Personality Measures

and

Academ ic Specialization

in Cegep Students

A thesis submitted to the Faculty of Graduate Studies and

Research in partial fulfilment of the requirements of the degree
of Master of Arts in Educational Psychology.
March, 1997

Department of Educational and Counselling Psychology

McGill University, Montreal
OAnabel Paquet-Gagnon, 1997
1*1 National Library
of Canada
Bibliothèque nationale
du Canada
Acquisitions and Acquisitions et
Bibliographie Services services bibliographiques
395 Wellington Street 395. me Wellington
OttawaON K1AON4 Ottawa ON K I A O N 4
Canada Canada

The author has granted a non- L'auteur a accordé une licence non
exclusive licence allowing the exclusive permettant a la
National Library of Canada to Bibliothèque nationale du Canada de
reproduce, loan, distribute or sell reproduire, prêter, distribuer ou
copies of this thesis in microform, vendre des copies de cette thèse sous
paper or electronic formats. la forme de microfiche/^, de
reproduction sur papier ou sur format
électronique.

The author retains ownership of the L'auteur conserve la propriété du

copyright in this thesis. Neither the droit d'auteur qui protège cette thèse.
thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels
may be printed or othenvise de celle-ci ne doivent être imprimés
reproduced without the author's ou autrement reproduits sans son
permission. autorisation.
Acknowledqement of financial support

This research project has been partially funded by the McGill

Social Sciences Research Grants Sub-Committee.
Acknowledqements

I would like to thank my thesis supervisor Professor F. Gillian

Rejskind, who has believed in me from the beginning and gave me the

opportunity to develop my own research project, for her unconditional

enthusiastic support and encouragements.

I would also like to thank Lawrance Kirmayer, Ph.D., and James

Robbins. Ph-D., who gave me the permission to use their unpublished

research translation of the NEO-Five Factor Inventory.

I also want to thank al1 students who participated in this project, as

well as their professors, who allowed them to complete the questionnaires

during class time.

Finally, the completion of this work wouldn't have been possible

without the tireless financial and moral support from my mother, Denise,

and al1 the love and encouragements frorn rny fiancé and best friend, Carl.
 iii

Table of contents

Title page
Acknowledgement of financial support ...................................................... i
..
Acknowledgements ...................................................................................II
...
Table of contents ...................................................................................... III

Abstract ....................................................................................................vi
...
Résumé ..................................................................................................VIII

Introduction ............................................................................................-1

Chapter 1: Theoretical considerations ................................................ -3

1.1 Theoretical background: Personality traits ...................................3

7 . 7 . 1 The Nature of Personality Traits .......................................... 3

1.7.2 The Five-Factor Mode1 of Personality .................................-4

1-1-2.1 Describing the Five Factors .................................... -6

Neuro ticism ........................................................... -8

Extraversion ............................................................9

Openness to experience ........................................ -9

Agreeableness ......................................................-10

Conscien tiousness ...............................................-1O

1-2Measuring Personality.. ............................................................... 1 1

 4.2 Discussion of the findings .......................................................... 46

4.2.1 The program main eflect ..................................................... 46

Cornparison of the findings to Kline and Lapham's ....48

4.2.2 The gender main effect ..................................................... -50

4.2.3 The interaction between program and gender .................... 52

4.2.4 The influence of parental occupation on personality .........-52

4.3 Implications of the findings...........................................................53

4.3.1 Implications for vocational counselling ...............................-53

4.3.2 implications for education ................................................... -55

Chapter 5: Summary and Conclusions ............................................... -58

References ............................................................................................ -61

Appendix 1: Sociodemographic Questionnaire ................................. 68

Appendix 2: Inforined Consent .......................................................... -70

Abstract

The main purpose of the present investigation was to compare the

personality traits of cegep students enrolled in arts, sciences, and social

sciences. In addition, the influence of gender on personality was

examined, as well as its interaction with academic specialization. Finally,

exploratory research was conducted to study the influence of selected

socio-dernographic variables on personality configuration.

Participants were asked to complete a socio-demographic

questionnaire as well as the French research translation of the NEO-Five

Factor lnver tory (NEO-FFI), which measures the five basic dimensions of

personality. Thus, another purpose of this study was to provide

validational nformation for the latter instrument.

Results indicated that arts students score significantly lower on

Extraversion than science and social science students. Furthermore,

females were shown to score significantly higher than males on measures

of Neuroticism, Extraversion, and Conscientiouness. In addition, the

interaction of prograrn specialization and gender was shown to be

significant on the Openness to experience scale. Exploratory research

 vii

indicated that students whose parents have a higher occupational level

scored significantly higher than other students on measures of Openness

to experience. Finally, the interna1 consistency of each scale and

correlations between the scales of the instrument were calculated. The

implications of our results were subsequently discussed, as well as

potential issues for future research.

 viii

Résumé

L'objectif principal de la présente enquête consistait à comparer les

traits de personnalité d'étudiants de niveau cégep en arts, en sciences, et

en sciences sociales. De plus. l'influence du sexe sur la personnalité a

été examinée, ainsi que son interaction avec le domaine de spécialisation

académique. Enfin, une enquête exploratoire a permis d'étudier

l'influence de certaines variables socio-démographiques sur la

configuration de la personnalité.

Les participants ont complété un questionnaire socio-

démographique ainsi que la traduction française provisoire du NEO-

Five Factor lnventory )> (NEO-FFI), mesurant les cinq grandes

dimensions de la personnalité. Ainsi, un autre objectif de ce projet était

de fournir des données de validation pour cet instrument.

Les résultats ont indiqué que les étudiants en arts ont un score

significativement plus bas sur l'échelle d'extraversion que les étudiants en

sciences et en sciences sociales. De plus, des scores significativement

plus hauts chez les femmes que chez les hommes ont été observées pour

les échelles de neuroticisme, d'extraversion et d'esprit consciencieux.

Aussi, l'interaction du domaine de spécialisation académique et du sexe

s'est démontrée significative sur l'échelle d'ouverture aux expériences.

La recherche exploratoire a indiqué que les étudiants pour lesquels le

score correspondant à l'occupation des parents est plus élevé, ont des

résultats significativement supérieurs aux autres étudiants sur l'échelle

d'ouverture aux expériences. Finalement, la consistance interne de

chaque échelle ainsi que les corrélations entre les échelles de

l'instrument ont été calculées. Les implications de nos résultats ont

ensuite été discutées. ainsi que les avenues potentielles pour de futures

recherches.
Introduction

One of the most important developments in personality psychology

in the past decade has been the wide recognition. by researchers from

many different traditions, that the Five-Factor Model, also refered to as

the Big Five, adequately represents the fundamental dimensions of

personality. Because of its cornprehensiveness. this model of personality

was shown to be appropriate for use in rnany different applications,

namely in counselling , clinical psychology, and psychiatry; in behavioral

medicine and health psychology; as well as in vocational counselling ,

industrial 1 organizational psychology, and educational research. Given

the wide recognition and applicability of this model, one purpose of the

present study is to introduce and evaluate an instrument in its

developmental stages - the French version of the NEO-Five Factor

lnventory (NEO-FFI) - which measures the five basic dimensions of

personality, so that it can eventually be used adequately on a common

basis with the Quebec Francophone population for many different

applications.

Researchers and educators generally agree that university

students enrolled in different academic programs do differ on the basis of

their personality configuration. Nevertheless, some questions siill remain

unanswered: Can students specializing in different academic programs at

the cegep level be distinguished on the basis of their personality profile?

And if so, which personality traits characterize each group of students?

The following chapters will explore these questions further and discuss

the usefulness of such information for students' benefit.

The project is an important extension of the current research in the

field. Interestingly, it will extend to Francophone populations studies that

have been restricted to Anglophone populations, and will examine effects

that are particularly pertinent to the Quebec educational system. In

addition, it picks up on very recent research concerning gender on the

relationship between personality and academic specialization. Thus, this

project has the potential to make a significant contribution to the field.

Chapter 1: Theoretical considerations

1.1 Theoretical background: Personality traits

This first section consists of a general overview of the current

literature on personality traits, as well as the influence of selected

variables on the development of these traits dimensions.

7.7.7 The Nature of Personality Traits

Personality variables called traits are generally conceptualized as

" relatively enduring dispositions (tendencies to act, feel, or think in a

certain manner in any given circumstance) that distinguish one individual

from another " (Kaplan & Saccuuo, 1993, p.19). These traits are interna1

dispositions which were shown to be relatively stable over time and

across situations (Costa & McCrae, 1988; Digman 1990; McAdams, 1994;

McCrae & Costa; I989,199O). Since personality traits represent only a

single aspect of the individual, they are usually inferred from cross-

situational consistencies in individual functioning. They are also t y p i ~ l

formulated in bipolar terms and are seen as additive and independent.

addition, these traits refer to very generalized individual differences in

socioemotional functioning (Conley, 1985a). Thus, personality traits may

be distinguished frorn variables such as values, attitudes, world views,

and schernas which are less socioemotional and more cognitive in nature

(Conley, 1985a).

