to be agreeable and sensitive
to other’s feelings, however, whether to tell the truth
or not can present something of a dilemma in social
interactions.
137
WEST & FALLON
138
Metts’ (1989) review of the literature revealed four
broad categories of reason individuals provide for
engaging in social deception – protecting one’s self,
protecting another person, protecting the relationship,
and to accomplish one’s goals during the interaction.
It remains open to debate whether typologies of
reasons for deception in social relationships
adequately represent the reasons for deception in
close relationships. Despite this, Metts (1989) found
in her own research, that partner-focused reasons were
the most frequently reported reason category,
followed by teller-focused, and relationship-focused.
To attempt to address the issue of how people in
relationships use deception with their partners, and
how this is related to their motivations for deceiving,
this research presents a study on deceptive strategies
and the reasons provided for their use.
Three research questions were asked: (1) What
kinds of reasons are given for engaging in deception,
and how frequently are they reported? (2) Do males
and females differ in the reasons they give for using
deception? (3) Are the deceptive strategies people
engage in differentially related to the reasons given
for engaging in deception?
Method
Participants
Participants were 152 individuals (73 male and 79
female), who were undergraduate psychology students
as well as members of the general population. The
mean age was 21.73 years, SD= 6.93 (range 18 to 52
years). Of the sample, 79.6% were exclusively dating,
7.9% were in defacto relationships, 7.2% were
married, 2% were engaged, and 3.3% reported ‘other’
relationships, most commonly reported as nonexclusive relationships. The mean length of
relationship was 30.5 months.
Materials and Procedure
Participants were presented with a questionnaire that
contained a number of relationship measures that
form part of a larger research study. Five strategies of
deception were presented to participants as part of this
questionnaire, and participants were asked to indicate
all those strategies they have used in their
relationship. The five strategies were based on
strategies devised by Boon and McLeod (2001). The
deceptive strategies were: (a) state something as true
that is not true; (b) make a true statement but say it in
a way to make your partner believe it is not true; (c)
communicate an untrue message non-verbally; (d)
deliberately omit information or fail to mention
something so as to lead your partner to a false belief;
and (e) exaggerate or distort information so as to lead
your partner to a false belief. Participants were
instructed to indicate as many options as applied to
them.
Participants were then asked, “What would be the
main reason/s that you used any of the above
strategies?” and given space to provide answers.
Results
The data presented in the current study was collected
as part of a wider study on relationships.
Of all the reasons provided by participants, there
were 28 types of reasons provided. On the basis of
previous research, seven broader categories of reasons
were identified: Conflict Avoidance (“To avoid a
fight”); Protecting Self (“Because I did something
wrong”); Protecting Relationship (“To protect the
relationship”); Protecting Partner (“To protect my
partner’s self-esteem”); Maintaining Impressions (“To
hide my true self”); Provoking Partner (“To annoy my
partner”); and Manipulation (e.g. “To get my own
way”). Table 1 shows the frequency with which each
reason was reported.
Table 1. Frequency of reasons
Reason
Conflict Avoidance
Protecting Self
Protecting Relationship
Protecting Partner
Maintaining Impressions
Provoking Partner
Manipulation
Frequency
35
22
12
66
22
7
10
Protecting One’s Partner was the most commonly
reported reason given for engaging in deception, with
over half of the sample reporting it. Conflict
Avoidance was the second most commonly reported
reason, followed by Protecting Self and Maintaining
Impressions, Protecting Relationship, Manipulation,
and Provoking Partner being the least commonly
reported reason.
A chi square analysis was run for each reason
category and gender. Results revealed no significant
differences between males and females in their
reasons given for engaging in deception.
For each of Boon and McLeod’s (2001) deceptive
strategies, chi square analyses were performed
between the deceptive strategy and each reason
category. For the strategy of lying, significant results
χ2 (1) = 6.46, p<.05 indicate that those who engaged
in this strategy, do not report Protecting Self as a
reason.
For the strategy of communicating an untrue
message verbally, significant results χ2 (1) = 6.84,
p<.01 indicate that those who used this strategy
tended to report Protecting Partner as their reason.
For the strategy of omission, significant results
were obtained for a number of reason categories.
Significant results indicate that those who used
omission did not report their reasons as Protecting
Self (χ2 (1) = 5.43, p<.05), Protecting the
Relationship (χ2 (1) = 7.42, p<.01), or Manipulation
(χ2 (1) =5.15, p<.05).
Lying in relationships
For the strategy of exaggeration/distortion,
significant results indicated that those who used this
deceptive strategy, did not report Protecting Self as a
reason (χ2 (1) =6.85, p<.01).
Discussion
The present study presents some exploratory work on
how the reasons people give for using deception are
related to the way they actually use deception. From
individual’s responses to open-ended questions, seven
categories of reasons for deceptive behaviour were
formed. Of these, Protecting One’s Partner was the
most commonly reported reason, while Provoking
One’s Partner was the least commonly reported
reason.
The finding that Protecting One’s Partner was the
most commonly reported reason is consistent with
Metts’ (1989) results. Metts (1989) did note however,
that the predominance of this category was largely a
result of the frequency of the specific reason “to avoid
hurting partner”. This finding was inconsistent with
the results presented by DePaulo et al. (1996),
however, as these researchers noted, social
relationship, rather than close relationships, were the
focus of their data collection.
The finding that Protecting One’s Partner was the
most common reason category would also seem to
reflect, from a Social Exchange Perspective, the
reciprocity involved in dyadic behaviour. As Cole
(2001) suggested, if the cost of being honest is too
high, for example causing hurt to one’s partner or
causing conflict, engaging in deception may be an
attractive alternative.
Surprisingly, the results did not show any
significant differences between males and females in
reasons they gave for engaging in deception. While
DePaulo et al. (1996) discussed social relationships,
they did find gender differences in reasons for lying.
Their results indicated that women told more otheroriented lies, and less self-centered lies than did the
men. Within close relationships however, the
interdependence, and the more intimate knowledge of
the partner, may mean that men and women are more
similar in their reasons for deceiving, particularly, if
as Metts (1989) suggests, the goal is to preserve the
relationship rather than either individual, and to
maintain relational cohesion.
In terms of how deceptive strategies were related to
the reasons provided, there were mixed results. Those
who used deceptive strategies such as lying,
exaggeration or omission, tended to not report that
they were protecting themselves. Further, those who
used omission further reported they were not
protecting their relationship, nor being manipulative.
The only deceptive category for which the results
indicated that people were providing a specific reason,
was for the strategy of communicating an untrue
message non-verbally, where individuals reported
they used this to protect their partner. These mixed
results need further clarification, which may involve
refining the categorization of reasons people provide
for their deception. It could be argued that even
though individuals may report using deception to
protect other people or to protect the relationship, the
139
underlying motivation is still to indirectly protect the
self from harm; either from harm to the self as a result
of upsetting one’s partner and the partner’s
subsequent reaction, or harm to the self as a result of
causing damage to an important relationship.
It may also be the categorization of the various
types of deception that needs refining. As Metts
(1989) pointed out, various researchers have provided
a number of different classifications of deceptive
strategies, and there has been no attempt to integrate
these.
While the results of the current study are not clear
in terms of elucidating how deceptive strategies may
be differentially related to the reasons people provide
for their use, they provide important information
about the reasons people give for deceiving those they
are close to, and provide a starting point from which
further analyses may be done.
References
Bok, S. (1978). Lying: Moral choices in public and
private life. New York: Pantheon.
Boon, S.D., & McLeod, B.A. (2001). Deception in
romantic relationships: Subjective estimates of
success at deceiving and attitudes towards
deception. Journal of Social and Personal
Relationships, 18, 463-476.
Cole, T. (2001). Lying to the one you love: The use of
deception in romantic relationships. Journal of
Social and Personal Relationships, 18, 107-129.
DePaulo, B.M., & Jordan, A. (1982). Age changes in
deceiving and detecting deceit. In R.S. Feldman
(Ed.), Development of nonverbal behavior in
children (pp.151-180). New York: Springer-Verlag.
DePaulo, B.M., & Kashy, D.A. (1998). Everyday lies
in close and casual relationships. Journal of
Personality and Social Psychology, 74, 63-79.
DePaulo, B.M., Kashy, D.A., Kirkendol, S.E., Wyer,
M.M., & Epstein, J.A. (1996). Lying in everyday
life. Journal of Personality and Social Psychology,
70, 979-995.
Fehr, B. (2001). The status of theory and research on
love and commitment. In G.J.O. Fletcher & M.S.
Clark (Eds.), Blackwell handbook of psychology:
Interpersonal processes (pp. 331-356). Malden,
MA: Blackwell.
Finkenauer, C., & Hazam, H. (2000). Disclosure and
secrecy in marriage: Do both contribute to marital
satisfaction? Journal of Social and Personal
Relationships, 17, 245-263.
Metts, S. (1989). An exploratory investigation of
deception in close relationships. Journal of Social
and Personal Relationships, 6, 159-179.
Millar, K.U., & Tesser, A. (1988). Deceptive behavior
in social relationships: A consequence of violated
expectations. The Journal of Psychology, 122, 263273.
Miller, G.P., Mongeau, P.A., & Sleight, C. (1986).
