Lkfi ‘- ~r , k . u. . ' 41‘.“ > ~ ‘ ‘ - u

314): , . . . 13:? 35..“ - .2 . ,. w;3}g§;;‘ .,~ 1%5flléa2“
a V. ' . l, ‘ ' u ‘ ‘1‘ l

‘ 2-". ""

"‘ ». .V a u.

.‘ j .
.,¢y.§yg‘g§§$~y., ‘ an?

.. \5~¢-“_;;¢ I: «.3 a: 4;:

~. ~ w‘ sirflwacfi‘i. “‘\\.733.,‘ N

u ‘ "3?? mérrfififi.

w

u- ,' 35-”:

w - , ~ .»
‘ ‘ ' , x‘ ‘ 'u: \

“ hl’”fi”‘j~:"9‘m';¥ Q; '4 luv? fl “‘3:

‘3 ‘31“: 4.! Ana.“ 33$?“ fret»? «gfiic‘i'ta Wm: .. 7:“ ‘ I“ '
W‘F3'P-i‘1aiiiwéii“ .‘- 0‘ ' & 2‘73: W‘xfisgh v? _...;,
Wyn-“t ' 3"" .1, - ;£&:%:;_-mh ' Ti‘RfiWt

u ~\- ' :3 fijfidfl‘xc 33%,;

1-144” ‘ ‘fifa

Unix,“ 1331‘

*m“‘“ ‘3-

1.x

‘1 . . u ‘ -
" z“ ‘ ’ . ~ A. r—aw 1“
.t‘f“-g:‘. ~. 7 u— 1‘ . m. ".flu . :,‘: 'T'L‘H f"
-. a... .LA‘LC‘ 1 <‘- u? *- ‘v‘ 1c: - :. . 333,} (“53‘
~ ‘
3% ~

J,
.x ... ‘ ..
ukh‘;

"pm-L

*3!

. , .
at ‘1'- . .
. h u. .: .
z '3. £15.. *rfwm
’~\~y. ~31-1.“- g“ n
" ’ 1}.“ tuft.
' '- K‘ *1
‘ ‘6.

”' M5

,
‘Lr

:
3‘

3
-?
.-l

 

 

AV .

1113'};- ‘ hf

. '4
'\ ‘1
. "u. v.

2.7;?) .4?» ‘ “3‘;
‘wéva .. y
-.. ., w
«r a
x’ ..

v. .
.

. cirfim . "IT

“fiat-"f: . 1;:-
frrggfiq, ' .

[.91

3‘53 .‘ :7: .‘ . “5.4.3
' ‘ ”3"“ "a?" '1 ”‘J‘E‘w’w “.45"; 3*" '9; hifg‘.;7 vfi'mfi " .. -~ £5???"

‘ ' - - ~. .q, " -' x , ' ; w _ ' ‘ ‘ _ r . ».
-§?€a‘rt’lfi% -‘ H 1 #36 fin!" : y .3 " ~ ‘1}? v . 2:? {Wm we
. YW‘!;1.' » .. ‘5‘ r-v ‘ ‘4 > H . ‘ 3 . .

z ”5892.. “1 o 2,: . . 3:339- ‘ $3? .5333:

“Niki-fr .2: TAHMWV‘FF‘Tfi.‘

74' y :1 z
R"‘~?n§:§ ’ .J' 3‘
1""..1 v’h‘f' .1; - - '
U—T ‘x. ‘ ,1

"7';

w,‘ "
“A

. "é.

cl. s~ ,.

r91 4

Lauri}!

.
. JJ
.va

writ-‘-
I w

.V

 

NWERSITY UBRARIE

lllllll’llllfll 1n Ill Hill

3 1293 01051 7773

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

Personality Factors, Defense
Mechanisms, and Social Support:
A Hierarchical Structural Model

presented by

Stephen Bruce Kincaid

has been accepted towards fulfillment
of the requirements for

Ph,D degree in m

ll... AM.

Major professor

 

[Mme February 17, 1994

 

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

tomwombchockoutMywrocord.

PLACE ll RETURN BOX
dd. duo.

TO AVOID FINES Mum on or

     
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
    

 

 

__,_.— _____ _4._ l__ __ 4 __4_ _

 

 

l

l ’—
I
l

r— 1s a - -

" usu IoAn Nflmativo Action/Equal opportunity Institution
WM!

 

 

 

 

 

 

PERSONALITY FACTORS, DEFENSE
MECHANISMS, AND SOCIAL SUPPORT:
A HIERARCHICAL STRUCTURAL MODEL

By

Stephen Bruce Kincaid

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1994

ABSTRACT
PERSONALITY FACTORS, DEFENSE
MECHANISMS, AND SOCIAL SUPPORT:
A HIERARCHICAL STRUCTURAL MODEL
by

Stephen Bruce Kincaid

A hierarchical structural model of the relationship between the Big Five
personality traits (operationalized by the NEO-PI-R), defense mechanisms
(operationalized by the Defense Mechanisms Inventory), and social support
(operationalized by the Norbeck Social Support Questionnaire) was examined in a
sample of 289 undergraduate students. Confirmatory factor analyses examined the
psychometric properties of the N EO-PI-R and the DMI; subsequent revisions greatly
improved the factors, and resulting coefficient alphas ranged from .84 to .88 for the
NEO—PI-R and .59 to .82 for the DMI. ANOVAS found scales N (F[l,285] = 26.4;
p_ <.001), E (F[l,285] = 22.9; p_ < .001), O (F[l,285] = 5.4; p < .05), and A
(F[l,285] = 6.4; p_ < .05) of the NEO-PI-R, scales PRO (F[l,274] = 29.6;

p_ < .001), and TAS (F[l,274] = 57.3; p_ < .001) of the DMI, and the Total Social
Support Score (F[l,287] = 8.0; p_ < .01) derived from the. NSSQ all varied with
gender; thus, gender was incorporated into the model. Twelve of 15 specific
hypotheses regarding the relationship of personality traits to defense mechanisms were
supported, and three of five specific hypotheses regarding the relationship of defense
mechanisms to social support were supported. High intercorrelations were found of

DMI scales; a second-order factor analysis revealed two related (-.56), labelled

"internalizing" and "externalizing" defensive styles, and these were incorporated into
the model prior to path analyses.

Although the majority of the hypotheses about the intercorrelations of
personality, defense mechanisms, and social support were in the correct direction and
statistically significant, there were significant differences between the hypothesized
structural model and the observed data (Chi-square = 133.54; df = 19; p < .001);
the largest errors involved the intercorrelations of scales of the revised N EO-PI-R. A
second-order factor analysis of the revised NEO-PI-R scales revealed two related
factors (-. 18), which were also labelled "internalizing" and "externalizing" personality
styles. These factors were incorporated into the revised Structural model, and specific
linkages in the model were revised. Using path analysis, the revised structural model
was found to fit the observed data (Chi-square = 9.24; df = 5; p < ns); individual
analyses of construct linkages and missing linkages further suggested that the model

was an accurate representation of the data obtained.

ACKNOWLEDGEMENTS

I would like to thank my Dissertation Committee for the thorough and
thoughtful attention they provided as this project proceeded; their criticism, insights,
and suggestions contributed to all aspects of my work. Dr. Norman Abeles,
Committee Chairperson, patiently steered the dissertation through several
incarnations, assisting in both the conceptualization of the hierarchical model, the
specific hypotheses proposed, and the integration of the results. Dr. Ralph Levine
was instrumental in my assimilation and utilization of the sophisticated statistical
methodology necessary to adequately test the model, and guided revisions to the
model following the second-order factor analyses. Dr. John Hurley and Dr. Dozier
Thornton critically reviewed versions of the dissertation at several stages, and
provided excellent direction regarding the theoretical background, the specific
hypotheses proposed, and the construction of the model. Although not a member of
the dissertation committee, Dr. John Hunter gave generously of his own time to work
with me on the statistical programs used during the analyses. Finally, Lauren, Adam,
and Jordan were patient, supportive, and encouraging throughout the completion of

both this dissertation and the clinical program as a. whole.

iv

TABLE OF CONTENTS

List Of Tables
List Of Figures
Introduction
Theoretical Approaches To Personality
Empirical Approaches To Personality
The Five-Factor Model Of Personality
Interpretation Of The Five Factors

Defense Mechanisms As An Intrapersonal
Process Influenced By Personality

Social Support As An Interpersonal
Process Influenced By Personality

Major Propositions
Operationalizing The Constructs
The NEO-PI-R
The Neuroticism Scale
The Extraversion Scale
The Openness Scale
The Agreeableness Scale

The Conscientiousness Scale

.................... 21

.................... 24

.................... 25

.................... 26

.................... 26

.................... 26

.................... 27

.................... 27

.................... 27

Use Of The NEO-PI In Prior Research
The Defense Mechanisms Inventory
The PRN Defensive Style
The PRO Defensive Style
The REV Defensive Style
The TAO Defensive Style
The TAS Defensive Style
Use Of The DMI In Prior Research
The Norbeck Social Support Questionnaire
Use Of The NSSQ In Prior Research
Specific Hypotheses
Hypotheses Concerning The Relationship
Of Personality Traits To The
Characteristic Use Of Defense
Mechanisms
Hypotheses Concerning The Relationship
Of The Characteristic Use Of
Defense Mechanisms To Social
Support

Integrating The Hypotheses Into
A Complex System

Method
Procedure
Measures
Demographic Background

The NEO-PI-R

vi

.................... 27

.................... 29

.................... 30

.................... 30

.................... 31

.................... 32

.................... 33

.................... 34

..................... 35

.................... 36

.................... 37

.................... 37

.................... 47

.................... 47

.................... 47

.................... 47

.................... 48

The Defense Mechanisms Inventory .................... 48

The N orbeck Social Support Questionnaire .................... 49

Results .................... 50
Demographic Characteristics of the Sample .................... 50
Responses To The NEO-PI-R .................... 50
Responses To The Defense Mechanisms Inventory .................... 56
Responses To The Norbeck Social Support Questionnaire .................... 66

Results Of Tests Of Hypotheses Concerning
The Relationship Of Personality Traits

To The Characteristic Use Of Defense Mechanisms .................... 66
Results Of The Test Of Hypothesis 1a .................... 67
Results Of The Test Of Hypothesis 1b .................... 67
Results Of The Test Of Hypothesis 1c .................... 67
Results Of The Test Of Hypothesis 1d .................... 69
Results Of The Test Of Hypothesis 1e .................... 69
Results Of The Test Of Hypothesis 1f .................... 70
Results Of The Test Of Hypothesis 1g .................... 70
Results Of The Test Of Hypothesis 1h .................... 70
Results Of The Test Of Hypothesis 1i .................... 71
Results Of The Test Of Hypothesis 1 j .................... 71
Results Of The Test Of Hypothesis 1k .................... 71
Results Of The Test Of Hypothesis ll .................... 72
Results Of The Test Of Hypothesis 1m .................... 72
Results Of The Test Of Hypothesis 1n .................... 72

vii

Results Of The Test Of Hypothesis lo .................... 73

Results Of Tests Of Hypotheses Concerning
The Relationship Of The Characteristic

Use Of Defense Mechanisms To Social Support .................... 73
Results Of The Test Of Hypothesis 2a .................... 75
Results Of The Test Of Hypothesis 2b .................... 75
Results Of The Test Of Hypothesis 2c .................... 75
Results Of The Test Of Hypothesis 2d .................... 76
Results Of The Test Of Hypothesis 2e .................... 76
Results From Tests Of The Complex System .................... 76
The Assessment And Inclusion Of Demographic
Variables .................... 77
Revising The DMI .................... 77
Results Of The Initial Integrational Analyses .................... 82
Developing And Testing An Alternative Integrational Model .................... 89
Revising The NEO-PI-R .................... 89
Reconstructing The Model .................... 90
Testing The Alternative Model .................... 94
Further Analyses Of The Alternative Model .................... 95

Analysis Of The Path Coefficient Between
Intemalizing Personality Traits And
Intemalizing Defenses .................... 95

Analysis Of The Path Coefficient Between
Extemalizing Personality Traits And
Intemalizing Defenses ................... 101

Analysis Of The Path Coefficient Between
Extemalizing Personality Traits And

\

viii

Social Support ................... 101

Analysis Of The Path Coefficient Between
Intemalizing Defenses And Externalizing
Defenses ................... 101

Analysis Of The Path Coefficient Between
Gender And Externalizing Defenses ................... 102

Analysis Of The Path Coefficient Between
Gender And Social Support ................... 102

Analysis Of The Path Coefficient Between
Externalizing Defenses And Social Support ................... 102

Analysis Of The Missing Link Between
Intemalizing Personality Traits And
Externalizing Defenses ................... 103

Analysis Of The Missing Link Between
Externalizing Personality Traits And
Externalizing Defenses ................... 103

Analysis Of The Missing Link Between
Gender And Intemalizing Defenses ................... 104

Analysis Of The Missing Link Between
Intemalizing Personality Traits And

Social Support ................... 104

Analysis Of The Missing Link Between
Intemalizing Defenses And Social Support ................... 104
Discussion .................. 106
Psychometric Qualities Of The Measures ‘ ................... 106
Reliability Of The Measures ................... 106
Issues Of The Factors Of The Scales ................... 107
Interpreting The Lower Order Analyses ................... 109
Interpreting The Original Path Model ................... 113

Interpreting The Alternative Path Model
Theoretical Implications

Clinical Implications
Summary
Appendices
Appendix A: Informed Consent Form
Appendix B: Demographic Questionnaire

Appendix C: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Neuroticism Scale Of The NEO-PI-R

Appendix D: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Extraversion Scale Of The NEO-PI-R

Appendix E: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Openness Scale Of The NEO-PI—R

Appendix F: Results Of A Confu'matory Factor
Analysis With Items Comprising The
Agreeableness Scale Of The NEO-PI-R

Appendix G: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Conscientiousness Scale Of The NEO-PI-R

Appendix H: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Neuroticism Scale Of The NEO-PI-R

Appendix I: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Extraversion Scale Of The NEO-PI—R

Appendix J: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Openness Scale Of The NEO-PI-R

................... 116

................... 119

................... 121

................... 122

................... 124

................... 125

................... 126

................... 128

................... 130

................... 132

.................. 134

................... 136

................... 138

................... 139

Appendix K: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Agreeableness Scale Of The NEO-PI-R ................... 141

Appendix L: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Conscientiousness Scale Of The NEO-PI-R .................... 143

Appendix M: Raw Scores For The Revised NEO-PI-R Scales,
By Gender ................... 145

Appendix N: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Principalization Scale Of The DMI ................... 153

Appendix 0: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Projection Scale Of The DMI ................... 155

Appendix P: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Reversal Scale Of The DMI ................... 157

Appendix Q: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Turning Against Others Scale Of The DMI ................... 159

Appendix R: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Turning Against Self Scale Of The DMI ................... 161

Appendix S: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Principalization Scale Of The DMI ................... 163

Appendix T: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Projection Scale Of The DMI ................... 164

Appendix U: Results Of A Confirmatory Factor

Analysis With Items Comprising The
Revised Reversal Scale Of The DMI ................... 165

xi

Appendix V: Results Of A Confirmatory Factor
Analysis With Items Comprising The
Revised Turning Against Others Scale Of The DMI ................... 166

Appendix W: Results Of A Confirmatory Factor
Analysis With Items Comprising The

Revised Turning Against Self Scale Of The DMI ................... 167
Appendix X: Raw Scores For The Revised DMI Scales,

By Gender ................... 168
Appendix Y: Raw Scores For The NSSQ Scales,

By Gender ................... 176
Appendix Z: Raw Scores For Scales Derived From

Second Order Factor Analyses, By Gender ................... 184

List Of References ................... 192
xii

 

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

LIST OF TABLES

The Hypothesized Direction Of Correlations
Among Scales Of The NEO-PI-R And The DMI

The Hypothesized Direction Of Correlations
Among Scales Of The DMI And The Total
Social Support Score Derived From The NSSQ

Sample And Normative Scores For Scales Of
The NEO-PI-R; Males

Sample And Normative Scores For Scales Of
The NEO—PI-R; Females

Alpha Reliabilities Of The Scales Of The NEO-PI-R

Intercorrelations of Scales Of The NEO-PI-R After
Correcting For Attenuation Due To
Measurement Error

Alpha Reliabilities Of The Revised Scales Of
The NEO-PI-R

Intercorrelations Of The Scales Of The Revised
NEO-PI-R After Correcting For
Attenuation Due To Measurement Error

Sample And Normative Scores For Scales Of
The DMI; Males

Table 10: Sample And Normative Scores For Scales Of

The DMI; Females

Table 11: Alpha Reliabilities Of The Scales Of The DMI

xiii

.................... 39

.................... 52

.................... 53

.................... 54

.................... 55

.................... 57

.................... 58

.................... 60

.................... 61

.................... 62

Table 12:

Table 13:

Table 14:

Table 15:

Table 16:

Table 17:

Table 18:

Table 19:

Table 20:

Table 21:

Table 22:

Table 23:

Table 24:

Intercorrelations Of The Scales Of The DMI After
Correcting For Attenuation Due To Measurement
Error

Alpha Reliabilities Of The Revised Scales Of
The DMI

Intercorrelations Of The Scales Of The Revised
DMI After Correcting For Attenuation
Due To Measurement Error

Correlations Among The NEO—PI-R And The DMI After
Correcting For Attenuation Due To
Measurement Error

Correlations Among The DMI And Total Social Support
Score Derived From The NSSQ After Correcting
For Attenuation Due To Measurement Error

Means, Standard Deviations, And Analyses of Variance
Of All Revised Scales As A Function Of Gender

Means, Standard Deviations, And Analyses of Variance

.................... 65

.................... 68

.................... 74

.................... 79

Of All Revised Scales As A Function Of Education Level ............... 80

Means, Standard Deviations, And Analyses of Variance
Of All Revised Scales As A Function Of Race

Scale Loadings Onto Second Order Factor Matrix
For The Revised DMI Scales

Second Order Factor Matrix For The Revised
DMI Scales

Correlation Matrix For All Measures In The Revised
Structural Equation Model After Correcting For
Attenuation Due To Measurement Error

Path Coefficients For All Linkages In The Revised
Structural Equation Model

Standard Errors For Path Coefficients In The Revised
Structural Equation Model

xiv

.................... 81

.................... 83

.................... 85

.................... 86

.................... 87

Table 25:

Table 26:

Table 27:

Table 28:

Table 29:

Table 30:

Table 31:

Errors (Actual - Predicted) For All Linkages In
The Revised Structural Equation Model

Scale Loadings Onto Second Order Factor Matrix
For The Revised NEO-PI-R Scales

Second Order Factor Matrix For The Revised
NEO-PI-R Scales

Correlation Matrix For All Measures In The
Alternative Structural Equation Model
After Correcting For Attenuation Due
To Measurement Error

Path Coefficients For All Linkages In The
Alternative Structural Equation Mode]

Standard Errors For Path Coefficients In The
Alternative Structural Equation Model

Errors (Actual - Predicted Correlations) For

All Linkages In The Alternative Structural
Equation Model

XV

.................... 88

.................... 92

.................... 93

.................... 97

.................... 98

.................... 99

................... 100

Figure 1:

Figure 2:

Figure 3:

Figure 4:

LIST OF FIGURES

Hypothesized General Structural Relationship
Of Personality, Defense Mechanisms, And
Social Support

Revised Hypothesized General Structural
Relationship of Personality, Defense
Mechanisms, and Social Support

Alternative Hypothesized Structural Relationship
Of Personality, Defense Mechanisms, And
Social Support

Path Coefficients For Alternative Hypothesized

Structural Relationship of Personality,
Defense Mechanisms, and Social Support

xvi

.................... 38

.................... 78

.................... 91

.................... 96

INTRODUCTION

"Personality" is an often used term to describe the mechanism of control of
human behavior. Within both theoretical and empirical psychology, personality is
often invoked as a causal explanation for individual differences in behavior. Despite
being widely used, the concept of personality is difficult to define, and a Single,
universal theory has not found acceptance. Allport (1937) suggested that personality
referred to the integration of all behaviors and tendencies that characterize an
individual’s orientation to the environment. Guilford (1959) saw personality as a
constellation of traits. Maddi (1980) described personality as the commonalities and
differences between individuals in their thoughts, feelings and action that are
independent of social and biological contexts.

Over the past decade there has been "an electrifying burst of interest in the
most fundamental problem [of personality research] - the search for a scientifically
compelling taxonomy of personality traits" (Goldberg, 1993; p. 26). Such interest has
led to the refinement of the five-factor theory of personality, or the Big Five. This
theory, an increasingly accepted model of personality (Buss, 1991; Costa & McCrae,
1992; Digman and Inouye, 1986), is derived using factor analytic techniques in efforts
to identify a relatively small number of stable, enduring dimensions that can be used

as a means of explaining larger numbers of more transient, situational dimensions

2
in human behavior. Although there are other factor-analytically derived models,

analyses which identify five factors (discussed below) have been recurrent and
replicated.

Recent Studies have begun to explore how personality traits, such as those of
the five-factor theory, may be associated with both intrapersonal and interpersonal
processes. Intrapersonal processes are those which occur within the individual;
interpersonal processes are those which occur between individuals. This study,
following a review of the current literature, proposes and examines a hierarchical
model of the relationship of the five-factor theory of personality to defense
mechanisms (as a mode of intrapersonal behavior) and to social support (as a mode of

interpersonal behavior).

Theoretical Approaches To Personality
The most widely known personality theory was developed by Sigmund Freud.

Freud’s personality theories, developed concomitantly with his theories of
psychoanalysis, attempted to provide causal explanations for the symptom
presentations he saw in his patients. Over the course of his writings, Freud proposed
two theories of personality. Freud’s (1920) first explanatory theory, often called the
topographical theory, focused on the instinctual nature of human behavior. Freud
proposed that human behavior was impacted on by instincts at three levels of
awareness: conscious, preconscious, and unconscious. In this theory, thoughts,
feelings and instincts which are at the conscious level are fully accessible to the

awareness and experience of the individual; thoughts, feelings and instincts at the

3

preconscious level are those that, although not immediately accessible, can be brought
into awareness and experience by the individual; thoughts feelings and instincts at the
unconscious level are those which are not immediately accessible and can not be
brought into awareness and experience at will by the individual. Human behavior, in
this theory, was a combination of instincts, thoughts and feelings at all three levels,
and to fully understand a person’s behavior the unconscious motivations as well as the
conscious motivations must be understood.

Freud soon found that there were conceptual problems with his topographical
theory concerning the role of affects other than anxiety (for example, guilt and
paranoia). Thus, Freud expanded his model and developed the structural theory of
personality (Freud, 1923), which divides the mind into three forces which traverse the
conscious, the preconscious and unconscious. This theory posed that personality
arises from the interplay or dynamics between three internal forces: primitive or
infantile desires and wishes, an internalized social value system, and a sense of reality
and the self. These forces were metaphorically represented as the id, superego and
ego, respectively. An integral component of the structural theory was the concept .of
drives. Arising from the id, drives were described as an energy source
(Freud, 1905), a mixture of psychic and somatic demands. These drives are can be
understood as an incorporation of aspects of his earlier instinctual theory
(Freud, 1915). Greenberg and Mitchell (1983) noted that these drives represented the
most fundamental human passions and urges, and as such, were at once both the
mechanism and the content of the mind. Freud (1926) theorized that all drives press

for discharge, that is, for expression. However, the expression of drives, and often

 

4

even awareness of drives, could be in opposition to the unique set of internalized
values and morals represented by the superego. Thus, the ego must integrate both the
impulsive demands of the id and the inhibitory demands of the superego

(Freud, 1923), and the characteristic manner in which it does so was described as
"personality" or "style. "

The Freudian view is not the only theoretical approach to personality, though
it was not until the early portion of this century that a body of knowledge began to
develop centered around the new concept of "personality" (Bumham, 1968).
Behaviorism, or learning theory, essentially rejects the existence of personality by
positing that behavior is modified and maintained through its consequences (Skinner,
1938; Watson, 1930). Thus, the development of personality is largely determined by
the manner in which the environment reacts to the behavior (Phares, 1984).
Phenomenological theories of personality (Kelly, 1955; Rogers, 1961) do not focus on
the interplay between hypothesized psychic structures and the integration of instincts
and drives but on the characteristic manner in which a person perceives, experiences,
and feels. In this view, the development of personality is largely determined by the
manner in which the environment is perceived by the individual (Phares, 1984).
Social learning theories of personality (Bandura & Walters, 1963; Mischel, 1973;
Rotter, 1982) generally focus on both cognitive and environmental factors, including
expectancies, reinforcement saliency, motivation, and observation. Thus, the
development of personality (though not likely to be defined as such) is interactional

(Phares, 1984).

Empirical Approaches To Personalig

Interest in personality does not reside solely within theoretical domains. While
the variety of human characteristics ascribed to personality is great, empirical efforts
have been made to impose order and organization. These efforts, typically based on
factor analytic methodology, attempt to separate the global concept of personality into
specific, orthogonal dimensions. Such an approach assumes that human behavior is
consistent across time and context, a view not universally embraced. The "trait"
hypothesis versus "state" hypothesis argument is woven throughout empirical
psychology, and is central to the study of personality structure (Wiggins & Pincus,
1992). "State" theories of personality typically posit that there is situational
specificity to human behavior, with few identifiable tendencies, and that human
behavior is ultimately derived from environmental forces. "Trait" theories of
personality typically posit that human behavior shows generalized tendencies,
predispositions, and consistencies, and that human behavior is ultimately derived from
internal forces understood as " personality. "

While this argument is ongoing, of interest to this study are factor-analytically
derived trait models of personality. Such models provide a manageable number of
factors, or building blocks, from which behavior can be described and understood.
These models attempt to hierarchically organize the multitude of variables of human
differences; each identified factor may incorporate "hundreds, if not thousands, of
traits" (Goldberg, 1993; p. 27). In an early empirical study of traits, Cattell (1943;
1947) used factor analysis to empirically derive over 35 bipolar dimensions of

personality, of which he ultimately reported 12 oblique factors (Peabody & Goldberg,

 

6
1989). Cattell first used peer ratings of college students, then later used both

questionnaire and objective test data in his factor analytic Studies (Digman, 1990).
Partly inspired by Cattell’s work, subsequent (and more sophisticated analyses) soon
identified five replicable factors, which have come to be known as the "Big Five "
personality theory or the "five-factor model" (Norman, 1963; Tupes &
Christal, 1992; Note: Tupes and Christal originally published their findings in 1961.
However, this report appeared in an Air Force manual with limited accessibility, and
most researchers have had to rely upon secondary sources to report their findings.
Recently, their original report was reprinted in the Journal of Personalig. Because of
the greater distribution and accessibility of this journal, it is the citation used here).
The five-factor model of personplitv. The five—factor model of personality
(alternatively known as the Big Five) has consistently been identified in empirical
studies of personality (Costa & Oliver, 1992; Digman and Inouye, 1986) and provides
"a good deal of the most important information one might expect to gather in order
describe an individual’s personality" (McAdams, 1992, p. 331). Although the labels
for these factors have not been universally accepted, the five factors generally assess
dimensions of Power, Love, Work, Affect and Intellect (Peabody & Goldberg, 1989).
Cattell (1943) and Norman (1963) are often cited as proximate influences of the
modern five-factor theory. However, Goldberg (1993) has traced the history of the
five-factor model to more distal influences, including Sir Frances Galton, who
attempted to estimate and categorize the number of terms in the English language
capable of describing personality. Descriptive terms of personality were chosen as a

Starting point for analysis because of the assumption that important characteristics of

7
personality would naturally be represented and encoded within language (Goldberg,

1993; Peabody & Goldberg, 1989). L. L. Thurman, working approximately 50 years
after Galton, applied pioneering factor analytic techniques to adjectives used for
describing people, and identified five independent factors (Thurstone, 1934), though
Goldberg suggested that his broad selection of adjectives used to describe the factors
disqualified identifying him as the discoverer of the five-factor theory as currently
understood. Instead, Goldberg points to Donald Fiske, who analyses led him to label
the five factors as (I) Confident Self-Expression, (H) Social Adaptability,

(III) Conformity, (IV) Emotional Control, and (V) Inquiring Intellect (Fiske, 1949;
cited in Goldberg, 1993). However, Fiske apparently did not follow up further on
these findings, and is subsequently displaced as an "accidental discoverer " in
Goldberg’s family tree. The "true fathers" are Cattell (1943) and Tupes and Christal
(1992), as both were the first to initiate and sustain research efforts toward describing
a factorial model of personality (Goldberg, 1993).

Not all investigators within the trait theory of personality embraced the concept
of five distinct factors. Norman (1963), an early critic of the model, believed that the
lack of computational resources limited earlier research from identifying more than
five factors, and proposed that as such limitations were overcome, additional factors
would be uncovered; this premise, however, was later tested and rejected (Goldberg,
1990). Others proposed that five factors was more than was needed to explain
personality adequately. For example, Eysenck (1960) used factor analysis to propose
a personality model containing two orthogonal, bipolar dimensions, with one axis

representing extraversion-introversion and one axis representing stability—instability.

 

8
Circumplex models have been proposed by Leary (1957), Benjamin (1974), and

Wiggins (1979), with dimensional bipolar scales of "affiliation" and "dominance"
being the common anchors for the axes of the circumplex (Wiggins, 1982).

Despite divergence in views, the initial findings in trait studies of personality
were encouraging. However, the Zeitgeist of personality theory changed during the
19605 as "state" research seemed to indicate that situational factors were more
predictive of behavior than personality factors (Digman, 1990). Mischel (1968) even
asserted that behavior was so variable and state-dependent that it was wrong to utilize
constructs such as ”personality". During this era, there was a diminished enthusiasm
for factor analytic models of personality, especially as some researchers claimed that
the five factors could be identified simply from semantic similarities among the items
loading onto each factor (Borkenau, 1988); this led to the concern that the traits
measured by factor analytic studies resided primarily in the rater.

Over the past decade, however, there has been a substantial increase in the
empirical support for the influence of pervasive individualistic traits. Numerous
researchers, using oblique rotations on widely varied data sets, have reported
identifying a five-factor solution similar those initially described during 19605
(Digman & Inouye, 1986; Piersma, 1986; Mche & Costa, 1987). Stepping outside
of academic, empirical terminology, Goldberg (1981) noted that these factors provide
information on five basic questions about a person: (1) Is the person active and
dominant or passive and submissive?; (2) Is the person warm and pleasant or cold and

distant?; (3) Is the person responsible and conscientious or undependable and

9

negligent?; (4) IS the person stable or crazy?; and, (5) Is the person oriented toward
intellectual pursuits or not?

Much of the recent revival of interest in the five-factor model has been led by
the research of Costa and McCrae, which Goldberg (1993) believes "did more to
form the modern consensus about personality structure than anything else that
occurred during the 19808" (p. 31). In their early work, Costa and McCrae (1976)
identified two factors (which they labelled "extraversion" and "neuroticism")
underlying the 16 PF, Cattell’s inventory of personality assessment, both of which
could be explained by earlier empirical efforts. They also identified a third factor
(which they labelled ”openness"). Since the 16 PF did not provide a good measure
for the third factor, they developed the NEO, or the "Neuroticism-Extraversion-
Openness Personality Inventory. " Responding to the work in the early 19805 of
Digman and Goldberg (Goldberg, 1993), Costa and McCrae (1985) subsequently
transformed their three-factor personality questionnaire into a five-factor questionnaire
by adding scales to reflect the dimensions of "agreeableness" and "conscientiousness"
(Costa & McCrae, 1985; Costa & McCrae, 1986); these factors will be discussed
further below. According to Costa & McCrae (1992) the Big Five ”represent the
most basic dimensions underlying the traits identified in both natural languages and
psychological questionnaires" (p. 14).