Individuals Vary in the degree to which they can be characterized

by specific traits. Most individuals obtain intermediate scores on

continuously distributed trait measures, but a minority score at low or high

extremes. High and intermediate scorers are more likely than low scorers

to possess the pattern of thoughts, feelings, and actions associated with

traits (Costa 8 McCrae, 1988; Digman, 1990; McCrae & Costa,

1989,1990)-

7.1.2 The Five-Factor Mode1 of Personalif y

Generally, personality psychologists agree that there are five

robust factors that can meaningfully classify personality constructs (e.g.,

Digman, 19SO;Golberg, 1990; McCrae & Costa, 1985; Rolland, 1993).

Indeed, these five factors were shown to have convergent and

discriminant validity across instruments and observers, and to rernain

stable throughout adulthood (McCrae & Costa, 1990). The structure of

personality in five factors have been identified in different semantic and

cultural contexts: Anglo-american, Chinese, Dutch, Finnish, Polish,

French (France), Japanese, Tagalog (Philipino), modern Hebrew,

German, and Russian (Angleitner & al., 1990;Rolland, 1993).

All in all, the value of the five-factor model was supported by

McCrae and John (1992),as they wrote:

'...the appeal of the model is threefold: It integrates a wide array of

personality constructs, thus facilitating communication among

researchers of many different orientations; it is comprehensive, giving

a basis for systernatic exploration of the relations between personality

and other phenornena; and it is efficient, providing at least a global

description of personality with as few as five scores (p.206).

"

One main goal of the present study is to explore the possible

relations between personality and academic specialization in cegep

students. Consequently, because of its comprehensiveness, the five-

factor rnodel of personality was selected as the underlying framework of

the present investigation.

1 1.2 1 Describing the Five Factors

While there is general agreement among researchers concerning

the number of personality factors, there is still some disagreement as to

their precise labeling. Yet, the most broadly accepted configuration was

described by McCrae and Costa (1987). According to them, each of the

five factors defines a domain of related traits and they are labeled as

follows: Extraversion (sociable - retiring, fun-loving - sober, affectionate

- reserved), Neuroticism (worrying - calm, insecure - secure, self-pitying

- self-satisfied), Openness to Experience (imaginative - down to earth,
preference for variety - preference for routine, independent -

conforrning), Agreeableness (soft-hearted - ruthless, trusting -

suspicious, helpful - uncooperative), and Conscientiousness (reliable -

undependable, well-organized - disorganized, persevering - quitting).

Longitudinal evidence has shown stability for each of these five

galaxies, leading to the conclusion that personality traits are enduring

dispositions (McCrae 8 Costa, 1990). Indeed, an important characteristic

common to both vocational interests and personality dispositions is their

high stability throughout adulthood. Cross-sectional and longitudinal

analyses of the personality scores of males and females aged 21 to 96

indicated that older adults may obtain slightly lower scores than younger
adults in Extraversion, Positive Emotions, and Openness to Experience

(though the difference is fairly small), which changes might be attributed

to maturation (Costa & McCrae, 1988). Indeed, studies that trace

individuals from college age into adulthood almost invariably report some

changes in the average levels of personality traits and more fluctuations

for younger individuals than are found in studies of individuals who are

initially older (e-g., Finn, 1986;Helson & Moane, 1987). Note that the

changes in personality traits observed in these studies were in the

average levels of the traits, which are much less important than changes

in patterns. In other words, although slight changes were observed in the

traits average levels, the relative importance of each trait dimension

compared to one another remained highly stable. All in all, these

observed changes are very small, and most studies support the position

that personality generally reaches stability between age 20 and 30.

Based on this evidence, one can assume that personality

dispositions in cegep students will be relatively stable across time.

However. one has to keep in mind that, in most Canadian and American

institutions, students enter college after graduating from high school.

Indeed, "college students" usually refers to students enrolled in an

undergraduate university program. In contrast, afler graduating from high

school, students in the Quebec educational system have to go through a

two-year cegep program before entering university. Thus, cegep students

are younger on the average than college students.

Neuro ticism (N)

Neuroticism is undoubtedly the best known personality factor. This

domain is generally described as a continuum ranging from emotional

instability or neuroticism to ernotional stability. This personality

dimension is mainly concerned with individual differences in the

experience of negative emotions and distress such as fear, sadness, guilt,

anxiety, and the like. High scorers on this scale tend to be distressed in

many aspects of their lives and can be described as worrying, nervous,

high-strung, insecure, self-pitying, and vulnerable, and tend to have a

very low opinion of themselves. In addition, because disruptive emotions

interfere with adaptation, individuals high in N tend to be less able to

control their impulses and cope with stress. In contrast, people who score

low on the Neuroticism scale are generally calm, at ease, relaxed, secure,

self-satisfied, and hardy (Costa 8 McCrae, 1992).

Ektraversion (E)

Extraversion is a superordinate trait primarily composed of the

lower-order traits of sociability, irnpulsivity, and assertiveness. Extraverts

are characterized by their sociability which is associated with

gregariousness, warmth, ta1kativeness, energy, cheerfulness, and

stimulation frorn others. These people are also adventurous, action-

oriented individuals who take chances,crave excitement, lose their

temper easily, and who tend to manifest dominance and boldness.

lndividuals who score low in the extraversion domain are, by definition,

introverted. They may be described as withdrawn, quiet, shy, inhibited,

reserved, unaggressive, passive, and pessimistic (e.g., McCrae 8 Costa,

1990).

Openness to experience (0)

The domain of Openness to Experience has been recently

conceptualized by McCrae & Costa (1980). High scorers on this

dimension can be described as original, creative, curious, independent,

artistic, and having broad interests. In addition, people high in O seem to

welcome new and unusual experiences without anxiety, which rnakes their

lives richer in experiences. In contrast, low scorers tend to be

conventional, simple, conforming, nonanalytical, traditional, conservative,

and having narrow interests.

Agreeableness (A)

The fourth trait domain is generally labeled Agreeableness (A) and

encompasses descriptors such as altruism, affection, and other humane

aspects of human personality (Digman, 1990; Goldberg, 1990; McCrae &

Costa, 1991). Hence, individuals at the high end of the continuum can be

described as interpersonally warm, generous, loving, helpful, patient,

ernpathic, understanding, and sincere. In contrast, people at the low end

of the continuum are described as being harsh, manipulative,

untrustworthy, malicious, and tend to disregard other people's feelings.

Agreeableness is a highly valued dimension which has shown to promote

psychological well-being (McCrae & Costa, 1991). Indeed, scores on this

scale have often been used to describe good versus evil individuals

(McCrae & Costa, 1987).

Conscientiousness (C)

The trait labeled Conscientiousness (C) is mainly concerned with

descriptors related to the areas of work, achievement, accomplishment,

and so forth. People who score high on C may be described as being

well-organized, etficient, dependable, persistent, steady, predictable,

conventional, and thrifty (Goldberg, 1990;McCrae & Costa, IWO).

Indeed, these individuals approach tasks and problems in a systernatic

and logical fashion, and their lives are planned carefully according to

principles and goals. They are also highly disciplined, reliable, and

responsible in their work as well as in their dealings with other people.

After reviewing many studies suggesting the importance of this trait in

determining school achievement, Digman (1990) has renamed

conscientiousness " the will to achieve ". Hence, at the low end of the C

continuum, people tend to be disorganized, inefficient, undependable,

unpredictable, lazy, and undecisive. In addition, unconscientious people

were shown to have Iittle respect for serious standards of work and

morality (McAdams, 1994).

1.2 Measuring Personality

Personality inventories consist of self-report questionnaires

containing a large number of items for which respondants must indicate

their level of agreement or adhesion (true or false). The different items

are then grouped in a number of trait factors, which scores are calculated

by summing up the scores obtained for each item in a specific subgroup.

Note that al1 trait factors are independent and determined by the author of

the instrument. Once the personality scores are calculated for each trait

dimension, the personality profile of the respondant can be established.

Several instruments have been developed over the years to

measure these personality characteristics. The most widely used

personality questionnaires are the Minnesota Multiphasic Personality

lnventory (MMP 1), the 16 Personality Factors (16 PF), the Myers-Briggs

Type lndicator (MBTI), and the Personality Research Form (PFR). In

addition, a recent instrument has received increasing attention in the past

decade: the NEO-Personality lnventory (NEO-PI), as well as its revised

version, the NEO-Personality Inventory-Revised (NEO-PI-R), and its

abrreviated version, the NEO-Five Factor lnventory (NEO-FFI). These

instruments were developed by McCrae and Costa to provide accurate

measures of the Big Five personality traits. Given the wide recognition

and usefulness of these instruments, the NEO-FFI was recently adapted

to the Quebec French semantic and cultural context by researchers at the

Montreal Jewish Hospital. Although their version is not available for

public use, the researchers gave me permission to use their instrument to

conduct the present study. Indeed, my results will provide data to

evaluate their findings.