Fudging with friends and lying to lovers: Deceptive
communication in personal relationships. Journal of
Social and Personal Relationships, 3, 495-512.
Peterson, C. (1996). Deception in intimate
relationships. International Journal of Psychology,
31 (6), 279-288.
140
WEST & FALLON
Roloff, M.E., & Cloven, D.H. (1990). The chilling
effect in interpersonal relationships: The reluctance
to speak one’s mind. In D.D. Cahn (Ed.), Intimates
in conflict: A communication perspective (pp. 4976). Hillsdale, NJ: Erlbaum.
Correspondence to:
Alexandra West
School of Psychology
Australian Catholic University National
Locked Bag 4115
Fitzroy MDC, VIC 3065
Australia
a.west@patrick.acu.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 141-144
Factors Affecting the Acquisition of Skills in PET (Parent Effectiveness
Training): Language, Gender and Education
Christine D. Wood (Christine.Wood@utas.edu.au)
School of Psychology
University of Tasmania Hobart TAS 7001
John A. Davidson (John.Davidson@utas.edu.au)
School of Psychology
University of Tasmania Hobart TAS 7001
providing a positive approach to parenting in a way
that can be developed as the family grows, and which
does not have to be changed to meet the demands of
adolescence.
An investigation (Wood, 2003) comparing these
skills in Australian parents after PET training using
either the standard US workbook (n=71) or the
vernacular Australian workbook (n=81) found that
with either book PET parents improved significantly
more in each of them than control parents (n=81) with
no PET. Qualitative information in the comparison
study showed that there were different effects for
male and female parents taking PET in terms of their
objectives and possibly their acquisition of PET skills.
Intuitively it seemed possible that there may have
been different effects according to both workbook
used and to education levels, and that some light
might be thrown on the small differences shown
between the standard American and Australian
workbook. The aim of the present study was to
investigate whether, in the two PET groups there
would be differential effects according to workbook
used, and to gender and levels of education of
participants.
Abstract
An investigation (Wood, 2003) comparing the
interpersonal skills of Australian parents after PET
courses using either the standard US (n=71) or the
vernacular Australian workbook (n=81) and control
parents (n=81) found that with either book PET parents
improved significantly more in empathic listening, nonantagonistic assertiveness and conflict resolution than
control parents. The present study explored the effects
of workbook linguistic style, gender, and education
levels on the acquisition of skills in the PET groups. A
significant interaction between and gender and
workbook was found for empathic listening. Males
performed significantly better with the Australian
version which is simpler and more straightforward. In
non-antagonistic assertiveness parents with university
education showed significantly more improvement, and
in conflict resolution females showed significantly
more improvement than males. On three issues related
to parenting stress - child management, relationships
and parent self-management there were trends for
greater improvement by parents without a university
education, and for females in child management and
relationships. Females showed significantly greater
improvement in parent self-management issues. These
findings suggest that PET is of benefit across a wide
spectrum of parents, and that the now general use of the
Australian workbook has particular value for fathers.
Method
Participants
One hundred and fifty one parents were enrolled in 25
PET courses, taught by 11 trained instructors in six
Australian states.
Parent Effectiveness Training (PET, Gordon, 1976) is
an eight-week course in which the interpersonal skills
of empathic listening, non-antagonistic assertiveness
and conflict resolution are the major part of a package
for parents aimed at improving both the parent-child
relationship and the unacceptable behaviours of their
children. Having worked with Carl Rogers and a team
of colleagues for five years in a program of intensive
teaching and research, Gordon acquired a deep
understanding of the workings of interpersonal
relationships. Based on this experience and his own
clinical practice, Gordon formulated his (1970)
Theory of Healthy Relationships from which he
developed a course he hoped would enable parents to
prevent the problems which seemed to arise so
frequently in family life.
Empathic listening in PET is the foundation of a
flexible system of family communication training.
The Gordon model provides operational skills training
for emotional self-regulation and tutoring – a skill
later termed “emotional coaching” by Gottman (1997)
– targeting individual goals set by the parent, and
Dependent Variable Measures In order to
demonstrate theoretical understanding of the
principles of PET, its practical application and the
subsequent reduction of parental stress concerning
behavioural and relationship issues, participating
parents were asked to complete a battery of measures
including the Parent-Child Response Sheet (PCRS,
Wood & Davidson, 1987, 1994/95, 2003), Issues of
Parental Concern (IPC, Gordon, 1976; Wood, 1997)
and the Subjective Units of Distress Scale (SUDS,
Wolpe, 1990).
Procedure
The measures were completed in the first session,
before any teaching had taken place, and immediately
collected by the instructor who placed them in a preaddressed express pack to be despatched back to the
experimenter as soon as possible. No feedback was
141
WOOD & DAVIDSON
142
given. The same procedure was followed at the end of
the last session. The parents first filled out the PCRS
which measures the acquisition of listening and
assertive skills, and the resolution of conflict by
eliciting the written responses of parents to six
standardised parent-child interactions. They then set
out their family concerns, generally about their
children’s behaviours and their relationships with the
children, on the workbook “Setting Objectives” (IPC)
form. They were also asked to assess the amount of
stress they felt about each issue on a SUDS (Wolpe,
1990) scale of 0 (little or no stress) to 10 (major
stress) and to include this figure on the form. The
SUDS was developed as a simple but sensitive
measure which can be readily understood and
calibrated but is essentially quick and non-intrusive.
It was first formulated by Wolpe (1969) for use in
desensitisation procedures, and is frequently used
both in clinical practice and experimental studies in
sensitive areas such as Post Traumatic Stress
Disorder.
Results
In order to investigate the effects of workbook, gender
and education on improvement in acquisition of PET
skills as assessed by the PCRS, further analyses of
variance were performed on the improvement scores
(Post – Pre) for the two PET groups across the three
measures. Because of small numbers within cells,
education was categorised in two levels only –
university and non-university. The improvements in
PET parents’ PCRS scores for Active Listening,
Assertiveness and Conflict Resolution respectively,
broken down by gender, education and workbook are
presented in Table 1.
Results for the ANOVA on the improvement scores
showed that for Active Listening there were
significant effects for Workbook, with greater
improvement overall using the Australian PET
workbook, F(1, 143) = 4.94, p = .028, and gender,
with females performing better than males F(1, 143) =
4.17, p = .043. There was also a significant
interaction between gender and workbook F(1,143) =
3.99, p = .048.
The means for females were slightly better with the
Australian version (US workbook M = 10.99, Aus
workbook M = 11.19) but for males there was a
striking difference between the two workbooks (US
workbook M = 7.34, Aus workbook M = 11.15).
In Appropriate Assertiveness the only significant
effect (or interaction) was that of a main effect for
education F(1, 143) = 12.46, p = 001, with a greater
mean improvement in the expression of appropriate,
that is, Gordon’s non-antagonistic assertiveness, for
university-educated parents (M = 10.46) as compared
with non-university educated parents (M = 7.37).
There were no significant interactions.
In Conflict Resolution skills the only significant
effect was a main effect for gender F(1, 138) = 8.98,
p = .003, with females (M = 5.39) showing a
significantly greater improvement than males (M =
3.26). There was a trend towards a gender x
education interaction, where the greatest improvement
was for university educated females, F(1, 138) = 3.79,
p = .053.
Table 1
Improvement in Parents’ Active Listening,
Assertiveness and Conflict Resolution Depending on
Workbook, Gender and Education
n
M
SE
Improvement in Active Listening
Workbook
Stan PET
Aus PET
70
81
9.17
11.17
0.67
0.60
Gender
Female
Male
117
34
11.09
9.25
0.43
0.79
Education
Uni
Non-uni
70
81
10.29
10.04
0.65
0.63
Improvement in Parent Assertiveness
Workbook
Stan PET
Aus PET
70
81
8.53
9.30
0.65
0.59
Gender
Female
Male
117
34
9.54
8.29
0.42
0.77
Education
Uni
Non-uni
70
81
10.46
7.37
0.63
0.61
Improvement in Conflict Resolution Skills
Workbook
Stan PET
Aus PET
67
79
3.90
4.79
0.53
0.48
Gender
Female
Male
112
34
5.39
3.26
0.34
0.62
Education
Uni
Non-uni
67
79
4.57
4.08
0.51
0.49
Behavioural and Relationship Concerns
In all, 1044 individual items were reported on the
Issues of Parental Concern (IPC) forms by the sample
of 232 Australian parents, both PET experimental
parents and controls. There was a good deal of
Factors affecting parent effectiveness training
similarity and overlap, and all the items were
carefully scrutinised and collapsed in 60 categories, as
far as possible according to the parents’ perspective.
Issues perceived by the parents as relating to
children’s physical and verbal behaviours were
categorised as Child Management (78.3%), whereas
those they perceived as interpersonal problems were
categorised as Relationship (16.4%). Parents’ selfperceived role frustrations and inadequacies of their
own behaviours were placed in the category of Parent
Self-Management (5.3%).
Improvement in SUDS Ratings The improvements
in SUDS ratings by PET parents following the PET
course are presented in Table 2.
In the analyses of PET parents’ improvement in
SUDS ratings, the effects of differences in workbook
were excluded because of limitations of sample size
and the greater interest in the effects of gender and
education in improvement of satisfaction with the
previously identified parenting issues.