Research on the five-factor model has extended beyond identifying factors
underlying personality assessment questionnaires. Analyses have found the five-factor
model to be robust across time, context, and type of analysis (Buss, 1991). The five-

factor model adequately accounts for earlier findings about the Dominance and

10
Warmth dimensions of the Interpersonal Circumplex (McAdams, 1992) The model

has been identified in longitudinal studies of children (Digman, 1989) and in
evaluations of adolescent adjustment as assessed by both teachers and school
counselors (Graziano & Ward, 1992). Meta-analytic studies of personnel selection
techniques have found that job performance can be associated with a five-factor model
(Barrick & Mount, 1991; Tett, Jackson, & Rothstein, 1991). A five—factor model
was found in lists of German adjectives (Ostendorf, 1990), as well as in Japanese
(Bond, Nakazato, & Shiraishi, 1975) and Chinese (Yang & Bond, 1990). Buss
(1991; 1992) has theorized the evolutionary advantages of developing these five facets
of personality, and has identified interpersonal tactics associated with different factors.
In fact, studies have so consistently replicated the findings of a five-factor model that
Digman and Inouye (1986) argued support was " consistent enough to approach the
status of law” (p. 116). Proponents of the five-factor model now argue that the
majority of all measured personality constructs can be interpreted as representing one
or more of the factors in the Big Five model (Digman, 1990).

This does not mean the theory has no critics. Goldberg (1993), in a review,
found arguments for both fewer and more than five (factors. Further, Waller and
Ben-Porath (1987) have argued that the five-factor model was accepted too quickly, as
Cattell’s original work has simply been replicated rather than subjected to conceptual
testing. More recently, McAdams (1992) outlined a comprehensive and severe
criticism of the Big Five on a theoretical level. McAdams noted that the five factors
were derived from lexical studies without a priori hypotheses, and thus, given the

subjective nature of interpreting factor analyses, do not represent a solid basis for

11

scientific inquiry. McAdams is also critical of the atheoretical nature of the five-
factor model, noting that the five factors are at best dimensions of the periphery in
personality rather than dimensions of the core (Note: The argument regarding
periphery and core dimensions of personality lies outside the scope of this study. In
brief, McAdams [1992] reviewed and drew on the concepts of Maddi [1980], who
delineated periphery dimensions from core dimensions of personality. Periphery
dimensions include trait constructs which can be used to identify particular
individuals, while core dimensions include constructs which deal with basic,
fundamental personality organization. Within Freudian theory, McAdams noted that
the id, ego and superego would be examples of core dimensions while oral and anal
personality types would be periphery dimensions). Because the five-factor model,
according to McAdams, does not address the basic nature of personality organization,
it can never be a complete theory. McAdams, drawing on the work of Revelle
(1987), questioned whether the five factors are actually a causal taxonomy, and
suggests that they may be only a descriptive taxonomy, and being conceptualized as
independent dimensions, lose sight of the person who holds the personality (of ,
Carlson, 1971). Finally, he raised issues about the specificity of information
generated by tools assessing the Big Five. McAdams noted that by using a five-point
Likert scale, subjects are asked both to generalize their experience into broad
categories and to remove situational or contextual aspects. He further noted that
inherent in the process of rating oneself is a comparison of oneself to other people;
thus, he believed the utility of the five factors to be limited by the breadth of

interpersonal experience of the subject. McAdams concluded that the Big Five

12

represent an important and positive development in personality theory, but is not the
definitive, integrative theory of personality as is often claimed.

Proponents of the five-factor model have responded to such criticisms.
McCrae and Costa (1984) noted that while contextual, situational aspects of life may
change the way in which traits are expressed, the traits themselves remain unchanged.
Peabody and Goldberg (1989), in response to criticisms that the strength of support
for the five factors varied across studies, identified three determinants of differences
between studies: (a) the selection of variables for inclusion in analyses, (b) whether
analyses were done on conceptual relationships between variables or on actual
measured differences, and (c) the degree of restriction imposed on the sample to
create a homogeneous sample. Further, they examined differences in the factor
structure of recent studies of the Big Five and found that clearly identifiable variations
of the Big Five personality factors were always found, particularly for factors H
(Agreeableness), III (Conscientiousness), and I (Extraversion). After a review,
Wiggins and Pincus (1992) found that the five-factor model has been generalized
across data sets sufficiently different from those utilized by Cattell, so that the theory
may be considered independently replicated. Pincus (1992) argued that the
comprehensive nature of the five-factor personality structure has been demonstrated
by the consistent results found across four major theoretical perspectives utilized: that
the factors represent an enduring characteristic disposition, a characteristic
interactional style, social competency, and lexical patterns. McCrae and John (1992)
have argued that the empirical validation of the model across methods and

instruments, the replication of the model across times and cultures, and the long

13

history of consistent findings of the model supplies more than adequate evidence that
the five-factor model has a viable place within personality psychology.

Thus, although not universally accepted, the five-factor model holds a central
place in current personality research. Much of the excitement over the robustness of
this model has emerged due to its ability to locate divergent constructs and empirical
findings in personality within the conceptual space of the five-factor model, and the
subsequent ability to create a meaningful and powerful assessment tool that is directly
linked with theory. Developing an assessment tool, however, is dependent upon the
interpretation of the five factors.

Intemretation of the five factors. While acceptance of a five-factor
model is widespread, a controversial aspect of the model, and thus of efforts to derive
assessment tools, lies in the interpretation (or meaning) of the five factors. Goldberg
(1993) suggests that there are actually two five-factor models: one based on the
original lexical hypotheses represented by N orrnan (1963), Goldberg (1992), and
Digman (1989), and one based on the empirical work of Costa and McCrae (1987).

McCrae and Costa (1986) used adjectives such as sociable, ftm-loving, and
affectionate to describe those individuals whose personalities would be characterized
as being high in Factor I, which they labelled Extraversion. They note that in
addition to sociability, individuals high in Extraversion are assertive, active, and
talkative, enjoy excitement and stimulation, and are optimistic (Costa &

McCrae, 1992). Individuals low in Extraversion are described as reserved,
independent, and free of social anxiety; they may not necessarily be pessimistic or

unhappy (Costa & McCrae, 1980; Costa & McCrae, 1992). While Costa and

14

McCrae (1986) use Extraversion as a label for Factor I, others have previously
labeled the factor "surgency" (Norman, 1963), " social activity" (Guilford, 1975) and
"interpersonal involvement" (Lorr, 1986, cited in Digman, 1990).

McCrae and Costa (1986) used adjectives such as worrying, insecure, and self-
pitying to describe those individuals whose personalities would be characterized as
being high in Factor IV, which they labelled Neuroticism. They noted that in
addition to worry and insecurity, individuals high in Neuroticism are likely to
experience any of a number of negative affects including as anger, guilt, fear, and
sadness (Costa & McCrae, 1992). Individuals low in Neuroticism are described as
emotionally stable, calm, and able to cope effectively with stressful situations (Costa
& McCrae, 1992). While Costa and McCrae (1986) use Neuroticism as a label for
Factor IV, others have previously labeled the factor " anxiety " (Cattell, 1957), and
"affect" (Peabody and Goldberg, 1989).

The interpretations of Factor I as representing a dimension of introversion-
extraversion and of Factor IV as a dimension of general psychopathology are widely
accepted (Digman, 1990; Bergeman, Chipuer, Plomin et. a1, 1993); earlier factor
analytic studies (e.g. Eysenck, 1960) had also captured these dimensions. The
remaining three factors have varying degrees of consensus regarding the label the
factor should receive. Factor II, which Costa and McCrae (1987) labelled
Agreeableness, appears to capture what Digman (1990) called "the more humane
aspects of humanity" (p. 442) and what previous studies have labelled as " friendly
compliance" (Digman, 1990) and "conformity" (Fiske, 1949). McCrae and Costa

(1986) used adjectives such as soft-hearted, trusting, and helpful to describe those

15
individuals whose personalities would be characterized as being high in

Agreeableness. Individuals low in agreeableness may be egocentric, skeptical, and
competitive, although Costa & McCrae do not view this as inherently problematic
since such traits are necessary for analytic and scientific thought (1992).

Conscientiousness, the label Costa and McCrae (1987) gave to Factor III, has
also been interpreted as "work ethic" (Peabody & Goldberg, 1989), "impulsive
sensation seeking" (Zuckerman, Kuhlman, & Carnac, 1988), and because of its
correlation with education, "the will to achieve" (Digman, 1988). McCrae and Costa
(1986) used adjectives such as well-organized, purposeful, careful, and self-
disciplined to describe those individuals whose personalities would be characterized as
being high in Conscientiousness. Individuals low in Conscientiousness, according to
Costa and McCrae (1992) are less exact in applying moral principles, and expend less
effort toward accomplishing their goals.

Costa and McCrae (1986; 1987) interpreted Factor V as Openness, and used
adjectives such as imaginative, independent, and preferring variety to describe those
individuals whose personalities would be characterized as being high in this factor.
Individuals low in Openness are described as conservative, conventional, emotionally
muted, and having a narrow scope of interests (Costa & McCrae, 1992). While
Costa and McCrae (1986) use Openness a label for Factor V, others have identified
this factor as "intellect" (Cattell, 1947; Digman, 1988), and descriptively addressed it
as capturing a personality dimension which allows for a flexibility of interests and

ideas (Digman, 1990).

16

Attempts to support the five-factor model utilizing exploratory orthogonal
rotations of existing measures of personality have met with success; the five-factor
model hypothesized by Costa & McCrae (1986; 1987) has been able to account for
the underlying factors of personality measures based upon Jungian functions,
Murray’s Needs, Traits on the Interpersonal Circumplex, and DSM-III-R personality
disorders (Costa & McCrae, 1992). Further, well-designed adoption studies utilizing
identical twins raised in either the same or separate households have been able to
isolate the effects of genetics and the environment on NEO-PI factors (Pedersen,
Plomin, McClearn & Friberg, 1988; Bergeman, Chipuer, Plomin, et. a1, 1993); these
studies estimated that 31% of the variance in neuroticism, 41% of the variance in
extraversion, 40% of the variance in openness, 12% of the variance in agreeableness,
and 29% of the variance in conscientiousness is attributable to genetic predispositions.

Thus, current thinking within trait research suggests that five factors can be
utilized to describe personality, the generalized tendencies, predispositions, and
consistencies from which human behavior is derived, including the utilization of
defense mechanisms, a mode of intrapersonal functioning, and the perception of social

support, a mode of interpersonal functioning.

Defense Mechanisms As An Intragrsonal Process Influenced By Personality

Sigmund Freud first presented the concept of defense mechanisms. In his
theory, instinctual drives arising from the id may be in conflict with the internalized
value system of the superego and result in the experience of anxiety. He used the

term "defense" to refer to the method by which the expression of a disturbing or

l7

anxiety-provoking instinctual drive was altered, at times to such a degree that the
original drive is completely outside of awareness and is expressed only in disguised,
symbolic ways. During this stage of the development of his theories, no specific
mechanisms of defense were proposed (Cramer, 1991). With his shift from
topographical to structural theories of the mind, Freud (1923) began to describe
defense mechanisms as all processes which protect the ego from instinctual demands
arising from the id. These processes are theorized to be motivated by the avoidance
of anxiety, guilt and loss (Cramer, 1991).

Freud (1936) initially suggested all defense mechanisms are pathological,
although he later conceded that they may have played a necessary developmental role.
Cramer (1991) has also argued that defense mechanisms may emerge as part of the
normal course of development cycle. As such, defense mechanisms carry no intrinsic
value but are defined as "normal" or "pathological" depending on their utility:
defenses which aid in adaptation may be considered healthy, and defenses which
detract from adaptation may be considered pathological. Although debate continues,
she reviewed evidence that suggests different defenses may first be displayed at
different times through childhood, suggesting that defenses are reflecting some
underlying process (e.g. , ego development, cognitive development); the display of
defenses in adulthood would then reflect the degree of ego development achieved.

Vaillant (1986) suggests that defense mechanisms can be categorized in one of
four categories, each reflective of a level of ego development: psychotic defenses
(including delusional projection, denial, and distortion), which are common in

children under age three but pathological at later ages; immature defenses (including

l8

projection, passive-aggressive behavior, and acting out), which are common in
children from age three to age 15 but pathological at later ages; neurotic defenses
(including repression, displacement, and intellectualization), which are pathological in
individuals from age three to age 90; and mature defenses (including altruism, humor
and sublimation), which are normal (or anticipated) in individuals from age three to
age 90. These developmental views of defenses reflecting underlying processes are
similar to the theories of Shapiro (1965; 1981; 1989), who argued that an
overarching, organizing, enduring personal style is formed in early the course of
normal development and serves as a matrix from which all subsequent characteristic
traits develop. He wrote that the formation of these personal styles preceded the
crystallization of other modes of ftmctioning, and that it is these styles which allow
consistency to be observed in individual behavior across domains of human
functioning (both adaptive and pathological). Shapiro believed these underlying
processes to be primarily cognitive ones, that is, patterns of attention and processing.
However, in his argument that these neurotic styles both underlie and cause the
clinical manifestations of some disorders, Shapiro noted "the neurotic problem is not
i_n the patient, it is the patient" (1989; p. xi). Thus, Shapiro believed that underlying
dimensions of cognitive style, as an aspect of personality, directly affect the
utilization of defense mechanisms.

Shapiro (1965) identified four "neurotic styles, " each of which represent a
relatively permanent (although not inflexible) patterns that generalize across time and
place: the obsessive-compulsive style, the paranoid style, the hysterical style, and the

impulsive style. He described the obsessive-compulsive individual as

 

l9

characteristically rigid in focus, i.e. an absence of attentional fluidity. In this Style,
individuals exhibit the ability to able to concentrate intensely, though they are unable
to shift focus smoothly between thoughts, and they are able to direct behavior with
great deliberation, though they are unable to act spontaneously. Similarly, distortions
in attention and autonomy can be found in a paranoid individual. However, unlike
the inability to shift focus seen in the obsessive—compulsive person, Shapiro argued
that the paranoid person exhibits a rigid attentional style of scanning and vigilance;
the paranoid individual is unable to concentrate intensely on any one stimuli.
However, like obsessive-compulsive individuals, the paranoid individual has a
distorted sense of autonomy. This deliberateness is not motivated by a desire for
achieving goals, but rather is an attempt to minimize expressive behaviors that might
reveal vulnerability.

The hysterical individual, as hypothesized by Shapiro, is characterized by
impressionistic thought, an incapacity for prolonged concentration, and affective
excitability. Because of a decreased ability to attend to cognitive, ideational stimuli,
the hysterical individual is dominated by affect that does not seem to possess any
depth. Whereas hysterics possess little cognitive control over affect, impulsive
individuals reveal a personality style marked by minimal cognitive control over
behavior, a disavowal of responsibility, minimal tolerance for frustration, and an
attentional style which is devoid of concentration or deliberation.

Thus, the characteristic use of certain defense mechanisms may be reflections
of underlying personality structure (Shapiro, 1965; 1981; 1989). Bettelheim (1982),

in suggesting that systematic mistrarislations of Freud’s work may have been made to

20
keep psychoanalytic theory in the hands of the medical community, noted that the

word "defense" does not adequately convey the lexical and emotional meaning of
Freud’s German term, "Abwehr. " He wrote that "to defend" connotates an active
effort to fight off, while a more accurate translation of Freud’s term would be "to
parry, " with connotations of a deflection or "turning aside through clever means."

He lamented that the term "defense mechanism" suggests that " inner processes, such
as reaction formations or denials, are something alien - something outside oneself" (p.
92). Similarly, Guntrip (1963) observed that defenses are, in part, a reflection of the
ego itself. Thus, the body of literature on defense mechanisms, while in whole larger
than the scope of this study, does suggest that personality variables are associated
with the utilization of defense mechanisms.

Empirical examinations of defense mechanisms have found support for such an
association. For example, Vickers et al. (1981) found that facets of Type A
personality are positively correlated with regression, intellectualization, and
projection. Personality changes in therapy have been related to movement from
immature to mature defenses (V aillant, 1966). The experience of anxiety has been
negatively associated with repression (Gleser & Ihilevich, 1969) and denial (Sarason
et. a1, 1972), and the striving for academic success has been positively associated with
intellectualization (Galinsky, 1971). The use of defense mechanisms seem better
predicted by measures of personality than of mood (Tauschke, Helrnes, & Merskey,
1991).

Although personality and the characteristic use of defense mechanisms have

been associated, it Should not be assumed that all individuals with a certain

 

21

personality structure will rely on the same defense mechanism under all
circumstances. Lazarus and Folkman (1984) have found that people use a wide
variety of coping strategies, and that they vary under different situations. Thus, there
is an association between personality traits and afi defense mechanisms, although the

strength of these associations will vary from person to person.

Social Support As An Intermrsonal Process Influenced By Personalig

Recent work also indicates that personality variables are associated with the
perception of social support. PsycLit indicates that there were 5159 articles published
in psychological journals between January, 1973 and June, 1993 that dealt, at least in
part, with the construct of social support. The role of social support has been
identified in numerous arenas, such as physical illness (for review, see Broadhead et
al., 1983), psychological disorders, including depression (Kessler & McLeod, 1985;
Lin & Ensel, 1984; Leavy, 1983; Eaton, 1978), the adjustment to teenage pregnancy
(Barrera, 1981), the adjustment to graduate school (Goplerud, 1980) and the
adjustment to marital separation (Kincaid & Caldwell, 1991; 1993).

Such broad applications of social support are indicative of the intuitive appeal
and comprehensive implications contained within the construct; however, since no
single theory drives the research, it is difficult to present a unitary definition of social
support. Most empirical studies attempt to further delineate the process of how social
support mediates the stress-psychopathology relationship; these typically utilize either
main (direct) effect hypotheses or interactional (buffering) effect hypotheses (for

reviews, see Cobb, 1976; Leavy, 1983; Cohen & Wills, 1984; Cohen &

22

Syme, 1985). However, as noted above, there is variation in the breadth of domains
in which social support is studied and consequently in the operationalization of the
term "social support. " Perhaps because of this, previous reviewers have criticized the
construct as having "a diversity of definitions, methodologies, and theories"

(Leavy, 1983), as "operationalized in a somewhat bewildering assortment of ways"
(Wilcox, 1981), and as "so vague or so broad that the concept is in danger of losing
its distinctiveness" (Barrera, 1986).

Despite differences in application and terminology, most studies involving
social support can be understood as involving one or both of two general social
support constructs outlined by House and Kahn (1985). House and Kahn categorized
operational definitions of social support as being either " structural " or "functional; "
research using structural definitions of social support typically focus on the number of
supporters in the social support network, the relationship of the supporter to the
individual, the degree of interrelationships among supporters in the network (density),
and the frequency of contact between the subject and members of the social support
network, while research using functional definitions of social support typically focus
on the perceived availability of emotional support, practical assistance, and tangible
support within the social support network.

Research utilizing the construct of social support has been criticized for
neglecting to examine the personal context within which social support operates.
Monroe and Steiner (1986) argued that studying the mediating effect of social support
on any dependent variable might best be conducted as an examination of the

incremental utility of social support within personality typologies. Vinokur et. al

23

(1987) noted that measurement of the perception of social support may reflect the
personal dispositions of the subject as much as the actual provision of support, and
found moderate evidence for this. Sarason et. a1 (1987) suggested that social support
could best be conceptualized " as a developmental personality characteristic, rather
than as simply an environmental provision" (p. 831). Others have argued that the
perception of social support is itself simply an expression of personality, because the
social support network itself (through the availability of people to include in the
support network and the availability of those within the network from whom to elicit
support) may be influenced by personality factors (Jung, 1988; Elliott &

Gramling, 1990).

There is some empirical evidence for the influence of personality upon the
social support network. Sarason and Sarason (1982) found conventionality, rigidity
and hostility to be correlated with the perception of inadequate social support.
Henderson (1984) found that people carefully create their social support networks,
and thus can not be viewed as passive recipients of the support it may offer. The
perception of emotional support has been linked to self-esteem, religiosity, and
authority (Dunkel-Schetter, Folkman, & Lazarus, 1987). Jung (1988) found that
interpersonal skills affected the process of developing and maintaining a social support
network. Elliott & Granrling (1990) found the perception of satisfactory social
support to be related to the effect communication of thoughts and feelings to others.
A path analysis revealed that some personality characteristics (help-seeking,
help-giving, affiliation) were associated with larger social support networks and the

perception of greater available social support (Connell & D’Augelli, 1990). Lakey

24
and Cassady (1990) found the perception of support to be inversely related to anxiety

and dysphoria, and argued that the perception of social support operates as a

personality variable.

Major Promsitions

Five personality variables have been consistently identified which are robust

 

across time, context and setting. Social support, an interpersonal phenomenon, is
related to intrapersonal processes (Jung, 1988; Elliott & Gramling, 1990). Similarly,
the utilization of defense mechanisms, an intrapersonal phenomenon, is related to
personality variables (Shapiro, 1965; Shapiro, 1981; Shapiro, 1989; Vaillant, 1966;
Vickers et a1. , 1981). Based upon the literature reviewed above, it is hypothesized
that personality is related to both intrapersonal and interpersonal functioning in a
hierarchical manner: personality traits first affect intrapersonal processes, which in
turn affect interpersonal processes.

One example of intrapersonal functioning is the characteristic utilization of
defense mechanisms. As one example, individuals who are open, receptive and
flexible to new ideas would be expected to be more accepting of the thoughts and
feelings of themselves as well as others, and therefore less likely to utilize defenses
based upon denial and repression. Similarly, individuals who are anxious,
self—absorbed and worried would be more likely to utilize defenses based upon self-
blame.

One example of interpersonal functioning is the perception of social support.

The perception of being understood and accepted by another person underlies the

25

construct of social support. The extent to which defense mechanisms are utilized
would be expected to either facilitate or inhibit relationships with other people. As an
example, individuals who frequently utilize projection as a defense should perceive
their environment as less emotionally supportive than individuals who rarely utilize
projection. Similarly, individuals who frequently utilize intellectualization as a
defense should have a larger social support network with which to seek stimulation
than should individuals who rarely utilized intellectualization.

Thus, the five-factor model of personality, the utilization of defense
mechanisms, and the perception of social support can be integrated into a causal
chain, with personality preceding and affecting the characteristic use of defense
mechanisms, and defense mechanisms preceding and affecting the perception of social
support. This study will propose and examine a hierarchical structural model of these
constructs, which will be operationalized by the N EO-PI-R, the Defense Mechanisms

Inventory, and the N orbeck Social Support Questionnaire.

Qpprationalizing The Construpts

The Neuroticism-Extraversion-Openness-Personality Inventory, Revised

 

(NEO-PI-R), initially developed and recently revised by Costa & McCrae (1985;
1992), provides a means of Operationalizing the personality variables of the five-factor
theory. The Defense Mechanisms Inventory (DMI; Ihilevich & Gleser, 1991)
provides a means of Operationalizing characteristic defensive style. The N orbeck

Social Support Questionnaire (NSSQ; Norbeck, Lindsey, & Carrieri, 1981; 1983)

26

provides a means of Operationalizing structural and functional aspects of social
support. Each of these measures is discussed further below.

The NEO-PI-R. The NEO-PI-R (Costa & McCrae, 1985) is a paper and
pencil measure developed specifically to assess the five-factor model of personality,
and has been called "one of the best state-of-the—art tools available for the general and
systematic assessment of normal personality" (p. 536; Leong & Dollinger, 1991).
The N EO-PI was revised to provide more subscale facets and to improved the internal
consistency of some scales (Costa & McCrae, 1992). Correlations between the old
and new version of the instrument have ranged from .93 to .95 (Costa & McCrae,
1992).

mNeuroticism Scale. Costa & McCrae’S ( 1992) review reported that
Scale N of the NEO-PI-R assesses the likelihood of feelings of negative affects such
as fear, sadness, guilt, and/or disgust. As such, Scale N is a general measure of
psychological distress and poor coping skills. High scores on Scale N are associated
with poor impulse control, irrational ideas, neurotic behavior, and disruptive
emotions. Low scores on Scale N are associated with calm, even-tempered
individuals who are able to cope with stressful situations without major psychological
disruption.

The Extraversion Scale. In their review, Costa & McCrae (1992)
reported that Scale E of the NEO-PI-R assesses sociability. High scores on Scale E
are associated with liking people, assertiveness, optimism, and excitability. Low

scores on Scale E are associated with reserved, independent individuals who, although

27

they prefer to be alone, are not unhappy and do not necessarily have any social
anxieties.

The Qppnness Scale. Costa & McCrae’s (1992) review reported that
Scale 0 of the NEO—PI-R assesses a broad domain including intellect, imagination,
aesthetic sensitivity, curiosity, and judgement. High scores on Scale 0 are associated
with self-examination of one’s inner experiences, unconventional values, a willingness
to challenge authority, and heightened emotional awareness. High scores on Scale O
are also linked to education, IQ, and creativity, although Scale 0 and intelligence are
not identical. Low scores on Scale 0 are associated with conventional behavior and
values, a preference for the familiar, a narrowed scope of interests, and muted
emotional responses.

The Aggeeableness Scale. In their review, Costa & McCrae (1992)
reported that Scale A of the NEO-PI-R assesses interpersonal tendencies. High scores
on Scale A are associated with altruism, the ability to empathize and sympathize, and
interpersonal cooperation. Low scores on Scale A are associated with individuals
who are antagonistic, egocentric, and competitive.

The Conscientiousness Scale. Costa & McCrae’s (1992) review
reported that Scale C of the NEO-PI-R assesses self—control. High scores on Scale C
are associated with good impulse control, planning, organization, and determination.
Low scores on Scale C are associated with a flexible moral code, decreased goal-
directed behavior, and increased hedonism and interest in sex.

Use of the N EO-PI in prior research. Because of the recency of

revisions conducted to form the N EO-PI-R, there is little psychometric data yet

 

28

available. However, given the high correlations between the NEO-PI and the
NEO-PI-R, and given that the revisions primarily involved subfacets of the scales,
there is little reason to believe that the N EO-PI-R will have significantly different
psychometric properties.

A considerable amount of information is available on the NEO-PI. It has been
used extensively to not only to assess the factors proposed in the five-factor model,
but to successful demonstrate the prevalence of the five factors throughout other
personality inventories (Digman, 1990). Factor analytic studies involving the
NEO-PI, using various techniques of rotation, have successfully identified a
five-factor model in the MMPI (Costa et. al, 1986), the California Psychological
Inventory (Mccrae, Costa, & Piedmont, 1993), the Eysenck Personality Inventory
(McCrae & Costa, 1985), the Interpersonal Style Inventory (Lorr, Youniss, & Kluth,
1992), the Jackson Personality Research Form (Costa & McCrae, 1988a), the
Myers-Briggs Type Indicator (McCrae & Costa, 1989b), the California Q-Set
(McCrae, Costa & Busch, 1986), and the Narcissistic Personality Inventory (Bradlee
& Emmons, 1992).

Construct validity has been demonstrated for all scales of the NEO—PI-R.
Neuroticism, as measured by the NEO-PI, has been positively associated with general
measures of neuroticism and psychological distress (Costa & McCrae, 1992) as well
as specific indicators of sexual dysfunctions (Pagan, Wise, Schmidt, et. a1, 1991;
Wise, Pagan, Schmidt, et. al, 1991). Extraversion and Agreeableness, as measured
by the NEO-PI, align with Wiggins’ interpersonal circle (McCrae & Costa, 1989a).

Low scores on NEO-PI Agreeableness, have been associated with indicators of sexual

29
dysfunctions (Fagan, Wise, Schmidt, et. a1, 1991; Wise, Fagan, Schmidt, et. a1,

1991). Openness, as measured by the NEO—PI, is positively correlated with cognitive
aspects of creativity (McCrae, 1987). Conscientiousness, as measured by the
NEO-PI, has been positively associated with both academic achievements (Dollinger
& Orf, 1991) and work achievements (Dye, 1991; cited in Costa & McCrae, 1992).

The five-factor model is associated with other psychological concepts. Watson
& Clark (1992) found that there were numerous positive and negative relations
between the factors measured by the N EO-PI-R and the expression of various affects.
Personality factors have been identified as an agent in some DSM-III-R Axis I
disorders (W idiger & Trull, 1992), and Axis II disorders within DSM-III-R have been
shown to correlate positively with Scales N, E, and A (Trull, 1992). Individuals
suffering from chronic physical pain have been found to vary in their experience of
coping with the pain depending on level of Scale N (Wade, Dougherty, Hart, &
Cook, 1992). The NEO—PI has been used to identify five—factor models in adjustment
studies of both children and adolescents (Graziano & Ward, 1992; Digman, 1989),
and recent research has begun on the relationship between the five-factors of
personality and physical health (Smith & Williams, 1992).

Given these findings, it has been suggested that the NEO-PI-R could become
an integral part of clinical personality assessments, and that, when used in
combination with projective techniques, could lead to more complete understanding of
clients (Ben-Porath & Waller, 1992; Costa, 1991; McCrae & Costa, 1986).

The Defense Mechanisms Inventog. The Defense Mechanisms Inventory

(Ihilevich & Gleser, 1986; Ihilevich & Gleser, 1991) is a pencil and paper measure

 

30

developed to categorize defensive styles. The five DMI scales (and corresponding
defensive style) are Principalization (PRN), Projection (PRO), Reversal (REV),
Turning Against Others (TAO), and Turning Against Self (TAS).

The Principalization Scale. Ihilevich and Gleser (1991) reported that
elevations on the Principalization (PRN) scale reflect individuals who " obscure,
reinterpret or generalize the meaning of a perceived threat, thereby casting it in
personally or socially more desirable terms" (p.44). They note that this defense relies
upon internalizing affect through splitting or repression. For these individuals, such a
posture allows thoughts, behavior and emotions to exist without integration, so that
anxiety which that might be aroused by a threat can be minimized, and releasing the
need for personal responsibility in behavior. They suggest that such people are often
accomplished in their careers, though lacking in meaningful interpersonal
relationships, relying on cognitive attempts at "understanding" and problem solving
and rarely showing outward manifestations of anxiety or depression. Their review of
previous research on those scoring high on the PRN scale of the DMI suggests such
individuals have good ego-strength, experience little anxiety, express high self—esteem
and are cognitively efficient. Their emotions tend to be stable, with high self-control,
and they tend to be oriented toward internal sources of satisfaction.

Ihilevich and Gleser (1991) found that elevations on the PRN scale occurred in
4.5 % of males and 5.9% of females in an outpatient psychiatric sample. Elevations
on the PRN scale were reported to be independent of age, gender, education or SES.

The Projection Scale. Ihilevich and Gleser (1991) reported that the

elevations on the Projection (PRO) scale reflect individuals who "overcome their fears

 

31
by constantly belittling and judging others. Their vigilance keeps them on guard

against anticipated attacks... The pervasive suspiciousness of PRO patients provides
them with the rationalizations they need to allay their anxiety about their sense of
powerlessness and fragile autonomy " (p.38). They suggest that such people maintain
chronic unresolved anger and jealousy, and frequently come into interpersonal
conflicts with others. Such individuals maintain a conviction of their own infallibility,
and become hostile when confronted with evidence to the contrary. Their review of
previous research on those scoring high on the PRO scale of the DMI suggests such
individuals are intolerant, tense, and are oriented toward external sources of
satisfaction.

Ihilevich and Gleser (1991) found that elevations on the PRO scale occurred in
4.5 % of males and 5.1% of females in an outpatient psychiatric sample. Elevations
on the PRO scale were reported to be positively correlated with gender (i.e. , males
were more likely to use a PRO style than females). Elevations in T-scores on the
PRO scale were reported to be independent of age, education or SES .