1.3 Personality and Academic Specialization

Several studies have examined the personality of students

pursuing different academic programs and they all confirrned the

hypothesis that students enrolled in different programs can be

differentiated on the basis of their personality profile (e-g., Kline and

Lapham, 1992; Corulla & Coghill, 1991; Clark, 1986; Harris, 1993;

Wankowski, 1968: see Kline & Lapham, 1992). Despite this general

concensus, researchers do not agree as to how these students differ. A

major cause of their divergence stems from the fact that researchers from

different traditions of psychology hardly understand andior communicate

with each other. Consequently, the personality measures collected in

different studies are organired according to diverse personality theories

and constructs, thus complicating their cornparison. The next paragraphs

will describe several studies on the topic in order to set a clear global

picture of the current scene.

Many researchers assume that student populations can be reliably

differentiated on the variables of neuroticism and extraversion. For

example, Kline & Lapharn (1992) referred to this assumption in their

discussion of personality and academic success as they indicated that

science students are usually found to be more introverted than arts

students (Wankowski, 1968:see Kline & Lapham, 1992),while social

studies students score higher on Neuroticisrn (Cattell et al., 1970: see

Kline 8 Lapham, 1992). However, as shown below, the results obtained

from other studies do not always support this hypothesis.

For example, in an investigation on the predictors of scientific

majors, Clark (1986) examined the personality measures of 200 college

students using the 16 Personality Factor lnventory (16 PF). His sample

was made up of 73 males and 127 females, ranging from 17 to 24 of age.

The findings of his study indicated that college major groups significantly

differed on 3 personality variables, with natural science majors being

more reserved (introverted), sober (as opposed to happy-go-lucky), and

practical than social and nonscience majors.

In contrast, Corulla and Coghill (1991) obtained different results in

their recent work on personality and educational streaming. In their first

study involving 17 fernale science students, 16 arts students (13 females,

3 males), and 13 male technical students from a Technical College,

personality information was collected using the Eysenck Personality

Questionnaire-Revised (EPQ-R) and the lmpuisiveness Questionnaire

( 7 ) Their findings indicated that the science group scored significantly

lower than arts and technical students on the Psychoticism scale. In

addition, a positive correlation between Neuroticism and arts students

was found, although not significant. When interpreting these findings,

however, one has to pay close attention to the confound between gender

and program in their sample. Indeed, while their science sample was

composed entirely of females, technical students were al1 males.

In their second study, they examined the personality correlates of

300 university students (100 arts students, 100 engineering science

students, and 100 social science students) using the same

instrumentation (Corulla 8 Coghill, 1991). The sample consisted of 215

males with a mean age of 21 -6,

and 85 females with a mean age of 20.3.

While Psychoticism and lmpulsivity were shown to correlate significantly

and positively with technical science studies, the results indicated a

significant positive correlation between Neuroticism and arts studies, and

a significant negative correlation between Psychoticism and engineering

students. These findings provide support for their first study in

suggesting that enrollment in hard sciences is negatively related to

Psychoticism. In addition these findings suggest that social conformers,

as translated by a low score on the lrnpulsiveness scale, are more likely

to specialize in sciences than individuals who score high on Psychoticism.

 Finally, the conclusion of their third study indicated that science

students tend to score lower on rneasures of Psychoticisrn than

psychology students, providing support for the two previous studies.

However, a close examination of their sample again revealed a gender

confound as their sample consisted mainly of males.

Along the sarne line of research, Kline and Lapharn (1992)

investigated the personality scores of students from different faculties in

different British universities. This sample of 1472 subjects with a mean

age of 19.2 years old was administered the Professional Personality

Questionnaire (PPQ), a measure of the "Big Five" personality variables.

Their results indicated that science and engineering students were

significantly more conscientious, higher in conformity (lower on Openness

to experience), and more tough-minded (lower on Agreeableness) than

arts. social science, and mixed major students. However, in contrast to

their original assumptions, scores on Neuroticism and Extraversion failed

to discriminate among students in these faculties.

More recently, Harris (1993) examined self ratings as well as peer

ratings on the Personality Research F o m (PRF), noting that students'

personalities differed by faculty. Her results indicated similar personality

profiles for arts and social science students, both characterized by being
significantly more affïliative, sociable, and sentient, whereas science

students were described as being more precise.

These studies mentioned above al1 suggest that students enrolled

in different academic programs at the college level differ on the basis of

their personality profile. However, because of the wide variety of the

instrumentation available to rneasure personality characteristics, no

consensus has been reached as to which traits distinguish students

enrolled in the different programs. In addition, these studies have

focused on college students, where the findings can hardly be

generalized to the cegep population. Not only are cegep students

younger than college students, but their scope of academic programs

differs greatly. Indeed, universities offer a much wider variety of

programs than cegep institutions do, because cegeps provide students

with a general preparation background to pursue university education. As

a result, cegep students enrolled in sciences include students in both

pure and applied sciences, while these students groups are segregated in

diffsrent academic specializations at the university or college level.

Thus, in contrast to the research that has been done in the past,

the present investigation will examine personality differences in cegep

students in terms of the five-facor personality model. In the light of the

studies mentioned above, personality differences are expected to be

found between arts, science, and social science students. Exploratory

analyses will then be performed in order to investigate the explicit nature

of these personality differences between these three student groups.

1.4 The relationship between Gender and Personality

The issue of personality differences between genders has been

studied extensively in the past. However, no concensus has been

reached on the topic. For example, a study investigating the predictors of

scientific majors in 200 college students indicated gender differences on

five personality variables as rneasured by the 16 Personality Factor

Inventory (16 PF) (Clark, 1986). Indeed, women were shown to be more

conscientious, toughminded, astute, and experimenting, with more tough-

poise than men. Nevertheless, these gender differences were not present

when the interaction between gender and college major was examined.

Along the same lines of research, Baker (1983) studied the

personality characteristics of 180 college students (30male and 30

female physical science majors, 30 male and 30 female biological science

majors, 30 male and 30 female nonscience majors) using the Myers-

Briggs Type lndicator (MBTI). According to his results, significant

personality differences between genders were observed on the Thinking-

Feeling scale. That is, gender differences were primarily reflected in

t e n s of decision making, with males prefering an analytical approach to

problern solving and females basing their decisions on personal values.

In addition, his analyses suggested that personality differences among

college majors were due to differences between males and females rather

than between majors.

Most relevant to the present investigation are the studies which

have looked at gender differences in terms of the NEO-Personality

Inventory. Indeed, recent studies using the NEO-PI-R scales with college

students agree that women tend to score higher than men on measures of

Agreeableness (e-g., Costa 8 McCrae, 1992; Jung, 1995). However,

Costa and McCrae (1992) also reported that females were significantly

more neurotic than males, while Jung (1995) observed that females were

significantly more extraverted. Furthermore, McCrae and Costa (1992)

indicated that their findings were supported by diverse studies, which

noted sirnilar patterns using other personality inventories (e.g., Eysenck 8

Eysenck, 1975: see McCrae & Costa, 1992).

The present study will examine gender differences among cegep

students in terms of the NEO-Five Factor Inventory, the abbreviated

version of the NEO-Personality Inventory. The short version of the

instrument was selected because of its availability and convenience.

Consequently, my hypotheses will be based on the literature looking at

gender differences in terms of the NEO-PI scales. Therefore, as a

cumulation of the studies mentioned above, gender differences are

expected to be found on Agreeableness, Neuroticism, and Extraversion,

with females scoring higher than males on each dimension. In contrast, I

predict that gender differences will not be observed on measures of

Openness to experience and Conscientiousness. Furthermore,

exploratory analyses will be performed in order to examine the interaction

between gender and academic specialization, since other studies looking

at the gender-program interaction didn't use the Big Five, while studies

with the Big Five didn't look at the interaction.

1.5 Hypotheses of the Present Study

The main purpose of this research project is to examine the

relationship between cegep students' personality profiles and their area of

academic specialization in terms of the five-factor model of personality. In

addition, personality measures of males and females will be compared

within each academic program. The provisory French version of the

NEO-FFI will be used for data collection as 1 assume that the tool will be
useful to the Canadian francophone communities for assessrnent in

different applied settings.

In the light of the literature reviewed earlier. rny first hypothesis is

that personality differences will be present between cegep students

enrolled in the arts, science, and social science programs. In addition, on

the basis of earlier reasearch with the five-factor model, 1 expect females

to score higher than males on measures of Agreeableness, Neuroticism,

and Extraversion. In contrast, I predict that gender differences will not be

observed on rneasures of Openness to experience and

Conscientiousness. Furthermore, exploratory analyses will be performed

in order to examine the interaction of gender and academic program on

personality differences, as well as the relationship between native

language, parental education and occupation, and the region where

subjects were raised and the five major personality categories. Another

issue of this research is the relative usefulness of the French translation

of the NEO-FFI in personality assessrnent, which will be examined in the

present study.
Chapter 2 : Method

2.1 Participants

The sample was entirely drawn from two French cegeps located in

the Montreal area, namely Cégep du Vieux-Montréal and Cégep de

Maisonneuve. The participants were recruited through classes on a

voluntary basis in the science, the social science, and the arts programs.