There were no significant effects for Child
Management issues but possibly a trend for greater
improvement by females F(1, 109) = 3.48, p = .065,
and by non- university parents F(1, 109) = 3.34, p =
.070.
In Relationship issues there was a trend for greater
improvement by non-university educated parents (M
= 3.43) as compared with university educated parents
(M = 1.95, F(1, 55) = 3.72, p = 059.
In Parent Self-Management issues females (M =
5.01) showed significantly greater improvement in
stress reduction than males (M = 1.88, F(1, 14) =
5.71, p = .032). There was a trend for non-university
educated parents, (M = 4.78) to show greater
improvement than university educated parents, M =
2.10, F(1,14) = 4.19, p = .060.
The study was constrained by the small numbers of
parents with low education levels, and possibly by the
preponderance of female to male parents.
Discussion
The results shown in this study are illuminating,
focusing as they do on the effects of workbook,
gender and education – that is upon how parents learn,
whether males or females learn better, and whether
previous education matters. Particularly important is
the finding that males acquire Active Listening skills
much better from the Australian workbook. The
levels of skill shown by males trained in Aus PET
relate well with results reported by Graham and Ickes
(1997), who found that men’s decoding ability
improved when they took into account not only facial
expression, but also body language, microexpressions and discrepancies between auditory and
visual cues, all of which are dealt with in PET
training, but are perhaps more easily extracted from
the simplified Australian version. Smith, Archer and
Constanzo (1991) found that with practice males were
able to reach levels of decoding ability which were at
least approximately equal to those of females, a
situation which is almost exactly reflected in Table 2,
showing males using PET equal females in terms of
their Active Listening improvement scores.
143
Table 2
Improvement in Parent SUDS Ratings on Child
Management, Relationship and Parent SelfManagement Issues Depending on Gender and
Education
n
M
SE
Child Management Issues
Gender
Female
Male
91
22
3.34
2.49
0.20
0.41
Education
Uni
Non-uni
56
57
2.49
3.33
0.30
0.34
Relationship Issues
Gender
Female
Male
45
14
2.91
2.46
0.37
0.67
Education
Uni
Non-uni
30
29
1.95
3.43
0.54
0.55
Parent Self-Management Issues
Gender
Female
Male
14
4
5.01
1.88
0.60
1.17
Education
Uni
Non-uni
13
5
2.10
4.78
0.67
1.13
It is not clear why university-educated parents show
higher scores for improvement in Assertiveness. It
must be remembered that greater improvement in
PET-style assertiveness relates to the acquisition of
non-antagonistic assertiveness, where the speaker
relies on honest personal feelings and concrete
description of behaviour rather than on blameful
statements. It does not necessarily imply change from
non-assertiveness to assertiveness. It would not be
correct to infer that that university educated parents
were found to be more assertive than non-university
parents, but rather that they had improved more in a
style of assertion which was less likely to lead to
defensiveness or opposition. It is possible that the
increased complexity involved in this kind of
response was a factor involved in the better
improvement shown by university educated parents.
Sharing one’s true feelings and reactions without
blocking others’ responses requires empathic skills
and leads to the establishment of authentic and true
relationships (Jakubowski & Lange, 1978). This skill
WOOD & DAVIDSON
144
is taught in PET as part of assertiveness (Changing
Gear) and requires the speaker to revert from
Assertiveness to Active Listening in dealing with
defensive replies. Several parents mentioned the
difficulty of acquiring this skill, but reported it was
extremely rewarding when it was successful. The
finding that female PET parents showed significantly
greater improvement in Conflict Resolution skills in
comparison with males is in accord with recent
research, and lends weight to the importance of
valuing women’s peacebuilding skills (McKay &
Mazurana, 2001).
The findings on parental satisfaction and reduction
of parents’ distress over their issues of concern are
worthy of note, although only one effect was
statistically significant. In relation to Parent-Self
Management, female parents showed significantly
greater improvement, and there was a trend for greater
improvement by non-university parents as compared
with those who were university educated. There was
a trend for greater improvement in relation to Child
Management issues by females and by non-university
parents. There was a similar trend for non-university
parents to show more improvement on Relationship
issues.
It is concluded that differential effects in the
acquisition of PET skills have been shown according
to the workbook used, and to the gender and
education levels of the participating parents.
Australian parents appear to benefit considerably
more from the use of the vernacular workbook, and
fathers in particular learn the foundational skill of
empathic listening significantly better. The trends for
non-university educated parents to improve more in
stress reduction about Child Management and Parent
Self-Management issues as well as Relationship
issues also suggest that the Australian workbook is
succeeding with a wide spectrum of parents. The
findings have implications for the teaching of PET in
Australia, and for the training of peace-related skills
in the broader community. PET skills relate closely
to the Conflict Resolution Model of Littlefield, Love,
Peck & Wertheim (1993) (Davidson & Wood, 2004;
Wertheim, Love, Peck & Littlefield, in press) and as
such should offer increased benefits not only to those
who acquire them but also to the communities in
which they live and work.
Further studies are needed to include a greater range
of parents, and to investigate the outcomes of PET
parent training for family relationships and the effects
on children in the school environment.
References
Davidson, J.A. & Wood, C.D. (2004). A conflict
resolution model. Theory into Practice, 43, 1, pp. 613.
Gordon, T. (1976). P.E.T.: Parent Effectiveness
Training Workbook. Solana Beach, CA:
Effectiveness Training Inc.
Gottman, J. (with De Claire, J.) (1997). The heart of
parenting: How to raise an emotionally intelligent
child. London: Bloomsbury Publishing.
Graham, T. & Ickes, W. (1997). When women’s
intuition isn’t greater than men’s. In W. Ickes (Ed.),
Empathic accuracy. New York: Guilford Press.
Jakubowski, P., & Lange, A. J. (1978). The assertive
option:
Your
rights
and
responsibilities.
Champaign, IL: Research Press.
McKay, S., & Mazurana, D. (2001). Gendering
peacebuilding. In D.J. Christie, R.V. Wagner &
D.D.N.Winter (Eds.), Peace, conflict and violence:
Peace psychology for the 21st century (pp. 223239). Upper Saddle River, NJ: Prentice Hall.
Smith, H.J., Archer,D., & Constanzo, M. (1991). “Just
a hunch”: Accuracy and awareness in person
perception. Journal of Nonverbal Behavior, 15, 318.
Wertheim, E. H., Love, A., Peck, C. & Littlefield, L.
(in press). Skills for resolving conflict: Creating
effective solutions through co-operative problem
solving (2nd ed.). Melbourne: Eruditions.
Wolpe, J. (1969/1990). The practice of behavior
therapy. 4th ed. New York: Pergamon Press.
Wood, C.D. (1997). Dr. Thomas Gordon’s Parent
Effectiveness Training workbook. Brisbane:
Effectiveness Training Institute of Australia Ltd.
Wood, C.D. (2003). How we talk to our children: An
evaluation of Parent Effectiveness Training for
emotional competence. University of Tasmania:
Unpublished doctoral thesis.
Wood, C. D., & Davidson, J. A. (1987). PET: an
outcome study. Australian Journal of Sex, Marriage
and Family, 8, 131-141.
Wood, C.D., & Davidson, J.A. (1994/1995).
Parenting for peace seven years on. London: Peace
Psychology Review, 1, 2, 123-129.
Wood, C.D., & Davidson, J.A. (2003). A paper and
pencil test for the assessment of PET
communication skills. Melbourne: Proceedings of
the 3rd Australasian Psychology of Relationships
Conference, 125-130.
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 145-149
Social Disconnection and Depression
Neil Welch (neil.welch@flinders.edu.au)
School of Education
Flinders University, Adelaide S.A. 5042 Australia
hypothesis. “It proposes that the extent to which one's
normal activities are restricted by a major life stressor
plays a central role in psychological adjustment, with
major disruptions in normal activities resulting in
poorer mental health outcomes” (Williamson &
Shaffer, 2000, p. 173). Since social disconnection
usually means activity restriction this model is
applicable to my hypothesis.
Since relationships are built around and upon shared
activities, activity restriction is likely to disrupt
peripheral and close relationships if the activity
restriction continues for any length of time. Given the
importance of activities in the development and
maintenance of friendships, activity restriction will
contribute to a constriction of active relationships
while impeding the replacement of lost friendships.
Activity restriction and social disconnection involve
a loss of interpersonal continuity. The importance of
continuity of relationships at times of crisis is
highlighted in the work of Omer & Alon (1994) and
is embedded in their ‘Continuity Principle’. This is a
generic principle for tackling the aftermath of
disasters and trauma. It “… stipulates that through all
stages of disaster, management and treatment should
aim at preserving and restoring functional, historical
and interpersonal continuities, at the individual,
family, organization, and community levels” (Omer &
Alon, 1994, p.273). This principle is promoted as a
guiding principle for dealing with trauma (Klingman
& Cohen, 2004). It is probably equally applicable to
individuals with depression and their families
although one wouldn’t want to try and preserve
chronically
dysfunctional
and
destructive
relationships.