The Reversal Scale. Ihilevich and Gleser (1991) reported that the

Reversal (REV) scale measures those individuals who " find it too emotionally
disn'essing to recognize their inner conflicts or confront external threats to their
interests. Their sense of basic security is so fragile and their fear of annihilation so
great, that it appears to require that they always maintain an image of physical and
psychological well-being" (p.73). They suggest that such people are outwardly
COOperative, pleasant and accommodating, though they are often oblivious to obvious

problems in their lives, require excessive reassurance, and may react with confusion

32

to stressful events. Their review of previous research on those scoring high on the
REV scale of the DMI suggests such individuals are self-controlled, trusting, and
have few resources available to assist in coping with prolonged stress. Their
emotions tend to be stable, with anxiety and depression rarely expressed. They tend
to be oriented toward internal sources of satisfaction.

Ihilevich and Gleser (1991) found that elevations on the REV scale occurred in
5.9% of males and 5.9% of females in an outpatient psychiatric sample. Elevations
on the REV scale were reported to be positively correlated with age. Elevations in
T-scores on the REV scale were reported to be independent of age, gender, education
or SES.

The Turning Against Others Scale. Ihilevich and Gleser (1991)
reported that the elevations on the Turning Against Others (TAO) scale reflect
individuals who use "aggression to negate their feelings of inadequacy and mask their
deep sense of inferiority. By dominating and aggressing, they create an illusion of
strength that conceals their profound fear of disapproval and rejection" (p. 26). They
suggest that such people are inconsiderate, competitive, more attentive to their own
needs than those of others, and lacking in truly meaningful or satisfying interpersonal
relationships. Such people deal with conflict by externalizing their frustration. Their
review of previous research on those scoring high on the TAO scale of the DMI
suggests such individuals are impulsive, aggressive, have poor self—control, and are
undersocialized. Their emotions tend to be labile, with externalized anger and

depression likely. They tend to be oriented toward external sources of satisfaction.

33

Ihilevich and Gleser (1991) found that elevations on the TAO scale occurred in
7.2% of males and 5.1% of females in an outpatient psychiatric sample. Elevations
on the TAO scale were reported to be negatively correlated with age, and positively
correlated with gender (i.e., males were more likely to use a TAO style than
females). Elevations on the TAO scale were reported to be independent of education
or SES.

The Turning Against Self Scale. Ihilevich and Gleser (1991) report
that the Turning Against Self (TAS) scale measures those individuals who "to
compensate for pervasive feelings of devaluation and worthlessness, ...ad0pt
perfectionistic standards... [and] ...fearing the devaluation of others, they establish
some sense of control through adopting positions such as ’I’ll judge myself before
others judge me’ " (p.61). They suggest that such people are pessimistic, unhappy,
and demoralized. They report excessively high expectations for themselves, and are
punitive when (inevitably) they do not meet these expectations. Their review of
previous research on those scoring high on the TAS scale of the DMI suggests such
individuals are introverted, filled with doubt, and guilt-ridden.

Ihilevich and Gleser (1991) found that elevations on the TAS scale occurred in
11.8% of males and 12.9% of females in an outpatient psychiatric sample, and was
the scale most likely to be elevated scale for both sexes. Elevations on the TAS scale
were reported to be positively correlated with gender (i.e. , females were more likely
to use a TAO style than males). There is evidence that elevations in T—scores on the

TAS scale may be inversely related to education and SES.

34

Use of the DMI in prior research. Although the Defense Mechanisms

Inventory has not yet been utilized with the NEO-PI-R, it has been used in a number
of other empirical studies. Within medical settings, the DMI has been used to show
associations between defensive style and increased menstrual discomfort (Greenberg &
Fisher, 1984). Peglar and Borgen (1984) found that varying DMI defensive profiles
were related to increased mortality during recovery from myocardial infarctions.
DMI profiles have been related to measures of physical symptoms, mood, pain, and
the use of pain medications in females recovering from major abdominal surgery
(Wilson, 1982), and to the use of contraceptives and unwanted pregnancies (Rader,
Bekker, Brown, & Richardt, 1978). Scholz (1973) found that suicide attempters were
significantly more likely to utilize the TAS defense than neuropsychiatric controls.

In psychological studies, Morelli and Andrews (1982) found that Ellis’ concept
of irrational beliefs correlated in predictable ways with defensive styles of the DMI.
The DMI has been used to determine consistent similarities within family dyads in the
utilization of defenses (Juni, 1992), and Klusman (1982) used the DMI when he
reported that defensive styles were able to contribute to the prediction of responses to
stress. Viney and Manton (1974) found individuals relying upon PRN as a defense
were significantly less anxious than those using other defenses. Schill, Rader, Evans,
& Segall ( 1976) found that male participants with a proclivity for guilt showed a high
preference for the inhibitory PRN and TAS defensive styles, while showing a low
preference for the impulsive TAO defensive style. The DMI has been used to show
that there are no significant differences between blind and sighted persons in the

characteristics selection and utilization of defense mechanisms (Minskoff & Curtis,

35
1984). In a review, Cramer (1988) concluded that, although there were some issues

of scale intercorrelations, the DMI was meaningfully related to personality variables.

The DMI was originally developed using an outpatient psychiatric sample.
However, Gordon (1979) used college students as participants while correlating DMI
and MMPI profiles, and produced findings which were similar to Gleser & Ihilevich’s
(1969) original study. This suggests that findings related to particular DMI profiles
may be generalized to the college population.

The Norbeglc Social Support Questionnaire. The Norbeck Social Support

 

Questionnaire (Norbeck, Lindsey, & Carrieri, 1981; 1983) is a self-report rating of
the support received from one’s social network. On this questionnaire, each
participant is asked to list up to 20 people who they believe currently provide some
form of support. They are then asked to describe their relationship (e.g. familial,
nonfamilial, professional, etc.) to each person, and to rate the amount of emotional
support, practical assistance, and advice/ informational support given to them by each
person. Subjects then indicate how long they have known the person, the frequency
of contact with each person, and which people in the network know each other. All
ratings are made on a 5-point Likert scale. Based on the responses to these questions,
a Total Functional Support score (based upon questions which ask about emotional
support, affiliation, and tangible aid) and a Total Structural Support score (based upon
questions which ask about the number of supporters, the frequency of contact, and the
duration of the relationship) is derived. For purposes of this study, the Total
Functional Support score and Total Structural Support score are multiplied to form a

Total Social Support score.

36

Content analyses of the N SSQ have found that the instrument is sensitive to the
major types of support empirically identified (Norbeck, Chaftez, Skodol-Wilson, &
Weiss, 1991). A recent examination of typical social support questionnaires (Sarason,
Shearin, Pierce, & Sarason, 1987) found that three social support measures
(representative of differing theoretical approaches to assessing social support) tend to
uniformly assess common constructs underlying the concept of social support, namely
that the individual is engaged in interpersonal relationships and feels accepted by
others. Although Sarason, et a1. (1987) did not directly examine the NSSQ, the
NSSQ is similar to two of the measures included, varying primarily in physical
format.

Use of the NSSQ in prior research. The NSSQ was initially used by its
developers to assess characteristics of the social support networks of nursing students
(Norbeck, Lindsey, & Carrieri, 1983; Norbeck, Lindsey, & Carrieri, 1981). More
recently, the NSSQ was used in a study which found that family members tend to
provide emotional support while friends provide more affirmation during prolonged
stress (Primomo, Yates, & Woods, 1990). Fisher et a1. (1989) included the NSSQ in
a battery designed to identify the effects of Large Group Awareness Training.
Quittner, Glueckauf, and Jackson (1990) used the N SSQ, in conjunction with other
measures, to contrast alternative models of how social support impacts upon parenting
stress, while Florian and Krulik (1991) used the NSSQ and found an inverse
relationship between loneliness and social support in the mothers of chronically and

mortally ill children. Kincaid and Caldwell (1991; In press) used the NSSQ to assess

——uu'--«- -. m. .J

37
both structural and functional aspects of the social support networks of maritally

separated individuals.
Spe_cific Hymtheses

This study proposes that a hierarchical relationship exists between personality
traits, the characteristic utilization of defense mechanisms (as a measure of
intrapersonal functioning) and the perception of social support (as a measure of
interpersonal functioning), with personality traits preceding and affecting defense
mechanisms, which in turn precedes and affects social support (see Figure 1). Each
of the specific hypotheses within the larger model are presented below, and are

individually identified for purpose of later discussion.

Hypotheses concerning the relationship of personality traits to the characteristic

use of defense mechanisms. It is hypothesized that each of the factors in the
five-factor model of personality will be associated with characteristic preferences of
defense mechanisms. An overview of the direction of these hypotheses is presented

in Table 1. Specifically, it is hypothesized that:

38
Figure 1
Hyppthesized General Structural Relationship Of Personality, Defense Mechanisms,

And Social Suppprt

 

PERSONALITY
TRAITS

V

DEFENSE
MECHANISMS

7

SOCIAL
SUPPORT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

39
Table l

The Hymthesized Direction Of Correlations Among Scales Of The
NEO-PI-R And The DMI

 

_S_ca;le N E O A C
PRN - + +
PRO -

REV - + - + +
TAO + - _ -
TAS + +

 

&: Abbreviations refer to the Neuroticism (N), Extraversion (E), Openness (O),
Agreeableness (A), and Conscientiousness (C) scales of the N EO—PI-R, and the
Principalization (PRN), Projection (PRO) Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales of the DMI.

(1a)

(1b)

(IC)

(1d)

40

Individuals who are chronically anxious and insecure would be
less likely to have good ego-strength, high self.esteem, and
stable emotions. Thus, it is hypothesized that the Neuroticism
scale of the NEO-PI-R will be negatively correlated with the
PRN (Principalization) scale of the DMI;

Individuals who are chronically anxious and insecure are, by
definition, experiencing difficult and troubling emotions and
would be likely to maintain self control and suppress their
experience of anxiety and depression. Thus, it is hypothesized
that the Neuroticism scale. of the N EO-PI-R will be negatively
correlated with the REV (Reversal) scale of the DMI;
Individuals who are chronically anxious and insecure would be
likely to use aggression to negate their feelings of inadequacy
and inferiority, concealing their interpersonal fears. Thus, it is
hypothesized that the Neuroticism scale of the NEO-PI-R will be
positively correlated with the TAO (Turning Against Others)
scale of the DMI;

Individuals who are chronically anxious and insecure would be
more likely to be unhappy, self-punitive, and guilt-ridden.
Thus, it is hypothesized that the Neuroticism scale of the
NEO-PI-R will be positively correlated with the TAS (Turning

Against Self) scale of the DMI;

(16)

(1f)

(1g)

(1h)

41

Individuals who are outgoing and enjoy the company of other
people would be more likely to suppress their emotions to
minimize difficulties in their relationships. Thus, it is
hypothesized that the Extraversion scale of the NEO-PI-R will
be positively correlated with the REV (Reversal) scale of the
DMI;

Individuals who are outgoing and enjoy other people would be
less likely to act in a manner that could jeopardize their
relationships with others. Thus, it is hypothesized that the
Extraversion scale of the NEO-PI-R will be negatively
correlated with the TAO (Turning Against Others) scale of the
DMI;

Individuals who are independent, broad-minded and flexible in
their thinking are more likely to have good ego-strength,
experience little anxiety, and be oriented toward internal sources
of satisfaction. Thus, it is hypothesized that the Openness scale
of the NEO-PI-R will be positively correlated with the PRN
(Principalization) scale of the DMI;

Individuals who are independent, broad-minded and flexible in
their thinking would be less likely to suppress their emotions.
Thus, it is hypothesized that the Openness scale of the
NEO-PI-R will be negatively correlated with the REV

(Reversal) scale of the DMI;

00

(11')

(1k)

(11)

(1m)

42

Individuals who are altruistic, sympathetic, and cooperative
would be less likely to be suspicious, hostile, and jealous of
others. Thus, it is hypothesized that the Agreeableness scale of
the NEO—PI-R will be negatively correlated with the PRO
(Projection) scale of the DMI;

Individuals who are altruistic, sympathetic, and cooperative
would be likely to minimize their own emotions and problems in
their relationships with others. Thus, it is hypothesized that the
Agreeableness scale of the NEO—PI-R will be positively
correlated with the REV (Reversal) scale of the DMI;
Individuals who are altruistic, sympathetic,and cooperative
would be unlikely to be dominating, aggressive, and judgmental
in their relationships with others. Thus, it is hypothesized that
the Agreeableness scale of the NEO-PI-R will be negatively
correlated with the TAO (Turning Against Others) of the DMI;
Individuals who are careful, self-disciplined and self-motivated
would be likely to good ego-strength, high self-esteem, and
stable emotions. Thus, it is hypothesized that the
Conscientiousness scale of the NEO-PI-R will be positively
correlated with the PRN (Principalization) scale of the DMI;
Individuals who are careful, self-disciplined and self-motivated
would be likely to maintain self control and suppress their

experience of anxiety and depression. Thus, it is hypothesized

 

43
that the Conscientiousness scale of the NEO-PI-R will be

positively correlated with the REV (Reversal) scale of the DMI;

(1n) Individuals who are careful, self-disciplined and self-motivated
would be likely to be impartial and non-judgmental in their
relationships with others. Thus, it is hypothesized that the
Conscientious scale of the NEO-PI-R will be negatively
correlated with TAO (Turning Against Others) scales of the
DMI; and,

(lo) Individuals who are careful, self-disciplined and self-motivated
would be likely to be self-critical, judgmental and
perfectionistic. Thus, it is hypothesized that the
Conscientiousness scale of the NEO-PI-R will be positively
correlated with the TAS (Turning Against Self) scale of the

DMI.

Hypptheses concerning the relationship of the characteristic; use of defenpe

 

mechanisms to social support. It is hypothesized that the characteristic preferences of
defense mechanisms will be associated with the ability to organize and utilize social
support. An overview of the direction of these hypotheses is presented in Table 2.
Specifically, it is hypothesized that:

 

Table 2

The Hymthesized Direction Of Correlations Among Scales Of The
DMI And The Total Social Support Score Derived From The NSSQ

 

_Sc_a1_g TSS
PRN +
PRO -
REV -
TAO -
TAS +

 

Note: Abbreviations refer to the Principalization (PRN), Projection (PRO) Reversal
(REV), Turning Against Others (TAO) and Turning Against Self (TAS) scales of the
DMI and the Total Social Support score (T SS) derived from the NSSQ.

(2a)

(213}

(2C)

(2d)

45

Individuals who minimize affect and rely upon cognitive
resources for coping would be likely to value and maintain
social relationships. Thus, it is hypothesized that the PRN
(Principalization) scale of the DMI will be positively correlated
with the Total Social Support rating derived from the NSSQ;
Individuals who are judgmental, suspicious, and reluctant to
express personal information to others would be unlikely to
value and maintain social relationships. Thus, it is hypothesized
that the PRO (Projection) scale of the DMI will be negatively
correlated with the Total Social Support rating given on the
NSSQ;

Individuals who are self-controlled, minimize affect, and
minimize their need for emotional support would be unlikely to
value and maintain social relationships. Thus, it is hypothesized
that the REV (Reversal) scale of the DMI will be negatively
correlated with the Total Social Support rating derived from the
NSSQ;

Individuals who are aggressive, competitive, and have
interpersonal fears would be unlikely to value and maintain
social relationships. Thus, it is hypothesized that the TAO
(Turning Against Others) scale of the DMI will be negatively
correlated with the Total Social Support rating derived from the

NSSQ; and,

 

46

(2e) Individuals who are self-critical, self-punitive, and guilt-ridden
would be unlikely to value and maintain social relationships.
Thus, it is hypothesized that the TAS (Turning Against Self)
scale of the DMI will be negatively correlated with the Total
Social Support rating derived from the NSSQ.
Integrating the hypotheses into a complex system. It is hypothesized that the
model presented through these hypotheses will be an acceptable best model of the
inter-relationships of these variables. This hypothesis will be tested utilizing path

analysis techniques, or structural equation modeling (Hunter, 1987).

 

METHOD

Procedure

Students were recruited through Introductory Psychology classes offered by the
Department of Psychology at Michigan State University; in return for participation,
they received credit toward the research participationcomponent of their class.
Efforts were made to recruit a minimum of 250 students; four different opportunities
for participation were offered. Separate sign-up sheets for male and female
participants were utilized in order to balance gender within both data collection
opportunities and the total sample. All measures were administered in a group
setting, with the maximum time required for completion of the measures
approximately two hours. Informed consent was obtained from each student prior to
administration of the psychological tests (see Appendix A). After completing a brief
demographic questionnaire, the NEO-PI-R was administered, followed by the DMI,
and finally the NSSQ. Each measure was scored as directed by the instrument’s

publisher.

Measures
Demographic Backggound. Participants were asked to complete a short,
nonstandardized questionnaire which asked about sex, age, educational level, marital

status, and race (see Appendix B).

47

 

48
The NEO-PI-R. The NEO-PI-R (Costa & McCrae, 1985) is a pencil and

paper test which assesses Neuroticism, Extraversion, Openness, Agreeableness, and
Conscientiousness. Subjects were asked to indicate their agreement with 240 items
regarding personality, using a 5—point Likert scale. The NEO-PI-R is normed for
both adults and college students.

Costa and McCrae (1992) reported coefficient alphas of .92 for the
Neuroticism Scale, .89 for the Extraversion Scale, .87 for the Openness Scale, .86 for
the Agreeableness Scale, and .90 for the Conscientiousness Scale. Similarly, Keyset
and Sweetland (1991) found coefficient alphas for the first three factors ranged from
.87 to .93, while for the last two factors coefficient alphas ranged from .76 to .86.
Costa and McCrae (1988b) reported that 6-year test-retest reliabilities of the first three
scales (Neuroticism, Extraversion and Openness) ranged from .82 to .83, while 3-year
test-retest reliabilities for the Agreeableness and Conscientiousness scales were .63
and .79, respectively.

The Defense Mechanisms Inventog. The Defense Mechanisms Inventory
(Gleser & Ihilevich, 1969) is a pencil and paper test which assesses five defensive
styles. Participants were asked to read 10 situations and select responses that would
be most and least like their own behavior, fantasies, droughts and feelings regarding
the situation. The DMI is normed for both adults and college students.

Gleser & Ihilevich (1969) reported one-week test-retest reliabilities for each of
the five scales ranged from r = .85 to r = .93; an independent reliability study
(Weissman, Ritter, & Gordon, 1971) approximated the findings of Gleser & Ihilevich

(1969), with reliabilities of the five scales ranging from r = .61 to r = .84. The

 

49

DMI has demonstrated construct validity through its significant correlations with a
number of personality measures, including the MMPI (Gleser & Ihilevich, 1969) and
Cattell’s 16PF (Gleser & Ihilevich, 1979). Although there has been some concern
about high correlations among some scales of the DMI (Gleser & Ihilevich, 1969;
Vickers & Hervig, 1981), particularly TAO and PRO, Cramer (1991) suggested that
the utilization of some defenses overlap in reality.

The Norbeck Social Support Questionnaire. Information on social support was
obtained by The N orbeck Social Support Questionnaire (Norbeck, Lindsey, &
Carrieri, 1981). Subjects were asked eight questions related to the provision of social
support and then rate members of their social support network in their ability to
provide that form of support using a five-point Likert scale. One-week test-retest
reliabilities have been reported as ranging from .87 to .92 (Norbeck, Lindsey, &
Carrieri, 1981); seven-month test—retest reliabilities have been reported as .73 for
Structural aspects of the network and .76 for functional aspects of the network
(Norbeck, Lindsey, & Carrieri, 1983). Byers and Mullis (1987) conducted reliability
and validity trials of the NSSQ on non-psychotic psychiatric inpatients, and found that
24-hour test-retest correlations ranged from .90 to .96 while coefficient alphas ranged
from .92 to .94; they further found that scores from their sample were significantly
lower than in a non-clinical sample reported by Norbeck et a1. (1983), which
supported the NSSQ’S construct validity. Bruhn and Phillips ( 1984), in reviewing l3
instruments available to measure social support, noted that the NSSQ has had
reliability and validity trials conducted (unlike some measures) and that the NSSQ was

acceptable for research applications.

RESULTS

Demographic Characteristics of the Sample

A total of 289 students were recruited and participated in this study; 287
completed the NEO-PI-R correctly and/or in a manner that could be scored, 276
completed the DMI correctly and/or in a manner that could be scored, and 289
completed the NSSQ correctly and/ or in a manner that could be scored. Participants
were nearly evenly split by gender (females = 51.6%; males = 48.4%), and had a
mean age of 19.2 (SE +/— 0.07). Nearly all participants were single (99.7%).
Participants were primarily freshmen (46.7%), though sophomores (24.6%), juniors
(20.1%), and seniors (8.7%) were also represented. The majority of participants
described their race as White (82.4%), followed by African-Americans (9.3%),
Asian-Americans (2.8%), Hispanics (0.7%), Native Americans (0.7 %), and other
races (1.7%); a small percentage of participants (2.4%) chose not to provide

information regarding race.

Resmnses To The NEO-PI-R

Of all participants, 287 completed the N EO-PI-R correctly and/ or in a manner
that could be scored; sample and normative data is presented in Tables 3 and 4. A
confirmatory factor analysis was conducted on each of the five factors of the

NEO-PI-R. Alpha reliabilities for the five factors of the NEO-PI-R ranged from .86

50

 

to .89 (see Table 5). These are consistent with those found previously (Costa &
McCrae, 1992; Keyser & Sweetland, 1991). Item loadings for each factor are
presented in Appendices C to G; the intercorrelations of the scales when corrected for
attenuation are presented in Table 6. Based on the these results, items were removed
from the scale; the criteria used in removing items were if the item demonstrated a
low loading onto its own factor, if the item loaded higher onto a factor other than its
own, or if the item violated the principal of parallellism (i.e., if all items measure the
same underlying construct, then they must also all relate to any other variable in the

same manner).

AS a result of the confirmatory factor analysis, 7 of 48 items were removed
from Scale N, 17 of 48 items were removed from Scale E, 14 of 48 items were
removed from Scale 0, 15 of 48 items were removed from Scale A, and 9 of 48
items were removed from Scale C. Nearly all items were removed for loading high
onto factors other than its own or by loading uniformly onto several factors; a small
number of items were removed for violating parallelism. The revised scales were
submitted to a confirmatory factor analysis and provide much stronger factors (see
Appendices H to L). The alpha reliabilities and the factor matrix for the revised
scales of the NEO—PI—R are presented in Table 7 and Table 8; raw scores for the
revised scales are presented in Appendix M. Comparing the revised factor

intercorrelations with those presented by Costa and McCrae (1992) found

51

52

 

 

 

Table 3
Sample And Normative Scores For Scales Of The NEO-PI-R: Mal_§
Sample Normative

Sege Mean SD Mean SD
Neuroticism 95 .2 17. 1 90.5 22. l
Extraversion 113.4 16.8 116.7 18.3
Openness 113.5 19.8 113.9 18.5
Agreeableness 103.5 17.7 107.4 16.2
Conscientiousness 106.5 18.6 1 13 .5 22.0

 

Note: Sample N = 139.

53
Table 4

Sample And Normative Scores For Scales Of The NEO-PI-R; Females

 

Sample Normative
_S_pale Mean SD Mean SD
Neuroticism 106.4 21.8 99. 8 20.9
Extraversion 123.1 20.0 123 .9 17.7
Openness 119.3 17.6 118.6 17.1
Agreeableness 110.3 17.3 117.2 15.7
Conscientiousness 105 .9 18. 6 1 15. 1 20. 6

 

Note: Sample N = 148.

54
Table 5

Alpha Reliabilities Of The Scales Of The NEO-PI-R

 

Scale N Items Alpha_
Neuroticism 287 48 .89
Extraversion 287 48 .88
Openness 287 48 .87
Agreeableness 287 48 .86
Conscientiousness 287 48 .88

 

55
Table 6

Intercprrelations of Scales Of The N EO-PI-R After Correcting For

Attenuation Due To Measurement Error

 

sc_a1e N E o A C
N ---
E -20
O 1 55
A -11 29 29
C -48 26 6 18

 

Note: Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

56

remarkable parallels; the mean difference in correlations was found to be .099, with
two intercorrelations involving Agreeableness (to Openness and to Extraversion)
differing by more than .25; removing these two correlations reduced the mean
difference in correlations between this sample with those presented by Costa and

McCrae to .067 points.

Remnses To Th_e Defense Mechanisms Inventog

Of all participants, 276 completed the DMI correctly and/or in a manner that
could be scored; sample and normative data is presented in Tables 9 and 10. A
confirmatory factor analysis was conducted on each of the five factors of the DMI.
Alpha reliabilities for subscales of the DMI ranged from .60 to .83 (see Table 11).
These reliabilities were more consistent with the findings of Weissman, Ritter, and
Gordon (1971) than with those of Gleser and Ihilevich (1969), suggesting that the
DMI possesses lower reliabilities than originally reported. Item loadings for each of
the factors are presented in Appendices N to R; the intercorrelations of the scales
when corrected for attenuation are presented in presented in Table 12. Based on the
these results, items were removed from the scale; the criteria used in removing items
were if the item demonstrated a low loading onto its own factor, if the item loaded

higher onto a factor other than its own, or if the item violated the principal

57
Table 7

Alpha Reliabilities Of The Revised Scales Of Th_e NEO-PI-R

 

S_eal_e N Items Alpha
Neuroticism 287 41 .88
Extraversion 287 31 .87
Openness 287 34 .87
Agreeableness 287 35 .84
Conscientiousness 287 39 . 88

 

58
Table 8

Intercorrelations Of The Scales Of The Revised N EO-PI—R After Correcting

For Attenuation Due To Measurement Error

 

Sage N E O A C
N ---
E -20 ---
O 02 46 ---
A -l7 14 15 «-
C -47 23 4 1 1 ---

 

Note: Decimals removed. Abbreviations refer to the N euroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

59

of parallellism (i.e. , if all items measure the same underlying construct, then they

must also all relate to any other variable in the same manner).

As a result of the confirmatory factor analysis, 21 of 40 items were removed
from the PRN scale, 20 of 40 items were removed from the PRO scale, 18 of 40
items were removed from the REV scale, 10 of 40 items were removed from the
TAO scale, and 11 of 40 items were removed from the PRN scale. Nearly all items
were removed for loading high onto factors other than its own or by loading
uniformly onto several factors; a small number of items were removed for violating
parallelism. The revised scales were again submitted to a factor analysis, which
yielded the revisions provided clearer factors but continued to Show high
intercorrelations of some scales (see Appendices S to W); further, alpha reliabilities
did not improve, although the reduction in items per may have contributed to this.
The factor matrix and the alpha reliabilities for the revised scales of the DMI are
presented in Table 13 and Table 14; raw scores for the revised scales are presented in

Appendix X.

60
Table 9

Sample And Normative Scores For Sgtles Of The DMI' Males

 

 

 

Sample Normative
gale Mean SD Mean SD
PRO 41.6 5.0 40.0 5.8
PRN 43.1 6.2 44.9 6.3
REV 36.7 6.7 37.1 7.4
TAO 43.6 9.6 41.6 9.2
TAS 35.1 6.3 36.4 6.8

 

Note: Sample N = 132.

61
Table 10

Sample And Normative Scores For Scales Of The DMI“ Females

 

 

Sample Normative
Scale Meaa SD Mean SD
Neuroticism 39.0 5.8 36.8 6.0
Extraversion 43 .9 6.6 46.6 6.2
Openness 35.3 7.4 38.0 8.0
Agreeableness 40.9 8.9 36.7 8.8
Conscientiousness 40.8 7.0 47.0 7.2

 

Note: Sample N = 139.

62
Table 11

Alpha Reliabilities Of Th5; Scales Of The DMI

 

Spale N Items Alpha_
Principalization 276 40 .71
Projection 276 40 .60
Reversal 276 40 .74
Turning Against Others 276 40 .83

Turning Against Self 276 40 .76

63
Table 12

Intercorrelations Of The Scales Of The DMI After Correcting For Attenuation

Due To Measurement Error

 

m: PRN PRO REV TAO TAS
PRN ---

PRO -65

REV 59 -81 ---

TAO -85 58 -84 —--

TAS -18 -57 -10 -49 ---

 

Note: Decimals removed. Abbreviations refer to the Principalization (PRN),
Projection (PRO), Reversal (REV), Turning Against Others (TAO) and Turning
Against Self (TAS) scales.

Table 13

Alpha Reliabilities Of Th; Revised Scales Of Th_e DMI

 

Me N Items Alpha
Principalization 276 19 .63
Projection 276 20 .59
Reversal 276 22 .72
Turning Against Others 276 30 .82

Turning Against Self 276 29 .78

 

Tl...\..h.vh in

65
Table 14

Intercorrelations Of The Scades Of The Revised DMI After Correcting

For Attenuation Due To Measurement Error

 

_S_c§e PRN PRO REV TAO TAS
PRN -—-

PRO -36 -_-

REV 27 -45 --_

TAO -75 29 -74 ---

TAS -23 -45 -19 -34 ---

 

Note: Decimals removed. Abbreviations refer to the Principalization (PRN),
Projection (PRO), Reversal (REV), Turning Against Others (TAO) and Turning
Against Self (TAS) scales.

66
Remnses To The Norbeck Social Support mestionnaire

All 289 participants completed the NSSQ correctly and in a manner that could
be scored. Separate ratings for Functional support (comprised of the Affective,
Affirmation, and Aid questions) and Structural support (comprised of the number of
supporters, the frequency of contact, and the duration of the relationship) were
obtained following published formulas (Norbeck, Lindsey, & Carrerri, 1981). As this
study is concerned with social support as a global concept with functional and
structural support equally essential, each of these scales were then multiplied together
in order to form a Total Social Support score. Given the nature of the derived Total
Social Support scale, however, no alpha reliabilities could be computed; raw scores

for the NSSQ are presented in Appendix Y.

Results Of Tests Of Hypotheses Concerning The Relationship Of Personalig Traits
To The Characteristic Use Of Defense Mechanisms

The first group of hypotheses concerned the interrelationship of personality
variables with the characteristic utilization of defense mechanisms. Following Levine
and Hunter (1983), each correlation was corrected for attenuation, and a 90%
confidence interval was computed. Further, computations were made of the
maximum likelihood estimate (or probability inference) that the correlation was in the
predicted direction.

Overall, 12 of 15 hypotheses concerning the relationship of personality traits to

the characteristic use of defense mechanisms were supported; the results of the tests

67
of Specific individual hypotheses are presented below. For an overview of the

correlations pertinant to these hypotheses, see Table 15.

Results Of The Test Of Hypothesis la A moderate negative correlation was
found between the revised Neuroticism scale of the NEO-PI-R and the revised
Principalization scale of the DMI (t; = -.32); this correlation rises to -.43 when
corrected for attenuation. There is a 90% confidence interval around the corrected
correlation of -.31 to -.55. Traditional statistical tests would reject the null hypothesis
(p < .001). As further evidence, there is a 100% inference probability that the true
correlation of these variables is negative, as predicted. Thus, this hypothesis was
supported.

Test of Hypothesis lb. A moderate negative correlation was found between
the revised Neuroticism scale of the NEO-PI-R and the revised Reversal scale of the
DMI (r; = -.29); this correlation rises to -.36 when corrected for attenuation. There
is a 90% confidence interval around the corrected correlation of -.25 to -.48.
Traditional statistical tests would reject the null hypothesis (p < .001). As further
evidence, there is a 100% inference probability that the true correlation of these
variables is negative, as predicted. Thus, this hypothesis was supported.