Prior to recruitement, al1 instructors agreed to allow students to fiIl in the

questionnaires during class time. A total number of 227 students agreed

to participate in this study. However, 7 were dropped from the study

because they submitted incomplete questionnaires. Thus, the sample

was made up of 220 participants (128 females and 92 males), with a

mean age of 18.72 ranging from 16 to 39. More specifically. complete

questionnaires were returned by 75 students in the arts prograrn (43

fernales, 32 males), 64 students in the science program (34 females, 30

males), and 81 students in the social science program (51 fernales, 30

males).

Note that cegep students enrolled in the arts program clearly do

specialize in fine arts such as photography, interior design, presentation

design, graphism, jewellery, cabinet work. textile impression, textile

construction, as well as general manual arts leading to university. In

contrast, social sciences students are mainly enrolled in the general

prograrn leading to university programs such as psychology, history. and

administration, although some of thern do specialize in administration at

the technical level. And finally, the science program embraces al1

students who are planning to pursue their studies in sciences at the

university level, as well as students completing a technical science

prgram at the cegep level, such as engineering, nursing, biochemistry,

and so on.

2.2 Measures

2.2.7 Personalif y Measures

Personality measures were obtained using the French research

translation of the NEO-Five Factor lnventory (French translation:

Kirmayer & Robbins, 1993). Aithough the instrument is not commercially

available yet, the bibliography of the NEO-PI-R professional manual

indicated that a French research translation had been developed by

Montreal researchers, for which validational research was currently in

progress. Consequently, these researchers gave me permission to use

their instrument under the condition that I share rny results with them for

validational purposes. I gladly accepted their proposition as I was

determined to give an innovative flavor to my research project.

The instrument is a direct translation of the original American

English version of the NEO-Five Factor lnventory (Form S) (NEO-FFI:

Costa 8 McCrae, 1992). The latter was developed through factor-analytic

methods as the abbreviated form of the NEO-Personality lnventory (NEO-

PI), which includes 240 items. The NEO-FFI is a 60 items self-report

questionnaire rated on a five-point Likert-type scale ranging from strongly

disagree to strongly agree. Items were balanced to reduce acquiescent

responding. Like the NEO-PI, the NEO-FFI is intended to measure the

five major dimensions of normal personality: Neuroticism (N), Extraversion

(E), Openness to Experience (O), Agreeableness (A), and

Conscientiousness (C). While there is no time limit to fiIl in the

questionnaire, most respondents require 10 to 15 minutes to complete it,

depending on their reading skills level.

The use of the NEO-FFI was selected for several reasons. First, a

study comparing the different instruments to operationalize the five-factor

mode1 of personality revealed that the NEO-FFI (English version) was the

most appropriate and comprehensive instrument to measure the "Big

Five" personality dimensions (Briggs, 1992; Ostendorf; 1994). In addition,

the basic structure in five dimensions was shown to encounter several

major personality inventory available today (McCrae & John, 1992).

Second, this instrument was designed for use with individuals who are 17

years of age or older (Costa & McCrae, 1992). Indeed, it appears to work

equally as well for college students as for adults. Consequently. the tool

seems to be potentially valuable to cegep students, as it is one of the few

tools that has been designed with the inciusion of this specific age range.

Based on the wide recognition of this instrument, it is my contention that a

French version adapted to the Quebec cultural and semantic context

could be a highly valuable toot for diverse applications, such as

counselling, personnel selection, and educational psychology, in our

province. Hence, this study will provide reliability data which will allow

improvement of the instrument which is presently in its developmental

stages.

Validity and reliabilify information

Costa and McCrae (1992) reported correlations between the NEO-

FFI (Form S) and the revised NEO-PI (NEO-PI-R; Costa & McCrae,

1992),which were .92,.90,.91, .77,and -87for the N, El O, A, and C

domains respectively. In addition, for a separate sample of 1539 adults,
the Cronbach's a interna1 consistency reliability coefficients for the

domain scales were 3 6 ,.77,.73,.68,and .81 for NI El O, A, and C

respectively with respect to the English version of the NEO-FFI. Evidence

of construct validity for the NEO-FFI (Form S) was obtained from self-

report adjective factors of the five-factor model. Costa and McCrae

(1992) reported convergent validity correlations ranging from -56to -82,

and absolute discriminant validity coefficients ranging from -00 to .20.

Further indications of construct validity was obtained by the pattern of

correlations between the NEO-FFI scores and spouse ratings of the NEO-

PI-R scales (Costa & McCrae, 1992). Such information on reliability and

validity is presently unavailable for the French version as it is in its

developmental stages. The present study will provide information that will

be of value for future research.

2.2.2 Sociodemographic Questionnaire

In addition to the NEO-FFI. subjects were asked to complete a

socio-demographic information sheet designed for this study which

requested the subjects to indicate their educational institution, gender.

age, academic program, native language, parents' educational level, and

parents' occupation.
 One-way analyses of variance (ANOVAs) were performed to test

the effects of subjects' program (pure and applied sciences, humanities,

arts and literature), gender (male, fernale), as well as their interaction on

each of the five personality dimensions. In addition, several other

statistical procedures were performed to explore the relationship between

selected sociodemographic variables and the Big Five personality

domains.
Chapter 3: Results

The following section will present the results obtained from this

investigation. Note that preliminary analyses have revealed no significant

differences in the subjects' scores on each personality dimension as a

function of native ianguage, parents' education, and the region where

subjects were raised (rural versus urban). For this reason, these

variables were not included in the subsequent analyses.

3.1 Program differences

The purpose of the first statistical analysis was to test the first

hypothesis, which stipulates that arts, science, and social science

students can be distinguished on the basis of their personality profile. In

addition, further exploratory analysis allowed us to examine the specific

nature of these differences in terms of personality traits.

Thus, one-way analyses of variance were performed to examine

the effects of program on each personality dimension, followed by a

TukeyJsStudentized Range (HSD) test to see which mean differences

were significant (Table 1 shows means). The analyses yielded a

significant value for Extraversion, f32.217)= 3.58, ~=0.0296
(Table 2),

indicating thît the student groups are characterized by a different level of

Extraversion. On subsequent analysis of the means of the three program

groups, the Tukey HSD revealed that social science students scored

5 ) arts students on the Extraversion

significantly higher (at ~ ~ 0 . 0than

scale. However, no significant differences were found between science

students and the two other groups on this dimension.

Table 1. Means for each program on the 5 NEO-FFI scales.

Program g N E O A C
Arts 75 33.36 42.71 44.16 43.28 43.37

Science 64 34.28 43.05 41 -78 41.86 42.83

Social 81 33.47 45.01 42.62 43.68 42.78

science

Table 2. Analyses of variance for program differences on the 5 NEO-FFI

scales.
Personality df -
SS
scale
 Overall, the results obtained in this analysis weakly support the

hypothesis that personality differences exist between the student groups,

since these differences were present only in terms of Extraversion.

3.2 Gender differences

The literature on the issue of personality differences between

genders using the NEO-PI-R suggests that personality differences do

exist between genders. More specifically, studies have indicated that

women tend to score higher than men on measures of Neuroticism.

Agreeableness, and Extraversion. Consequently, my second hypothesis

predicts that personality differences will be found between genders, with

females scoring significantly higher than males on measures of a)

Neuroticism, b) Agreeableness, and c) Extraversion. In contrast, no sex

differences are expected to be found on the dimensions of d) Openness

to experience and e) Conscientiousness.

Thus, one-way analyses of variance were performed on the gender

means for each personality dimension (Table 3). These analyses yielded

significant F values for three of the five personality dimensions, namely

Neuroticism; E(1,218)= 14.08, p=0.0002,Extraversion; E(1,218) = 6.90,

 and Conscientiousness; E(1.218)= 5.78, g=0.0171 (Table 4),
~=0.0092,

with females scoring significantly higher than males on each dimension.

Table 3. Means for each aender on the 5 NEO-FFI scales.

Gender -
n N E O A C
Female 128 35.52 44.52 42.99 43.59 44.01

Male 92 31.10 42.45 42.77 42.22 41.59

Table 4. Analyses of variance for sex differences on the 5 NEO-FFI

scales.
Personality a -
SS -
MS -
F -
P
scale

The results obtained in this second set of analyses provide partial

support to my hypotheses. As was expected, personality differences were

found between males and females on measures of Neuroticisrn and

Extraversion, with females scoring significantly higher than males. In

contrast, the hypotheses forrnulated for the other dimensions were not

supported by the results. Surprisingly, not only were no group differences

observed on Agreeableness. but unexpected gender differences also

turned out to be significant on the Conscientiouness scale.