Since this study focuses on individuals diagnosed
with clinical depression, it is important to make a few
comments about depression, its diagnosis and
treatment. Depression is generally viewed as a
medical disorder, as is evident from the diagnostic
criteria outlined in the Diagnostic and Statistical
Manual of Mental Disorders DSM-1V-TR (American
Psychiatric Association, 2000). The DSM criteria is in
terms of an individual’s symptoms not their social or
cultural circumstances, even though there is an
extensive research literature that highlights the
importance and role played by social events,
processes and relationships in the etiology of
depression (Halford & Bouma, 1997; Hops, Perry,
Davis, 1997; Joiner, Coyne, Blalock, 1999; Kiesler,
1996; Rudolph, Hammen, Burge, Lindberg, Herzberg,
Daley, 2000); and evidence of the effectiveness of
socially-focused interventions such as problem
solving counselling (Mynors-Wallis, 1996), marriage
counselling (Beach, 2001), and interpersonal
psychotherapy (Kopta, Lueger, Saunders, & Howard,
1999) to name but a few. It is therefore not surprising
Abstract
This paper explores the hypothesis that social
disconnection and its consequences underlie many of
the difficulties individuals and their families have in
adapting to negative life circumstances; and that
identifying the processes leading to and maintaining
social disconnection advances our understanding and
ability to design effective interventions. Social
disconnection is not seen merely as a consequence of
negative circumstances but as a process that actively
shapes events, relationships and emotional affect.
Results will be presented about processes identified as
promoting social disconnection in 12 case studies
involving individuals diagnosed with clinical
depression. Some of the processes identified relate to
relationship and life structure changes, social stigma
and rejection, treatment practices, and patterns of
interaction.
Negative life circumstances and events are frequently
associated with the development of depression, social
withdrawal and disconnection, and with difficulties in
adapting to change (Brown & Harris, 1978;
Williamson, 1998, 2000). Research into the interplay
between these factors and the processes underlying
their association is still at a relatively early stage and
yet would seem to be of critical importance in
developing a more holistic understanding of
depression, the development of symptoms and the
healing or recovery process.
In the research literature there are few references to
‘social disconnection’. Baldwin (1998) in her study of
a 12 year old lad, referred to ‘social disconnection’ in
terms of disrupted relationships with primary
caregivers, family of origin, friends, and associates. In
this current study, social disconnection is defined in
terms of constriction of life structure and
interpersonal range; where life structure refers to a
person’s weekly activity routine; and interpersonal
range to the number of individuals with whom that
person regularly interacts. Both life structure and
interpersonal range are expected to influence the
nature and quality of a person’s close relationships.
In the author’s work as a practicing psychologist
and researcher he had long been interested in the
processes that promote and inhibit social
disconnection at times of trauma and the influence of
these processes on mental health. This research is
concerned with exploring these processes in the lives
of individuals who have been diagnosed with clinical
depression. The hypothesis underpinning this research
is that social disconnection is not just a consequence
of depression but a process that shapes the course of
the ‘illness’.
The activity restriction model of depressed affect
(Williamson & Shaffer, 2000) is compatible with this
145
146
WELCH
that there is an overemphasis on medical treatments at
the expense of alternative approaches that endeavour
to understand depression from an interpersonal and
social perspective. Drug therapy may assist in
managing symptoms but can’t directly help people to
solve chronic social problems or to construct a more
meaningful and satisfying life for themselves and
their families. Failure to help people to change their
social circumstances may simply increase the
probability of future depressive ‘episodes’, and the
likelihood that other family members will develop
affective disorders (O'Connor, 2001).
Most of us live in partnerships, families and social
networks where our decisions and actions are made
with other people in mind. We live interdependent
lives, where our problems, difficulties and actions are
inextricably linked. Yet common treatments for
depression generally focus on individual clients,
ignoring the impact on partners, families and social
networks (Beck, Rush, Shaw, Emery, 1979; Holmes,
2002; Lee, 1998; Thase, 1996) e.g., recent research
has been found that children with a depressed parent
or parents have a much greater probability of
developing an affective disorder than those with nondepressed parents (Kim, Capaldi, & Stoolmiller,
2003). Similarly, treatments for depression generally
overlook the role partners and families play, or indeed
could play, in the treatment and recovery process,
even though there is research evidence that
improvements in key relationships, such as the marital
and parent-child relationship, are associated with
recovery from depression (Beach, 2001), whereas
ongoing marital problems are associated with a poor
prognosis for recovery (Rousanville, Weissman,
Prusoff, & Herceg-Baron, 1979). Thus while there is
evidence that the relationship context shapes recovery
from depression, we generally devalue its role in our
treatment practices. We usually persevere with
individual approaches at the expense of couple, family
and community based interventions.
In the context of this study it needs to be mentioned
that the two most widely reported and recommended
treatments for depression are psychotropic medication
and cognitive behaviour therapy. Yet there is evidence
that these treatments are not universal panaceas and
that there are frequent difficulties with treatment
acceptance, compliance and outcome e.g., “as many
as 68% of depressed patients refuse or non-comply
with pharmacological management (Ramana, Paykel,
Melzer, Mehta, & Surtees, 2003)” cited in (Burke,
2003, p.250). I have been unable to find statistics
dealing with rejection and non-compliance in relation
to Cognitive Behaviour Therapy.
The Study
The present study involved 12 case studies where
information was gathered by means of semi-structured
interviews and observations from multiple sources, at
multiple points in time (Yin, 2003). Participation in
this study was voluntary and based on informed
consent. These case studies focused on the
experiences of nine women and three men, aged from
29 to 58 years (Mean age = 36.1; SD = 8.6), who had
been diagnosed with clinical depression, and had been
recruited for this study via mental health workers and
general medical practitioners. Only one of the focus
individuals was from a non-English speaking
background and he had lived in Australia for more
than 30 years.
Information was collected through conversational
interviews using prepared protocols that were
designed to allow interviewees to talk freely about
their experiences and issues. Multiple interviews were
held with each focus individual over a period ranging
from 3 months to 2 years. Whenever possible
interviews were also held, on at least one occasion,
with a partner or a close friend; and in several
situations there were interviews with adult and
teenage offspring, co-workers and worksite managers.
The interviewer’s many years of clinical experience in
interviewing and counselling adults with depression
was a major strength of this study.
While the same set of information was sought from
each focus individual, the interview process had to be
responsive to fluctuating moods and setbacks being
experienced by the interviewees. This, at times,
necessitated considerable flexibility on the part of the
interviewer in terms of the order of questions, topics
covered, and the pace and length of interviews.
Information gathered was analysed in terms of :- the
temporal sequencing of events; the changing patterns
of activity and social interaction; the presence and
prioritisation of problems; the social consequences of
actions and events; the association between
interpersonal events and intrapersonal experiences;
patterns and differences across the case studies;
alternative explanations; and corroborating evidence.
The focus individuals lived in a diverse range of
household structures:
• two were living alone and had never married
• one was living with her school age child
• one was living with her pre-school child and de
facto
• three were living as married couples with preschool children
• three were living as married couples with schoolage children
• two were living alone as married couples, their
adult children having left home.
Their occupations were nurse, self employed
plumber, clerical worker, service counter assistant,
full time university student, truck driver, property
manager, cleaner, home duties. Six were currently on
WorkCover i.e., had experienced a work place injury
or trauma. Ten were parents.
The six on WorkCover were not clinically
depressed prior to the work related injury or trauma,
and only one of the six had had a prior episode of
clinical depression. This had been associated with a
break-up with her partner.
Three of the 12 focus individuals were hospitalised
for depression during the period of this study. This
enabled an exploration of the impact of
hospitalization from a concurrent perspective.
Social disconnection & depression
Results
As the time available for this presentation is limited,
the author will confine himself to reporting nine
observations.
1) In these case studies negative life circumstances
led to escalating social disconnection as a means
of reducing exposure to stressors and of avoiding
social problems. While social disconnection
reduced
exposure to some stressors and
enabled individuals to exercise a high level of
control over their ownpersonal lives, it exposed
them to other problematic experiences and
stressors. It negatively impacted on the lives of
partners, families, friends and workmates, and
their reactions were not always supportive and
positive; and because their lives were restricted it
resulted in these focus individuals having reduced
access and interaction with friends and relatives,
less companionship, less opportunities for social
support, less involvement in positive social
activities, and a life structure that centred around
life at home.
The advantages of social disconnection for
these focus individuals were achieved at the cost
of living an atypical life, being deprived of many
of life’s social pleasures, and an exacerbation of
feelings of loneliness, difference, boredom, threat
and depressed affect.
2) Fear of social rejection, ostracism, stigma and
harassment was a key factor shaping the everyday
lives of the focus individuals. Each had a personal
map of the places and people they sought to avoid
to minimise the risk of negative social
experiences. When they couldn’t avoid certain
public places it shaped how they operated in those
areas. e.g., individuals reported only quick visits
to shopping centres for very specific purchases or
transactions but not for general browsing or
socializing. The goal was to minimise contact
with people they knew to avoid embarrassing
questions, black looks, criticism, ostracism and
social pressure.
Stigma was perceived by this sample group as
endemic in the workplace and community. One
informant reported that she would rather have her
arm cut off than have depression, as this would be
more visible and more easily accepted by others
as a disability.