Test of Hypothesis 1c. A moderate positive correlation was found between the
revised Neuroticism scale of the NEO-PI-R and the revised Turning Against Others

scale of the DMI (r: = .18); this correlation rises to .21 when corrected for

 

68
Table 15

Correlations Among The NEO-PI-R And The DMI After Correcting For

Attenuation Due To MeasuremerL Error

 

SQLIC PRN PRO REV TAO TAS
Scale N ~43 ~01 ~36 21 44
Scale E 10 ~19 ~03 ~05 06
Scale 0 24 ~18 01 ~11 01
Scale A 33 ~44 39 ~47 29
Seal; C 20 ~21 23 ~09 ~13

 

_N_o;e;: Decimals removed. Abbreviations refer to the N euroticism (N),

Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the NEO-PR-R, and the Principalization (PRN), Projection (PRO), Reversal
(REV), Turning Against Others (TAO) and Turning Against Self (TAS) scales of the
DMI.

69

attenuation. There is a 90 % confidence interval around the corrected correlation of
.10 to .32; Traditional statistical tests would reject the null hypothesis (p < .01).
Use of maximum likelihood estimates indicates that there is a 100% inference
probability that the true correlation of these variables is positive, as predicted. Thus,
the hypothesis of the direction of this correlation was supported, although the
magnitude did not reach statistical significance.

Test of Hypothesis 1d. A moderate positive correlation was found between the
revised Neuroticism scale of the NEO—PI-R and the revised Turning Against Self scale
of the DMI (r = .37); this correlation rises to .44 when corrected for attenuation.
There is a 90% confidence interval around the corrected correlation of .34 to .55.
Traditional statistical tests would reject the null hypothesis (p < .001). As further
evidence, there is a 100% inference probability that the true correlation of these
variables is positive, as predicted. Thus, this hypothesis was supported.

Test of Hyppthesis 1e. A small negative correlation was found between the
revised Extraversion scale of the NEO-PI—R and the revised Reversal scale of the
DMI (r = ~.02); this correlation rises to ~03 when corrected for attenuation. There
is a 90% confidence interval around the corrected correlation of -. 15 to .10.
Traditional statistical tests would not reject the null hypothesis (p = NS). Use of
maximum likelihood estimates indicates that there is only a 37 % inference probability
that the true correlation of these variables is positive, as predicted. Thus, this

hypothesis was not supported.

70
Test of Hyppthesis If. A small negative correlation was found between the

revised Extraversion scale of the N EO-PI-R and the revised Turning Against Others
scale of the DMI (r = ~.04); this correlation rises to -.05 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
-.16 to .07. Traditional statistical tests would not reject the null hypothesis

(p = NS). Use of maximum likelihood estimates indicates that there is a 75%
inference probability that the true correlation of these variables is positive, as
predicted. Thus, the hypothesis of the direction of this correlation was supported,
although the magnitude did not reach statistical significance.

Test of Hypothesis 1g. A small positive correlation was found between the
revised Openness scale of the NEO-PI-R and the revised Principalization scale of the
DMI Q" = .18); this correlation rises to .24 when corrected for attenuation. There is
a 90% confidence interval around the corrected correlation of .11 to .37. Traditional
statistical tests would reject the null hypothesis (p < .01). As further evidence, there
is a 100% inference probability that the true correlation of these variables is positive,
as predicted. Thus, this hypothesis was supported.

Test of Hypothesis 1h. A small positive correlation was found between the
revised Openness scale of the NEO-PI—R and the revised Reversal scale of the DMI
(r = .01); this correlation remains at .01 when corrected for attenuation. There is a
90% confidence interval around the corrected correlation of -.11 to .14. Traditional
statistical tests would not reject the null hypothesis (p = NS). There is only a 43 %
inference probability that the true correlation of these variables is negative, as

predicted. Thus, this hypothesis was not supported.

71

Test of Hymthesis 1i. A moderate negative correlation was found between the
revised Agreeableness scale of the N EO-PI-R and the revised Projection scale of the
DMI (; = -.31); this correlation rises to ~.44 when corrected for attenuation. There
is a 90% confidence interval around the corrected correlation of -.31 to -.51.
Traditional statistical tests would reject the null hypothesis (p < .001). As further
evidence, there is a 100% inference probability that the true correlation of these
variables is negative, as predicted. Thus, this hypothesis was supported.

Test of Hypothesis lj. A moderate positive correlation was found between the
revised Agreeableness scale of the NEO-PI-R and the revised Reversal scale of the
DMI (r = .30); this correlation rises to .39 when corrected for attenuation. There is
a 90% confidence interval around the corrected correlation of .27 to .50. Traditional
statistical tests would reject the null hypothesis (p < .001). As further evidence,
there is a 100% inference probability that the true correlation of these variables is
positive, as predicted. Thus, this hypothesis was supported.

Test of Hymthesis 1k. A moderate negative correlation was found between
the revised Agreeableness scale of the NEO-PI-R and the revised Turning Against
Others scale of the DMI Lr = ~.39); this correlation rises to ~.47 when corrected for
attenuation. There is a 90 % confidence interval around the corrected correlation of
~37 to ~57. Traditional statistical tests would reject the null hypothesis (p < .001).
As further evidence, there is- a 100% inference probability that the true correlation of

these variables is negative, as predicted. Thus, this hypothesis was supported.

 

72
Test of Hymthesis 11. A small positive correlation was found between the

revised Conscientiousness scale of the NEO-PI-R and the revised Principalization
scale of the DMI (r = .15); this correlation rises to .20 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
.07 to .33. Traditional statistical tests would reject the null hypothesis (p < .05).
As further evidence, there is a 99% inference probability that the true correlation of
these variables is positive, as predicted. Thus, this hypothesis was supported.

Test of Hyppthesis 1m. A small positive correlation was found between the
revised Conscientiousness scale of the NEO-PI-R and the revised Reversal scale of the
DMI (1; = .18); this correlation rises to .23 when corrected for attenuation. There is
a 90% confidence interval around the corrected correlation of .11 to .35. Traditional
statistical tests would reject the null hypothesis (p < = .01). As further evidence,
there is a 100% inference probability that the true correlation of these variables is
positive, as predicted. Thus, this hypothesis was supported.

Test of Hyppthesis 1n. A small negative correlation was found between the
revised Conscientiousness scale of the NEO-PI-R and the revised Turning Against
Others scale of the DMI (1; = ~.08); this correlation rises to ~09 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
-.21 to .02. Traditional statistical methods would not reject the null hypothesis
(p = NS). Use of the maximum likelihood approach indicates that there is a 91%
inference probability that the true correlation of these variables is negative, as
predicted. Thus, the hypothesis of the direction of this correlation was supported,

although the magnitude did not reach statistical significance.

73

Test pf Hypothesis lo. A small negative correlation was found between the
revised Conscientiousness scale of the NEO-PI-R and the revised Turning Against
Self scale of the DMI (r = ~.11). This correlation rises to ~.13 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
~01 to -.25. Traditional statistical tests would reject the null hypothesis (p = NS).
As further evidence, there is only a 3 % inference probability that the true correlation

of these variables is positive, as predicted. Thus, this hypothesis was not supported.

Results Of Tests Of Hypotheses Concerning The Relationship Of The Characteristic

Use Of Defense Mechanisms To Social Support

The second group of hypotheses concerned the interrelationship of the
characteristic utilization of defense mechanisms with the availability of social support.
Following Levine and Hunter (1983), each correlation was corrected for attenuation,
and a 90 % confidence interval was computed. Further, computations were made of
the maximum likelihood estimate (or probability inference) that the correlation was in
the predicted direction.

Overall, 3 of the 5 hypotheses concerning the relationship of the characteristic
use of defense mechanisms to social support were supported; the results of the tests of
specific individual hypotheses are presented below. For an overview of correlations

pertinant to these hypotheses, see Table 16.

74
Table 16

Correlations Among The DMI And Total Social Support Score
Derived From The NSSQ After Correcting For Attenuation Due To

Measurement Error

 

 

_S_gal_e TSS
PRO 1 l
PRN 01
REV 01
TAO ~09
TAS 05

 

Nata: Decimals removed. Abbreviations refer to the Principalization (PRN),
Projection (PRO), Reversal (REV), Turning Against Others (TAO) and Turning
Against Self (T AS) scales of the DMI, and the Total Social Support score derived
from the NSSQ.

 

75
Test of Hymthesis 2a. A small positive correlation was found between the

revised Principalization scale of the DMI and the Total Social Support scale derived
from the NSSQ (; = .09); this correlation rises to .11 when corrected for attenuation.
There is a 90% confidence interval around the corrected correlation of ~01 and .24.
Traditional statistical tests would not reject the null hypothesis (p = NS). Use of the
maximum likelihood approach indicates that there is a 93 % inference probability that
the true correlation of these variables is positive, as predicted. Thus, the hypothesis
of the direction of this correlation was supported, although the magnitude did not
reach statistical significance.

Test of Hymthesis 2b. A small positive correlation was found between the
revised Projection scale of the DMI and the Total Social Support scale derived from
the NSSQ (; = .01); this correlation remains at .01 when corrected for attenuation.
There is a 90% confidence interval around the corrected correlation of ~. 12 to .14.
Traditional statistical tests would not reject the null hypothesis (p = NS). As further
evidence, there is a only a 57% inference probability that the true correlation of these
variables is negative, as predicted. Thus, this hypothesis was not supported.

Test of Hyp_othesis 2c. A small positive correlation was found between the
revised Reversal scale of the DMI and the Total Social Support scale derived from the
NSSQ (r = .01); this correlation remains at .01 when corrected for attenuation.
There is a 90% confidence interval around the corrected correlation of -. 10 to .13.
Traditional statistical tests would not reject the null hypothesis (p = NS). As further
evidence, there is only a 57 % inference probability that the true correlation of these

variables is negative, as predicted. Thus, this hypothesis was not supported.

 

76
Test of Hypothesis 2d. A small negative correlation was found between the

revised Turning Against Others scale of the DMI and the Total Social Support scale
derived from the NSSQ (g = ~.08); this correlation rises to -.09 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
~20 to .02. Traditional statistical tests would not reject the null hypothesis

(p = NS). Use of the maximum likelihood approach indicates that there is a 91%
inference probability that the true correlation of these variables is negative, as
predicted. Thus, the hypothesis of the direction of this correlation was supported,
although the magnitude did not reach statistical significance.

Test of Hyp_othesis 2e. A small positive correlation was found between the
revised Turning Against Self scale of the DMI and the Total Social Support scale
derived from the NSSQ (; = .04); this correlation rises to .05 when corrected for
attenuation. There is a 90% confidence interval around the corrected correlation of
~07 to .16. Traditional statistical tests would not reject the null hypothesis
(p = NS). Use of the maximum likelihood approach indicates that there is a 75%
inference probability that the true correlation of these variables is positive, as
predicted. Thus, the hypothesis of the direction of this correlation was supported,

although the magnitude did not reach statistical significance.

Results From Tests Of The Complex System

Before integrating the hypotheses and conducting a path analysis, the model
was revised based on prior analyses. These revisions served primarily to respond to

indications in the data that the original model did not capture meaningful aspects of

77

the relationship of these constructs. The revised general hypothesized model is

presented in Figure 2.

The assessment and inclusion of demoggaphic variables. Following the

confirmatory factor analysis and revision of scales of the NEO-PI-R and DMI,
oneway analyses of variance for each scale were conducted using demographic
information as the independent variables (see Tables 13 to 15). Generally, these
analyses revealed consistent significant differences in the scales of the NEO-PR—R, the
DMI, and the N SSQ based upon sex. There were no differences based on educational
level, and, while some differences based on race were obtained, the homogeneity of

the sample precluded incorporating this into the path model.

Revising the DMI. Given the correlation matrix of the revised DMI scales,
which suggested low construct validity (see Table 12), a second order factor analysis
was conducted. This analysis confirmed the presence of two underlying factors,
which were labelled "internalizing defenses" and "externalizing defenses" (see Tables
20 and 21). Intemalizing defenses were considered to be the Principalization and
Reversal scales of the DMI (which correlated .27) while externalizing defenses were

considered to be the Projection and Turning Against Others scales of the DMI (which

 

Figure 2

78

Revised Hyppthesized General Structural Relationship of Personalig, Defense

Mechanisms, and Social Support

 

 

GENDER

 

 

 

V

 

 

 

 

 

 

 

 

 

 

 

 

PERSONALITY
TRAITS
INTERNALIZING EXTERNALIZING
DEFENSE DEFENSE
MECHANISMS MECHANISMS
SOCIAL
SUPPORT

 

 

 

 

 

79
Table 17

Means, Standard Deviations, And Analyses of Variance Of All Revised Sgtles As

A Faction Of Gender

 

MEAN/SD

Saale Female Male df F eta_

N 91.5/ 81.0/ 1,285 26.4*** .29
19.8 14.9

E 82.6/ 74.6/ 1,285 22.9*** .27
15.2 13.2

0 85.7/ 81.5/ 1,285 5.4* .13
14.1 16.6

A 80.8/ 76.4/ 1,285 6.4* .15
14.6 14.2

C 87.1/ 87.3/ 1,285 0.0 .00
17.4 16.6

PRN 22.5/ 22.4/ 1,274 0.0 .32
4.1 3.9

PR0 17.8/ 20.2/ 1,274 29.6*** .00
3.7 3.5

REV 20.6/ 21.4/ 1,274 1.8 .07
5.3 4.7

TAO 31.7/ 33.1/ 1,274 2.0 .08
7.5 8.2

TAS 30.2/ 24.9/ 1,274 57.3*** .41
6.1 5.5

TSS 362/ 283/ 1,287 8.0** .16
251 222

 

*p <.05 **p_ <.01 ***p <.001

Note: Abbreviations refer to the Neuroticism (N), Extraversion (E), Openness (O),
Agreeableness (A), and Conscientiousness (C) scales of the NEO-PR-R, the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales of the DMI, and the Total Social
Support Score derived from the NSSQ. Social Support expressed in hundreds. Eta is
the unbiased estimator.

80
Table 18

Means, Standard Deviations, And Analyses of Variance Of All Revised Scales As

A Function Of Educational level

 

MEAN/SD

_S_ca_le Frsh. Soph. Jr. Sr. df F eta_

N 87.5/ 86.5/ 86.9/ 79.2/ 3,283 1.5 .10
19.6 15.6 17.4 14.8

E 80.4/ 77.6/ 76.2/ 79.0/ 3,283 1.24 .08
15.2 15.2 13.2 14.5

o 84.5/ 84.2/ 82.3/ 80.9/ 3,283 0.58 .04
15.1 16.4 16.7 12.1

A 78.9/ 79.6/ 76.5/ 79.4/ 3,283 0.56 .03
15.1 13.2 14.6 15.2

C 87.0/ 86.3/ 85.3/ 94.7/ 3,283 2.04 .12
16.6 14.4 18.6 17.2

PRN 21.9/ 22.9/ 22.9/ 23.1/ 3,272 1.51 .07
3.8 4.2 4.2 3.9

PR0 18.7/ 18.9/ 19.7/ 19.2/ 3,272 1.01 .10
3.7 4.4 3.7 3.4

REV 21.1/ 21.3/ 20.4/ 20.7/ 3,272 0.34 .01
5.0 5.2 5.1 5.1

TAO 32.9/ 31.7/ 32.6/ 30.8/ 3,272 0.70 .05
7.8 7.6 8.6 6.7

TAS 22.8/ 26.9/ 26.3/ 26.7/ 3,272 2.84 .15
5.8 5.7 7.8 6.7

TSS 327/ 347/ 282/ 341/ 3,286 0.85 .06
236 245 223 275

 

*9 <.05 **p <.01 ***p <.001

_N_p_t_e_; Abbreviations refer to the Neuroticism (N), Extraversion (E), Openness (O),
Agreeableness (A), and Conscientiousness (C) scales of the NEO-PR—R, the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales of the DMI, and the Total Social
Support Score derived from the N SSQ. Social Support expressed in hundreds. Eta is
the unbiased estimator.

8 1
Table 19

Means Standard Deviations And Analyses of Variance Of All Revised Scales As

 

 

 

A Function Of Race
MEAN/SD
Scale (1) (2) (3) (4) (5) (6) df F eta_
N 86.4/ 83.8/ 77.0/ 87.9/ 82.0/ 89.6/ 5,281 0.29 .00
18.5 11.4 22.6 8.7 29.7 9.6
E 79.6/ 71.6/ 77.5/ 74.0/ 83.5/ 71.8/ 5,281 1.89 .14
15.0 11.2 13.4 12.6 5.0 15.8
0 84.7/ 72.4/ 80.0/ 79.5/ 92.0/ 87.8/ 5,281 3.50** .21
15.5 12.7 14.1 11.3 1.4 13.9
A 79.4/ 72.2/ 82.0/ 81.2/ 56.5/ 78.2/ 5,281 225* .16
14.9 12.2 4.2 14.4 3.5 11.3
C 87.0/ 88.1/ 70.5/ 88.4/ 89.0/ 89.4/ 5,281 0.45 .03
17.1 12.5 14.8 13.6 4.2 26.9
PRN 22.5/ 21.5/ 21.5/ 23.0/ 24.0/ 22.0/ 5,270 0.40 .02
4.0 4.1 5.0 3.6 2.8 1.7
PR0 18.8/ 20.5/ 21.5/ 19.6/ 20.5/ 19.7/ 5,270 1.13 .10
3.8 4.7 4.9 2.3 3.5 0.6
REV 20.7/ 22.9/ 19.0/ 23.1/ 20.5/ 22.7/ 5,270 1.11 .09
5.1 3.7 4.2 4.5 0.7 11.6
TAO 32.3/ 34.8/ 37.5/ 26.1/ 32.5/ 30.7/ 5,270 1.60 .13
7.8 6.3 13.4 10.8 3.5 7.8
TAS 28.0/ 23.7/ 22.5/ 30.8/ 24.5/ 27.0/ 5,270 2.61* .18
6.3 6.0 9.2 5.5 4.9 10.4
TSS 331/ 355/ 131/ 187/ 467/ 230/ 5,283 1.19 .10
239 284 68 180 300 158
*p<.05 **p<.01 ***p<.001

Note: (1) White; (2) African-American; (3) Hispanic; (4) Asian—American; (5)
Native-American; (6) Other. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the NEO-PR-R, the Principalization (PRN), Projection (PRO), Reversal (REV),
Turning Against Others (TAO) and Turning Against Self (TAS) scales of the DMI,
and the Total Social Support Score derived from the NSSQ. Social Support expressed
in hundreds. Eta is the unbiased estimator.

82
correlated .29) and the reflection of the Turning Against Self scale of the DMI (which

correlated with Projection and Turning Against Others scales at -.45 and ~.34,
respectively). Coefficient alpha for the new Intemalizing Defenses scale (DMI-I) was
found to be .73 while coefficient alpha for the new Externalizing Defense scale
(DMI-E) was found to be .83; raw scores for these derived variables can be found in
Appendix Z. This revision is not without precedence. Anderson and Leitner (1991)
noted that anxiety, depression, and psychiatric symptomatology was related to an
internalizing defensive style in college females, while Noam and Recklitis (1990)
found that attributions of conflict as within or external to the self were associated with

internalizing and externalizing defense mechanisms.

Results of the initial integrational analyses. The correlation matrix of all
variables utilized in the revised model is presented in Table 22. Path coefficients for
all hypothesized linkages are presented in Table. 23, with the standard errors presented
in Table 24. Errors in the model (i.e. , actual minus predicted correlations) are

presented in Table 25.

83
Table 20

Scale Loadings Onto Second Order Factor Matrix For The Revised DMI Scafi

 

Factor
Scale 1 2
PRN 52 ~48
REV 52 ~56
PRO ~78 62
TAO ~42 47
TAS ~40 72

 

Nata: Decimals removed. Factor 1 interpreted as Intemalizing Defenses, comprised
of the Principalization and Reversal scales of the DMI; Factor 2 interpreted as
Externalizing Defenses, comprised of the Projection, Turning Against Others, and
Turning Against Self scales of the DMI (with a reflection of the TAS scale being
used).

84
Table 21

Second Qfler Factor Matrix For Tlae Revised DMI Sca_le_s

 

Factor 1 2
1 --_
2 ~56 ~~~

 

Nata: Decimals removed. Factor 1 interpreted as Intemalizing Defenses, comprised
of the Principalization and Reversal scales of the DW; Factor 2 interpreted as
Externalizing Defenses, comprised of the Projection, Turning Against Others, and
Turning Against Self scales of the DMI (with a reflection of the TAS scale being
used).

 

85
Table 22

Correlation Matrix For All Measures In The Revised Structural uation

Model After Correcting For Attenuation Due To Measurement Error

 

Scale A B C D E F G H I
(A) Gender --—

(B) Scale N ~31 ~~~

(C) Scale E ~29 -20 ---

(D) Scale 0 ~15 02 46 ~~-

(E) Scale A ~16 ~17 14 15 ~~~

(F) Scale C 01 47 23 04 11 ~~~

(G) DMI-Intemalizing 06 ~49 04 14 43 26 ~~~

(H) DMI-Extemalizing 42 ~13 ~09 ~13 ~54 ~02 ~56 ~~~

(I)_So_c_ialSupport ~17 ~04 18 13 03 03 ~11 02 ~~~
Note: Decimals removed. Abbreviations refer to the Neuroticism (N),

Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the NEO-PR-R.

86
Table 23

Path Coefficients For All Linkages In The Revised Structural @uation Model

 

Szala A B C D E F G H I
(A) Gender -..-

(B) Scale N ~31 ~~-

(C) Scale E -29 --_ -__

(D) Scale 0 ~15 —-- --- ---

(E) Scale A —16 --- --- --- --_

(F) Scale C 01 m m --- --.. _--

(G) DMI-Intemalizing ~~~ ~45 ~19 17 35 05 ~~~

(H) DMI-Externalizing ~~~ ~27 ~04 ~03 ~56 ~07 ~~~ -~-
Mal Suyport 17 11

 

Note: Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the NEO-PR-R

87
Table 24

SM Errors For Path Coefficients In The Revised Structural uation
Model

 

 

 

Scale A B C D E F G H I
(A) Gender 00

(B) Scale N 05 ---

(C) Scale E 05 --- ---

(D) Scale 0 06 m --- -__

(E) Scale A 06 m --- --- -__

(F) Scale C 06 m m --- --- ---

(G) DMI-Intemalizing ~~~ 06 O7 07 05 06 ~~-

(H) DMI-Externalizing ~~~ 07 07 07 04 07 ~~~ ~~-
(.'I_)_So_ci_al Support 07 07

Note: Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the N EO-PR-R

 

88
Table 25

Errors (Actual '~ Predicted) For All Linkages In The Revised Structural
Eapation Model

 

 

 

Spale A B C D E F G H I
(A) Gender ---

(B) Scale N 00 __-

(C) Scale E 00 ~29 ---

(D) Scale O 00 ~05 42 ~~~

(E) Scale A 00 -22 09 13

(F) Scale C 00 ~47 23 04 11 ~~~

(G) DMI-Intemalizing ~05 ~05 24 ~01 11 21 ~~-

(H) DMI-Extemalizing 23 17 00 ~08 04 05 ~46 ~~-

(1) Social Support ~21 O7 22 11 04 03 ~05 ~08 ~~~
E3: Decimals removed. Abbreviations refer to the Neuroticism (N),

Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C) scales
of the NEO-PR-R.

89
Using path analysis, the model was found to be significant

(Chi-square = 133.54, df = 19, p < .001), indicating that the overall fit of the
proposed structural model and the obtained data differed significantly. Thus, this

hypothesis was not supported.

Developing And Testing An Alternative Integrational Model

Path analytical results of the first approximation of the model indicated the
need for modification, and an understanding of the relationship between these
constructs was obtained that could be used to guide further explorations. Examination
of the model as originally proposed showed that errors in path prediction were found
to primarily involve the inter-relationship of five factors of the N EO-PI-R, and the
relationship of internalizing defenses to externalizing defenses (see Table 25). Thus,
an effort was made to revise the model in a manner that could include and explain
these findings.

Revisra' g the N EO-PI—R. Given the factor matrix of the revised NEO scales
(see Table 8), particularly the relationship between Scale N and Scale C, and Scale E
and Scale 0, a second-order factor analysis was completed to confirm the presence of
two underlying factors, which were labelled "internalizing personality traits" and
"externalizing personality traits. " Intemalizing personality traits were considered to
be Scale N and a reflection of Scale C of the NEO~-PI~R (which had a factor
intercorrelation of ~.47) while externalizing personality traits were considered to be
Scale E, Scale 0, and Scale A of the NEO-PI-R (Scale E and Scale 0 had a factor

intercorrelation of .46; Scale E and Scale A had a factor intercorrelation of .14; Scale

90
O and Scale A had a factor intercorrelation of .15). Results of this analysis are

presented in Tables 26 and 27. Coefficient alpha for the new Intemalizing Personality
Traits scale (NEO—INT') was found to be .91 , while coefficient alpha for the new
Externalizing Personality Traits scale (N EO-EXT) was found to be .90; raw scores

for these derived variables can be found in Appendix CC.

Reconstructing the model. In addition to revising operationalization of
personality traits, the model was reconstructed in several ways. Although guided by
the empirical findings of the first path analysis, a conceptual framework was also
utilized which, when given a high intercorrelation, placed internal processes earlier in
the hierarchical model than external processes. This was conducted both within
constructs and across constructs, as described below. For an overview of the

alternative model, see Figure 3.

First, the relationship between internalizing and externalizing personality traits
was examined. Given the moderate negative correlation between these variables, and
the pattern of correlations with other variables in the model, it was predicted that
these variables were not hierarchically related. Thus, no direct path was predicted

between internalizing and externalizing personality traits.

91
Figure 3
Altemtive Hyppthesized Structural Relationship Of Personality, Defense

Mechanisms, And Social Suppprt

 

 

INTERNALIZING EXTERNALIZING
PERSONALITY PERSONALITY
TRAITS TRAITS

INTERNALIZING

DEFENSE
MECHANISMS

 

 

 

 

 

 

 

 

 

 

 

 

V

 

EXTERNALIZING
DEFENSE
MECHANISMS

 

 

 

 

 

 

GENDER

 

 

 

 

 

SOCIAL
SUPPORT

   

 

 

 

 

92
Table 26

Loadings For Second Order Factor Matrix For The Revised NEO-PI-R 8cm

 

Factor
Scale 1 2
N 69 ~24
C 69 ~23
E ~31 66
O ~03 70
A ~20 21

 

Nata: Decimals removed. Factor 1 interpreted as Intemalizing Personality, comprised
of Scale N and Scale C of the N EO-PI-R (with a reflection of Scale C being used);
Factor 2 interpreted as Externalizing Personality, comprised of the Scale E, Scale O,
and Scale A of the NEO-PI-R.

93
Table 27

Second Order Factor Matrix For The Revised NEO-PI-R Scal_es_

Eaator 1 2
1

2 ~18 ~~~

M: Decimals removed. Factor 1 interpreted as Intemalizing Personality, comprised
of Scale N and Scale C of the NEO-PI-R (with a reflection of Scale C being used);
Factor 2 interpreted as Externalizing Personality, comprised of the Scale E, Scale 0,
and Scale C of the NEO-PI-R.

94

Second, the relationship between internalizing and externalizing defenses was
examined. Given the high correlation between these variables, it was hypothesized
that they are associated in hierarchical rather than orthogonal manner. Following the
conceptual framework, the model was structured so that externalizing defenses was
preceded and predicted by internalizing defenses.

Third, given that internalizing defenses now are hierarchically associated with
externalizing defenses, the model was changed so that both internalizing and
externalizing personality traits both precede and predict internalizing defenses, but
that only internalizing defenses predicts externalizing defenses.

Fourth, to more clearly delineate internal and external experiences, the model
was changed so that externalizing personality traits directly predicts social support.
Similarly, the model was changed so that externalizing defenses directly predicts
social support but internalizing defenses does not.

Finally, given empirical findings, the model was changed so that gender
predicted both externalizing defenses and social support.

Testing the alternative model. The correlation matrix of all variables utilized
in the revised model is presented in Table 28. Path coefficients for all hypothesized
linkages are presented in Table 29, with the standard errors presented in Table 30.
Errors in the model (i.e. , actual minus reproduced correlations) are presented in Table
31; these errors are generally much smaller than in the original model. Overall, the
model was found to be nonsignificant (Chi-square = 9.24, df = 5, p < NS),

indicating that the overall fit of the proposed structural model and the obtained data

95

do not differ significantly. Thus, this hypothesis was supported. For an overview,

see Figure 4.

Further Analyses Of The Alternative Model

Given that the structural model adequately describes the relationship between
the constructs, additional analyses were undertaken to examine the relationship among
constructs that were directly and indirectly linked in the alternative model.

Analysis of the path coefficient between Intemalizing Personalng Traits and

Intemalizing Defenses. Analyses indicated a path coefficient of .40 between
internalizing personality traits and internalizing defenses. With a standard error of
.05, the 95% confidence interval around this path coefficient is from .30 to .50, and
the 68% confidence interval around this path coefficient is from .35 to .45; since zero
is not included in this interval, traditional statistical analyses would reject the null
hypothesis (p < .001). Further, there is a greater than 99.9% inference probability

that the true association between these constructs is positive.

Figure 4

Path Coefficients For Alternative H othesized Structural Relationshi of Personali

 

96

Defense Mechanisms, and Social Support

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INTERNALIZING EXTERNALIZING
PERSONALITY PERSONALITY
TRAITS TRAITS
.40 T ' -.21

INTERNALIZING
DEFENSE
MECHANISMS
V -5.
EXTERNALIZING
DEFENSE
MECHANISMS

 

 

 

 

 

 

GENDER

 

 

 

 

SOCIAL
SUPPORT

   

 

 

 

97
Table 28

Cprrelation Matrix For All Measures In The Alternative Structural uation

Model After Correcting For Attenuation Due To Measurement Error

 

5946 A B C D E F
(A) Gender ---
(B) NED-Intemalizing -19 ---

(C) NED-Externalizing ~28 —18 ---
(D) DMI-Intemalizing ~06 44 ~28 ---
(E) DMI-Externalizing ~41 06 36 ~56 ~~~

Wanton -17 -03 17 ~06 07
Note: Decimals removed.

98
Table 29

Path Coefficients For All Linka es In The Alternative Structuatl Equation
Model

.39-416 A B C D E 1:

(A) Gender ---

(B) NEG-Intemalizing --- ---

(C) NBC-Externalizing ——- --- -__

(D) DMI-Intemalizing ~~~ 4O -21 ---
(E) DMI-Externalizing ~44 --— --- ~56 -__
(E) Social Support ~15 ~~~ 14 ~~~ ~04 ~~-

 

Note: Decimals removed.

99
Table 30

Standard Errors For Path Coefficients In The Alternative Structural uation

Model

Saale A B C D E F
(A) Gender m

(B) NED-Intemalizing --- -__

(C) NEG-Externalizing --- --- -_-

(D) DMI-Intemalizing ~~~ 05 06 ~~~
(E) DMI-Externalizing 05 ~~~ m 04 --_
(F) Socia Support 06 ~~~ 06 ~~~ 07 ---

 

Note: Decimals removed.

100
Table 31

Errprs (Actual ~ Predicted Correlations) For All Linkages In The Alteraative

Structural uation Model

M A B C D E F
(A) Gender --_

(B) NEG-Intemalizing 00 -__

(C) NED-Externalizing 00 00 ---

'8

(D) DMI—Intemalizing 00 00 __-
(E) DMI-Externalizing 02 22 08 02 ---

LF) Social Support 00 ~04 00 ~05 01 ~~-
Note: Decimals removed.

101

Aalysis of the path coefficient between Externalizing Personalig Traits and
Intemalizing Defenses. Analyses indicated a path coefficient of ~.21 between

externalizing personality traits and internalizing defenses. With a standard error of
.06, the 95% confidence interval around this path coefficient is from ~09 to -.33, and
the 68% confidence interval around this path coefficient is from ~.15 to ~.27; since
zero is not included in this interval, traditional statistical analyses would reject the
null hypothesis (p < .001). Further, there is a greater than 99.9% inference

probability that the true association between these constructs is negative.