3.3 The interaction between program and gender

Studies investigating the personality profiles of college students

consistently indicated significant gender differences (e-g., Clark, 1983.

1986). Nevertheless, these gender differences were not present when

the interaction between gender and college major was examined. Indeed,

some analyses suggested that personality differences among college

majors were due to differences between males and females rather than

between majors. Thus, one purpose of the present study was to explore

gender differences among cegep students within the different academic

programs.

One-way analyses of variance were performed to test the effects of

the gender by program interaction on each of the five personality

variables. Interestingly, the results indicated the presence of a significant

gender by program interaction (p0.05)only on the Openness to

experience scale, with a trend on Agreeableness. Table 5 below shows

the means on which the analyses of variance were performed to explore

this issue.

Table 5. Means for program bv aender groups on the 5 NEO-FFI scales.

Program G fi N E O A C
MsJM~sD~M~M~
- - - -
Arts F 43 34.98 7.55 43.70 4.52 44.14 5.34 43.09 5.92 44.35 5.77

Science F 34 35.88 7.62 44.18 6.65 43.56 6.12 43.74 5.76 43.94 7.09

Soc. sci. F 51 35.73 8.83 45.45 5.65 41.65 6.69 43.90 5.29 43.77 7.89

Firstly, the ANOVA with Neuroticism as the dependent variable

yielded a significant value for the model; E(5,2l4)= 3.16, ~=0.0089.

In

addition, the partitioning of the sum of squares (type III SS) revealed the

independent significant contribution of the variable gender (~=0.0003)to

explain group differences on Neuroticism (Table 6), which finding is

consistent with the previous analyses (Table 4).

Table 6. Analvsis of variance with Neuroticism as the dependent
variable.

Source -
df TypeIIISS -
MS -F -
P

Model 5 1184.806 236.961 3.16 0.0089"

Program 2 81,097 40.549 0.54 0.5828

Gender 1 1038.900 1038.900 13.87 0.0003"

Program by 2 75,888 37.944 0.5 1 0.6033

aender

Secondly, the ANOVA with Extraversion as the dependent variable

yielded a significant value for the model; E(5,214) = 2.77, ~=0.0189.In

addition, the partitioning of the sum of squares (type III SS) revealed the

independent significant contribution of both gender (~=0.0134)and

program (p=0.0348)
to explain group differences on Extraversion (Table

7), whcih finding is consistent with the two previous sets of analyses

(Tables 2 and 4).

Table 7. Analysis of variance with Extraversion as the dependent

variable.

Source -
df TvpeIIlSS -
MS -
F -
P

Model 5 458.374 91.675 2.77 0.01 89"

Program 2 225.443 112.722 3.41 0.0348"

Gender 1 205.700 205.700 6.23 0.01 34'

Program by 2 16.990 8.495 0.26 0.7735

gender
 Thirdly, the ANOVA with Openness to Experience as the

dependent variable yielded a significant value for the model;

-
F(5,214)=2.61, ~=0.0259.
In addition, the partitioning of the sum of

squares (type III SS) revealed the independent significant contribution of

the interaction between gender and program (p=0.0175) to explain group

differences on Openness to experience (Table 8). As depicted in Table 5,

the low male score in the science group seems to account for this effect,

as male scores varied most on this dimension - from least open in

science, a traditionally male dominated field (

I
V
& 39-97), to most open in

social science, a traditionally female dominated area (M=44.27).

Table 8. Analysis of variance with Openness to Experience as the

dependent variable.

Source -
df TypeIIISS -
MS -
F -
P

Model 5 564.499 112.900 2.61 0.0259"

Program 2 213.445 106.722 2.46 0.0875

Gender 1 7.431 7.431 0.17 0.6791

Program by 2 356.943 178.472 4.12 0.0175"

gender

Fourthly, the ANOVA with Agreeableness as the dependent

variable yielded a significant value for the model; E(5,214)=2.32,

~=0.0442.The partitioning of the sum of squares (type III SS) for this

variable didn't show any independent significant contributions to the

model, although a trend towards significance was present for both,

gender and program (Table 9). Again, Table 5 shows science males as

the discrepant score.

Table 9. Analysis of variance wlth Acireeableness as the dependent

variable.

Source -
df TypeIIISS -
MS -
F -
P

Mode1 5 392.068 78.414 2.32 0.0442*

Gender 1 101.961 1O1-961 3.02 0.0837

Program by 2 180.914 90.457 2.68 0.0709

gender

Finally, the ANOVA with Conscientiousness as the dependent

variable did not yield a signifiant value for the model. However, the

partitioning of the sum of squares (type III SS) revealed the independent

significant contribution of the variable gender @=0.0177) to explain group

differences on Conscientiousness (Table 1O), which finding is consistent

with the previous analyses (Table 4).

Table 1O. Analysis of variance wlth Conscientiousness as the
dependent variable.

Source -
df Type III SS -
MS -F -
P

Program 2 22.421 11-210 0.20 0.81 63

Gender 1 315.31 3 315.313 5.71 0.0177'

Program by 2 1,453 0.726 0.01 0.9869

gender

The results obtained in this exploratory analysis revealed a

significant program by gender interaction only on Openness to

experience, with a trend towards significance on Agreeableness. The low

male science scores on these two dimensions seem to account for both of

these effects.

3.4 Personality differences based on parental occupation

On the sociodemographic questionnaire. subjects were asked the

occupation of both parental figures. The information was then translated

or coded into numbers, and then averaged, in such a way as to get a

single parental occupation score for each subject. Accordingly, a

systematic coding systern based on occupations' skill level was

developed as foilows:

a) low unskilled occupations, as well as unemployment were coded as 1;

b) office work and middle skilled occupations (e.g., carpenters, plumbers)

were coded as 3; and

c) high professionals (e-g., doctors, dentists) and top management

positions were coded as 5.

The occupational information was coded along a continuum

ranging from low to high socioeconomic status. Thus, some occupations

were also coded as 4 or 2, although no label were given to these

categories. For example, occupations coded as 2 included

welder,fireman, policeman, sewers, etc..., while occupations coded as a 4

included educators, accountants, and so on. Then, the average code for

both parents was computed as to get one code representing the

occupation of both parents. Because observations with rnissing values

were not included in this analysis, only 201 observations were used in this

analysis.

Diverse statistical procedures were performed to explore possible

relationships. These analyses indicated significance when a one-way

analysis of variance was performed with parental occupation as the

classifying variable and Openness to experience as the dependant

variable; c(4,196) = 2.63, ~=0.0355(Table 11). Furthermore, a Tukey's

Studentized Range (HSD) test indicated that subjects for whorn parental

occupation was coded as 5 (high professional and top managers) scored

on measures of Openness to experience

significantly higher (~e0.05)

than students whose parental occupation catergory was coded as 1, 2, or

3.

Table 11. Analysis of variance with Openness to Experience as the

dependent variable.
Source -
df -
SS -
MS -
F -
P

parental 4 452.081 113.020 2.63 0.0355"

occu~ation

3.5 Reliability of the instrument

As mentioned in the previous section, the French translation of the

instrument used in this study is currently under deveiopment. Thus, one

of the purposes of this investigation was to provide reliability information

for the test. Hence, the split-half procedure was selected as a means to

approximately determine the interna1 consistency of the scales using the

Pearson correlation coeffkients. The results obtained from this analysis

are wmpared to the reliability measures calculated using coefficient

alpha for the English version of the instrument (Table 12).

Table 12. Interna1 consistencies of the NEO-FFI scales as calculated in

this study wmpared to the E n ~ l i s h
version.

Personality French English

scale translation (1) version (a)'

1. Source: Costa, P. T., Jr., & McCrae, R., R. (1992). NEO-PI-R

Professional Manual. Odessa, Florida: Psychological Assessrnent
Resources. p.53

3.6 Correlations between the scales

Table 13 presents the correlations between the scales of the

French translation of the NEO-FFI given the data from this study. Weak

to moderate correlations among the five trait dimensions can be

obsewed.
Table 13. Pearson correlation coefficients between the scales of the
French translation of the NEO-FFI given the data from this
study.

Summary of the results

AH in all, the results of this investigation point to limited personality

differences between students enrolled in arts, science, and social science

at the cegep level. Indeed, these student groups were shown to differ on

measures of Extraversion, with social sciences scoring significantly higher

than science students. In addition, personality differences were found

between genders on measures of Neuroticism and Extraversion, with

fernales scoring significantly higher than males. Furthermore, a

significant program by gender interaction was found on Openness to

experience, with a trend towards significance on Agreeableness. Finally,

students whose parents' occupations require high skills showed

significantly higher scores on Openness to experience than students

whose parents' occupations require only lirnited skills.