While focus individuals wanted to avoid people
they knew, this didn’t extend to strangers.
Everyday contact with strangers wasn’t a problem
providing this didn’t increase the risk of being
seen by familiar others.
It is noteworthy that focus individuals generally
reported a history of social rejection and
harassment that left them highly sensitized to
such experiences and prone to distrusting others.
3) Focus individuals had a profound sense of
difference from the others around them, and this
had shaped their lives and continued to do so.
They generally chose to avoid social situations
147
where they were too visibly different and in a
numerical minority of one. This severely
restricted their choices of activities, friends,
partners, and lifestyle, and constrained their
socializing and membership in groups. By
restricting their potential friendship pool and
proximity to others, it increased the likelihood
that they would form friendships and partnerships
with people they perceived as similar to
themselves.
This sense of difference was based very much
on self observation. Focus individuals saw
themselves as being different from others in a
wide range of ways including their experiences,
achievements, skills, circumstances, relationships
and history. Non-normative experiences and
normal experiences at non-normative ages
contributed to a sense of being a misfit
(Neugarten, 1976).
The presence of at least one similar companion
could make a critical difference as to whether
they participated in some social settings or not.
Unfortunately being labeled as depressed along
with its associated experiences added to their
sense of difference and increased their difficulty
in finding a similar companion. Even former
friends and companions were often seen
differently. As one person said “I couldn’t talk
with my friends in the same way I did before”.
An increasing sense of difference and lack of
companions for activities appears to be an
important factor in the process of social
disconnection.
4) Focus individuals were reliant on their partners
and immediate families for the great bulk of the
companionship and support they received during
depressive episodes. They generally lacked close
confidants outside the home and even when such
relationships existed they had very limited if any
ontact with them during depressive episodes.
Partners played a key role in this. Friends and
relatives frequently avoided visiting the home
when they disliked the partner, and a few partners
actively discouraged such visits and interactions
as unwanted intrusions.
How well focus individuals were supported was
thus dependent very much on the capacities and
functioning of partners and their willingness and
ability to provide companionship and support,
and manage family matters. When partners
themselves were depressed, ill, or preoccupied
with work or other matters, as they frequently
were in these case studies, this posed additional
problems for everyone and this contributed to the
hospitalization for depression of three of the
focus individuals.
Frequently dysfunctional relationships with
partners combined with difficulties in managing
children were an ongoing source of stress,
conflict and dissatisfaction. Yet little if anything
was done to help focus individuals work through
these issues. No doubt in part because they
weren’t directly asking for this assistance.
148
WELCH
Several women thought they would be much
better off in most ways if they separated from
their partner, but fear of life alone without a
partner and loss of the security of the family
home dissuaded them from taking this course of
action.
5) Lack of support from their own parents was
generally reported as a major source of
discouragement and interpersonal stress. This was
especially the case when there were young
children and provision of child care was a
problem. Lack of adequate child care support was
a major challenge for all depressed mothers.
Frequently their own parents declined to be
involved in ways that were expected and desired.
This resulted in family tension and rifts. Dislike
of the partner may have been one of reasons for
this reluctance and disengagement.
6) Focus individuals displayed interaction patterns
that impeded the development of close
friendships and the formation of extensive social
networks. They were generally very private
individuals who made frequent use of forms of
avoidance and disengagement to distance
themselves from social difficulties and conflict.
They usually didn’t seek help from others but this
simply reflects the nature of their close
relationships and coping style. Some might even
describe them as loners.
7) Primary responsibilities for child care, child
discipline and general family management were a
major source of stress for depressed mothers and
this severely impeded their recovery. Depressed
men in this study did not have the same primary
responsibilities and thus were able to avoid many
of the associated stressors.
Children who posed challenging behaviours
and were generally non-compliant caused much
parental distress. Failure to adequately manage
their child’s behaviour exposes parents to much
social criticism and embarrassment and adds to
their difficulties in obtaining child care support
and arranging play opportunities for their child.
This adds to the pressure on them to stay at home
to avoid social criticism, embarrassment and
feelings of being an incompetent parent.
Unfortunately if children become too homebound
this usually exacerbates their challenging
behaviours.
8) It was not just focus individuals that shaped the
process of social disconnection. Focus individuals
generally perceived significant others as
disengaging and avoiding them, but didn’t have
any good understanding as to why this happened
except possibly because they were not real
friends.
Irrespective of the reasons for this
disengagement and avoidance, failure of close
friends and family to stay closely connected
contributed to a sense of abandonment, isolation
and personal unworthiness.
9) Treatment practices added to the difficulties being
experienced by married couples and their
families, by ignoring the practical and
relationship needs of focus individuals and family
members. This exacerbated pre-existing problems
and undermined treatment cooperation and
compliance.
A few of the difficulties reported were:
• partner’s resentment at being excluded from an
active involvement in the treatment process
especially when they had to rearrange their
lives to manage the needs of the family. This
increased the tension within the home
• the doctor inadvertently encouraging social
disconnection at early stages in the treatment
through their advice and sickness certificates
• psychotropic medications affecting their social
functioning, dulling their senses and reducing
their libido
• adverse drug interactions between psychotropic
and recreational drugs interfering with a return
to their normal socializing
• hospitalisation seriously disrupting family
relationships and functioning, the lives of
partners and children, as well as increasing the
anxieties and insecurities of offspring, without
any steps being taken to ameliorate the
situation. Even when focus individuals were
discharged there was little if any prior
preparation of the family or provision of
immediate transition support
These practices added to the pressures on families,
eroding their relationships, undermining family
stability and cooperation with regard to the treatment
process, and reinforcing their sense of powerlessness,
isolation and social disconnection.
Conclusion
This study identified processes that led to social
disconnection in the lives of 12 individuals diagnosed
with clinical depression. It provides support evidence
for the hypothesis that social disconnection and its
consequences underlie many of the difficulties
individuals and their families have in adapting to
negative life circumstances. This has practical
implications especially for our understanding and
treatment of depression.
‘Treatment’ for depression needs to be seen as
involving a number of tasks of which symptom relief
and management is but one. Unless we assist
clinically depressed individuals to overcome issues
relating to social disconnection, loneliness, and
relationship problems, there is a strong likelihood that
they and their family members will live socially
restricted unhappy lives at high risk of recurring
episodes of depression.
While this study involved only a small sample and
there are limitations in generalising the results, it
nevertheless raises important issues worthy of
attention by both researchers and practitioners.
Social disconnection & depression
References
American Psychiatric Association. (2000). Diagnostic
and Statistical Manual of Mental Disorders: DSM1V-TR (4th ed.). Washington, DC: Author.
Anooshian, L. J. (2003). Social Isolation and
Rejection of Homeless Children. Journal of
Children & Poverty, 9(2), 115-134.
Baldwin, N. (1998). A Qualitative Case Study of a
Socially Disconnected Preadolescent in Foster
Care. Unpublished Honours, Flinders, Adelaide.
Beach, S. R. H. (2001). Marital Therapy for CoOccurring Marital Discord and Depression. In S. R.
H. Beach (Ed.), Marital and Family Processes in
Depression: A Scientific Foundation for Clinical
Practice (pp. 205-224). Washington, DC: American
Psychological Association.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G.
(1979). Cognitive Therapy of Depression. New
York: Guilford Press.
Brown, G. W., & Harris, T. (1978). Social Origins of
Depression: A Study of Psychiatric Disorder in
Women. London: Tavistock Publications.
Burke, L. (2003). The Impact of Maternal Depression
on Familial Relationships. International Review of
Psychiatry, 15(August), 243-255.
Halford, W. K., & Bouma, R. (1997). Individual
Psychopathology and Marital Distress. In W. K.
Halford & H. J. Markman (Eds.), Clinical
Handbook of Marriage and Couples Intervention
(pp. 291-321). Chichester: John Wiley & Sons.
Holmes, J. (2002). All You Need is Cognitive
Behaviour Therapy ? British Medical Journal, 324,
288-290.
Hops, H., Perry, B. A., & Davis, E. (1997). Marital
Discord and Depression. In W. K. Halford & H. J.
Markman (Eds.), Clinical Handbook of Marriage
and Couples Intervention (pp. 537-554). Chichester:
John Wiley & Sons.
Joiner, T., Coyne, J. C., & Blalock, J. (1999). On the
Interpersonal Nature of Depression: Overview and
Synthesis. In T. Joiner & J. C. coyne (Eds.), The
Interactional Nature of Depression: Advances in
Interpersonal Approaches (pp. 3-19). Washington,
DC: American Psychological Association.
Kiesler, D. J. (1996). Contemporary Interpersonal
Theory
and
Research:
Personality,
Psychopathology, and Psychotherapy. New York:
John Wiley & Sons, Inc.
Kim, H. K., Capaldi, D. M., & Stoolmiller, M. (2003).
Depressive Symptoms Across Adolescence and
Young Adulthood in Men: Predictions From
Parental and Contextual Risk Factors. Development
and Psychopathology, 15, 469-495.
Klingman, A., & Cohen, E. (2004). School-based
Multisystemic Interventions for Mass Trauma.
London: Kluwer Academic/Plenum Publishers.