Analysis of the path coefficien_t between Externalizing Personalig Traits and

 

Social Supmrt. Analyses indicated a path coefficient of .14 between externalizing
personality traits and social support. With a standard error of .06, the 95 %
confidence interval around this path coefficient is from .04 to .24, and the 68%
confidence interval around this path coefficient is from .11 to .17; since zero is not
included in this interval, traditional statistical analyses would reject the null hypothesis
(p < .01). Further, there is a 99% inference probability that the true association
between these constructs is negative.

Analysis of the path coefficient between Intemalizing Defenses and
Externalizing Defenses. Analyses indicated a path coefficient of ~56 between
internalizing defenses and externalizing defenses. With a standard error of .04, the
95 % confidence interval around this path coefficient is from ~.48 to ~64, and the
68% confidence interval around this path coefficient is from ~52 to ~60; since zero is

not included in this interval, traditional statistical analyses would reject the null

102
hypothesis (p < .001). Further, there is a greater than 99.9% inference probability

that the true association between these constructs is negative.

Analysis of the path coefficient between Gender and Externalizing Defenses.
Analyses indicated a path coefficient of ~.44 between gender and externalizing
defenses. With a standard error of .05, the 95 % confidence interval around this path
coefficient is from ~35 to -.55, and the 68% confidence interval around this path ‘
coefficient is from ~.40 to ~50; Since zero is not included in this interval, traditional
statistical analyses would reject the null hypothesis (p < .001). Further, there is a
greater than 99.9% inference probability that the true association between these
constructs is negative.

Analysis of the path coefmzient between Gen_der and: Social Support. Analyses
indicated a path coefficient of ~. 15 between gender and social support. With a
standard error of .06, the 95 % confidence interval around this path coefficient is from
~01 to ~25, and the 68% confidence interval around this path coefficient is from ~07
to ~. 19; since zero is not included in this interval, traditional statistical analyses would
reject the null hypothesis (p < .01). Further, there is a 98% inference probability

that the true association between these constructs is negative.

Analysis of the path coefficient between Externalizing Defenses and Social

 

Suppprt. Analyses indicated a path coefficient of ~04 between externalizing defenses
and social support. With a standard error of .07, the 95 % confidence interval around
this path coefficient is from -. 18 to .10; since zero is included in this interval,

traditional statistical analyses would not reject the null hypothesis (p = NS).

103

However, there is a 72 % inference probability that the true association between these
constructs is negative.

Analysis of the missing link between Intemalizing Personalig Traits and
Externalizing Defenses. Analyses indicated a correlation (corrected for attenuation) of
.06 between internalizing personality traits and externalizing defenses. With a
standard error of .06, the 95 % confidence interval around this correlation is from ~04
to .16; since zero is included in this interval, traditional statistical analyses would not
reject the null hypothesis (p = NS). There is an 85% inference probability that the
true relationship between these constructs is positive. However, the difference
between the actual and predicted correlation (corrected for attenuation) was .22
(p < .01). Thus, it appears that the model correctly predicts the direction but
overestimates the strength of the relationship between these constructs, even though
they are not directly linked.

Analysis of the missing link between Externalizing Personality Traits and
Externalizing Defenses. Analyses indicated a correlation (corrected for attenuation) of
.36 between externalizing personality traits and externalizing defenses. With a
standard error of .05 , the 95 % confidence interval around this correlation (corrected
for attenuation) is from .28 to .44; since zero is not included in this interval,
traditional statistical analyses would reject the null hypothesis (p < .001). There is an
greater than 99.9 % inference probability that the true relationship between these
constructs is positive. The difference between the actual and predicted correlation

(corrected for attenuation) was .08 (p = NS). Thus, it appears that the model

104

accurately accounts for the relationship between these constructs, even though they
are not directly linked.

Analysis of the missing link between Geade_r and Intemalizing Defenses.
Analyses indicated a correlation (corrected for attenuation) of ~06 between gender
and internalizing defenses. With a standard error of .06, the 95% confidence interval
around this correlation is from ~. 16 to .04; since zero is included in this interval,
traditional statistical analyses would not reject the null hypothesis (p = NS). The
difference between the actual and predicted correlation (corrected for attenuation) was
~.04 (p = NS). Thus, it appears that the model accurately accounts for the
relationship between these constructs, even though they are not directly linked.

Analysis of the missing link between Intemalizing Personalig Traits and Social
Sapppat, Analyses indicated a correlation (corrected for attenuation) of ~03 between
internalizing personality traits and social support. With a standard error of .06, the
95% confidence interval around this correlation is from ~. 13 to .07; since zero is
included in this interval, traditional statistical analyses would not reject the null
hypothesis (p = NS). The difference between the actual and predicted correlation
(corrected for attenuation) was ~.04 (p = NS). Thus, it appears that the model
accurately accounts for the relationship between these constructs, even though they
are not directly linked.

Analysis of the missing link between Intemalizing Defenses and Social
Sappara Analyses indicated a correlation (corrected for attenuation) of ~06 between
internalizing defenses and social support. With a standard error of .06, the 95 %

confidence interval around this correlation is from -. 16 to ~04; since zero is included

105

in this interval, traditional statistical analyses would not reject the null hypothesis
(p = NS). The difference between the actual and predicted correlation (corrected for
attenuation) was ~.05 (p = NS). Thus, it appears that the model accurately accounts

for the relationship between these constructs, even though they are not directly linked.

 

DISCUSSION

The findings of this study support the hypothesis that personality traits, defense
mechanisms and social support are associated in a predictable pattern in college-aged
students. Both simple (or lower order) and complex (or higher order) analyses of the
relationships among measures of these constructs supported numerous hypotheses
about the pattern of relationships between specific personality traits and specific
defense mechanisms, while higher order path-analyses supported the hypothesis of a
hierarchical, causal relationship. However, separate second-order factor analyses
indicated that restructuring both the NEO-PI-R and DMI into measures of
"internalizing" and "externalizing" personality traits and defenses, respectively, was
warranted, and in fact facilitated the creation of a hierarchical model. Although this
this study’s primary purpose was not to assess the psychometric qualities of the

instruments used to measure the constructs, several related findings are noteworthy.

Psychometric Qualities Of The Measures

Reliabilig of the measures. These analyses highlight the importance of not relying
solely on alpha as a measure of internal consistency. The NEO-PI-R was found to
have strong internal consistency, as evidenced by alphas ranging from .87 to .89 for
each of its factors. The DMI scales were less impressive, with alphas ranging only

from .60 to .83. However, confirmatory factor analyses found that, as indicated by

106

 

107

the number of items that were removed, both scales contained items which loaded
poorly onto its assigned factor. Since the computation of alpha is in part dependent
upon the number of items comprising the scale, removing the peripheral items from
factors of the NEO-PI—R and DMI measures did not notably increase their alphas,
although it improved the consistency of the factor loadings. Since the psychometric
qualities of both instruments, and particularly the NEO-PI-R, have been thoroughly
examined, these findings may represent anomalies in this study, perhaps related to the
homogeneity of the sample. It was not possible to compute alpha for the N SSQ.
Since ratings for each person in the network are made independently of all other
ratings, there are no " items " as such in which to assess internal consistency.

Taken as a whole, the reliabilities found in this study mirror the degree to
which the addressed construct being measured is considered to be "stable and
enduring". The highest alphas were found for the Big-Five personality traits, which
were assumed to be stable and enduring. Scales measuring defense mechanisms,
assumed to be less stable and enduring, had less reliability. While defenses may be
an intrapersonal process, the use of defense mechanisms is likely to be situationally
based, thus lowering the alphas for these scales. Some support for this poSition can
be found in the evidence that there is a tendency to utilize defense mechanisms from
either an "internalizing" group or an "externalizing" group. Finally, the reliabilities
for the employed measure of social support could not be determined.

Issues of the factors of the scales. Both the NEO-PI-R and the DMI were found to
have intercorrelations between subscales which presumably measure relatively

independent constructs, and the NSSQ Total Social Support Score is worth further discussion.

108

, There are theoretical reasons that defense mechanisms would be
intercorrelated. As noted by Cramer (1988), there is no theoretical or empirical
reason to expect that individuals rely upon only one defensive style and rigidly apply
it in all situations. Thus, although a decline in the specificity of the theory, this
study’s reduction of the five defense mechanisms to two defensive styles may
represent an increase in the external validity (Campbell & Stanley, 1966) of this
study, that is, the significance and utility of the theory. An ipsative reason for the
intercorrelations found on the DMI may be due to the manner in which the test is
administered, which requires subjects to choose between five alternative options when
describing their reactions to various stories (see Appendix D). Thus, because only
two choices can be made (i.e., most like me, least like me), three other choices must
be ignored. The scoring format, which sums all related responses scored on a single
factor, is set up so that as one defensive style is selected and mathematically
increased, another defensive style is not and can not be selected, and thus is not
mathematically increased. Such a reduction in the freedom of responses insures that
the scales will be correlated to some extent. It is reasonable to expect that, were
subjects to rate the likelihood of the possibility that they would use each defensive
. strategy on a Likert scale, a different pattern of relationships among the scales would
have been found; however, previous comparisions of free~ and fixed-format responses
have suggested that scale scores did not differ significantly (Ihilevich & Gleser,
1986).

The correlations among scales of the NEO-PI-R are more problematic, and is

discussed below as part of the theoretical implications of this study.

109

This study chose to focus on Total Social Support as a dependent variables,
and derived this variable from the product of the Functional and Structural Support
scores of the NSSQ. While Functional and Structural Support are theoretically
different constructs, the total amount of functional support available to an individual
will be directly related to structural issues such as network size and density. The
total amount of support available can be no greater than the maximum amount the
network is capable of providing. Given this, a small social support network will not
be able to provide as much functional support as a larger social support network. To
adjust for this, previous studies have attempted to analyze the mean amount of
functional social support available (c.f. , Kincaid & Caldwell, in press), however, this
study was concerned with the overall ability of an individual to organize and access
social support, and thus, both the total amount of functional social support and total

structural support were of use for analyses.

Interpreting The Lower Order Analys_ea

The majority of the hypotheses regarding the relationship of personality traits
to the characteristic use of defense mechanisms were supported. Examination of the
magnitude of these relationships (see Table 11) indicated Scales N and Scales A
contributed the strongest associations between personality and defenses.
Agreeableness was significantly related to all five defense mechanisms, and
Neuroticism was significantly related to four of the five defense mechanisms assessed
by the DMI. These findings partially refute earlier works which suggested that

measures of neuroticism and extraversion had the most predictive utility (Wiggins, 1968).

110

Agreeableness was found to be most strongly associated in a negative manner
with the PRO, TAO scale of the DMI, and in a positive manner with the REV scales
of the DMI. Scale A of the NEO-PI-R is a measure of interpersonal tendencies, with
high scores associated with altruism, the ability to empathize and sympathize, and
interpersonal cooperation, and low scores associated with individuals who are
antagonistic, egocentric, and competitive (Costa & McCrae, 1992). The PRO and
TAO both assess the tendency to project the source of one’s problems out into the
social environment (Ihilevich & Gleser, 1991). Thus, the negative association found
between Scale A and PRO/TAO defenses indicates that individuals who are socially
oriented strive to maintain a pleasant homeostasis in their interpersonal relationships
are less likely to engage in the externalization of their problems. Since the REV scale
assesses the degree to which such problems are internalized (Ihilevich & Gleser,
1991), it can be concluded that such socially oriented individuals internalize their
problems instead of externalizing them.

Neuroticism was found to be positively associated with the TAO and TAS
scales of the DMI, and negatively associated with the PRN, and REV scales of the
DMI. Scale N is a measure of psychological distress and poor coping skills, with
high scores associated with poor impulse control, irrational ideas, neurotic behavior,
and disruptive emotion, and low scores associated with calm, even-tempered
individuals who are able to cope with stressful situations without major psychological
disruption (Costa & McCrae, 1992). Both the TAO and TAS scales measure coping
efforts taken by individuals thought to feel inadequate, inferior, or worthless, with

TAO representing a more impulsive, under-socialized effort and TAS representing a

1 1 1
more controlled, over-socialized (inhibited) effort (Ihilevich & Gleser, 1991). Thus

the positive association found between Scale N and TAO/TAS defenses indicates that
individuals with increasing levels of psychological distress and feelings of inadequacy
may attempt to compensate or cope with these feelings by increasing reliance on the
blame of self or others. Either response (TAO/TAS) represents an unhealthy attempt
at coping, for blaming either oneself or others for problems and emotions (a) is not
likely to be a fully realistic appraisal, and (b) offers no attempts at solving the
problem.

Similarly, the PRN and REV scales of the DMI both measure the tendency to
minimize the expression of strong affects (through intellectualization or even
repression) and to avoid stimulating anxiety (Ihilevich & Gleser, 1991). Thus, the
negative association found between Scale N and PRN / REV defenses indicates that
individuals with increasing levels of psychopathology are less likely to utilized
defenses which might further heighten their anxiety. Taken together, these results
suggest that neurotic individuals seek to minimize their own sense of anxiety and
search for the source of their problems rather than a solution to their problems. Such
a finding fits Freud’s (1936) theoretical description of neurosis.

Extraversion was found to be negatively associated with the TAO scale of the
DMI. Scale E is a measure of sociability, with high scores associated with liking
people, optimism, and excitability, and low scores associated with reserved,
independent individuals (Costa & McCrae, 1992). As noted, the TAO scale measures
an aggressive, impulsive, under-socialized effort to control feelings of inadequacy and

doubt. Thus, the negative association found between Scale E and TAO defenses

112

indicates that individuals who are socially oriented and attentive to the quality of
interpersonal relationships are less likely to engage in the aggressive blame of others.
Conscientiousness was found to be positively associated with the PRN, REV
scales of the DMI, and negatively associated with the TAO, TAS scales of the DMI.
Scale C of the NEO—PI-R is a measure of self-control, with high scores associated
with good impulse control, planning, organization, and determination, and low scores
associated with a flexible ethics and decreased goal-directed behavior (Costa &
McCrae, 1992). As noted, PRN / REV defenses measure the tendency to avoid strong
affects. Thus, the positive association between Scale C and PRN/REV defenses
indicates that individuals who are self-controlled and see themselves as a causal agent
in the determination of significant aspects of their lives also tend to minimize the
affective component of their coping efforts and increase rationale, dispassionate
efforts. Similarly, since TAO/TAS defenses assess over~ and under-controlled efforts
at assessing identifying sources of problems and assessing blame, the negative
association between Scale C and TAO/TAS defenses indicates that individuals who
believe they Can control all aspects of their lives do not attempt to blame themselves
or others for their problems. The positive association between Scale C and TAO
defenses makes theoretical sense; however, the positive association between Scale C
and TAS defenses was contrary to the predicted relationship and is difficult to
interpret theoretically. Perhaps this finding indicates that conscientious and controlled
individuals tend not to assign blame to themselves or others, and approach all

problems in a rationale, dispassionate manner.

113

An unexpected positive association found between Extraversion and REV
defenses appears to indicate that individuals who value social relationships repress
their affects, perhaps in order to maintain the relationships. There was no relationship
found between Scale O and REV defenses, although a negative relationship was
anticipated. This appears to indicate that the extent to which an individual is
imaginative, sensitive, and curious is not associated with the degree of repression

utilized as a coping style.

Integpretmg' The Original Path Model

Following the tests of the simple relationships of the constructs, the original
path model was revised to include the effects of gender, and to account for the high
intercorrelations of the scales of the DMI.

A second-order factor analysis suggested that the intercorrelations of the scales
could be accounted for by two underlying factors, which were interpreted as an
"internalizing" and "externalizing" defensive style. AS noted, the Strongest
associations between personality traits and the characteristic utilization of defense
mechanisms were found in Scale N and Scale A; such findings facilitated
interpretation of the later second-order factor analysis of the DMI. The Neuroticism
Scale, which correlated strongly with each defense mechanisms, has previously been
discussed as a measure of general psychopathology (Digman, 1990; Bergeman,
Chipuer, Plomin et. a1, 1993); thus, Scale N would be expected to be globally
correlated with defense mechanisms, for heightened expressions of both neurotic

personality traits and defense mechanisms would be considered evidence of

 

114

psychopathology. Further, the Agreeableness scale, which correlated strongly with
defenses that involve other people, has previously been discussed as a measure of
interpersonal tendencies (Costa & McCrae, 1992); thus, Scale A would be expected to
be globally correlated with interpersonal defense mechanisms, for heightened
tendencies for interpersonal orientations should lead to reduced use of interpersonal
defense mechanisms. Therefore, it appears that the second order factor analyses of
both the NEO-PI-R and the DMI serve to reduce both measures to the two general
dimensions representing the strongest observed correlations among personality and
defense mechanisms. Although reducing the five defense mechanisms to two styles
would reduce predictive ability within specific aspects of the model, it eased the
prediction and interpretation of the overall model.

The original hierarchical structural model was not supported, although the
majority of the hypotheses about the relationships among the NEO-PI—R, the DMI,
and the NSSQ were in both the expected directions and of magnitude sufficient to
reach statistical significance. There are several possible reasons for this.

First, it is possible that although specific correlations between factors of the
NEO-PI—R, DMI, and N SSQ were correctly predicted, these correlations were either
spurious. Spurious correlations would add to the error variance of a structural model
(Costner, 1989), and lead to a significant Chi-square when the structural, causal
interrelationships were assessed through the path analysis. However, an argument
against this can be mounted. While correlations are not causation, causation creates
correlation (Campbell & Stanley, 1963); both theory and results of additional analyses

of specific relationships between personality traits, defense mechanisms, and social

115

support suggested that these relationships were not spurious. This suggests the
original model was incorrectly constructed.

Second, it is possible that specific linkages in the model were predicted
incorrectly. Again, however, theory and results of additional analyses of specific
relationships suggest that this model, with the exception of a few linkages, was
properly formed.

Finally, it is possible that the original model was too complex. Since
increasing the degrees of freedom allow even small differences between the predicted
and observed models to be detected in a Chi-square analysis, particularly with
increasing sample size, it is possible that the model was largely correctly formed but
imperfect; thus, given the statistical power of the study, the model was rejected
through the Chi-square analyses.

In hindsight, it appears most likely that the failure of the original path model
was due to a combination of incorrect prediction of a small number of the linkages in
a highly complex model with analyses conducted that were sensitive to minute
differences between the predicted and observed models. Clearly, some linkages were
predicted incorrectly, given the failure of three of 15 hypotheses regarding the
relationship of personality traits to the characteristic utilization of defense
mechanisms, and the failure of two of the five hypotheses regarding the relationship
of the characteristic use of defense mechanisms to social support. Since over 15
linkages were proposed and the study obtained an n of nearly 300, the model had a
large number of degrees of freedom, and thus a great deal of statistical power to

detect the impact of these errors, and reject the model as a whole.

1 16
Interpreting The Alternative Path Model

In examining the analyses of the original model, it became clear that
cosiderable error in the model may be due to the high correlations among personality
traits that were not accounted for by the model. A second-order factor analysis
conducted on the scales of the NEO-PI-R found evidence of two underlying factors,
which were again interpreted as representing " internalizing " and "externalizing"
personality styles. As with the DMI, the reduction of the personality traits to two
styles also reduced the predictive ability of the model while easing the prediction and
interpretation of the overall model. However, this also creates theoretical issues with
the five-factor model.

Having reorganized the personality traits into two overarching styles, the path
model was reconstructed, using both theory and the results of the first path analysis,
to account for the relationship between all the constructs. Both internalizing and
externalizing personality traits were found to predict internalizing defense
mechanisms, while personality traits are associated with externalizing defense
mechanisms only through internalizing defense mechanisms. It appears that a
hierarchical relationship occurs within constructs as well as between constructs;
support was found for locating intrapersonal processes prior to interpersonal processes
in the model. Most notably, the utilization of internalizing defenses such as denial
and repression seems to preclude the use of externalizing defenses such as projection
and aggression. Further, externalizing personality traits were found to be directly

linked to social support; this suggests that, regardless of the characteristic defenses

_7 _.._ Mq— _._au--

117

utilized, the formation and management of relationships with other people is partially
dependent upon one’s orientation toward the world.

Although gender effects are reported by the developers of both the NEO-PI—R
and the DMI, these analyses suggest that the influence of gender is most evident in
interpersonal settings, particularly in the use of defense mechanisms which involve
other people. Gender was not found to be directly associated with personality traits.
Women were found to have less reliance on externalizing defense mechanisms and
utilized greater levels of social support, findings consistent with prior literature.
Surprisingly, however, the use of internalizing defenses was not directly associated
with gender. Implicit rule structures in society could account for this: men and
women may both be free to develop defensive styles which impact primarily upon
themselves, but women are encouraged to be caretaking and nurturing of others, and
thus implicitly restricted from developing defensive styles which could impact
negatively upon others.

Overall, the Chi-square analyses indicated that the revised model fits the data,
in that it adequately accounts for observed correlational relationships found among the
constructs (both those with direct links and those without direct links). Errors
between the predicted and observed correlations were generally low; the weakest
aspect of the model involved the construct of social support.

The construct of social support was the most problematic aspect of this study.
Independent of the model developed, social support was not found to be strongly
correlated with any other construct measured. This may be because the psychometric

properties of the NSSQ measure may not be adequate for empirical studies as has

1 18
been previously reported (a contradiction of previous research; c.f. Bruhn & Phillips,

1984). This may also be because there is in fact no true association of social support
to more internal processes, a position difficult to defend theoretically. However, a
likely alternative explanation is that the NSSQ measures the perception of social
support but does not measure the actual quality of interpersonal relationships. Both
the N EO-PI-R and the DMI are self-report measures of oneself only (and, incidently,
have a more opaque face validity, or a less clear association between the items on the
scale and the underlying constructs being measured). The NSSQ, however, in
addition to being transparent in its objectives, measures only the perceptions of one
half of a dyadic relationship. Were it possible to assess both members’ perceptions
of the social support given for each and every dyadic relationship reported by each
and every subject (a daunting task), a more accurate measure of the quality of these
relationships could be achieved and thus, with less error variance in the measure, the
magnitude of the association of social support to personality traits and defensive styles
would be likely to increase. Further, since there is no way to assess the internal
consistency of the NSSQ, no corrections could be made for attenuation due to
measurement error. This meant that the NSSQ was entered into the path model with
a perfect reliability, a proposition doubtful in reality. By minimally the estimating
alpha to be near .90, the magnitude of these relationships would increase; if alpha
were further reduced (that is, if estimates of the error in measurement were
increased), the magnitudes would continue to rise. Since no estimate of measurement
error was made for the NSSQ while actual error in measurement seems certain, it is

likely that the underlying true association between these constructs is stronger than

119
found here. Thus, within this framework, the portions of the model involving social

support are likely to be a conservative estimate of the relationship to personality traits

and defense mechanisms.

Theoretical Implications
This study found that the factors of the N EO—PI-R could be reduced to two

dimensions, labeled "internalizing" and "externalizing" traits; clearly, however, these
findings replicate and validate in other populations. If the five factors are truly
independent and orthogonal, how can two underlying factors account for the
correlations found? Perhaps, as suggested by McAdams (1992), the five factors are
actually trait categories rather than distinct entities, and further, that these categories
have overlapping boundaries. In this view, the five-factor theory operates at "the
broadest level of hierarchy, and in that sense they are to personality what the
categories of ’plant’ and ’anirnal’ are to the world of natural objects ~ extremely
umful for some initial rough distinctions but of less value for predicting Specific
behaviors" (John, 1989; p. 267). Or perhaps, as suggested by Wiggins and Pincus
(1992), the five-factor model is complementary to other circumplex models of traits
(c.f. Benjamin, 1974; Leary, 1957; Wiggins, 1979). In this view, the factors of
extraversion and agreeableness (in the five-factor model) are seen as more stable
across contexts, while circumplex dimensions of dominance and nurturance (in the
circumplex model) are seen as more influenced by context. An additional,
parsimonious explanation can be found by conceiving of the two underlying

dimensions of the NEO-PI-R as "intrapersonal" traits and "interpersonal" traits. By

fi .
.-«-—_._n_———-— w.—

120

conceiving of the dimensions in this manner, a parallel can be drawn to suggestions
that the Big Five factors supplement circumplex classifications with the additional
dimensions of conscientiousness, neuroticism, and openness to experience (Trapnell &
Wiggins, 1990; Wiggins & Trapnell, 1992).

As noted by McAdams (1992), the primarily utility of trait theory is the
explanation of consistency in human behavior. Thus, Zuroff (1986) has distinguished
three different types of behavioral consistency in which traits could be expressed, with
each position accepted and argued by different theorists using different studies as
evidence. Zuroff’s first position is that traits cause behavioral consistency across all
situations and contexts. His second position is that traits cause behavioral consistency
within a finite number of Situations and contexts otherwise have no effect. His third
position is that traits cause behavioral consistencies averaged across all situations and
contexts.

The results of this study support Zurost third position. First, the reduction
of the five scales of the NEO-PI-R to two dimensions represents a move from higher
predictive specificity to lower predictive specificity; they become, therefore, a
measure of the general tendencies of the subject. Similarly, the reduction of the DMI
from five scales to two dimensions allows the assessment of general behavioral
tendencies as well. Second, the DMI presents a number of stimulus scenarios which
vary in contextual features, and determines the relative strength of each defense
mechanisms by aggregating tendencies across all these contexts. Finally, correlations
between the specific personality traits and specific defense mechanisms as well as the

magnitude of the path coefficients indicates that there is not a perfect causal

121
relationship. Taken together, the results of this study indicated that specific

personality traits are generally related to specific defense mechanisms, that specific
defense mechanisms are generally related to social support, and that the path model
globally describes the hierarchical relationship among the constructs. Thus, it appears
the path model allows, with a high degree of accuracy, the probability of the
characteristic use of certain defense mechanisms when given information about
personality traits, and the probability of the availability of social support when given
information about the characteristic use of defense mechanisms. Variance in the
actual usage of defense mechanisms and the actual availability of social support would
therefore vary with context and other situational variables. This is highly
complementary with the view of Buss (1989), who, in comparing trait theory with the
experimental manipulation of specific responses, argued that traits are mostly clearly
expressed in familiar contexts where considerable latitude about the choice of
response is given over a extensive period of time. This view can be applied to these
findings in statistical terms: increasing the degrees of freedom (i.e. , contextual and
situational factors) also increases the likelihood of the expression of personality traits
through the characteristic use of defense mechanisms, and subsequently, the

availability of social support.

Clinical Implications

This study is able to contribute to knowledge in clinical settings, both
diagnostic and treatment. Costa (1986; 1991) has expressed his h0pe that the

five-factor model, as assessed by the NEO-PI-R, could become an integral part of

122

clinical personality assessments, and has suggested that the combination of current
clinical assessment tools with personality measures such as the NEO-PI-R could lead
to more complete understanding of clients. This study further increases the heuristic
knowledge that can be derived from the N EO-PI-R, and expands upon the data
provided by the DMI. Since DSM-HI-R (American Psychiatric Association, 1987)
describes Axis H disorders, in part, in terms of personality traits, the association of
such traits with defense mechanisms suggests the NEO-PI-R may have utility in
increasing diagnostic accuracy. Further, while not predicting any diagnostic outcome,
this study makes plausible that many of the constructs of interest to clinicians (i.e.,
defensive style, interactional patterns with others) are clearly related to personality
traits, and may be caused by these traits. Efforts to explain or intervene at either the
level of defense mechanisms or of social interactions (both of which frequently occur
in psychotherapy) may be misguided without an awareness of the underlying
personality traits that are driving such processes. While such information may have
emerged during the course of therapy , this Study suggests a model which can provide
early insights and a framework for understanding the client. Thus, understanding the
pathways between intrapersonal and interpersonal behavior can assist in answering

diagnostic questions and can refine the application of treatment techniques.

S_u_mtm
Personality theory suggests that traits, and particularly those traits described in
the five-factor theory, are stable and enduring characteristics of individuals that

underlie many aspects of both intrapersonal and interpersonal behavior. To examine

..A —_ .

123

this proposition, this study proposed and tested aspects of the relationship between
personality traits, the characteristic use of defense mechanisms, and social support.
Preliminary analyses suggested that specific personality traits could be linked to
specific defense mechanisms, and, similarly, that specific defense mechanisms could
be linked to social support. When integrating these constructs into a structural model,
however, it was found that these constructs were best represented by a reduction to
"internalizing" and "externalizing" dimensions. Using these dimensions a hierarchical
structural model was developed that was found to represent the observed data
adequately. This model suggests that interpersonal processes are influenced by the
characteristic use of defense mechanisms, which are themselves influenced by
underlying personality traits; thus, as predicted, it provides support for a hierarchical

relationship.

APPENDICES

APPENDIX A

124
APPENDIX A

Informed Consent Form

This research project is being undertaken to understand the relationship between
several different aspects of personality and human behavior. The surveys you will
complete are commonly used in psychological research. Following completion of the
surveys, you will be fully informed about the exact details of this project and any
questions you have will be answered.

Your participation is completely voluntary and you may withdraw your agreement to
participate at any time without penalty. If you choose to participate, you will be
asked to complete three surveys and a brief demographic profile. These surveys will
take approximately two hours to complete.

Since you will be awarded research credits in return for participation, total anonymity
can not be provided; however, all answers you provide to the surveys are anonymous
and confidential. That is, your name will not appear in the same place as any survey
results. However, because the answers to surveys will be anonymous, no personal

information about results can be given to you. All information you provide is strictly
confidential, and you not will not be identified in any report of the research findings.

By completing and returning these questionnaires, you are indicating your
voluntary agreement to participate in this project.