The interna1 consistencies calculated for the five scales of the

.54, .56,.73 for Neuroticism, Extraversion,

instrument are -74, -50,

Openness to experience, Agreeableness, and Conscientiousness

respectively. Furthermore, weak to moderate correlations between the

domain scales of the instrument were found, ranging from -0.298 to 0.037.
Chapter 4: Discussion

4.1 The French version of the NEO-FFI

Weak to moderate correlations beiween the five domain scales of

the French research translation were found in this study, ranging from -

0.298 to 0.037. These numbers are comparable to those of the English

version which range from 0.00 to 0.41, as obtained by Tokar and

Swanson (1995). Furthermore, in accord with previous studies (Jung,

1995; Costa 8 McCrae, 1985: see Jung, 1995), the findings indicated that

Neuroticisrn and Extraversion are inversely related (r =-O.144), so are

and Extraversion and

Neuroticism and Agreeableness (r =-0.283),

Agreeableness are positively related (r =0.256).These numbers suggest

that there is either a slight overlap in the items or in the factors

themselves. Nevertheless, the associations between the factors in the

French and English versions seem to be of the same nature.

As seen in Table 12, interna1 consistencies obtained for the NEO-

FFI scales were .74, .50,.54, .56,and, .73 in this study employing the

French translation, compared to .86,.77,.73,.68,and .81 for the English

version, for NIEl O, A, and C, respectively (Costa & McCrae, 1992). The
fact that the coefficients are substantially lower than those calculated for

the English version point to the low reliability of three of the five scales,

namely Extraversion, Openness, and Agreeableness. However, an

important element to take into consideration when examining these

differences is the relatively small sample size of the present study

(N=220) compared to Costa and McCrae's (N =1539),which might have

negatively affected the results.

The lack of interna1scale consistencies suggest potential

limitations in the adaptation of the inventory to the Quebec semantic and

cultural context. Thus, in order to get some insights into the possible

causes of this limitation, the developmental procedures of the French and

the English versions of the NEO-FFI will be both briefly described.

On the one hand, the English version of the NEO-FFI was

developed as an abbreviated version of the NEO-PI. Indeed, the items

selected for the NEO-FFI were the validimax factors from the NEO-PI.

That is, the validimax method was used to maximize convergent and

discriminant validity between the validimax factors of the NEO-PI. Then,

for each trait dimension, the 12 items with the most negative or positive

correlation on the corresponding scale were selected as prelirninary NEO-

FFI items. After examination of these items, substitutions were made to

Vary item content, cancel items with similar loading, and make sure that

no more than two-thirds of the items on any scale were keyed in the same

direction, as a control for acquiescent responding.

In contrast, the approach used to adapt the NEO-FFI for the

French Quebec population was not based on the validimax method. In

fact, the French version of the instrument was a direct translation of its

corresponding English version. Indeed, the English version (Form S) was

first translated in French by a researcher in the field. Then, a

professional translator perforrned a backtranslation of the document. And

finally, the researcher and the translator examined together the

differences between the backtranslation and the original English version,

and discussed the choice of words for the French version. That is,

instead of translating the English version into French word by word,

Kirmayer and Robbins (1993) tried to keep the essence or the meaning of

each item by putting them in current Quebec French language.

In the light of these considerations, a number of weaknesses are

important to mention when examining the latter approach. First, the 60

items selected using the validimax method from the NEO-PI in the English

version were sirnply translated to get a French version. However, there is

a major conceptual problem since the 60 items, as selected by the

validimax method do not necessarily correspond to the ones that would

have been obtained for the French version. Two ways to remediate to the

situation could be explored further in future research. First, the

translation process could be started over and initiated by a translation of

the NEO-PI-R, from which the validimax rnethod could be used in

selecting the 60 items. Or else, the present translation could be modified

so that the items with low correlations with their corresponding domain

scale be replaced by new items with higher factor loadings.

4.2 Discussion of the findings

4.2.1 The program main effect

The first set of analyses provided weak support to the hypothesis

that arts, science, and social science students at the cegep level differ on

the basis of their personality profile. Indeed, exploratory analyses

indicated that group differences were found only on the Extraversion

dimension. which effect was mainly due to differences between social

science and arts students.

This finding is supported by the literature on occupational choice

suggesting that extraverts tend to prefer and to excel in occupations that

involve frequent interaction or cooperation with other people (e.g., Barrick

& Mount, 1991; Costa et al.,1984; Wilson, 1W8),which occupations

usually require studies in social science at the cegep level. In contrast,

artists tend to have more solitary interests and are consequently more

introverted (Bendig, 1963), which occupations correspond to studies in

arts at the cegep level.

In contrast to the present findings, however, Bendig (1963) also

indicated that artists are not the only category of people who tend to be

more introverted in nature, but so are research scientists, engineers, and

mathematicians. Although the findings didn't reveal that science and

social science students significantly differ on Extraversion, a

reexamination of the analyses allowed us to notice a trend towards

significance. The means presented in Table 5 suggest that this trend

might be true of male students. We could therefore suppose that a

balanced ratio of males and females in the sample might have resulted in

a significant difference on Extraversion between science and social

science students.

An alternative interpretation could be that students considering a

career in pure and applied sciences, such as chemist, engineer, or

physician, rnay fol low the stereotype of being introverted. In contrast,

students more inclined toward health sciences rnay be more extraverted,

given their distinct future career lines, such as medical doctors,

pharmacist, or dieticians, which occupations often involve frequent

interactions with other people. If this is the case, then personality

rneasures may be more useful in differentiating career paths within fields

rather than by their academic precursors. In the present investigation, the

sample included both purelapplied and health sciences in the sarne

student group because these programs are highly similar at the cegep

level (with the exception of 1 or 2 courses). The fact that the term

"program" used in the literature doesn't apply here points to the

usefulness of the present investigation. In addition, future research

involving cegep students is needed to provide replicative data to validate

the findings.

Cornparison of the findings to Kline and Lapham's

As mentioned earlier, Kline and Lapham (1992) recently conducted

a study similar to the present one. Indeed, they examined the personality

profiles of students in the same age range using the Professional

Personality Questionnaire (PPQ), a different instrument to measure the

Big Five personality traits. Interestingly, although both studies involved

students' personality scores in terms of the five-factor model, the h o sets

of findings are highly inconsistent with one another. Indeed, Kline and

Lapharn (1992) found that science and engineering students were

significantly higher on Conscientiousness, Conformity, and Tough-

mindedness than arts, social science, and mixed major students. In

contrast, the findings revealed that arts, science, and social science

students could only be differentiated on the basis of their Extraversion

level, with social science students scoring significantly higher on

Extraversion than arts students.

The discrepancy between these two sets of findings can be

explained by several factors. Firstly, although both studies provided

personality measures in terms of the five-factor rnodel, different

instruments were used to collect the data. Thus, each set of results might

have been affected by the weaknesses of their respective

instrumentation. Indeed, in a recent study, Angleitner and Ostendorf

(1994) showed that the Professional Personality Questionnaire (PPQ), as

used by Kline and Lapham, was totally inappropriate to measure the Big

Five personality traits. Indeed, only one of the assumed relationships

arnong the instrument's scales and the five factors was supportedl namely

the convergence between the Conventionality scale and the Openness to

experience factor. Similarly, the limitations of the instrument, as

described earlier, may also have affected the results. Furthermore, the
language differences between both instruments rnay also constitute a

possible source of explanation.

Secondly, program differences due to distinct educational systems

were translated into different student populations, which may also

contribute to the discrepancy between both sets of findings. In addition,

not only were both sarnples recruited through volunteering, thus not

randomly chosen from their respective population, but the relatively small

sample size might also have affected the results. All in all, the limitations

in both studies point to the idea that although Kline and Lapham rnay be

totally wrong, so may we. In addition, this cornparison illustrates well why

no concensus has been reached on this topic, and why more research is

needed in order to demistify this issue.

4.2.2 The en der main effect

The second set of hypotheses was formulated on the basis of

earlier reasearch with the NEO-PI scales. Thus, we predicted that

females would score higher than males on measures of Agreeableness,

Neuroticism, and Extraversion, while no gender differences would be

present on measures of Openness to experience and Conscientiousness.

The analyses provided support for only two of the five hypotheses,
namely for Neuroticism and Openness to experience. Indeed, while

females scored higher than males on the Neuroticisrn scale, no gender

differences were observed on measures of Openness. It is also

noteworthy to mention that although males and females did not

significantly differ on Agreeableness at pc0.05, the difference was close

to significance (p=0.0896).

In contrast to what was expected, however, groups differences

were found on Extraversion and Conscientiousness, with females scoring

significantly higher than males on both dimensions. The gender

differences on Conscientiousness as observed in the sample coincide

with Clark's (1986) findings as measured by the 16 PF. In addition, the

significance on Extraversion is supported by Jung (1995), who found

slightly higher scores for females on this dimension, which he explained

by the interpersonal orientation of females. As mentioned earlier,

previous research investigating gender differences in personality do not

allow to settle the question, and nor do these findings. Thus, further

research is needed in order to explore this issue, with particular emphasis

on the effects of societal gender role changes on personality.