Kopta, S. M., Lueger, R. J., Saunders, S. M., &
Howard, K. I. (1999). Individual Psychotherapy
Outcome and Process Research: Challenges
Leading to Greater Turmoil or a Positive Transition
? Annual Review of Psychology, 50, 441-469.
Lee, C. (1998). Alternatives to Cognition: A New
Look At Explaining Human Social Behavior.
149
Mahwah, New Jersey: Lawrence Erlbaum Assoc
Inc.
Mynors-Wallis, L. M. (1996). Problem-Solving
Treatment: Evidence for Effectiveness and
Feasibility in Primary Care. International Journal
of Psychiatry in Medicine, 26, 249-262.
O'Connor, R. (2001). Active Treatment of Depression.
New York: W. W. Norton & Company.
Omer, H., & Alon, N. (1994). The Continuity
Principle: A Unified Approach to Disaster and
Trauma. American Journal of Community
Psychology, 22(2), 273-287.
Ramana, R., Paykel, E. S., Melzer, D., Mehta, M. A.,
& Surtees, P. G. (2003). Aftercare of Depressed
Inpatients, Service Delivery, and Unmet Needs.
Social Psychiatry and Psychiatric Epidemiology,
38, 109-115.
Rousanville, B. J., Weissman, M. M., Prusoff, B. A.,
& Herceg-Baron, R. L. (1979). Marital Disputes
and Treatment Outcome in Depressed Women.
Comprehensive Psychiatry, 20(5), 483-490.
Rudolph, K. D., Hammen, C., Burge, D., Lindberg,
N., Herzberg, D., & Daley, S. E. (2000). Toward an
Interpersonal Life-Stress Model of Depression: The
Developmental Context of Stress Generation.
Development and Psychopathology, 12, 215-234.
Thase, M. E. (1996). Cognitive Behavior Therapy:
Manual for Treatment of Depressed Inpatients. In
V. B. Van Hasselt & M. Hersen (Eds.), Sourcebook
of Psychological Treatment Manuals for Adult
Disorders (pp. 201-231). New York: Plenum.
Williamson, G. M. (1998). The Central Role of
Restricted Normal Activities in Adjustment to
Illness and Disability: A Model of Depressed
Affect. Rehabilitation Psychology, 43(4), 327-347.
Williamson, G. M., & Shaffer, D. R. (2000). The
Activity Restriction Model of Depressed Affect:
Antecedents and Consequences of Restricted
Normal Activities. In G. M. Williamson, D. R.
Shaffer & P. A. Parmelee (Eds.), Physical Illness
and Depression in Older Adults: A Handbook of
Theory, Research, and Practice (pp. 173-200). New
York: Kluwer Academic/Plenum.
Yin, R. K. (2003). Case Study Research: Design and
Methods (Third ed.). Thousand Oaks, California:
Sage Publications.
Correspondence to:
Neil Welch
School of Education, Flinders University
GPO Box 2100 Adelaide 5001
Neil.Welch@flinders.edu.au
RELATIONSHIP TRANSITIONS
Proceedings of the APS Psychology of Relationships Interest Group 4th Annual Conference pp. 150-155
Adolescent Psychological Health and School Attitudes: The Impact of
Attachment Relationships
Ross B. Wilkinson (Ross.Wilkinson@anu.edu.au)
Monika Kraljevic
School of Psychology
The Australian National University, Canberra ACT 0200 Australia
internal ‘working models’ (Ainsworth, 1989; Bowlby,
1973). Parent-infant experiences provide a template or
set of “rules” for the kind of experience one might
expect in an adult relationship. That is, working
models are developed as guides for future
expectations and behaviour in interpersonal
relationships (Bowlby, 1973).
Early attachment relationships that are centred on
love and responsiveness are argued to lead to the
development of working models that contain beliefs
about the self as worthy of love and attention. In
contrast, insecure attachment relationships that are
based on inconsistency and unresponsiveness result in
working models that contain beliefs about the self as
unworthy and unlovable, and beliefs about others as
unavailable and antagonistic (Dekovic & Meeus,
1997). These beliefs have implications for how the
individual approaches relationships and social
interactions. The type of attachment an individual
develops with their primary attachment figure, and
therefore the information stored within the internal
working model, differ across individuals depending
on early childhood experience and attachment quality.
In this way, the quality and pattern of adult and
adolescent relationships is related to individual infant
relationship events (Collins & Read, 1990).
Abstract
The simultaneous and differential effects of parental,
peer, and intimate friendship attachment on attitudes to
school and psychological health were examined in a
sample of 520 adolescents aged from 13 to 19 years.
Based on recent extensions to attachment theory it was
predicted that parental, peer and intimate friend
attachment would impact on psychological health
variables, while peer attachment and intimate
friendship attachment along with depression and selfesteem would influence attitudes towards school. The
hypothesised pattern of relationships was evaluated
using structural equation modelling techniques.
Overall, there was only partial support for the
hypotheses. Results showed that parental attachment
and peer attachment were related to psychological
health but intimate friendship attachment was not.
Intimate friendship was the only attachment variable
that directly influenced school attitudes. The three
attachment variables were only weakly inter-correlated.
Implications with regard to extensions of attachment
theory are discussed.
The effect that our relationships with lovers, friends,
parents, and acquaintances have on our psychological
wellbeing has been a central concern of research
examining adult psychological adjustment for many
years. More recently, adolescents and their
interpersonal relationships have become the focus of
research into these important associations (e.g.
Armsden & Greenberg, 1987; Meeus, Osterwegel, &
Vollebergh, 2002; Wilkinson, in press). Adolescent
attachment research has continually shown that
relationship quality has an impact on an individual’s
psycho-social adjustment. A high quality of
interpersonal relationships in adolescence and
adulthood is strongly associated with higher levels of
self-esteem, less depression and better social
adjustment (Armsden & Greenberg, 1987; Engels,
Finkenauer, Meeus, & Dekovic, 2001; Greenberg,
Siegel, & Leitch, 1983). In the present study, the selfreported quality of adolescent relationships with
parents, peers and intimate friends is considered in
regards to the degree of association these have with
depression, self-esteem and school attitudes.
Attachment in Adolescence
Adolescent attachment research differs from research
focused on infants and adults because it has tended to
address relationship quality rather than specific
categories of attachment such as anxious or
ambivalent attachments (e.g. Ainsworth, 1989; Hazan
& Shaver, 1987). Researchers focus on the quality of
relationships in adolescence and how these impact on
psychological
outcomes.
Adolescent-parent
relationships are often subject to investigation, similar
to infant-parent relationship research. However, in
addition to this, adolescent peer attachments have
become of interest. Both parent and peer attachment
quality have been shown to be associated with
psychological health and adjustment in adolescence
(Armsden & Greenberg, 1987).
During adolescence, changes in attachment bonds
occur as individuals learn to develop and value nonfamilial relationships. Independence and associations
with others becomes increasingly important and
young adolescents begin to identify with and seek
support from peers more frequently. However,
parental attachments continue to remain salient and
constant throughout adolescence. Overall, the
majority of research has shown that parental
attachment is a significantly more powerful predictor
Attachment Continuity
The link between infant and adult attachment is based
on the belief that early attachment experiences
become cognitively encoded and referred to in
relationships throughout the lifetime. These past
experiences are stored in the form of selfrepresentations and expectancies, which develop into
150
Adolescent psychological health, school attitudes and attachment
of adolescent well-being than peer attachment quality,
although this varies depending on the definition of
well-being employed (Wilkinson & Walford, 2001;
Wilkinson, in press). Generally, studies tend to
indicate that insecure attachment is associated with
vulnerability for depression symptoms and low levels
of self-esteem (Engels et al., 2001; Heaven &
Goldstein, 2001; Muris, Meesters, Melick, &
Zwambag, 2001).
Adolescents have been found to be more susceptible
to symptoms of depression when they perceive low
levels of trust and communication in their attachment
to their parents, and high levels of alienation (Milne &
Lancaster, 2001; Muris et al., 2001). Adolescents are
more likely to have high self-esteem when parents are
supportive and interested in their activities, and low
self-esteem when parents are perceived as rejecting
(Collins & Read, 1990).
Attachment quality not only impacts on adolescent
psychological health, but also on other forms of
adjustment, such as attitudes towards school. Insecure
attachment has been found to be associated with
significantly lower reports of academic achievement
and ambition than secure attachment (Cooper, Shaver,
& Collins, 1998) and secure attachment has been
found to be a predictor of positive school adjustment
(Soucy & Larose, 2000). Cotterell (1992) found that
attachments with school mentors had a greater
positive relationship with school adjustment, than
parental attachments. He suggested that relationships
with peers and other adults at school offer support that
is different from support provided by parents. A more
recent study showed that the relationship between
parent attachment quality and school adjustment is
mediated by social support (Soucy & Larose, 2000). It
would seem from these findings that satisfaction with
support provided by peers and intimate friends has a
greater association with school attitudes and
adjustment than does parental attachment.
Intimate Friendship Attachment
Adolescent attachment research has tended to focus
on attachment to ‘peers’ with very few studies
concentrating on attachments to close or intimate
friends. However, Schneider and colleagues
(Schneider, Atkinson, & Tardif, 2001) have claimed
that intimate attachment relationships must be
considered as separate from broad peer friendships.