APPENDIX B

125
APPENDIX B

DEMOGRAPHIC QUESTIONNAIRE

SEX: _______ Female
_ Male
AGE: _
EDUCATIONAL LEVEL: __ Freshman
_ SOphomore
__ Junior
_ Senior
_ Graduate/ Returning Student
MARITAL STATUS: __ Single

_ Married
_ Divorced or Separated

_ Widowed

ETHNIC BACKGROUND:

 

APPENDIX C

126
APPENDIX C

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE NEUROTICISM SCALE

 

OF THE NEO-PI-R
ITEM N E O A C REMOVED?
1 35 ~3 -1 6 ~1
6 45 ~14 ~2 ~21 ~12
11 59 ~10 13 ~4 ~14 YES
16 47 ~26 ~17 10 ~10
21 26 ll 26 2 ~13 YES
26 40 ~26 ~14 ~9 ~36
31 40 1 ~10 11 ~12
36 36 ~6 0 ~41 ~22
41 59 ~26 -8 ~1 ~32
46 41 ~10 6 16 ~14
51 50 6 8 ~6 ~22
56 27 ~21 ~17 ~14 ~28
61 44 ~2 5 9 ~14
66 27 ~15 ~16 ~51 ~12
71 50 ~6 20 3 ~17
76 47 ~10 2 8 ~15
81 24 ~3 15 22 ~16 YES
86 53 10 10 6 ~19
91 48 ~8 ~7 ~9 ~17
96 26 ~12 ~5 ~35 ~12
101 38 3 1 5 ~16
106 31 ~1 ~2 11 1
111 39 27 29 18 ~7
116 37 ~25 ~12 ~10 ~29
121 29 7 22 2 ~13
126 25 ~19 ~10 ~45 ~13
131 40 -8 -5 ~1 ~20
136 52 ~23 ~16 4 ~27
141 13 20 31 5 ~8 YES
146 45 -8 6 8 ~36
151 44 3 ~1 11 ~3

156 40 -4 ~1 ~35 ~11

 

ITEM N E O A C REMOVED?
161 50 ~31 ~20 2 ~41

166 21 ~42 ~16 ~11 ~33 YES
171 36 ~13 0 ~8 ~25

176 30 ~25 ~21 ~16 ~39

181 39 ~13 -4 12 ~3

186 16 18 27 -3 7 YES
191 52 ~15 ~7 ~16 ~32

196 29 ~14 ~15 ~1 ~25

201 21 19 7 ~7 ~14

206 32 ~27 ~20 ~9 ~40 YES
211 31 14 21 21 3

216 35 4 12 ~16 4

221 61 ~26 ~8 ~12 ~49

226 19 9 4 9 0

231 28 12 18 ~6 ~16

236 41 ~26 -8 ~22 ~35

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX D

APPENDIX D

128

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS

WITH ITEMS COMPRISING
THE EXTRAVERSION SCALE

 

OF TIE NEO-PI-R
ITEM N E O A C REMOVED?
2 ~8 32 16 40 4 YES
7 ~26 53 19 -4 5
12 ~4 8 ~14 ~49 19 YES
17 O 14 22 0 ~25 YES
22 4 35 30 ~4 3 YES
27 l 36 28 22 6
32 ~3 47 31 28 15
37 ~6 49 7 6 -3
42 ~26 ~l ~1 1 ~13 20 YES
47 ~ 16 19 9 8 38 YES
52 5 22 17 5 9
57 ~1 47 35 19 12
62 ~8 46 26 46 23 YES
67 1 39 17 26 ~12
72 ~17 34 10 -4 14
77 ~18 27 12 9 17
82 5 35 40 2 -5 YES
87 ~22 48 32 40 12
92 ~20 40 27 57 14 YES
97 23 28 1 1 15 ~5
102 ~21 28 1 1 ~16 10
107 ~3 51 23 7 4
1 12 2 32 40 l 5 YES
1 17 ~2 58 38 15 1 1
122 ~6 56 31 25 9
127 -9 41 21 30 5
132 ~13 30 14 6 23
137 ~33 50 31 13 19
142 ~1 39 20 ~2 -4
147 -4 23 20 42 ~8 YES
152 ~ 19 47 23 17 24
157 ~5 11 ~22 ~17 ~2 YES

129

 

ITEM N E O A C REMOVED}
162 ~29 39 6 3 22
167 8 ll -2 2 13 YES
172 0 34 36 17 ~1 YES
177 ~12 63 25 24 16
182 5 37 45 25 17 YES
187 -1 53 27 23 14
192 -4 37 8 ~18 2
197 -6 23 11 0 10
202 -2 24 13 -3 0
207 ~14 28 30 10 16 YES
212 ~2 42 29 19 10
217 ~2 57 23 0 3
222 ~31 25 20 ~3 32 YES
227 ~7 62 33 9 17
232 3 40 7 14 7

___2_37 ~2 53 28 18 12

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX E

130
APPENDIX E

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE OPENNESS SCALE

 

OF THE NEO-PI-R
ITEM N E O A C REMOVED?
3 5 13 31 —8 ~5
8 ~5 1 1 45 4 2
l3 7 33 37 7 12
18 ~10 4 25 25 ~17
23 -6 1 37 ~13 ~3
28 5 22 37 12 4
33 l 1 25 40 4 ~ 19
38 12 23 42 9 6
43 14 35 44 18 1
48 ~6 29 32 20 21
53 -8 24 50 2 9
58 7 33 47 20 3
63 23 29 36 -1 ~14
68 7 23 41 24 3
73 l 31 35 20 18 YES
78 ~7 ~26 ~10 —9 ~19 YES
83 ~13 l 1 l7 -3 22 YES
88 ~8 12 23 ~1 1 ~15
93 19 22 39 18 ~19
98 ~5 13 55 14 6
103 7 31 46 20 9
108 ~12 13 14 7 14 YES
1 13 ~17 1 l 39 0 9
1 18 1 28 36 7 10
123 24 30 43 8 0
128 ~1 31 56 20 0
133 15 39 39 22 5 YES
138 ~20 2 11 -5 ~6 YES
143 ~17 3 20 3 10
148 ~3 0 28 14 ~5
153 21 10 20 -1 ~27 YES

158 16 26 44 10 4

 

ITEM N E O A C REMOVEQ?
163 6 24 54 19 9

168 -1 24 30 7 2 YES
173 ~3 19 49 16 7

178 ~5 25 39 38 1 1 YES
183 ~1 36 44 22 13 YES
188 0 21 56 16 1

193 3 32 30 28 22 YES
198 ~8 19 29 10 ~13

203 ~2 30 43 12 30

208 2 4 25 31 2 YES
213 5 17 48 16 ~2

218 0 1 1 36 5 - 10

223 9 29 49 1 1 ~2

228 ~19 7 15 9 —4 YES
233 ~8 33 45 2 33

238 -1 1 6 l3 ~3 -6 YES

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX F

APPENDIX F

132

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS

WITH ITEMS COMPRISING
THE AGREEABLENESS SCALE

OF THE NEO-PI-R

 

ITEM N E O A C REMOVED?
4 ~28 27 16 40 8
9 1 ~12 ~18 l9 ~l
14 ~17 15 7 56 11
19 4 ~2 11 41 4
24 5 ~11 ~5 35 ~11
29 I3 21 25 20 -1 YES
34 ~10 26 19 34 7
39 ~21 4 7 55 8
44 —4 31 23 46 16
49 ~13 ~25 ~36 15 ~5 YES
54 5 ~19 -3 21 ~8
59 ~22 2 11 28 ~12
64 ~26 3 3 30 ~1
69 ~2 ~10 ~20 21 1 YES
74 ~13 40 24 55 I8
79 5 ~14 ~7 34 ~10
84 16 5 8 44 ~7
89 I6 18 10 26 9
94 ~12 l7 7 39 14
99 -1 9 16 43 8
104 ~1 43 38 38 24 YES
109 9 ~l 2 37 ~8
114 1 ~6 3 37 10
119 4 19 19 23 12
124 ~26 27 24 44 13
129 0 38 33 28 23 YES
134 ~1 37 38 41 19 YES
139 ~16 1 -1 27 7
144 38 ~25 ~19 6 ~32 YES
149 4 16 18 29 14
154 ~17 16 11 37 10
159 ~20 ~5 ~2 47 14

133

 

ITEM N E O A C REMOVED?
164 ~14 46 32 26 24 YES
169 ~3 ~19 ~ 12 31 0

174 17 ~17 ~22 19 ~20

179 3 35 38 38 26 YES
184 -1 1 32 28 S4 1 1

189 ~20 -6 6 48 1 1

194 -2 33 36 40 22 YES
199 ~21 0 9 32 -3

204 5 ~12 2 l6 ~1

209 6 36 37 44 24 YES
214 ~8 22 16 33 10

219 5 ~12 ~6 33 ~1 1

224 1 34 30 42 29

229 ~26 25 16 26 31 YES
234 6 ~ 12 0 30 ~19

239 5 ~1 ~5 9 1 YES

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)

scales.

APPENDIX G

1 34
APPENDIX G

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE CONSCIENTIOUSNESS SCALE

 

OF THE NEO-PI-R
ITEM N E O A C REMOVED?
5 ~13 2 3 ~1 1 22
10 ~l 2 ~15 ~5 23
15 ~4 15 19 25 40
20 2 ~17 ~21 ~21 15 YES
25 ~27 3 ~15 ~3 55
30 ~27 ~8 ~13 5 22
35 ~2 26 33 16 18 YES
40 ~9 1 ~13 6 43
45 ~26 6 O 1 34
50 ~23 21 ~7 ~5 55
55 ~35 ~2 ~12 -4 39
60 -5 9 12 6 32
65 -9 20 23 5 33
70 -7 ~9 1 5 28
75 ~11 16 13 11 32
80 ~44 17 1 5 36
85 ~24 20 6 12 55
90 ~25 ~1 —4 3 28
95 ~38 1 0 8 42
100 -5 2 ~8 9 44
105 ~12 1 0 6 3 YES
1 10 ~16 31 22 21 55
115 ~54 12 0 16 54
120 ~24 ~8 ~6 3 42
125 ~17 17 15 ~3 34
130 ~28 15 6 12 59
135 ~ 12 29 17 18 46
140 ~20 28 17 14 41
145 ~25 27 10 12 51
150 ~13 ~30 ~24 ~l 8 YES
155 ~40 21 13 9 38
160 7 2 ~6 ~19 l 1 YES

135

 

ITEM N E O A C REMOVED?
165 ~5 0 ~10 4 17

170 ~15 32 19 10 48

175 ~38 27 20 16 41 YES
180 ~23 ~6 ~8 13 30

185 ~13 35 23 12 45

190 9 ~9 ~10 1 16 YES
195 ~10 24 18 15 41

200 ~20 30 19 19 53

205 ~29 5 ~1 1 1 1 40

210 ~3 14 ~l 12 43

215 ~10 34 28 18 44 YES
220 ~27 0 ~9 5 33

225 ~20 ~13 ~9 18 35

230 9 ~13 ~20 ~13 22

235 ~34 22 3 4 54

240 -4 ~12 ~9 4 19 YES

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX H

136
APPENDIX H

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED NEUROTICISM SCALE
OF THE NEO-PI-R

 

ITEM N E O A C
1 35 ~3 ~l 4 ~3
6 47 ~16 0 ~23 ~14

11 58 ~10 17 -8 ~17
16 48 ~25 ~15 9 ~11
26 43 ~23 ~12 ~7 ~37
31 42 6 ~10 9 ~13
36 36 ~3 2 ~43 ~25
41 60 ~23 ~5 0 ~32
46 41 ~10 7 17 ~17
51 47 7 10 ~11 ~22
56 28 ~19 ~17 ~13 ~29
61 43 0 7 7 ~15
66 29 ~13 ~13 ~52 ~14
71 49 ~6 23 0 ~19
76 46 ~11 5 6 ~16
86 53 8 11 3 ~19
91 47 ~9 ~5 ~10 ~16
96 27 -9 -5 ~36 ~13

101 38 4 3 1 ~16

106 30 ~2 ~1 9 0

116 37 ~23 ~12 ~8 ~30

121 26 8 22 1 ~12

126 26 ~18 ~8 ~47 ~12

131 42 -5 ~3 ~5 ~20

136 53 ~20 ~12 5 ~28 -

146 44 ~9 6 8 ~35

151 45 2 1 5 ~4

156 39 -l 1 ~38 ~13

161 51 ~30 ~19 4 ~41

171 34 ~12 2 ~6 ~25

176 30 ~21 ~19 ~12 ~41

181 37 ~11 ~2 12 ~4

137

 

ITEM N E O A C
191 53 ~14 -6 ~17 ~33
196 30 ~12 ~14 ~2 ~24
201 21 18 7 ~12 ~12
211 32 14 20 16 4
216 35 3 13 ~23 5
221 62 ~24 ~4 ~11 ~48
226 20 11 5 7 0
231 26 11 20 ~8 ~17
236 41 ~26 ~6 ~18 ~37

 

Note: n = 287. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX I

138
APPENDIX I

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED EXTRAVERSION SCALE

 

OF THE NEO-PI-R

ITEM N E O A C
7 ~28 54 18 ~7 6
27 0 38 27 19 5
32 -5 46 30 22 I4
37 ~6 49 7 3 ~2
52 4 18 14 2 1 1
57 4 44 35 14 13
67 0 40 16 23 ~1 1
72 ~17 38 10 ~1 1 14
77 ~19 27 11 7 18
87 ~24 48 29 37 14
97 23 30 1 1 1 1 ~5
102 ~23 30 13 ~19 9
107 ~2 52 24 2 4
1 l7 ~3 60 35 7 13
122 -8 55 29 17 10
127 ~1 l 40 19 27 5
132 ~14 28 14 0 22
137 ~33 49 31 8 20
142 ~2 35 15 -6 ~2
152 ~19 47 22 10 25
162 ~28 40 5 ~3 23
177 ~13 63 21 17 19
187 ~2 52 26 15 13
192 ~5 38 9 ~24 2
197 ~7 25 13 -4 1 1
202 ~2 24 9 ~8 1
212 ~3 40 28 12 1 1
217 ~3 57 22 ~6 7
227 ~7 62 32 l 18
232 4 41 5 11 9
237 ~2 5; 29 12 12

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX J

139
APPENDIX J

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED OPENNESS SCALE
OF THE NEO-PI-R

 

ITEM N E O A C
3 3 6 35 ~10 ~4
8 ~8 10 47 3 2
13 6 30 39 0 12
18 ~12 3 21 28 ~16
23 ~8 ~1 40 ~13 ~3
28 1 22 36 11 4
33 8 24 41 1 ~19
38 10 20 43 5 5
43 12 34 43 13 0
48 ~9 27 31 16 21
53 ~12 19 54 0 8
58 4 30 47 12 4
63 21 26 36 -8 ~13
68 6 23 41 19 2
88 ~10 13 21 ~13 ~15
93 16 18 38 17 ~20
98 ~7 9 56 10 5
103 3 26 45 14 8
113 ~19 9 41 —1 7
118 ~1 28 33 1 13
123 23 25 42 ~l 0
128 -4 31 56 15 ~1
143 ~17 2 l7 3 10
148 ~7 ~2 26 14 ~5
158 14 24 44 5 6
163 3 20 54 15 9
173 ~5 18 48 15 6
188 ~3 19 59 12 0
198 ~11 17 25 10 ~14
203 ~3 25 46 4 30

213 1 11 47 15 ~4

140

 

ITEM N E O A C
218 -2 1 1 37 2 ~9
223 8 27 48 3 ~l
233 ~9 29 44 -4 30

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)

scales.

APPENDIX K

141
APPENDIX K

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED AGREEABLENESS SCALE

 

OF THE NEO-PI-R

ITEM N E O A C
4 ~31 27 14 43 9
9 0 ~9 ~ 19 19 ~1
14 ~18 13 4 57 12
19 4 ~4 1 1 43 5
24 4 ~11 ~6 38 ~10
34 -9 23 18 29 9
39 ~22 5 4 58 9
44 ~7 26 20 41 17
54 5 ~21 ~4 23 ~7
59 ~23 2 8 32 ~1 1
64 ~26 5 1 35 ~3
74 ~15 36 22 53 19
79 6 ~16 -9 35 -9
84 14 2 7 45 -6
89 15 19 10 20 9
94 ~12 15 5 38 15
99 ~2 10 15 42 6
109 7 ~2 ~2 39 ~10
114 l ~7 0 36 10
119 0 19 17 19 12
124 ~28 24 23 43 12
139 ~14 3 ~1 29 9
149 3 13 15 23 14
154 ~16 15 9 37 10
159 ~20 ~7 —4 52 14
169 ~2 ~17 ~13 34 -1
174 18 ~17 ~22 23 ~19
184 ~12 30 25 52 12
189 ~21 ~8 3 51 12
199 ~22 1 5 35 ~3
204 6 ~12 4 l6 ~1

214 ~10 22 14 32 11

142

 

ITEM N E O A C
219 5 ~12 ~6 37 ~9
224 ~2 31 30 32 28
234 4 ~13 0 34 ~19

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)

scales.

APPENDIX L

143
APPENDIX L

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED CONSCIENTIOUSNESS SCALE

 

OF THE NEO-PI-R

ITEM N E O A C
5 ~12 1 3 ~ 15 23
10 2 4 ~16 ~8 18
15 ~4 13 19 20 39
25 ~25 4 ~17 4 57
30 ~25 ~7 ~14 8 19
40 ~7 ~1 ~15 5 40
45 ~25 2 -3 0 35
50 ~21 21 ~7 ~1 1 56
55 ~34 ~3 ~14 —4 39
60 ~5 6 12 5 34
65 ~8 19 23 0 37
70 ~9 ~11 1 4 26
75 ~12 16 13 7 34
80 ~44 16 1 4 35
85 ~24 20 3 6 54
90 ~24 ~5 ~5 2 26
95 ~39 ~1 ~1 6 40
100 —4 -1 ~7 7 43
110 ~17 28 19 13 57
115 ~53 11 -2 16 53
120 ~21 ~10 ~3 4 43
125 ~15 16 15 ~8 36
130 ~28 15 2 8 58
135 ~11 25 15 12 49
140 ~21 26 17 10 40
145 ~25 27 8 7 52
155 ~41 19 11 9 37
165 ~3 ~l ~8 2 17
170 ~14 30 21 5 48
180 ~20 ~9 ~7 15 31
185 ~14 32 22 4 46
195 ~10 23 18 11 43

144

 

ITEM N E O A C
200 ~18 26 18 1 1 52
205 ~29 5 ~ 12 12 41
210 ~3 14 ~2 6 44
220 ~26 1 ~13 4 33
225 ~20 ~14 -9 18 36
230 12 . ~1 1 ~20 ~ 16 20
235 ~33 18 1 1 55

 

Note: n = 287. Decimals removed. Abbreviations refer to the Neuroticism (N),
Extraversion (E), Openness (O), Agreeableness (A), and Conscientiousness (C)
scales.

APPENDIX M

145

AUEPETHDIXILI

RAW SCORES FOR THE REVISED NEO-PI-R SCALES,

 

BY GENDER

ID Scale N Scale E Scale 0 Scale A Scale C
FEMALE 1 87 90 83 100 108
2 111 89 84 70 73
3 108 107 81 55 89
9 102 64 72 105 78
15 96 73 100 66 93
17 91 102 90 81 105
18 101 78 84 70 80
19 99 75 78 80 62
22 89 98 76 87 73
23 91 76 73 73 94
29 74 104 97 102 110
33 69 95 101 64 114
34 81 88 96 96 85
36 86 66 78 65 77
40 102 76 78 86 91
41 61 87 103 60 61
42 129 78 91 76 71
43 92 83 81 71 99
44 83 114 90 88 101
45 84 102 101 92 81
47 110 109 99 79 57
48 73 92 92 84 119
50 103 87 91 54 92
52 86 90 117 86 84
55 59 101 73 105 111
56 104 76 107 . 97 103
57 94 55 74 84 97
60 92 80 69 91 63
62 124 88 79 65 34
63 100 86 91 92 72
64 77 89 79 84 99
65 105 106 75 68 55
66 79 69 62 89 86

146

 

ID Scale N Scale E Scale 0 Scale A Scale C
71 98 84 84 95 88
72 104 97 114 81 54
73 63 110 70 91 101
75 70 107 81 84 82
77 53 68 82 89 80
80 150 86 104 90 58
81 84 96 86 57 81
82 86 64 70 76 110
83 80 72 78 74 97
84 104 64 82 57 91
85 66 97 85 97 105
86 109 62 75 80 86
87 125 80 79 67 61
90 87 91 98 101 90
92 93 71 69 79 108
101 89 82 86 97 74
104 97 96 92 83 92
108 94 90 87 74 81
119 81 90 96 86 84
123 94 75 67 86 99
124 86 65 69 62 83
125 MISSING MISSING MISSING MISSING MISSING
126 80 85 79 81 92
127 87 96 79 101 128
129 76 92 102 88 94
130 76 77 69 87 95
131 131 104 94 95 82
133 82 104 92 95 97
134 114 64 85 75 103
135 102 92 110 80 84
137 114 80 102 98 74
138 81 74 87 83 96
139 91 81 72 87 96
140 101 94 111 75 64
146 91 68 97 100 102
147 81 76 103 91 97
151 101 59 76 77 92
153 84 76 67 66 78
154 85 98 92 84 104
155 120 116 119 85 93
156 30 109 118 34 142

147
ID Scale N Scale E Scale 0 Scale A Scale C

157 81 66 86 98 88
158 98 65 72 70 84
160 90 79 70 74 91
163 83 82 78 53 104
164 100 96 98 67 97
166 85 80 92 91 102
167 78 70 78 95 82
I72 92 78 96 64 80
173 105 72 75 88 89
174 80 102 91 97 81
175 66 82 96 77 71
178 118 112 90 39 72
179 70 101 92 85 99
181 87 54 85 77 88
182 112 47 91 78 85
183 120 66 87 72 61
190 84 76 73 67 85
191 114 54 67 76 111
192 99 64 79 78 89
194 104 90 85 85 91
195 95 75 83 63 101
198 66 92 83 104 105
200 110 95 94 87 91
201 75 94 75 88 81
205 63 94 93 107 104
206 106 83 76 75 93
207 92 90 109 93 61
208 104 63 119 76 63
210 107 86 84 74 75
211 109 89 101 87 85
214 124 88 55 33 105
215 130 61 114 62 70
216 82 87 100 74 80
217 121 71 95 93 65
221 99 76 74 68 95
225 85 62 73 61 81
227 118 68 86 89 84
228 83 74 60 63 77
229 128 83 68 56 55
230 111 94 81 85 85

231 132 75 70 85 88

148
ID Scale N Scal_e E Scale 0 Saale A Scalae C

 

232 78 70 60 77 83
233 115 79 100 74 88
234 142 36 90 64 67
235 80 60 65 81 101
239 84 83 104 95 69
242 90 68 74 67 86
243 67 90 89 74 92
246 84 92 90 70 136
247 68 88 110 69 84
254 67 102 87 89 102
255 78 114 88 89 109
259 112 85 97 97 77
260 97 94 84 102 88
261 71 51 51 72 96
262 95 65 78 103 86
263 68 85 103 98 92
264 80 83 78 67 78
266 78 55 58 90 111
267 80 95 70 99 75
268 84 68 62 77 70
269 69 73 71 77 80
270 104 69 110 78 58
271 112 75 77 69 64
272 61 87 90 85 60
274 97 102 110 93 54
275 69 91 86 84 96
276 73 92 78 95 87
280 93 87 91 58 95
281 67 77 84 72 105
283 68 94 86 115 108
285 85 65 73 79 84
286 86 70 75 82 84
288 80 94 80 99 89
291 66 88 109 104 103
IFEmAIULEIFII49 148 148 148 148 148
FEMALE NIEAN 91.5 82.6 85.7 80.8 87.1

FEMALE SD 19.8 15.2 14.1 14.6 17 .4

149

 

ID Saale N Scale E Sade O Scale A Scale C
MALE 4 64 93 88 97 105
5 66 89 93 63 84
6 69 74 66 65 92
7 72 80 81 89 90
8 74 61 63 83 66
10 102 84 84 73 84
11 76 86 70 82 110
12 91 64 68 71 61
13 54 92 97 50 110
14 81 62 69 69 59
16 82 63 100 80 83
20 80 66 59 70 94
21 75 78 82 58 97
24 80 73 63 70 92
25 76 70 70 78 87
26 48 77 68 84 108
27 99 77 106 57 100
28 78 99 83 76 94
30 60 94 103 93 126
31 66 99 95 77 104
32 47 92 111 108 72
35 84 69 70 62 86
37 88 68 66 72 78
38 74 76 75 73 76
39 71 79 69 71 86
46 88 81 99 70 79
49 84 62 70 81 73
51 79 67 86 55 106
53 67 92 87 89 96
54 103 92 112 73 66
58 77 64 76 91 79
59 88 68 97 73 121
67 47 96 87 104 94
68 44 67 103 74 113
69 105 73 95 92 67
70 105 63 70 71 76
74 112 83 91 71 64
76 76 82 83 74 101
78 99 64 52 99 63
79 88 83 94 76 87
88 77 75 112 94 67

150
II) Scak31J ScakaEi Scak:() Shane/A Scak:C3

89 75 65 64 75 84
91 85 62 69 75 82
93 96 73 128 87 68
94 97 65 66 70 67
95 82 67 65 75 81
96 80 58 57 68 77
97 61 80 93 59 86
98 109 80 94 58 74
100 105 84 78 90 84
102 93 68 70 79 81
103 75 64 100 99 85
105 92 99 78 94 123
106 130 70 121 47 28
107 83 74 89 48 103
109 74 69 118 60 110
110 84 64 72 79 82
111 80 66 68 70 78
112 66 57 62 103 75
113 70 103 73 67 106
114 90 69 92 84 74
115 89 66 67 77 79
116 105 60 66 56 90
117 90 71 75 72 79
118 90 54 59 77 72
120 94 101 106 89 77
121 75 90 87 48 86
122 91 85 72 83 93
128 86 74 77 68 72
132 82 75 89 77 75
136 83 67 75 63 70
141 62 90 98 95 88
142 82 55 87 78 87
143 92 73 67 70 60
144 82 68 63 69 91
145 76 84 124 101 80
148 91 72 77 66 78
149 76 78 110 68 104
150 82 59 61 72 85
159 70 82 77 56 109
161 107 76 89 86 104

162 91 59 87 70 82

151
ID Scale N Satle E Scale 0 Scale A Scale C

165 71 63 92 94 79
168 78 55 67 74 90
169 74 73 73 68 71
170 66 91 84 89 114
171 77 59 65 85 95
176 84 74 75 70 88
177 89 94 74 71 106
180 89 74 107 73 76
184 99 79 65 79 83
185 93 92 80 70 90
186 77 67 74 81 92
187 77 63 74 91 82
188 98 67 67 71 94
189 92 97 101 64 70
193 77 62 60 64 95
196 47 93 124 90 103
197 59 83 93 80 123
199 88 39 111 96 101
202 MISSING MISSING MISSING MISSING MISSING
203 96 66 66 99 85
204 69 97 91 81 130
209 98 76 76 93 122
212 65 71 70 84 106
213 83 70 69 76 67
218 98 74 94 72 82
219 85 73 65 83 80
220 91 80 73 50 97
222 71 59 71 71 103
223 95 82 66 60 82
224 84 77 75 71 80
226 107 89 64 74 52
236 50 86 101 90 89
237 78 79 87 113 111
238 80 63 73 79 77
240 112 36 105 119 81
241 84 65 77 69 86
244 66 108 102 42 90
245 83 63 76 68 81
248 65 99 86 90 108
249 68 74 68 89 75

250 94 53 65 70 85

152
ID Scale N Scale E Scale 0 Scale A Scale C

 

251 75 69 68 77 84
252 56 53 64 89 62
253 87 64 74 73 86
256 91 88 84 77 93
257 77 82 85 73 99
258 72 79 79 70 121
265 97 57 60 75 80
273 46 92 107 1 10 128
277 82 77 66 77 80
278 83 76 73 82 86
279 77 72 80 72 84
282 78 68 62 78 68
284 75 79 106 91 74
287 67 76 112 33 82
289 76 97 66 71 75
290 79 59 83 67 111
292 70 65 69 70 90
MALE N 140 139 139 139 139 139
MALE MEAN 81.0 74.6 81.5 76.4 87.3
MALE SD 14.9 13.2 16.6 14.2 17.6
TOTAL N 289 287 287 287 287 287
TOTAL NIEAN 86.4 78.7 83.7 78.7 87.1

TOTAL SD 18.3 15.4 16.0 14.9 17 .3

APPENDIX N

153
APPENDIX N

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE PRINCIPALIZATION SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS REMOVED?
5 29 ~1 l 5 ~22 ~3
10 21 ~3 ~2 ~13 ~5
14 28 ~38 26 ~25 3 YES
18 38 ~ 19 28 ~40 1
21 29 ~24 38 ~30 ~ 14 YES
30 ll ~5 7 ~15 1 YES
32 37 ~30 26 ~25 ~8
40 28 ~20 31 ~21 ~20 YES
43 20 ~14 ~6 ~9 0
50 27 ~5 1 ~16 ~8
54 27 ~10 21 ~22 ~14 YES
59 38 ~22 29 ~31 ~10
64 14 ~16 12 ~ 14 ~2 YES
68 31 ~3 17 ~23 ~17
74 29 ~34 15 ~31 16
78 27 ~35 18 ~26 9 YES
85 15 4 ~9 ~1 ~1 1
89 34 ~21 17 ~20 ~14
94 35 ~21 30 ~38 —2 YES
96 5 ~5 3 ~4 ~9 YES
103 28 ~24 9 ~22 6
106 34 ~19 26 ~38 1 YES
1 15 32 ~26 24 ~27 ~5 YES
1 18 18 ~15 23 ~29 5 YES
125 6 ~ 15 ~4 ~7 7 YES
129 18 ~6 1 1 ~22 ~3 YES
132 32 ~21 28 ~25 ~12 YES
139 28 ~27 23 ~16 ~ 14 YES
141 14 ~10 1 ~18 9
148 31 ~23 21 ~29 1
152 23 ~4 7 ~23 0

159 26 ~13 15 ~22 ~6

154
ITEM PRN PRO REV TAO TAS REMOVED?

162 11 ~15 11 ~17 5 YES
170 ~l 6 ~6 7 ~14 YES
172 12 ~25 11 ~11 2 YES
178 16 ~13 5 ~27 19 YES
182 5 ~7 ~9 ~2 3 YES
186 18 0 5 ~13 ~12
194 43 ~6 34 ~31 ~31
199 40 ~16 17 ~18 ~27

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX 0

155
APPENDIX 0

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE PROJECTION SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS REMOVED?
1 ~16 22 ~8 17 ~25
7 ~10 6 ~12 7 0 YES
15 7 12 -2 ~4 ~18
19 ~12 19 ~24 l 1 ~2 YES
22 ~10 22 ~17 18 ~24 YES
27 ~7 15 ~ 12 5 ~6
33 ~19 25 ~19 8 0
36 ~30 42 ~32 29 ~ 18
42 ~13 6 ~22 10 7 YES
47 ~10 10 -1 1 2 2 YES
52 ~25 23 ~19 11 0
56 ~14 9 ~31 15 5 YES
65 ~26 34 ~22 31 ~28 YES
69 ~14 31 ~7 7 ~19
71 ~15 25 —2 13 ~30
79 ~20 26 ~20 14 ~5
82 ~16 13 -9 14 ~11 YES
86 ~l l 31 ~21 12 ~14
93 ~15 24 ~14 13 ~13
99 ~17 13 ~25 14 2 YES
105 -6 6 ~2 8 ~17 YES
110 ~30 50 ~36 40 ~34
111 ~5 25 ~1 1 ~22
117 ~36 41 ~29 44 ~35 YES
123 ~9 17 ~8 8 ~13
126 ~1 23 ~14 3 ~12
135 ~12 ' 14 ~29 13 ~3 YES
137 0 22 ~15 ~12 6
144 ~15 15 ~13 9 ~5 YES
149 ~2 7 ~17 9 ~7 YES
155 ~22 21 ~20 1 1 ~3

156 ~15 12 ~26 28 ~19 YES

ITEM PRN PRO REV TAO TAS REMOVED?