4.2.3 The interaction between program and aender

Personality differences between genders within each program were

found on Openness to experience. The male scores seem to account for

this effect, as they varied most on this dimension - from least open in

science, a traditionally male dominated field, to most open in social

science, a traditionally female dominated area. This observation

suggests for males a relationship between gender stereotypes and

Openness to experience, which might be examined further in future

research. In addition, as opposed to what has been hypothesized in the

past, our results suggest that wornen do not appear to avoid science

because their personality doesn't match the requirements of the field

since the gender by program interaction was not significant for

Extraversion. Thus, future research is needed in order to move toward a

better understanding of personality differences between genders within

academic programs in terms of the five-factor mode1 of personality.

4.2.4 The influence of parental occupation on personality

Exploratory research was carried out to investigate the influence of

selected sociodernographic variables on the five major personality

variables. Results indicated that parents' education is positively

associated with the Openness trait in the cegep population. Similarly,

Eysenck and Allsopp (1986) found the sarne relationship between

students' socioeconomic status and measures of Unconventionality.

Indeed, previous studies have indicated that Openness to experience and

years of education are moderately correlated (r = .19)(Costa & McCrae,

in press-c and Costa et al., 1991: see McCrae & Costa, 1992). Thus,

future research might attempt to determine whether the level of Openness

itself can be increased by education.

4.3 Implications of the findings

for vocafional counselling

4.3.7 Im~lications

As McCrae 8 Costa (1991) pointed out, the five-factor model rnay

be particularly appropriate for use in vocational counselling because it is

concise, designed for nonpathological individuals, and sensitive to

people's strengths and weaknesses. Indeed, it was shown that

personality assessment using the five-factor model may provide a useful

complement to vocational interests information, as it might reinforce or

modify the information yielded by vocational interests scores. For

example, it was shown that people who score high on Openness to

experience tend to be interested in a wide variety of occupations,

including those that low scorers on this dimension prefer (McCrae &

Costa, 1992). Thus, this information should be taken into consideration

when interpreting vocational interests scores, as very open individuals

might indicate an initial interest in occupations that are unlikely to sustain

their interest. Similarly, someone who scores high on measures of sales

interest but low on Extraversion may indicate carelessness in completing

the vocational interest inventory or else a poor understanding of the

nature of the work. Indeed, Super (1957, p.193) stated that '... the best

way to promote vocational adjustment is to give people information about

themselves..." Accordingly, the NEO-FFI gives information on emotional,

interpersonal, experiential, attitudinal, and motivational styles, which

should also be taken into account in evaluating the fit between individuals

and vocations (McCrae & Costa, 1991).

The findings seem to suggest that personality, particuiarly

Extraversion, may be one of the factors contributing to the selection of

academic program at the cegep level. However, the general lack of

personality differences suggests that it may be more fruitful to use

personality to guide students to particular career paths within their

respective fields of study.

4.3.2 Implications for education

Two of the five factor domains, namely Openness and

Conscientiousness, are of particuiar interest to the field of educational

psychoiogy. Indeeu, Openness to experience was shown to be

moderately related to measures of intelligence and more strongly related

to measures of divergent thinking, which ability is thought to contribute to

creativity (McCrae, 1987). As a result, educators might investigate

whether education can promote the development of Openness in their

students, or whether Openness lead students to pursue more education.

As McCrae and Costa (1992) noted, another application of this

dimension in an educational setting could be to investigate whether

unconventionality, as translated by a high level of Openness, is a source

of misunderstanding or frustration in traditional classrooms. If that is the

case, educators might attempt to adopt teaching strategies that would

promote students' autonomy, such as independent learning in open

schools. Indeed, McPartland and Epstein (1975) observed that students

from the higher social classes showed a more positive relationship

between school openness and acadernic achievement than did students

from the lower social classes. Consequentiy, future research may

investigate whether the use of these strategies promotes the development

of Openness in their students.

The measure of Conscientiousness was also shown to be

positively associated with college-level acadernic performance (e-g.,

Wolfe 8 Johnson, 1995; Dollinger 8 Orf, 1991). Conscientious students

tend to be well-organized, purposeful, and persistent. Indeed, some

studies have suggested that these traits lead to higher academic

achievernent (Digman 8 Takemoto-Chock, 1981: see McCrae 8 Costa,

1992). In addition. conscientious people consider themselves, and are

rated by others, to be more intelligent (McCrae & Costa, 1987: see

McCrae & Costa, 1992). Thus, personality assessment on this domain

scale may be a useful supplement to ability and intelligence measures as

predictors of academic and later-life success.

The assumption that college students enrolled in different

academic programs do differ in terms of personality implies that distinct

insructional strategies rnay have greater benefit than others for the

different student populations. For example, based on the findings

indicating that social science students are more extroverted than arts

students, we could suppose that while a social science group would enjoy

class discussions and team work, an arts group would prefer traditional
lectures and individual work. One question needs to be explored further,

namely, do teaching methods characteristic of different fields attract and

retain different people, or are people shaped by the program they

choose?
Chapter 5: Summary and Conclusions

The present investigation provides weak support for the hypothesis

that cegep students enrolled in arts, sciences, and social sciences can be

distinguished on the basis of their personality. Indeed, individuals in

social sciences were shown to be more extraverted than arts students,

and a trend, more particularly for males, also suggested that social

sciences tend to be more extraverted than science students. This finding

has implications in the selection of teaching strategies that will allow

students to learn more effectively the course material. In addition, this

information can be useful in vocational counselling, as a supplement to

vocational interests.

In addition, gender differences were found on Neuroticism,

Extraversion, and Conscientiousness, with females scoring higher than

males on each dimension. Yet, more research is needed to explore this

point further and is beyond the scope of this study.

Furthermore, exploratory analyses have indicated that gender

differences were also present within each program concentration on

measures of Openness to experience. The low male scores on this

dimension seem to account for this effect, as they varied from least open

in science, a traditionally male dominated field, to most open in social

science, a traditionally female dominated area. This observation

suggests for males a relationship between gender stereotypes and

Openness to experience, which might be examined further in future

research. In addition, more research is needed to move toward a better

understanding of personality differences between genders within

academic programs in terms of the five-factor model of personality, and to

study how this information could be used in applied settings.

Finally, the ski11 level of parents' occupations was shown to be

associated with high scores on the measure of Openness, a relationship

that was supported by previous research. However, future investigations

need to address the question of whether the level of Openness can be

increased as a result of education, or whether Openness leads students

to pursue more education.

Despite our positive findings, some limitations of the present study

have to be considered when interpreting our results, namely the

weaknesses of Our instrumentation, our volunteer sampling rnethodology,

and our relatively small, unbalanced sample.

 Al1 in ail, this study suggests that more research is needed to

determine the educational correlates of personality traits, and how the

information could be used in an educational setting. Furthemore. future

research is needed to improve the French version of the NEO-FFI, sa that

it may eventuaily be used reliably for many different applications involving

the francophone population in Quebec and the rest of Canada.

 References

Angleitner, A. & Ostendorf, F. (1994). A cornparison of different

instruements proposed to measure the Big Five. European Review of

Applied Psychology, 44, 45-53.

Angleitner, A.. Ostendorf, F. & John, O. P. (1990). Towards a

taxonomy of personality descriptors in G e n a n : A psycholexical study.

Special Issue: Personality Language. European Journal of Personality, 4,

89-1 18.

Baker, D. R. (1983). Can the difference between male and female

science majors account for the low number of women at the doctoral level

in science? Research in Colleqe Students Teaching, 13, 102-107.

Barrick, M. R., & Mount, M. K. (1991). The Big Five personality

dimensions and job performance: a meta-analysis. Personnel

44,Q 1-25.
PSYC~O~O Y,

Bendig, A. W. (1963). The relation of temperamant traits of social

extraversion and emotionality to vocational interests. Journal of General

PSYC~O~O Q 31
69, Y ,1-318.

Briggs, S. R. (1992). Assessing the five-factor model of personality

description. Journal of Personality. 60, 271-291.

 Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook

for the Sixteen Personalitv Factor Questionnaire. Champaign, IL: lnstitute

for Personality and Ability Testing.

Clark, M. L. (1986). Predictors of scientific majors for black ans

white college students. Adolescence, 12, 205-213.

Conley, J. J. (1985a). Longitudinal stability of personality traits: A

multitrait-multimethod-multioccasionanalysis. Journal of Personality and

Social Psycholo~y,49, 1266-1282.

Corulla, W. J., & Coghill, K. R. (1991). Can educational streaming

be linked to personality? A possible Iink between extraversion,

neuroticism, psychoticism and choice of subjects. Personalitv and

Individual Differences. 12, 367-374.