They argue that early and later intimate relationships
(for example, parent and romantic partners) should be
more similar to each other than they are to peer group
friendships. Intimate relationships in adolescence may
be developed with peers, romantic partners or siblings
and during this period friendships start to become
more significant. Self-disclosure becomes more
frequent as adolescents realise that their friends
understand the experience of adolescence and are
valuable sources of information and advice.
Adolescents with intimate friendships that are
satisfying, disclosing, and sources of companionship
report being less depressed and have higher selfesteem than adolescents whose relationships with
their friends are not as intimate (Buhrmester, 1992).
151
The Current Study
While adolescent attachment research has often
explored the relationship between parent and peer
attachments on a variety of aspects of psychological
well-being, studies have not often investigated
attachment relationships simultaneously and their
specific impacts on well-being and adjustment. By
incorporating and evaluating the results of studies of
adolescent attachment quality on psychological health
and school adjustment, it can be concluded that
different attachment relationships play important and
differing roles in adolescent well-being and
adjustment. The central proposal of the current study
is that the quality of parental attachments has a more
important role in psychological health, whereas
support provided by peers and close friends is more
influential in other areas of adjustment, such as
attitudes towards school.
The overall pattern of hypothesised relationships
are presented as a model in Figure 1. Based on
previous research, it is expected that parental
attachment will be directly related to depression
symptoms and self-esteem and indirectly related to
school attitudes via the psychological health variables.
Higher levels of parental attachment should lead to
less depression, higher self-esteem, and indirectly,
better school attitudes. Peer and intimate friendship
attachment, on the other hand, will be directly related
to self-esteem and school attitudes but only indirectly
School
Attitudes
Self-Esteem
Depression
+
+
+
+
Parent
Attachment
+
+
+
Peer
Attachment
Friend
Attachment
Note: Covariance paths and error terms not indicated
Figure 1. The Hypothesised Model
related to depression symptoms via self-esteem.
Better peer group and intimate friendships should
result in better self-esteem, a more positive school
attitude, and lower levels of depression. High levels of
depression and lower levels of self-esteem are
expected to negatively impact on school attitudes.
Method
Participants
527 ACT high school and college students
participated in the study during scheduled class times.
Of the participants, 248 (47 %) were female and 279
(53 %) were male. Participant age ranged from 13 to
19 years with a mean of 16.4 years (SD = 0.91). The
majority of the sample were identified from parent
WILKINSON & KRALJEVIC
152
Depression A 10-item depression scale consisting of
items from previously published scales was used to
measure symptoms of depression (Wilkinson, in
press). Participants responded to questions involving a
range of typical depression symptoms. Examples of
items include “I’ve felt too tired to do things”, “I’ve
felt unhappy or sad”, and “I’ve felt hopeless about the
future”. Items were responded to on a scale from 1 (a
lot of the time) to 3 (never). Items were recoded and
summed so that higher scores indicated more
depression. Internal consistency of the scale was high
at .92 (see Table 1).
occupation as of middle to upper socio-economic
status.
Procedure & Measures
Questionnaire booklets were distributed and
completed during scheduled class times and
participants were given instructions sheets outlining
ethical issues such as consent, confidentiality,
voluntary participation and anonymity.
Parent and Peer Attachment A short form of the
Inventory of Parent and Peer Attachment (IPPA;
Armsden & Greenberg, 1987) was used to assess
parent and peer attachment quality. Fifteen items from
the original scale were used for each of the Parent and
Peer Attachment scales. Participants were asked to
rate items on a five-point scale (1 = almost always or
always true, 5 = almost never or never true). Armsden
and Greenberg (1987) report high internal consistency
and test-retest reliability for the IPPA scales.
School Attitude A 10-item school attitude scale was
used to measure general attitudes towards teachers,
schoolwork, and ‘going to school or college’. Items
were taken from previously published scales and these
were responded to on a four-point scale of 1 (strongly
agree) to 4 (strongly disagree). Examples of items
include ‘I like being at school’, ‘Teachers often treat
you like you were kids’ and ‘I find school work easy’.
The scores for the items were coded and summed so
that higher scores indicated a more positive school
attitude. Internal consistency was high at .85 (see
Table 1).
Intimate Friendship Attachment A shortened
version of the Intimate Friendship Scale (IFS) was
used to measure intimate friendship attachment. The
IFS is a measure developed to assess the quality of
adolescent’s relationship with an intimate friend who
is close in age, such as a peer or sibling (Wilkinson,
Haigh, & Kraljevic, 2003). The short version of the
IFS consists of 15 items that are responded to on a
five-point scale ranging from 1 (strongly disagree) to
5 (strongly agree). Items include ‘When I have a bad
day my friend cheers me up’, ‘I am there for my
friend when he/she needs support’ and ‘I don’t need
to rely on my friend’. Participants were asked to keep
in mind one close friend when responding to the
items. Internal consistency and test-retest reliability
for the IFS is high (Wilkinson et al., 2003).
Results
Initially, the data were screened for missing values
and outliers using the Statistical Package for Social
Sciences (SPSS). Using the procedure outlined by
Tabachnick and Fidell (1996) with a Mahalonabis
distance criteria of p < .001, seven multivariate
outliers were identified and were deleted, resulting in
520 cases remaining.
Descriptive Statistics and Correlations
Means, standard deviations, reliability coefficients,
and correlations are presented in Table 1. Contrary to
expectations, the correlations between the attachment
variables, although significant, are rather weak. With
regard to the outcome variables of Depression, SelfEsteem and School Attitude, these all correlate
significantly with each other. In particular, Depression
has a moderate negative relationship with SelfEsteem. However, the relationship between School
Attitude and the two psychological health variables is
quite small. While both Parent and Peer Attachment
are correlated with the psychological health variables,
Self-Esteem
The
16-item
Self-Liking/SelfCompetence Scale – Revised Version (SLCS-R) was
used to assess self-esteem (Tafarodi & Swann, 2001).
Items are rated on a five-point scale from 1 (strongly
disagree) to 5 (strongly agree). After coding, the items
were summed so that higher scores indicate a higher
level of self-esteem. Internal consistency and testretest reliability for the SLCS-R are high (Tafarodi &
Swann, 2001).
Table 1
Correlations, Means, and Standard Deviations
1. Depression
2. Self-Esteem
3. School Attitude
4. Parent Attach.
5. Peer Attachment
6. Friend Attachment
Mean
SD
1.
.921
-.599*
-.280*
-.313*
-.247*
-.073
18.81
4.04
2.
3.
4.
5.
6.
.902
.229*
.256*
.290*
.032
52.04
10.72
.850
.141*
.239*
.370*
25.21
3.96
.915
.257*
.183*
48.63
12.55
.854
.248*
53.93
12.24
.910
53.42
9.76
* p < .01; Cronbach’s alpha presented on the diagonal
Adolescent psychological health, school attitudes and attachment
Friend Attachment is not. All of the attachment
variables were positively related to School Attitude.
of Depression as expected. Both Depression and SelfEsteem evinced small effects on School Attitudes.
Structural Equation Modeling (SEM)
The hypothesized model of the relationship between
the attachment measures and outcomes measures was
evaluated using (SEM) techniques implemented in the
AMOS 4 computer program (Arbuckle & Wothke,
1999). The model was evaluated using Maximum
Likelihood Estimation and model fit indices.
Modification indices were examined and parameters
were freed or fixed as appropriate to generate a final
model that maximized fit while retaining theoretical
coherence. In accordance with current practice in
reporting the assessment of SEM models, a number
of different fit statistics were employed to evaluate
the fit of the models (Tabachnick & Fidell, 1996).
The χ2, Adjusted Goodness of Fit Index (AGFI),
Comparative Fit Index (CFI), and Root Mean Square
Approximation Error (RMSEA) were selected as
appropriate fit indices.
The hypothesised model tested is presented in
Figure 1. Note that errors terms and covariances
between exogenous variables are not shown in the
figure but were included in the tested model. After
analysis, estimated paths with standardized weights
of less than .10 were deemed to be of minor
importance and were deleted from the model
irrespective of whether they were significant. The
reduced final model was then reanalyzed and is
presented in Figure 2.
Overall, the fit statistics indicated that the final
model was a good fit to the data. The AGFI (.975) and
CFI (.991) statistics were both above 0.95 indicating
that the model fitted the data. The χ2 (11.80, df = 5, p
= .10) was non-significant, indicating a good fit. The
RMSEA (.041) was below the recommended value of
0.05 that would indicate a good fit (Arbuckle &
Wothke, 1999). Squared multiple correlation
coefficients from the final model indicated that 12%
of the variance in Self-Esteem, 34.3% of the variance
in Depression, and 21.4% of the variance in School
Attitudes were accounted for. Similar to the zero
order correlations, the exogenous attachment variables
produced low, though significant, inter-correlations in
the final model. Parent and Peer Attachment were
correlated at .257, Parent and Friend Attachment at
.183, and Peer and Friend Attachment at .248.
Examination of the path coefficients in the final
model (Figure 2) reveal that Parent and Peer
Attachment evinced small direct and positive effects
on Self-Esteem with Parent Attachment also having a
small ameliorating effect on Depression. Both Parent
and Peer Attachment produced negligible, though
positive indirect effects on School Attitude (< .1).