 

161 4 l4 ~7 ~2 ~14
169 ~13 13 ~13 12 ~11 YES
173 9 -4 ~6 ~6 ~2 YES
179 32 ~20 15 ~25 ~4 YES
181 ~32 31 ~19 18 ~8
188 ~2 10 ~11 1 ~1 YES
193 ~17 23 ~5 6 ~10
200 ~26 32 ~32 30 ~17 YES

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX P

157
APPENDIX P

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING

 

THE REVERSAL SCALE
OF THE DMI
ITEM PRN PRO REV TAO TAS REMOVED?
2 1 1 ~26 21 ~30 19 YES
9 18 ~20 33 ~28 ~3
13 30 ~34 39 ~34 ~l
16 26 ~28 35 ~34 4 YES
23 6 ~8 5 ~4 ~7 YES
26 31 ~36 46 ~41 1
34 19 ~18 32 ~30 ~2
38 19 ~34 41 ~25 ~7
41 3 - 10 11 ~16 7 YES
48 17 ~l 1 36 ~26 ~ 14
53 17 ~21 22 ~22 0 YES
58 13 ~16 31 ~22 ~8
62 8 ~39 0 ~24 42 YES
66 29 ~33 38 ~39 6
73 32 ~38 19 ~32 16 YES
80 26 ~22 30 ~27 ~7 YES
83 6 ~12 15 ~17 0
87 ~2 ~25 32 ~19 4
91 40 ~35 32 ~37 -3 YES
98 19 ~28 28 ~21 -5
104 27 ~35 24 ~34 13 YES
107 8 ~43 30 ~23 15
1 12 27 ~34 36 ~30 ~5
120 39 ~44 31 ~41 10 YES
124 26 ~18 11 ~25 5 YES
128 28 ~16 30 ~27 ~13 YES
131 6 ~3 21 ~ 16 ~12
140 6 ~12 25 ~9 ~18
145 9 ~26 34 ~13 ~ 12
146 1 1 ~24 31 ~27 8
154 25 ~25 32 ~19 ~17 YES

157 9 ~29 36 ~19 ~1

158
ITEM PRN PRO REV TAO TAS REMOVED?

I64 ~10 15 6 10 ~29 YES
168 8 1 26 ~10 ~26
175 ~6 15 13 7 ~36 YES
176 ~25 20 2 19 ~29 YES
185 14 ~15 12 ~19 1 YES
189 26 ~28 38 ~27 ~12
192 6 ~13 20 ~17 3
196 11 ~41 40 ~27 10

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX Q

159
APPENDIX Q

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE TURNING AGAINST OTHERS SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS REMOVED?
3 ~32 4 ~25 36 -9
6 ~38 10 ~36 53 ~21
12 ~48 36 ~37 54 ~28
20 ~39 20 ~33 46 ~21
24 ~39 22 ~31 39 -1 l
28 ~22 24 ~36 43 ~31
31 ~38 35 ~31 51 ~39
37 ~22 14 ~34 28 ~4
44 ~16 1 1 ~13 21 ~23
49 ~23 7 ~30 24 0
51 ~29 14 ~38 32 ~3
60 ~24 22 ~33 34 ~18
61 ~14 19 ~4 20 ~35
67 ~53 27 ~40 58 ~21
75 ~45 30 ~32 44 ~19
77 ~18 27 ~12 30 ~41 YES
84 34 23 ~27 31 ~ 10 YES
88 29 23 ~41 37 ~12
92 ~41 32 ~44 46 ~15
97 ~09 12 ~ 15 07 ~04 YES
101 ~38 48 ~26 42 ~38 YES
108 ~19 15 ~28 28 ~13
1 13 ~43 32 ~49 54 ~16
1 16 ~16 8 ~16 l8 ~8
122 ~18 18 ~8 21 ~29 YES
130 ~36 13 ~37 41 -9
133 ~18 15 ~ 18 30 ~23
136 ~33 19 ~33 39 ~16
143 ~21 18 ~31 30 ~16
150 ~22 32 ~31 27 ~23 YES
151 ~23 16 ~31 30 ~10

158 ~24 35 ~27 20 ~14 YES

160
ITEM PRN PRO REV TAO TAS REMOVED?

163 ~24 10 ~16 30 ~19
167 ~5 ~17 ~10 2 1 1 YES
174 ~20 2 - 15 15 O
180 ~3 l 6 ~27 22 8
184 ~2 6 ~5 10 ~20 YES
190 ~43 24 ~42 46 ~12
195 ~43 28 ~43 50 ~16
198 i 21 ~18 24 ~17 YES

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX R

161
APPENDIX R

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE TURNING AGAINST SELF SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS REMOVED?
4 ~1 ~5 ~3 ~15 23
8 13 -7 15 ~39 24
11 ~ 15 -2 ~23 ~5 34
17 ~13 ~7 ~9 0 13 YES
25 1 ~23 ~5 ~25 44
29 ~17 ~ 12 ~7 ~6 27
35 4 ~30 -7 ~23 47
39 ~l ~13 ~8 ~19 34
45 2 ~13 28 ~24 5 YES
46 ~15 ~13 0 4 10 YES
55 9 ~19 15 ~17 7 YES
57 ~17 ~7 ~1 ~6 20
63 12 ~8 6 ~24 13 YES
70 10 ~31 —9 ~22 44
72 -7 ~3 ~8 ~4 14 YES
76 ~23 ~10 ~22 ~1 41
81 17 ~30 19 ~34 23 YES
90 ~6 ~16 0 ~15 29
95 ~25 ~15 ~8 0 30
100 ~14 ~3 ~4 0 10 YES
102 ~14 ~13 ~14 ~6 36
109 4 ~26 5 ~23 33
1 14 ~15 ~14 ~15 ~9 43
119 ~19 1 ~21 3 24
121 ~12 ~16 3 ~13 28 .
127 -9 ~20 8 ~13 26
134 ~14 ~20 ~11 ~5 39
138 -3 ~20 0 ~19 35
142 9 ~15 8 ~22 15 YES
147 ~18 ~16 ~2 2 17
153 ~6 ~23 3 ~12 26

160 ~7 ~20 -5 ~14 35

162

 

ITEM PRN PRO REV TAO TAS REMOVED?
165 6 ~31 ~3 ~23 40
166 3 ~13 ~5 ~20 30
171 ~6 ~2 ~1 1 ~13 27
177 3 ~8 ~6 ~2 4
183 8 ~36 13 ~19 22
187 ~1 ~ 15 3 ~24 34
191 3 ~37 ~6 ~17 41
197 - 19 ~2 ~20 ~8 40

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others

(TAO) and Turning Against Self (TAS) scales.

APPENDIX S

163
APPENDIX S

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED PRINCIPALIZATION SCALE

 

OF THE DMI

ITEM m PRO REV TAO TAS
5 25 ~13 0 ~23 ~3
10 20 ~6 ~2 ~13 ~4
18 30 ~10 17 ~40 2
32 28 ~25 23 ~27 ~12
43 23 ~10 1 ~9 0
50 34 ~6 -4 ~16 ~8
59 35 ~15 28 ~34 ~14
68 31 ~7 13 ~24 ~16
74 20 ~36 12 ~28 19
85 20 ~9 ~15 2 ~7
89 37 ~24 9 ~21 ~17
103 20 ~17 4 ~21 6
141 18 4 -4 ~19 9
148 36 ~17 19 ~27 2
152 36 0 5 ~24 ~3
159 25 —4 5 ~21 -9
186 22 4 2 ~14 ~14
194 47 3 28 ~34 ~33
199 4_3 ~10 9 ~17 ~26

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX T

164
APPENDIX T

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED PROJECTION SCALE

 

OF THE DMI
ITEM PRN 130 REV TAO TAS
1 ~19 25 0 10 ~24
15 9 l6 2 ~4 ~21
27 ~l 19 ~10 1 ~3
33 ~13 29 ~11 5 1
36 ~17 33 ~33 27 ~16
52 ~23 30 ~12 10 4
69 ~10 31 -9 2 ~17
71 ~7 33 1 6 ~28
79 ~13 23 ~16 12 ~5
86 ~11 31 ~21 10 ~11
93 -6 32 ~13 10 ~10
110 ~17 43 ~33 38 ~35
111 -1 36 ~2 1 ~22
123 ~5 18 ~5 7 ~10
126 ~2 19 ~13 2 ~9
137 9 14 ~17 ~11 8
155 ~23 9 ~23 8 ~l
161 14 16 ~4 ~3 ~14
181 ~29 37 ~13 17 ~5
193 ~24 27 ~3 3 ~15

 

Note: n = 27 6. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX U

165
APPENDIX U

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED REVERSAL SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS
9 9 4 30 ~28 -7
13 26 ~21 35 ~37 ~2
26 27 ~29 44 ~39 -3
34 12 ~13 32 ~32 -3
38 13 ~26 44 ~25 ~ 13
48 15 ~2 34 ~30 ~14
58 5 ~3 27 ~24 - 14
66 19 ~31 29 ~40 5
83 2 ~2 18 ~15 —1
87 ~16 ~13 39 ~17 -1
98 16 ~28 25 ~18 ~3
107 0 ~27 30 ~24 10
1 12 23 ~34 31 ~31 ~7
131 3 13 24 ~20 ~14
140 ~2 ~5 27 ~12 ~21
145 0 ~18 38 ~1 1 ~ 12
146 7 -8 32 ~27 1
157 2 ~17 38 ~17 ~4
168 8 6 24 ~13 ~29
189 22 ~19 40 ~27 ~ 12
192 3 ~8 22 ~18 1
196 1 ~35 49 ~24 6

 

N___ote: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX V

1 66
APPENDIX V

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED TURNING AGAINST OTHERS SCALE

 

OF THE DMI
ITEM PRN PRO REV TAO TAS
3 ~33 ~9 ~15 41 ~8
6 ~34 2 ~29 55 ~19
12 ~43 25 ~31 53 ~28
20 ~35 17 ~22 47 ~18
24 ~33 27 ~26 37 ~9
28 ~10 9 ~40 43 ~28
31 ~35 28 ~34 50 ~37
37 ~22 4 ~33 30 0
44 ~21 8 ~14 22 ~17
49 ~24 4 ~28 26 6
51 ~23 1 ~33 33 4
60 ~18 20 ~31 34 ~ 12
61 ~10 20 0 16 -29
67 ~48 21 ~28 58 ~20
75 ~40 24 ~23 43 ~ 17
88 ~25 5 ~40 37 ~6
92 ~37 18 ~37 42 ~ 12
108 ~1 I ~3 ~26 30 ~1 1
1 13 ~36 14 ~42 54 ~16
1 16 ~8 6 -8 17 ~8
130 ~23 0 ~37 42 ~5
133 ~20 8 ~22 30 ~21
136 ~22 7 ~32 37 ~12
143 ~26 15 ~30 27 ~12
151 ~24 13 ~22 28 -9
163 ~18 8 ~13 28 ~ 17
174 ~17 ~1 ~15 l7 ~1
180 ~32 2 ~20 26 1 1
190 ~47 18 ~41 44 ~ 12
195 ~39 7 ~43 51 ~13

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX W

167
APPENDIX W

RESULTS OF A CONFIRMATORY FACTOR ANALYSIS
WITH ITEMS COMPRISING
THE REVISED TURNING AGAINST SELF SCALE

 

OF THE DMI

ITEM PRN PRO REV TAO TAS
4 ~1 -4 ~2 ~15 23
8 l2 ~2 9 ~40 23
11 ~11 ~11 ~26 ~3 35
25 2 ~11 ~3 ~22 46
29 ~25 ~14 ~6 -4 26
35 3 ~40 ~12 ~19 46
39 -1 ~13 ~10 ~17 36
57 ~19 ~1 ~1 ~3 15
70 6 ~27 ~14 ~19 40
76 ~17 ~18 ~19 4 40
90 ~5 ~11 ~2 ~15 29
95 ~24 ~14 ~2 3 29
102 ~15 ~15 ~15 —4 37
109 2 ~25 1 ~22 36
114 ~10 ~16 ~19 ~5 46
119 ~15 ~5 ~16 5 27
121 ~7 ~13 5 ~10 28
127 ~9 ~5 10 ~15 23
134 ~10 ~20 ~13 ~2 37
138 ~4 ~18 ~2 ~17 35
147 ~26 -9 0 4 15
153 ~12 ~22 6 ~9 27
160 ~7 ~14 ~9 ~9 36
165 4 ~39 -3 ~20 38
166 5 ~13 ~6 ~17 30
171 ~6 4 ~3 ~15 25
187 6 ~11 3 ~22 34
191 -4 ~36 -7 ~14 43
197 ~17 1 ~27 ~4 41

 

Note: n = 276. Decimals removed. Abbreviations refer to the
Principalization (PRN), Projection (PRO), Reversal (REV), Turning Against Others
(TAO) and Turning Against Self (TAS) scales.

APPENDIX X

168
AJHHENHIEXIX

RAW SCORES FOR REVISED DMI SCALES,

 

BY GENDER

ID PRN PRO REV TAO TAS
FEMALE 1 22 14 16 42 34
2 20 20 11 42 33
3 17 23 8 45 32
9 19 18 24 42 27
15 19 26 16 32 33
17 21 23 22 33 17
18 25 22 22 36 19
19 26 20 14 33 31
22 28 21 17 26 30
23 21 20 10 38 33
29 23 18 26 22 29
33 15 17 24 34 28
34 17 17 24 37 30
36 23 19 25 28 26
4O 24 20 20 24 37
41 24 17 16 39 28
42 28 I8 12 32 33
43 26 21 26 27 25
44 27 26 17 25 28
45 24 12 29 28 33
47 19 21 7 44 3O
48 29 17 28 29 15
50 22 18 20 35 28
52 32 16 27 21 18
55 26 21 24 24 25
56 23 14 23 31 29
57 27 18 16 28 38
60 26 16 32 9 33
62 21 13 19 29 39
63 21 17 26 28 30
64 21 14 28 32 25
65 20 19 25 28 34
66 27 8 33 22 26

169

 

ID PRN PRO REV TAO TAS
71 25 14 28 24 33
72 23 21 21 30 36
73 19 20 15 36 29
75 29 18 21 33 29
77 23 24 20 28 26
80 20 17 20 38 27
81 24 14 23 30 33
82 24 16 27 30 20
83 MISSING MISSING MISSING MISSING MISSING
84 21 20 19 45 25
85 22 12 22 30 37
86 15 21 20 35 29
87 16 21 17 41 27
90 19 18 16 37 38
92 18 17 23 44 24
101 27 15 25 19 31
104 24 21 20 36 28
108 20 19 20 38 31
119 23 17 17 27 35
123 24 21 24 22 37
124 21 13 19 28 40
125 18 18 11 42 38
126 19 19 23 32 33
127 18 22 20 35 33
129 22 10 27 31 31
130 18 14 25 34 35
131 18 18 14 30 44
133 20 13 32 20 30
134 23 19 20 34 36
135 21 17 21 34 37
137 22 13 22 33 39
138 25 19 24 28 26
139 23 18 24 23 29
140 21 18 15 . 38 31
146 23 13 26 31 33
147 25 12 30 19 30
151 MISSING MISSING MISSING MISSING MISSING
153 19 19 21 37 27
154 19 25 11 48 17
155 18 13 22 36 29

156 30 19 16 34 24

170

 

ID PRN PRO REV TAO TAS
157 22 21 22 27 38
158 20 28 8 47 27
160 MISSING MISSING MISSING MISSING MISSING
163 30 19 17 32 23
164 24 20 13 37 29
166 26 12 13 40 28
167 26 17 25 26 26
172 26 16 28 20 24
173 33 16 24 23 29
174 21 16 28 24 34
175 28 21 25 27 24
178 18 21 10 41 34
179 28 13 22 26 24
181 24 9 24 40 28
182 16 19 14 42 34
183 18 14 16 39 34
190 20 23 16 44 19
191 20 14 23 24 45
192 19 18 18 46 30
194 MISSING MISSING MISSING MISSING MISSING
195 26 17 17 32 26
198 27 13 32 24 31
200 26 15 20 29 39
201 26 16 17 34 28
205 27 19 22 24 27
206 13 18 24 31 34
207 23 20 22 30 29
208 17 21 20 41 30
210 17 15 23 34 37
211 22 19 26 25 36
214 19 23 14 MISSING 34
215 18 24 13 37 32
216 27 17 30 29 16
217 26 14 17 23 38
221 18 23 21 39 28
225 18 19 24 36 24
227 16 22 19 36 30
228 MISSING MISSING MISSING MISSING MISSING
229 21 15 21 39 37
230 19 13 24 28 33

231 22 15 19 37 33

171

 

 

ID PRN PRO REV TAO TAS
232 25 18 17 28 28
233 19 21 22 34 30
234 23 18 10 36 37
235 21 19 18 37 28
239 22 12 22 22 47
242 18 17 21 38 31
243 18 26 16 47 22
246 23 19 29 27 25
247 23 21 25 23 27
254 27 13 19 32 28
255 28 12 22 32 26
259 14 21 19 40 30
260 18 15 19 27 45
261 26 12 16 30 36
262 26 21 25 15 34
263 30 13 28 25 24
264 27 19 12 32 27
266 25 15 24 26 27
267 23 15 25 25 35
268 16 25 21 44 21
269 24 17 24 23 33
270 21 20 15 37 34
271 18 15 23 26 38
272 18 25 16 47 16
274 30 20 18 17 40
275 30 20 18 17 40
276 24 13 22 34 29
280 21 19 15 39 28
281 22 23 21 31 22
283 20 16 26 34 20
285 27 21 19 27 29
286 27 17 21 27 23
288 24 18 16 28 38
291 31 17 21 22 24

FEMALE N 149 144 144 144 143 144
FEMALE NIEAN 22.5 17.8 20.6 31.7 30.2

FEMALE SD 4.1 3.7 5.3 7.5 6.1

172

 

ID PRN PRO REV TAO TAS
4 24 22 24 27 28
5 26 16 24 30 21
6 20 31 25 32 16
7 26 17 23 30 21
8 MISSING MISSING MISSING MISSING MISSING

10 19 16 23 40 24

11 22 16 31 28 32

12 22 25 25 22 27

13 24 27 17 25 24

14 23 20 18 40 20

16 26 18 22 21 29

20 28 16 24 23 25

21 26 20 26 30 14

24 26 22 24 39 16

25 16 26 22 34 . 24

26 26 15 19 32 32

27 22 19 23 28 24

28 13 18 12 53 24

30 MISSING MISSING MISSING MISSING MISSING

31 23 16 33 24 23

32 28 13 33 18 25

35 17 27 23 30 25

37 17 26 24 32 31

38 19 22 20 28 26

39 24 25 24 27 29

46 27 21 20 33 20

49 26 21 20 40 18

51 19 16 20 33 27

53 28 24 19 27 22

54 22 19 23 39 21

58 25 24 23 24 25

59 23 22 22 32 27

67 33 18 28 24 12

68 23 17 29 25 26

69 25 22 26 13 28

70 25 21 21 29 30

74 23 19 18 47 23

76 21 15 25 35 28

78 27 16 18 36 29

79 22 22 11 42 24

88 23 25 12 39 22

173

 

ID PRN PRO REV TAO TAS
89 22 24 23 29 18
91 28 23 21 36 18
93 22 20 20 38 25
94 25 20 20 36 9
95 18 23 22 38 26
96 20 28 27 29 24
97 26 23 21 30 21
98 19 24 I3 41 29
100 19 21 18 40 24
102 25 18 22 28 29
103 26 20 22 23 27
105 24 19 I7 33 32
106 21 24 14 40 24
107 23 21 18 36 28
109 23 23 21 40 15
110 24 19 22 35 25
111 20 23 17 35 29
112 19 21 29 16 34
113 25 21 19 38 23
114 21 20 36 22 15
115 23 23 19 21 33
116 26 22 15 28 31
117 19 23 6 47 27
118 29 22 21 24 22
120 25 22 15 31 29
121 22 17 14 47 20
122 16 25 21 34 32
128 25 29 21 29 12
132 23 21 20 32 23
136 27 23 19 33 17
141 27 13 33 17 26
142 23 21 17 43 25
143 21 24 17 34 30
144 MISSING MISSING MISSING MISSING MISSING
145 23 20 28 17 34
148 16 20 20 41 27
149 25 16 28 26 19
150 19 20 22 37 26
159 MISSING MISSING MISSING MISSING MISSING
161 27 15 27 16 34

162 18 19 19 43 22

174

 

ID PRN PRO REV TAO TAS
165 25 17 26 13 34
168 24 19 22 39 23
169 26 15 21 39 20
170 28 19 18 33 20
171 23 20 23 27 32
176 14 19 14 49 26
177 17 20 21 45 24
180 24 23 15 29 26
184 22 15 26 34 24
185 14 24 19 37 28
186 21 16 31 19 31
187 25 17 17 39 28
188 19 17 23 45 13
189 20 22 19 40 22
193 23 19 24 40 18
196 32 17 24 27 21
197 25 18 24 30 24
199 25 21 21 35 22
202 25 15 20 33 30
203 21 20 25 23 36
204 27 14 22 32 19
209 22 17 20 43 24
212 21 17 17 41 29
213 14 18 23 45 26
218 24 16 29 25 25
219 20 18 23 20 33
220 16 20 17 44 22
222 18 17 21 39 29
223 16 23 22 41 23
224 18 24 22 38 28
226 16 21 12 49 27
236 26 22 22 36 19
237 24 21 24 20 27
238 19 20 20 40 24
240 25 17 22 30 32
241 24 26 20 37 16
244 21 21 22 39 24
245 MISSING MISSING MISSING MISSING MISSING
248 28 20 20 29 22
249 24 20 25 34 26

250 20 21 18 37 34

175

 

ID PRN' PRO REV 'TAO TAS
251 17 21 2O 37 29
252 23 29 16 36 22
253 18 16 21 38 26
256 23 25 20 26 30
257 26 18 22 38 15
258 20 18 26 30 28
265 MISSING MISSING MISSING MISSING MISSING
273 21 13 19 37 35
277 17 22 30 33 29
278 21 17 20 36 23
279 22 20 22 28 35
282 17 20 20 40 25
284 29 17 19 28 35
287 20 21 22 45 16
289 16 23 14 45 20

290 MISSING MISSING MISSING MISSING MISSING
292 MISSING MISSING MISSING MISSING MISSING

MALE N 140 132 132 132 132 132
MALE MEAN 22.4 20.2 21.4 33.1 24.9
MALE SD 3.9 3.5 4.7 8.2 5.5
TOTAL N 289 276 276 27 6 275 27 6
TOTAL MEAN 22.5 18.9 20.9 32.4 27.7

TOTAL SD 4.8 4.0 5.3 7 .8 6.4

APPENDIX Y

176

AJHHENHIEK‘Y

RAW SCORES FOR NSSQ SCALES,

 

BY GENDER
Functional Structural Total Social
ID Support Support Support
FEMALE 1 201 82 16482
2 253 99 25047
3 266 105 27930
9 530 184 97520
15 365 140 51100
17 390 135 52650
18 254 90 22860
19 300 93 27900
22 158 68 10744
23 189 83 15687
29 333 110 36630
33 220 71 15620
34 210 79 16590
36 335 121 40535
40 261 95 24795
41 270 100 27000
42 222 82 18204
43 167 65 10855
44 526 173 90998
45 396 145 57420
47 275 96 26400
48 294 109 32046
50 279 92 25668
52 389 146 56794
55 313 120 37560
56 309 99 30591
57 107 33 3531
60 151 44 6644
62 290 104 30160
63 152 53 8056
64 211 80 16880
65 439 149 65411

177

 

Functional Structural Total Social
ID Support Support Support
66 291 96 27936
71 481 166 79846
72 486 156 75816
73 291 110 32010
75 434 153 66402
77 173 60 10380
80 231 77 17787
81 209 80 16720
82 387 135 52245
83 149 77 11473
84 419 181 75839
85 338 118 39884
86 157 62 9734
87 251 96 24096
90 317 106 33602
92 236 94 22184
101 196 59 11564
104 280 86 24080
108 209 87 18183
119 276 97 26772
123 160 53 8480
124 546 177 96642
125 346 141 48786
126 282 108 30456
127 398 147 58506
129 282 89 25098
130 228 70 15960
131 187 66 12342
133 319 116 37004
134 232 90 20880
135 211 83 17513
137 304 122 37088
138 163 64 10432
139 183 66 12078
140 274 99 27126
146 179 75 13425
147 308 123 37884
151 218 102 22236
153 328 129 42312
154 359 123 44157

178

 

Functional Structural Total Social
ID Support Support Support
155 364 120 43680
156 574 181 103894
157 249 97 24153
158 161 60 9660
160 574 199 114226
163 150 56 8400
164 369 128 47232
166 175 59 10325
167 214 70 14980
172 323 109 35207
173 162 66 10692
174 349 130 45370
175 259 87 22533
178 365 132 48180
179 365 131 47815
181 142 55 7810
182 212 86 18232
183 222 84 18648
190 374 125 46750
191 222 74 16428
192 267 84 22428
194 201 88 17688
195 284 115 32660
198 326 107 34882
200 447 173 77331
201 414 153 63342
205 505 191 96455
206 407 172 70004
207 340 117 39780
208 320 111 35520
210 218 76 16568
211 406 155 62930
214 470 149 70030
215 281 98 27538
216 308 106 32648
217 388 155 60140
221 417 178 74226
225 188 83 15604
227 344 124 42656
228 408 127 51816

179

 

 

Functional Structural Total Social

ID Support Support Support
229 263 124 32612
230 221 88 19448
231 301 108 32508
232 292 98 28616
233 247 80 19760
234 165 53 8745
235 185 74 13690
239 374 141 52734
242 422 137 57814
243 284 94 26696
246 487 175 85225
247 300 124 37200
254 347 120 41640
255 304 102 31008
259 443 179 79297
260 410 136 55760
261 337 127 42799
262 396 146 57816
263 238 94 22372
264 313 125 39125
266 53 22 1166
267 202 77 15554
268 291 109 31719
269 444 172 76368
270 230 83 19090
271 347 115 39905
272 226 79 17854
274 427 179 76433
275 272 101 27472
276 158 63 9954
280 132 50 6600
281 580 199 115420
283 287 101 28987
285 456 185 84360
286 490 172 84280
288 164 60 9840
291 166 53 8798
FEMALE N 149 149 149 149
FEMALE MEAN 298 108 36231
FEMALE SD 108 38 25074

180

 

Functional Structural Total Social

ID Support Support Support
MALE 4 219 73 15987
5 143 65 9295
6 239 83 19837
7 334 157 52438
8 215 95 20425
10 285 99 28215
11 262 101 26462
12 136 62 8432
13 309 114 35226
14 489 161 78729
16 258 87 22446
20 151 53 8003
21 196 77 15092
24 150 47 7050
25 118 41 4838
26 237 101 23937
27 228 82 18696
28 252 98 24696
30 289 92 26588
31 207 72 14904
32 144 49 7056
35 198 84 16632
37 86 28 2408
38 100 44 4400
39 133 50 6650
46 248 92 22816
49 501 187 93687
51 193 102 19686
53 294 111 32634
54 295 110 32450
58 122 44 5368
59 284 117 33228
67 228 85 19380
68 125 50 6250
69 355 142 50410
70 232 96 22272
74 192 67 12864
76 324 91 29484
78 268 97 25996
79 253 85 21505

181

 

Functional Structural Total Social
ID Support Support Support
88 427 142 60634
89 227 100 22700
91 438 154 67452
93 313 115 35995
94 106 33 3498
95 137 50 6850
96 159 64 10176
97 421 161 67781
98 214 99 21186
100 194 75 14550
102 132 62 8184
103 223 102 22746
105 131 45 5895
106 250 131 32750
107 85 30 2550
109 131 49 6419
110 198 93 18414
111 476 190 90440
112 185 92 17020
113 227 98 22246
114 193 97 18721
115 242 89 21538
116 206 81 16686
117 327 130 42510
118 250 81 20250
120 234 86 20124
121 261 98 25578
122 226 79 17854
128 173 64 11072
132 185 86 15910
136 258 112 28896
141 465 183 85095
142 106 56 5936
143 310 110 34100
144 403 177 71331
145 554 163 90302
148 242 83 20086
149 267 75 20025
150 364 163 59332
159 193 71 13703

182

 

Functional Structural Total Social
ID Support Support Support
161 266 93 24738
162 214 75 16050
165 228 72 16416
168 214 77 16478
169 180 79 14220
170 319 94 29986
171 224 97 21728
176 263 128 33664
177 186 66 12276
180 278 106 29468
184 232 80 18560
185 178 65 11570
186 231 91 21021
187 283 110 31130
188 357 105 37485
189 417 124 51708
193 166 60 9960
196 456 164 74784
197 333 101 33633
199 165 70 11550
202 186 67 12462
203 388 151 58588
204 279 91 25389
209 238 84 19992
212 164 66 10824
213 354 118 41772
218 350 133 46550
219 216 86 18576
220 406 145 58870
222 117 38 4446
223 52 18 936
224 485 167 80995
226 275 103 . 28325
236 457 156 71292
237 439 128 56192
238 344 167 57448
240 289 105 30345
241 513 166 85158
244 315 108 34020
245 245 91 22295

 

 

Functional Structural Total Social

ID Support Support Support

248 526 179 94154

249 329 109 35861

250 170 70 1 1900

251 200 66 13200

252 108 42 4536

253 252 99 24948

256 324 118 38232

257 227 90 20430

258 321 98 31458

265 27 9 1 1 1 30969

273 296 87 25752

277 158 63 9954

278 MISSING MISSING MISSING

279 191 76 14516

282 211 87 18357

284 230 89 20470

287 244 81 19764

289 85 40 3400

290 373 131 48863

292 472 174 82128

MALE N 139 139 139
MALE MEAN 257 96 28294
MALE SD 106 37 22198
TOTAL N 288 288 288
TOTAL MEAN 278 102 32400
TOTAL SD 109 38 24019

APPENDIX Z

1 84
APPENDIX Z

RAW SCORES FOR SCALES DERIVED
FROM SECOND ORDER FACTOR ANALYSES,

 

BY GENDER
NEO-PI-R/ NEO-PI—R/ DMI/ DMT/
ID Intemalizing Externalizing Intemalizing Externalizing
FEMALE 1 137 273 38 84
2 195 243 31 91
3 176 243 25 97
9 181 241 43 92
15 160 239 35 86
17 143 273 43 99
18 178 232 47 101
19 194 233 40 84
22 173 261 45 77
23 154 222 31 85
29 122 303 49 73
33 113 260 39 86
34 154 280 41 84
36 167 209 48 82
40 168 240 44 66
41 157 250 40 90
42 217 245 40 76
43 151 235 52 83
44 139 292 44 85
45 160 295 53 69
47 209 287 26 95
48 112 268 57 91
50 168 232 42 84
52 161 293 59 82
55 107 279 50 84
56 159 280 46 78
57 154 213 43 71
60 186 240 58 57
62 248 232 40 62
63 184 269 47 76

64 135 252 49 81

185

 

NEO-PI-R/ N EO-PI-R/ DMI/ DMI/
ID Intemalizing Externalizing Intemalizing Externalizing
65 206 249 45 74
66 153 220 60 69
71 169 263 53 66
72 208 292 44 76
73 119 271 34 89
75 145 272 50 81
77 130 239 43 86
80 250 280 40 90
81 160 239 47 73
82 135 210 51 87
83 141 224 MISSING MISSING
84 170 203 40 99
85 120 279 44 67
86 181 217 35 87
87 221 226 33 95
90 155 290 35 78
92 144 219 41 100
101 172 265 52 65
104 162 271 44 90
108 170 251 40 87
119 154 272 40 68
123 154 228 48 67
124 161 196 40 62
125 MISSING MISSING 29 83
126 147 245 42 81
127 118 276 38 82
129 140 282 49 72
130 139 233 43 72
131 206 293 32 64
133 142 291 52 67
134 170 224 43 74
135 176 282 42 75
137 197 280 44 68
138 142 244 49 85
139 152 240 47 73
140 194 280 36 85
146 146 265 49 71
147 141 270 55 63
151 165 212 MISSING MISSING

153 165 209 40 91

186

 

 

NEO-PI-R/ NEO-PI-R/ DMI/ DMI/
ID Intemalizing Externalizing Intemalizing Externalizing
154 138 274 30 119
155 183 320 40 82
156 48 261 46 91
157 152 250 44 69
158 173 207 28 110
160 157 223 MISSING MISSING
163 136 213 47 91
164 160 261 37 90
166 140 263 39 85
167 155 243 51 78
172 172 238 54 75
173 173 235 57 73
174 156 290 49 65
175 152 255 53 85
178 203 241 28 89
179 128 278 50 77
181 156 216 48 82
182 184 216 30 86
183 216 225 34 81
190 157 216 36 109
191 160 197 43 55
192 167 221 37 94
194 170 260 MISSING MISSING
195 152 221 43 86
198 117 279 59 70
200 176 276 46 68
201 151 257 43 84
205 115 294 49 78
206 170 234 37 76
207 188 292 45 81
208 198 258 37 93
210 189 244 40 72
211 183 277 48 67
214 175 176 33 MISSING
215 218 237 31 88
216 159 261 57 91
217 213 259 43 60
221 161 218 39 95
225 163 196 42 92
227 193 243 35 86

 

187

 

 