Costa, P. T. Jr., McCrae, R. R.,8 Holland, J. R. (1984).

Personality and vocational interests in an adult sample. Journal of

Applied Ps~choloav.63, 390-400.

Costa, P.T. Jr., & McCrae, R. R. (1988). Personality in adulthood:

A six-year longitudinal study of self-reports and spouse ratings on the

NE0 Personality Inventory. Journal of Personality and Social

Psvcholoay, 54, 853-863.

Costa, P. T. Jr., & McCrae, R. R. (1992). The Revised NE0

Personalitv Inventon/ Professional Manual. Odessa, FL: Psychological

Assessment Resources.
 Costa, P. T. Jr., 8 McCrae, R. R. (in pressc). Trait psychology

cornes of age. In T. B. Sonderegger (Ed.), Nebraska Symposium on

Motivation: Psycholo~yand Aaing. Lincoln, NE: University of Nebraska

Press.

Costa, P. T. Jr., McCrae, R. R., & Dye, D. A. (1991). Facet scales

for Agreeableness and Conscientiousness: A revision of the NE0

Personality Inventory. Personality and Individual Differences, 12, 887-

898.

Digman, J. M. (1990). Personality structure: Emergence of the

five-factor model. In M. R. Rosenzweig and L. DW. Porter (Eds.), Annual

Review of Psychology (Vol. 41, pp. 417-440). Palo Alto: Annual Reviews,

lnc.

Digman, J. M., & Takemoto-Chock, N. K. (1981). Factors in the

natural language of Personality: Re-analysis, cornparison, and

interpretation of six major studies. Multivariate behavioral research, 16,

149-170.

Dollinger, S. J., & Orf, L. A. (1991). Personality and performance

in 'personality": Conscientiousness and openness. Journal of Research

in Personality, 25, 276-284.

Eysenck, S. B., 8 Allsopp, J. F. (1986). Personality differences

between students and craftsmen. Personality and Individual Differences,

-
7,439-441.
 Eysenck, H. J., & Eysenck, S. B. G.(1975). Manual of the Eysenck
Personalitv Questionnaire. San Diego: EdlTS Publishers.

Finn, S. E. (1986). Stability of personality self-ratings over 30

years: Evidence for agelcohort interaction. Journal of Personality and

Social Psychologv, 50, 813-818.

Harris, J. A. (1993). Personalities of students in three faculties:

perception and accuracy. Personality and Individual Differences, 15, 351-

352.

Helson, R., & Moane, G. (1987). Personality change in women

from college to midlife. Journal of Personality and Social Psycholo~y.53,

176-186.

Goldberg, L. R. (1990). An alternative "description of personality" :

The Big Five factor structure. Journal of Personality and Social

Psvcholoav, 59, 1216-1 229.

Jung, J. (1995). Ethnic group and gender differences in the

relationship between personality and coping. Anxiety, Stress. and

Copin~,8, 113-126.

Kaplan, R. M., & Saccuuo, D. P. (1993). Psycholocjcal Testincr:

Principles, Applications. and Issues. Third Edition. Pacific Grove,

California: BrookslCole Publishing Company.

Kirmayer, L. J., & Robbins, J. M. (1993). French research

translation of the NEO-FFI.

 Kline, P., 8 Lapham, S. L. (1992). Personality and faculty in

British universities. Personality and Individual Differences, 13, 855-857.

Lowman, R. L. (1993). The inter-domain model of career

assessrnent and counseling. Journal of Counselino and Development,

71,549-557.
-
McAdarns, D. P. (1994). The Person: An Introduction to

Personality Psycholoqy. Second Edition. Orlando, FL: Harcourt Brace &

Company.

McCrae, R. R. (1987). Creativity, divergent thinking, and openness

to experience. Journal of Personalitv and Social Psycholo~y.52, 1258-

1265.

McCrae, R. R., & Costa, P. T. Jr. (1980). Openness to experience

and ego level in Loevinger's Sentence Completion Test: Dispositional

contributions to developmental rnodels of personality. Journal of

Personality and Social Psycholocry, 39, 1179-1190.

McCrae, R. R., & Costa, P. T. Jr. (1985). Updating Norman's

"adequate taxonomy": Intelligence and personality dimensions in natural

language and in questionnaires. Journal of Personalitv and Social

Psvcholony, 49, 710-721.

McCrae, R. R., & Costa, P. T. Jr. (1987). Validation of the five-

factor model of personality across instruments and observers. Journal of

Personalitv and Social Psycholo~v,52, 81-90.

 McCrae, R. R., 8 Costa, P. T. Jr. (1989). The structore of

interpersonal traits: Wiggins' circumplex and the five-factor model.

Journal of Personality and Social Psycholoav. 56, 586-595.

McCrae, R. R., 8 Costa, P. T. Jr. (1990). Personality in Adulthood.

New York: Guilford Press.

McCrae, R. R., 8 Costa, P. T. Jr. (1991). The NE0 Personality

Inventory: Using the five-factor model in counseling. Journal of

Counselin~and Development, 69, 376-372.

McCrae, R. R. 8 John, O. P. (1992). An introduction to the five-

factor model and its applications. Journal of Personalitv. 60, 175-215.

McPartland, J. M., & Epstein, J. L. (1975). Social class differences

in the effects of open schools on student achievement. [On-line].

Baltimore, MD:John Hopkins Univertsity, Center for Social Organization of

Schools Report (Vol 1 93). Abstract frorn PsychlNFO Item: 54-08414

Ostendorf, F., & Angleitner, A. (1994). A comparison of different

instruments proposed to measure the Big Five. European Review of

Applied Psycholo~y,44, 45-53.

Rolland, J-P. (1993). Validité de construct de << marqueurs » des

dimensions de personnalité du modèle en cinq facteurs. Revue

européenne de Psychologie Appliquée, 43, 317-337.

Super, D. E. (1957). The Psvcholoav of Careers. New York :

Harper B Brothers.
 Tokar, D. M., & Swanson, J. L. (1995). Evaluation of the

correspondance between Holland's vocational personality typology and

the five-factor mode1 of personality. Journal of Vocational Behavior. 46,

89-108.

Wankowski, J. A. (1968). Some aspects of motivation in success

and failure at University. In Proceedinas of the Fourth Annual

Conference of the Society for Research into Hiqher Education. London:

SHRE.

Wilson, G.D. (1978). Introversion-extraversion. In H. London and

J. E. Exner, Jr. (Eds.), Dimensions of Personality (pp.217-261). New

York: John Wiley & Sons.

Wolfe, R. N., & Johnson, S. D. (1995). Personality as a predictor

of school performance. Educational and Psychological Measurement. 55,

177-185.
Appendix 1: Sociodemographic Questionnaire
 Information socio-dernographique

Établissement scolaire:

Sexe: masculin féminin

Âge:

Programme: Est-ce votre le", 2'"' ou 3*"' année dans

ce prggramrne?

Considérez-vous votre ville d'origine une région rurale, semi-rurale, ou

urbaine?

Quelle est votre ville d'origine?

Langue maternelle:

Quel est le niveau scolarité de vos parents? Mère:

Père:

Quelle est l'occupation de vos parents? Mère:

Père:
Appendix 2: lnformed Consent
 Consentement éclairé

Cher étudiant,

Chère étudiante,

Dans le cadre de mes études de maitrise, je conduis

un projet de recherche visant à étudier la personnalité des étudiants et
étudiantes de niveau cégep inscrits dans les programmes de sciences
pures, sciences humaines et arts afin de déterminer s'il existe des
différences entre les individus poursuivant des programmes différents.
De plus, je tiens à examiner les différences de personnalité entre les deux
sexes, ainsi que l'influence de certaines variables socio-démographiques-
Le but de cette étude est de déterminer si l'information révélée par les
mesures de personnalité pourraient éclairer les étudiants effectuant leur
choix de programme collégial.

Si vous êtes intéressé(e) à participer à ce projet, vous aurez a

remplir le questionnaire intitulé NEO-FFI (traduction française du << NE0
Five Factor lnventory D), lequel devrait vous prendre de 10 à 15 minutes
à compléter. Les réponses que vous indiquerez seront analysées et
traitées en toute confidentialité. En aucune façon sera-t-il possible de
vous identifier personnellement. Vos résultats individuels seront cumulés
à ceux des autres participants de manière à ce que les données soient
analysées en groupe. Votre participation est entièrement volontaire et
vous pouvez vous retirer du projet en tout temps. Je vous remercie à
l'avance de bien vouloir participer à cette recherche.

Nom:

Signature:

Date:
 lMAGE EVALUATION
TEST TARGET (QA-3)

APPLIED 4 IMAGE. lnc

-.- Rochester.
--
1653 East Main Street
--
-
--
--
- - - , NY 14609 USA
Phone: 7 16/482-0300
Fax: i l 61288-5989

O 1993. & p l i i Image. Inc.. All Rights Reserved