There was a small to moderate positive effect of
Friend Attachment on School Attitudes, although this
variable did not influence Self-Esteem as predicted
and this path was removed from the model.
Interestingly, and contrary to expectations, the path
from Peer Attachment to School Attachment was also
not large enough to warrant retaining in the final
model. Self-esteem was a moderate negative predictor
153
Discussion
Overall, the hypotheses in relation to parental
attachment were supported while there was only
partial support for the hypothesized pattern of
relationships between peer and friend attachment and
the outcome variables. The quality of the parental
attachment relationship influenced both of the
psychological health variables and attitudes towards
school as predicted. Although, the weight of the paths
generated in the model were quite modest.
School
Attitudes
-.19
.11
Depression
-.51
Self-Esteem
.35
-.18
Parent
Attachment
.19
.24
Peer
Attachment
Friend
Attachment
Note: Covariance and error terms are not indicated
Figure 2
Final Model
The results with regard to the quality of peer and
intimate friendship attachment were not quite as
expected. While peer attachment did influence selfesteem as anticipated, the direct relationship with
school attitudes was not supported. Further,
examination of the indirect effects, of peer attachment
on school attitudes via self-esteem and depression,
indicated no significant relationship. The quality of
intimate friend attachment, on the other hand, had a
direct and positive impact on school attitudes as
predicted, but did not have the positive impact on selfesteem that was expected.
Generally, the results of the SEM analysis are
consistent with what was revealed by the zero order
correlations. Essentially, there are quite modest
relations between the variables, with the exception of
the moderate relationship between depression and
self-esteem. The weak relationships found in this
study between the three attachment measures is
somewhat surprising and may be the key to
understanding why the original model with regard to
peer and friend attachment did not function as
predicted. The relative lack of relationship between
the attachment measures is not consistent with the
notion that cognitive internal working models
strongly influence global attachment patterns.
Although there was some relationship between what
might be seen as the core attachment construct of
parental attachment and the other two measures of an
expanded attachment network, peer and intimate
friend attachment, it is clear that these constructs are
characterized here more by their independence than
154
WILKINSON & KRALJEVIC
their interdependence. It is difficult to conclude,
therefore, that participants in this study could be
characterized as having a particular attachment
‘style’. Rather, the quality of the attachment
relationship seems to depend a great deal on whether
adolescents are asked to think about their parents, best
friends, or peers.
The different attachment relationships also seem to
each play somewhat of a different role in relation to
the particular outcome variables. Only parent
attachment had a direct role in relation to depressive
symptoms while only friend attachment had a direct
role on attitudes towards school. Consistent with
previous findings (Wilkinson, in press) the quality of
peer relationships was particularly related to
adolescent self-esteem. It seems to be the case that as
we begin to move beyond the often simplistic
application of attachment theory and examine
different forms of attachment relationships we are
seeing the underlying complex nature of attachment
relationships and their implications for adolescent
adjustment (Meeus, et al., 2002).
The results of this study indicate that the quality of
relationships with parents and a group of friends is
more important for psychological health in
adolescence then having a high quality intimate
friendship with someone similar in age. Yet, the
results also indicate that having a close intimate friend
is associated with a more positive school attitude. To
some extent this finding can be seen as supporting
Cotterell’s (1992) earlier research indicating that
mentors play a more important role in positive school
adjustment than parents. In this case, having a close
friend that can be confided in, relied upon, and turned
to when times are stressful is associated with a more
positive attitude towards the school environment.
The data presented here are, of course, correlational
and cross-sectional in nature, and thus this limits the
extent to which any claims of causal relationships can
be justified. Longitudinal research would certainly
overcome some of these limitations and would have
further benefits in enabling an examination of the
changes in attachment patterns and networks in
adolescence over time. It is also worth noting that the
work presented here did not explicitly examine sex
differences in adolescent attachment patterns. There is
some evidence that there are significant differences in
attachment to mothers and attachment to fathers for
female and male adolescents (Wilkinson & Parry, in
press). Given the present findings in relation to the
relative independence of measures of attachment to
different significant others, exploration of sex
differences and interactions may add further to our
understanding of the complexity of attachment in
adolescence.
No matter at what stage of the life-span we are in,
the quality of our interpersonal relationships impacts
on our psycho-social functioning. In recent times
attachment theory has been strongly promoted as an
organising framework for understanding why this is
so. There is mounting evidence, however, that the
simplistic application of attachment theory tenets to
developmental periods outside of infancy is
unsatisfactory. Adolescence, as a period of expanding
social networks and shifting interpersonal needs, is an
area of research in which the complexity of multiple
interpersonal attachments is becoming apparent. The
challenge is to evolve attachment theory to account
for this level of complexity.
References
Ainsworth, M. D. S. (1989). Attachments beyond
infancy. American Psychologist, 44, 709-716.
Arbuckle, J. L., & Wothke, W. (1999). AMOS 4.0
user's guide. Chicago: SPSS Inc.
Armsden, G. C., & Greenberg, M. T. (1987). The
inventory of parent and peer attachment: Individual
differences and their relationship to psychological
well-being in adolescence. Journal of Youth and
Adolescence, 16, 427-454.
Bowlby, J. (1973). Attachment and loss, Volume 2,
Separation. New York: Basic Books.
Burhmester, D. (1992). Intimacy of friendship,
interpersonal competence, and adjustment during
preadolescence
and
adolescence.
Child
Development, 61, 1101-1111.
Collins, N. L., & Read, S. J. (1990). Adult
attachment, working models, and relationship
quality in dating couples. Journal of Personality
and Social Psychology, 58, 644-663.
Cooper, M. L., Shaver, P. R., & Collins, N. L. (1998).
Attachment styles, emotion regulation, and
adjustment in adolescence. Journal of Personality
and Social Psychology, 74, 1380-1397.
Cotterell, J. L. (1992). The relation of attachments and
supports to adolescent well-being and school
adjustment. Journal of Adolescent Research, 7, 2842.
Dekovic, M., & Meeus, W. (1997). Peer relations in
adolescence: Effects of parenting and adolescent's
self-concept. Journal of Adolescence, 20, 163-176.
Engels, R. C. M. E., Finkenauer, C., Meeus, W. H. J.,
& Dekovic, M. (2001). Parental attachment and
adolescents' emotional adjustment: The associations
with social skills and relational competence.
Journal of Counseling Psychology, 48, 428-439.
Greenberg, M. T., Siegel, J. M., & Leitch, C. J.
(1983). The nature and importance of attachment
relationships to parents and peers during
adolescence. Journal of Youth and Adolescence, 12,
373-386.
Hazan, C., & Shaver, P. (1987). Romantic love
conceptualized as an attachment process. Journal of
Personality and Social Psychology, 52, 511-524.
Heaven, P. C. L., & Goldstein, M. (2001). Parental
influences and mental health among some
Australian youth: Crosscultural analysis. Australian
Journal of Psychology, 53, 170-175.
Meeus, W., Osterwegel, A., & Vollebergh, W. (2002).
Parental and peer attachment and identity
development in adolescence. Journal of Adolesence,
25, 93-106.
Milne, L. C., & Lancaster, S. (2001). Predictors of
depression in female adolescents. Adolescence, 36,
207-223.
Muris, P., Meesters, C., van Melick, M., & Zwambag,
L. (2001). Self-reported attachment style,
attachment quality, and symptoms of anxiety and
Adolescent psychological health, school attitudes and attachment
depression in young adolescents. Personality and
Individual Differences, 30, 809-818.
Schneider, B. H., Atkinson, L., & Tardif, C. (2001).
Child-parent attachment and children's peer
relations: A quantitative review. Developmental
Psychology, 37, 86-100.
Soucy, N., & Larose, S. (2000). Attachment and
control in family and mentoring contexts as
determinants of adolescent adjustment to college.
Journal of Family Psychology, 14, 125-143.
Tabachnick, B. G., & Fidell, L. S. (1996). Using
multivariate statistics (3rd ed.). New York: Harper
Collins.
Tafarodi, R. W., & Swann, W. B. (2001). Twodimensional self-esteem: Theory and measurement.
Personality and Individual Differences, 31, 653673.
Wilkinson, R. B. (in press). The role of parental and
peer attachment in the psychological health and
self-esteem of adolescents. Journal of Youth &
Adolescence.
Wilkinson, R. B., Haigh, E., & Kraljevik, M. (2003).
Psychometric properties of the Intimate Friendship
Scale: A preliminary analysis. Unpublished
manuscript.
Wilkinson, R. B., & Parry, M. M. (in press).
Attachment styles, Quality of Attachment
Relationships, and Components of Self-Esteem in
Adolescence. Australian Journal of Psychology.
Wilkinson, R. B., & Walford, W. A. (2001).
Attachment and personality in the psychological
health of adolescents. Personality and Individual
Differences, 31, 473-484.
155
Keep reading this paper — and 50 million others — with a free Academia account
Used by leading Academics
Kenneth Pakenham
The University of Queensland, Australia
Irina Malkina-Pykh
Saint-Petersburg State University
Elmar Brähler
Universität Leipzig
Michael Neale
Virginia Commonwealth University