NEO-PI-R/ NEO-PI-R/ DNH/ DMI/

ID Intemalizing Externalizing Intemalizing Externalizing
228 163 197 MISSING MISSING
229 231 207 42 77
230 183 260 43 68
231 203 230 41 79
232 152 207 42 82
233 184 253 41 86
234 232 190 33 79
235 138 206 39 89
239 173 282 44 47
242 162 209 39 86
243 132 253 34 112
246 107 252 52 83
247 141 267 48 78
254 122 278 46 78
255 126 291 50 81
259 191 279 33 92
260 166 280 37 56
261 135 174 42 68
262 168 246 51 63
263 133 286 58 78
264 162 228 39 86
266 124 203 49 76
267 164 264 48 65
268 172 207 37 110
269 146 221 48 69
270 205 257 36 84
271 205 221 41 64
272 161 262 34 117
274 199 305 48 56
275 131 261 48 56
276 144 265 46 78
280 156 236 36 92
281 119 233 43 96
283 117 295 46 91
285 161 217 46 80
286 160 227 48 81
288 148 273 40 71
291 120 301 52 78
FEMALE N 148 148 144 143
FEMALE MEAN 162.3 248.7 43.1 81.0

FEMALE SD 29.9 30.6 7.2 13.3

 

OO\IO\Ut-§e

10

12
13
14
16
20
21
24
25
26
27
28
30
31
32
35
37
38
39

49
51
53
54
58
59
67
68
69
70
74
76
78
79

NEO-Pl-R/ NEO-PI-R/ DMI/ DMI/
Intemalizing Externalizing Intemalizing Externalizing
116 278 48 81
139 245 50 89
135 205 45 109
139 250 49 87
166 207 MISSING MISSING
175 241 42 95
123 238 53 76
187 203 47 82
102 239 41 90
180 200 41 101
156 243 48 72
143 195 52 75
136 218 52 100
146 206 50 107
147 218 38 98
98 229 45 78
156 240 45 86
141 258 25 109
93 290 NHSSING MISSING
118 271 56 80
131 311 61 70
156 201 40 94
168 206 41 90
157 224 39 86
142 219 48 83
167 250 47 97
171 213 46 104
131 208 39 83
128 268 47 91
196 277 45 98
157 231 48 85
125 238 45 86
112 287 61 94
89 244 52 80
195 260 51 69
186 204 46 82
204 245 41 104
134 239 46 85
194 215 45 83
158 253 33 100

188

189

 

NEO-PI-R/ NEO-PI-R/ DNH/ DMI/
ID Intemalizing Externalizing Intemalizing Externalizing
88 168 281 35 104
89 151 204 45 97
91 162 206 49 102
93 186 288 42 96
94 189 201 45 109
95 159 207 40 98
96 162 183 47 95
97 132 232 47 97
98 192 232 32 98
100 179 252 37 99
102 169 217 47 79
103 149 263 48 76
105 126 271 41 81
106 258 238 35 103
107 137 211 41 89
109 122 247 44 111
110 161 215 46 92
111 159 204 37 91
112 148 222 48 65
113 121 243 44 98
114 173 245 57 88
115 168 210 42 75
116 172 182 41 79
117 170 218 25 105
118 177 190 50 84
120 175 296 40 85
121 146 225 36 108
122 155 240 37 87
128 172 219 46 109
132 163 241 43 93
136 171 205 46 102
141 131 283 60 68
142 152 220 40 102
143 189 210 38 89
144 149 200 MISSING MISSING
145 152 309 51 63
148 170 215 36 94
149 129 256 53 89
150 157 192 41 94

159 1 18 215 MISSING MISSING

190

 

NEO-PI-R/ NEO-PI—R/ DMI/ DMI/
ID Intemalizing Externalizing Intemalizing Externalizing
161 160 251 54 59
162 166 216 37 101
165 151 249 51 59
168 145 196 46 97
169 161 214 47 98
170 109 264 46 94
171 139 209 46 76
176 154 219 28 103
177 139 239 38 103
180 170 254 39 85
184 172 223 48 87
185 159 242 33 95
186 145 222 52 66
187 153 228 42 89
188 160 205 42 112
189 178 262 39 102
193 140 186 47 105
196 100 307 56 88
197 93 256 49 85
199 144 246 46 97
202 MISSING MISSING 45 79
203 168 231 46 70
204 98 269 49 91
209 134 245 42 98
212 118 225 38 90
213 175 215 37 98
218 173 240 53 79
219 163 221 43 68
220 152 203 33 104
222 128 201 39 91
223 170 208 38 103
224 163 223 40 95
226 213 227 28 104
236 118 277 48 102
237 124 279 48 75
238 160 215 39 97
240 191 260 47 75
241 155 211 44 108
244 132 252 43 97

245 160 207 MISSING MISSING

191

 

 

NEO-PI-R/ NEO-PI-R/ DMI/ DMI/

ID Intemalizing Externalizing Intemalizing Externalizing

248 1 14 275 48 92

249 152 231 49 90

250 168 188 38 85

251 149 214 37 90

252 150 206 39 103

253 157 211 39 90

256 157 249 43 83

257 136 240 48 105

258 109 228 46 84

265 177 192 MISSING MISSING

273 77 309 40 76

277 159 220 47 86

278 154 231 41 93

279 152 224 44 75

282 169 208 37 96

284 158 276 48 72

287 142 221 42 1 1 1

289 158 234 30 l 1 1

290 128 209 MISSING MISSING

292 140 204 MISSING MISSING

MALE N 139 139 * 132 132
MALE MEAN 152.2 232.5 44.6 90. 1
MALE SD 27.2 28.8 7 .0 12.7
TOTAL N 287 287 276 275
TOTAL MEAN 157.3 241.3 43.6 85.7

TOTAL SD 29.2 31.0 7.3 13.2

LIST OF REFERENCES

 

192

LIST OF REFERENCES

Allport, G. (1937). Personalig: A psychological interpretation. New York: Holt.

American Psychiatric Association (1987). Diagpostic and statistical manual of
mental disorders; Third edition. Revised. Washington, DC: Publisher.

Anderson, T ., & Leitner, L. M. (1991). The relationship between defense
mechanisms and reported symptomatology in college females. Personaligy and
_Iinvidual Differenaas, 1_2, 967-969.

Bandura, A., & Walters, R. (1963). Social learning theog and parsonalig
development. New York: Holt, Rinehart, Winston.

Barrera, M. (1981). Social support in the adjustment of pregnant adolescents:
Assessment issues. In B. H. Gottlieb (Ed.), Social networks and social
support (pp. 69-96). Beverly Hills: Sage.

Barrera, M. (1986). Distinctions between social support concepts, measures, and
models. American Journal of Commqu Psychology, 15, 413-445.

Barrick, M. R., & Mount, M. K. (1991). The Big Five personality dimensions and
job performance: A meta-analysis. Personnel Psychology, _4_4, 1~26.

Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of
depression. New York: Guilford Press.

Benjamin, L. S. (1974). Structural analysis of social behavior. Psychological
Review, S1_, 392-425.

Ben-Porath, Y. S., & Waller, N. G. (1992). "Normal"personality inventories in
clinical assessment: General requirements and the potential for using the NEO
Personality Inventory. Psychological Assessment, _4_, 14-19.

 

 

193

Bergeman, C. S., Chipuer, H. M., Plomin, R., Pedersen, N. L.,McClearn, G. E.,
Nesselroade, J. R., Costa, P. T. Jr.,& McCrae, R. P. (1993). Genetic and
environmental effects on Openness To Experience, Agreeableness, and

Conscientiousness: An adoption/twin study. Journal of Personalig, Q,
159-179.

Bettelheim, B. (1982). Freud and man’s soul. New York: Vintage Books.

Bond, M. H., Nakazato, H., & Shiraishi, D. (1975).Universality and distinctiveness
in dimensions of Japanese person perception. Journal of Cross-Cultural
Psychology, E, 346-357.

Borkenau, P. (1988). The multiple classifications of acts and the Big Five factors of
personality. Journal of Research in Personaligy, _2_2_, 337-352.

Bradlee, P. M., & Emmons, R. A. (1992). Locating narcissism within the
interpersonal circumplex and the Five-Factor model. Personalig and
Iadividual Differences. 1;, 821-830.

Broadhead, W. E., Kaplan, B. H., James, S. A., Wagner, E. H., Schoenbach, V. J.,
Grimson, R., Heyden, S., Tibblin, G., & Gehlback, S. H. (1983). The
epidemiological evidence for a relationship between social support and health.

American Journal of Epidemiology, 11_7, 521-537.

Bruhn, J. G., & Phillips, B. U. (1984). Measuring social support: A synthesis of
current approaches. Journal of Behavioral Medicine, 1, 151-169.

Bumham, J. C. (1968). Historical background for the study of personality. In
E. F. Borgatta & W. W. Lambert (Eds), Handbook of firsonalig theog and
research. Chicago: Rand McNally.

 

Buss, D. M. (1992). Manipulation in close relationships: Five personality factors in
interactional context. Journal of Personalipg, @, 493-499.

Byers, P. H., & Mullis, M. R. (1987). Reliability and validity of the Norbeck
Social Support Questionnaire in psychiatric inpatients. Educational and
Psychological Measurement, 41, 445-448.

Campbell, D. T., & Stanley, J. C. (1966). Exmrimental and guasi-exparimental
desigg for research. Chicago: Rand McNally.

Carlson, R. (1971). Where is the person in personality research? PsychologicaL
Bulletin, 75, 203-219.

194

Cattell, R. B. (1943). The description of personality:Basic traits resolved into
clusters. Jouaal of Abnormal and Social Psychology, S8, 476-506.

Cattell, R. B. (1947). Confirmation and clarification of primary personality factors.
Psychometrika, L2, 197-220.

Cattell, R. B. (1957). Perscaality and motivation structure ad measurement.
New York: World Book.

Cobb, S. (1976). Social support as a moderator of life stress. Psychological
Medicine, 3_8_, 300-314.

Cohen, S., & Syme, S. (Eds) (1985). Social support and health. Orlando, FL:
Academic Press.

Cohen, S., & Wills, T. A. (1985). Stress, social support,and the buffering
hypothesis. Psychological Bulletin, a, 310-357.

Connell, C. M., & D’Augelli, A. R. (1990). The contribution of personality
characteristics to the relationship between social support and perceived
physical health. Health Psychology, 2, 192-207.

Cooper, C., & Kline, P. (1982). A validation of the Defense Mechanisms
Inventory. British Journal of Medical Psychology, SS, 209-214.

Costa, P. T., Jr. (1991). Clinical use of the five-factor model: An introduction.
Journal of Personality Assessment, S], 393—398.

Costa, P. T., Jr., Busch, C. M., Zonderman, A. B.,& McCrae, R. R. (1986).
Correlations of the MMPI factor scales with measures of the five factor model

of personality. Journal of Personaligy Assessment, _5_0, 640-650.

Costa, P. T., Jr., & McCrae, R. R. (1976). Age differences in personality
structure: a cluster analytic approach. Journal of Gerontology, _3__1_, 564-570.

Costa, P. T., Jr., & McCrae, R. R. (1980). Influence of extraversion and
neuroticism on subjective well-being: Happy and unhappy peOple. Journal of

Personalng and Social Psychology, S8, 668—678.

Costa, P. T., Jr., & McCrae, R. R. (1985). _T_h_e NEO Personality Inventogy.
Odessa, FL: Psychological Assessment Resources.

Costa, P. T., Jr., & McCrae, R. R. (1988a). From catalog to classification:
Murray’s needs and the five-factor model. Journal of Personalig and Social
Psychology, S5, 258-265.

 

195

Costa, P. T., Jr., & McCrae, R. R. (1988b). Personality in adulthood: A six—year
longitudinal study of self-reports and spouse ratings on the NEO Personality

Inventory. Journal of Personalig and Social Psychology, _5_4_, 583-863.

Costa, P. T., Jr., & McCrae, R. R. (1992). NEO-PI-R Professional Manual.
Odessa, FL: Psychological Assessment Resources.

Costner, H. L. (1989). The validity of conclusions in evaluation research: A
further development of Chen and Rossi’s theory-driven approach. Evaluation

and Program Planning, 1_2, 345-353.

Cramer, P. (1988). The Defense Mechanisms Inventory: A review of the research
and discussion of the scales. Journal of Personality Assessment, 5_2, 142-164.

Digman, J. M. (1988). Classical theories of_tr_ait organization and the Big Five
flctors of personality. Presented at the Annual Meeting of the American
Psychological Association, Atlanta.

Digman, J. M. (1989). Five robust trait dimensions: Development, stability and
utility. Journal of Personalig, fl, 195-214.

Digman, J. M. (1990). Personality structure: Emergence of the five—factor model.
Annual Review of Psychology, 4_1, 417-440.

Digman, J. M. & Inouye, J. (1986). Further specification of the five robust factors
of personality. Journal of Personality and Social Psychology, _5_Q, 116-123.

Dollinger, S. J ., & Orf, L. A. (1991). Personality and performance in
"personality": Conscientiousness and Openness. Journal of Research in
Personalig, SS, 276-284.

Dunkel-Schetter, C., Folkman, S., & Lazarus, R. S.(1987). Correlates of social
support receipt. Journal of Personalig and Social Psychology, fl, 71-80.

Dye, D. A. (1991). Construct validig of the Individual Achievement Record: Phase
H - A clarification of the underlying factors. (Technical Report PRD-91-04).

Washington DC: Office of Personnel Management.

Eaton, W. W. (1978). Life events, social supports, and psychiatric symptoms: A
reanalysis of the New Haven data. Journal of Health and Social Behavior, 1_9,
230-234.

Elliott, T. R., & Gramling, S. E. (1990). Personal assertiveness and the effects of

social support among college students. Journal of Counseling Psychology, SZ,
427-436.

196

Eysenck, H. J. (1960). Expariments in parsonalig. London: Routledge & Kegan
Paul.

Fagan, P. J., Wise, T. N., Schmidt, C. W., Ponticas, Y. (1991). A comparison
of five-factor personality dimensions in males with sexual dysfunction and
males with paraphilias. Journal of Personally Assessment, S2, 434-448.

Fisher, J. D., Silver, R. C., Chinsky, J. M., Goff, B., Klar, Y., & Zagieboylo, C.
(1989). Psychological effects of participation in a large group awareness
training. Jougal of Conaulting and Clinical Psychology, 17, 747-755.

Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings
from different sources. Jogal of Abnormal Social Psychology, _4__4, 329-344.

Florian, V., & Krulik, T. (1991). Loneliness and social support of mothers of
chronically ill children. Soaial Science and Medicine, S2, 1291-1296.

Freud, S. (1905). Three essays on the theory of sexuality. The Standard Edition.
London: Hogarth Press.

Freud, S. (1915). Instincts and their vicissitudes. The Standard Edition. London:
Hogarth Press.

Freud, S. (1920). Beyond the pleasure principle. The Standard Edition. London:
Hogarth Press.

Freud, S. (1923). The ego and the id. The Standard Edition. London: Hogarth
Press.

Freud, S. (1926). Inhibitions, symptoms, and anxiety. The Standard Edition.
London: Hogarth Press.

Freud, S. (1936). The problems of anxieg. New York: Norton.

Galinsky, M. D. (1971). Relationship among personality, defense, and academic
failure. Journal of Personality Assessment, SS, 359—363.

Gleser, G. C. , & Ihilevich, D. (1969). An objective instrument for measuring

defense mechanisms. Journal of Consulting and Clinical Psychology, 3,
51-60.

Gleser, G. C., & Hrilevich, D. (1979). Personality factors and ego defenses.
Academic Psychology Bulletin, 1, 171-179.

197

Goldberg, L. R. (1990). An "alternative description of personality": The big
five-factor structure. JournaI of Personalig and Social Psychology, _5_9,
1216-1229.

Goldberg, L. R. (1992). The development of markers of the Big-Five factor
structure. _PSychological Assessment, 5. 26—42.

Goldberg, L. R. (1981). Language and individual differences: The search for
universals in personality lexicons. In L. Wheeler (Ed.), Review of personalig
and social psychology (Vol. 2, pp. 141-166). Beverly Hills: Sage.

Goldberg, L. R. (1993). The structure of phenotypic personality traits. American
Psychologist, _4_8, 26-34.

Goplerud, E. N. (1980). Social support and stress during the first year of graduate
school. Professional Psychology, N, 283-290.

Gordon, N. G. (1979). Defense mechanism preference and dimensions of
psychopathology. Psychological Reports, 451, 188-190.

Graziano, W. G., & Ward, D. (1992). Probing the Big Five in adolescence:
Personality and adjustment during a developmental transition. Journal of
Personalig, a0, 425-439.

Greenberg, J. R., & Mitchell, S. A. (1983). Object relations in psychoanalytic
theogg. Cambridge: Harvard University Press.

Greenberg, R. P., & Fisher, S. (1984). Menstrual discomfort, psychological
defenses, and feminine identification. Journal of Personalig Assessment, 11S,
643-648.

Guilford, J. P. (1957). Personalig. New York: McGraw-Hill.

Guilford, J. P. (1975). Factors and factors of personality.Psychological Bulletin, Q,
802-814.

Guntrip, H. (1963). Four phases of psychodynamic theory.In Guntrip, H., Schizoid
phenomena, object relations and the self. Madison, CT: International
Universities Press, Inc.

Henderson, A. S. (1984). Interpreting the evidence on social support. Social
Psychiam, 1_9, 49-52.

 

 

198

House, J ., & Kahn, R. (1985). Measures and concepts of social support. In S.
Cohen & S. L. Syme (Eds), Social Support and Health. Orlando, FL:
Academic Press.

Hunter, J. E. (1987). Multiple dependent variables in program evaluation.
In M. M. Mark & R. L. Shotland (Eds), Multiple methods in program
evaluation: New directions for program evaluation Number 35. San
Francisco: Jossey-Bass.

 

Hunter, I. E., & Gerbing, D. W. (1982). Unidimensional measurement, second

order factor analysis, and causal models. Research in Organizational
Behavior, 4, 267-320.

Ihilevich, D., & Gleser, G. C. (1986). Defense mechanisms: Their classification
correlates and measurement with the Defease Mechanisms Inventog.
Owosso, NH: DMI Associates.

 

 

Ihilevich, D., & Gleser, G. C. (1991). Defenses in psychotherapy: The clinical
application of the Defense Mechanisms Inventog. Owosso, MI: DMI
Associates.

John, O. P. (1989). Towards a taxonomy of personality descriptors. In D. M. Buss

and N. Cantor (Eds), Personalig psychology: Recent trends and emerging
directions (pp. 261-271). New York: Springer-Verlag.

Jung, J. (1988). Toward a social psychology of social support. Basic and Applied
Social Psychology, S, 57-83.

Juni, S. (1992). Familial dyadic patterns in defenses and object relations.
Contem or Famil Thera , a, 259-268.

Kelly, G. A. (1955). The psychology of mrsonal constructs:A theog of parsonality.
New York: Norton.

Kessler, R. C., & McLeod, J. D. (1985). Social support and mental health in
community samples. In S. Cohen & S. L. Syme (Eds), Social support and
health (pp. 219-240). Orlando: Academic Press.

Kincaid, S. B., & Caldwell, R. A. (1991). Initiator status and family support
during marital separation: A test of an interaction hypothesis. American

Journal of Comqu Psychology, _12, 71-79.

Kincaid, S. B., & Caldwell, R. A. (in press). Marital separation: Cause, coping
and consequences. Journal of Divorce and Remarriage.

199

Klusman, L. E. (1982). The Defense Mechanisms Inventory as a predictor of
affective responses to threat. Journal of Clinical Psychology, fl, 151-155.

Lakey, B., & Cassady, P. B. (1990). Cognitive processes in perceived social
support. Journal of Personality aad Social Psychology, 5_9, 337-343.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal,and coping. New York:
Springer.

Leary, T. (1957). Interpersonal diagnosis of personalig. New York: Roland Press.

Leavy, R. (1983). Social support and psychological disorder:A review. Journal of
Comqu Psychology, N, 3-21.

Leong, F. T. L., & Dollinger, S. J. (1991). In D. J. Keyser& R. C. Sweetland
(Eds), Test critiques. Volume VHI (pp. 527-539). Austin, TX: Pro-Ed.

Levine, R. L., & Hunter, J. E. (1983). Regression methodology, correlation,
meta-analysis, confidence intervals, and reliability. Journal of Leisure
Research, 1S, 323-343.

Lin, N., & Ensel, W. M. (1984). Depression-mobility and its social etiology: The
role of life events and social support. Journal of Health aad Social Behavior,
a, 176-188.

Lorr, M., Youniss, R. P., & Kluth, C. (1992). The Interpersonal Style Inventory
and the five-factor model. Journal of Clinical Psychology, 4S, 202-206.

Maddi, S. R. (1980). Personalipt theories: A comparative evaluation. Homewood,
IL: Dorsey.

McAdams, D. P. (1992). The five-factor model in personality:A critical appraisal.
Journal of Personality, 6_0, 329-361.

McCrae, R. R. (1987). Creativity, divergent thinking,and openness to experience.
Journal of Personalig and Social Psychology, S_2_, 1258-1265.

McCrae, R. R., & Costa, P. T., Jr. (1984). Personality is transcontextual: A reply
to Veroff. Personality and Social Psychology Bulletin, m, 175-179.

McCrae, R. R., & Costa, P. T., Jr. (1985). Personality in adulthood. New York:
Guilford.

200

McCrae, R. R., & Costa, P. T., Jr. (1985). Comparison of EPI and psychoticism
scales with measures of the five-factor model of personality. Personality arm
Individual Differences, S, 587—597.

McCrae, R. R., & Costa, P. T., Jr. (1986). Clinical assessment can benefit from
recent advances in personality psychology. American Psychologist, £11.
1001-1003.

McCrae, R. R., & Costa, P. T., Jr. (1987). Validation of the five-factor model of
personality across instruments and observers. Journal of Personalig and

Social Psychology, S_2_, 81-90.

McCrae, R. R., & Costa, P. T., Jr. (1989a). The structure of interpersonal traits:
Wiggins’ circumplex and the five-factor model. Jouraal of Personality aria
Social Psychology, SS, 586-595.

McCrae, R. R., & Costa, P. T., Jr. (1989b). Reinterpreting the Myers-Briggs type
Indicator from the perspective of the five-factor model of personality. Journal

of Personalipz, fl, 1740.

 

McCrae, R. R., Costa, P. T ., Jr., & Busch, C. M. (1986). Evaluating
comprehensiveness in personality systems: The California Q-Set and the
five-factor model. Journal of Personalig, 21, 430—446.

McCrae, R. R., Costa, P. T., Jr., & Piedmont, R. L. (1993). Folk concepts,
natural language, and psychological constructs: The California Psychological
Inventory and the five-factor model. Journal of Personality, fl, 1~26.

McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model
and its applications. Journal of Personalig, Q, 175-215.

Minskoff, S., & Curtis, J. M. (1984). Comparison of defense mechanisms utilized
in perception by congenitally blind and sighted respondents. Psychological
Reports, SS, 228-230.

Mischel, W. (1968). Personality and assessment. New York: Wiley.

Mischel, W. (1973). On the interface of cognition and personality: Beyond the
person-situation debate. ‘American Psychologist, S5, 740-754.

Monroe, S. M., & Steiner, S. C. (1986). Social support and psychopathology:
Interrelations with preexisting disorder, stress and personality. Journal of

Abnormal Psychology, SS, 29—39.

201

Morelli, G., & Andrews, L. (1982). Cognitive irrationality and defensiveness.
Psychological Reports, S1 , 387-393.

Noam, G. G., & Recklitis, C. J. (1990). The relationship between defenses and
symptoms in adolescent psychopathology. Journal of Personalitv Assessment,
S5, 311-327.

Norbeck, J. S., Chaftez, L., Skodol-Wilson, H., & Weiss, S. J .(1991). Social
support needs of family caregivers of psychiatric patients from three age
groups. Nursing Research, 59, 208—213.

Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1981).The development of an
instrument to measure social support. Nursing Research, SQ, 264-269.

Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1983).The further development
of the Norbeck Social Support Questionnaire: Normative data and validity
testing. Nursing Research, 2, 4-9.

Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes:
Replicated factor Structure in peer nomination personality ratings. Journal of

Abnormal and Social Psychology, Q, 574-583.

Ostendorf, F. (1990). [Language and personality structurezToward the validation of
the five-factor model of personality] Regensburg: S. Roderer.

Peabody, D., & Goldberg, L. R. (1989). Some determinants of factor
representations of trait adjectives. Journal of Personality and Social
Psychology, S2, 552-567.

Pedersen, N. L., McClearn, G. E., Plomin, R., & Nesselroade,]. R. (1988).
Neuroticism, extraversion and related traits in adult twins reared apart and

reared together. Journal of Personality and Social Psychology, SS, 950-957.

Peglar, M., & Borgen, F. H. (1984). The defense mechanisms of coronary patients.
Journal of Clinical Psychology, L0, 669-679.

Phares, E. J. (1984). Introduction to mrsonaligg. Columbus, OH: Merrill
Publishing Co.

Piersma, H. L. (1986). The factor structure of the Millon Clinical Multiaxial
Inventory (MCMI) for psychiatric inpatients. Journal of Personalng
Assessment, SQ, 578-584.

202

Primomo, J., Yates, B. C., & Woods, N. F. (1990). Social support for women
during chronic illness: The relationship among sources and types to
adjustment. Research in Nursing and Health, Li, 153-161.

Quittner, A. L., Glueckauf, R. L., & Jackson, D. N. (1990).Chronic parenting
Stress: Moderating versus mediating effects of social support. Journal of
Personality and Social Psychology, _5_9, 1266-1278.

Rader, G. E., DeMoyne—Bekker, L., Brown, L., &Richardt, C. (1978).
Psychological correlates of unwanted pregnancy. Journal of Abnormal
Psychology, S1, 373-376.

Revelle, W. (1987). Personality and motivation; sources of inefficiency in
cognitive performance. Journal of Research in Personality, a, 436-452.

Rogers, C. R. (1961). On becoming a person: A therapist’s view of psychotherapy.
Boston: Houghton Mifflin.

 

Rotter, J. B. (1982). The deveIOpment and application of soaial learning theog:
Selected papers. New York: Praeger.

Sarason, I., Ganzer, N. J., & Singer, M. (1972). Effects of modeled self-disclosure
on the verbal behavior of persons differing in defensiveness. Journal of

Consulting and Clinical Psychology, 3, 483-490.

Sarason, 1., & Sarason, B. (1982). Concomitants of social support: Attitudes,
personality characteristics, and life experiences. Journal of Personalig, SQ,
331-344. .

Schill, T., & Bekker, D. (1976). Sexual responsivity and ego defenses. Journal of
Clinical Psychology, SS, 79-80.

Schill, T., Rader, G., Evans, R., and Segall, S. (1976).Defense preferences of
high- and low-hostility subjects. Journal of Consulting and Clinical
Psychology, g, 867.

Scholz, J. A. (1973). Defense styles in suicide attempters. Journal of Consulting
and Clinical Psychology, 4_1, 70-73.

Shapiro, D. (1965). Neurotic sales. New York: Basic Books.
Shapiro, D. (1981). Autonomy and rigid character. New York: Basic Books.

Shapiro, D. (1989). Psychotherapy of neurotic character. New York: Basic Books.

203
Skinner, B. F. (1938). The behavior of organisms. New York: Appleton.

 

Smith, T. W., & Williams, P. G. (1992). Personality and health: Advantages and
limitations of the five-factor model. Journal of Personalig, S_O, 395-423.

Tett, R. R, Jackson, D. N., & Rothstein, M. (1991).Personality measures as

predictors of job performance: A meta-analytic review. Personnel Psychology,
fl, 703-742.

Thurstone, L. L. (1934). The vectors of the mind. Psychological Review, 5;, 1-32.

Trapnell, P. D., & Wiggins, J. S. (1990). Extension of the Interpersonal Adjectives
Scale to include the Big Five dimensions of personality. Journal of Personality
and Social Psychology, S2, 781—790.

Trull, T. J. (1992). DSM-HI-R personality disorders and the five-factor model of
personality: An empirical comparison. Journal of Abnormal Psychology, 191,
553-560.

Vaillant, G. E. (1966). A twelve year follow-up of New York narcotic addicts.
IV: Some characteristics and determinants of abstinence. American Journal of
Psychiag, 1g, 573-584.

Vaillant, G. E. (1986). Empirical studies of ego mechanisms of defeng.
Washington DC: American Psychiatric Press, Inc.

Vickers, R. R., & Hervig, L. K. (1981). Comparison of three psychological
defense mechanism questionnaires. Journal of Personalig Assessment, 4S,
630-638.

Vickers, R. R., Hervig, L. K., Rahe, R. H., & Rosenman, R. H.(1981). Type A
behavior pattern and coping and defense. Psychosomatic Medicine, SS,
381-396.

Viney, L. L., & Menton, M. (1974). Defense mechanism preferences and the
expression of anxiety. Social Belayior and Personality, a, 50-55.

 

Vinokur, A., Schul, Y., & Caplan, R. (1987). Determinants of perceived social
support: Interpersonal transactions, personal outlook, and transient affective

states. Journal of Personality and Social Psychology, 5_3, 1137-1145.

Wade, J. B., Dougherty, L. M., Hart, R. P., & Cook, D. B. (1992). Patterns of
normal personality structure among chronic pain patients. Pain, fl, 37—43.

204

Waller, N. G., & Ben-Porath, Y. (1987). Is it time for clinical psychology to
embrace the five-factor model of personality? American Psychologist, S_2_,
887-889.

Watson, D., & Clark, L. A. (1992). On traits and temperament: General and
specific factors of emotional experience and their relation to the five-factor
model. Journal of Personality, @, 441-476.

Watson, J. B. (1930). Behaviorism. Chicago: University of Chicago Press.

Weissman, H. N ., Ritter, K., & Gordon, R. M. (1971). Reliability study of the
Defense Mechanisms Inventory. Psychological Reports, S2, 1237-1238.

Widiger, T. A., & Trull, T. J. (1992). Personality and psychopathology: An
application of the five-factor model. Journal of Personality, 11, 363-393.

Wiggins, J. S. (1968). Personality Structure. Annual Review of Psychology, 12,
293-350.

Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms:
The interpersonal domain. Journal of Personality and Social Psychology, fl,
395-412.

Wiggins, J. S. (1982). Circumplex models of interpersonal behavior in clinical
psychology. In Kendall, P. C., and Butcher, J. N. (Eds), Handbook of
Research Methods in Clinical Psychology. New York: Wiley.

Wiggins, J. S., & Pincus, A. L. (1992). Personality: Structure and assessment.
Annual Review of Pachology, 4_3, 473-504.

Wiggins, J. S., & Trapnell, P. D. (1992). Personalin structure: The return of the
Big Five. In Briggs, S. R., Hogan, R., Jones, W. H. (Eds), Handbook of
Personalig Psychology. Orlando, FL: Academic.

Wilcox, B. (1981). The role of social support in adjustment to marital disruption:
A network analysis. In B. Gottlieb (Ed.), Social networks and social support.
Beverly Hills: Sage.

Wilson, J. F. (1982). Recovery from surgery and scores on the Defense
Mechanisms Inventory. Journal of Personality Assessment, SS, 312-319.

Wise, T. N., Fagan, P. J., Schmidt, C. W., Ponticas, Y. (1991). Personality and
sexual functioning of transvestitic fetishists and other paraphliacs. Journal of
Nervous and Menal Diseaae, m, 694-698.

 

205

Wolfe, M. M. (1978). Social validity: The case for subjective measurement, or
how applied behavior analysis is finding its heart. Journal of Applied

Behavioral Analysis, _1_l, 203-214.

Yang, K. S., & Bond, M. H. (1990). Exploring implicit personality theories with
indigenous or imported constructs: The Chinese case. Journal of Personalig

and Social Psychology, 3, 1087-1095.

Zuckerman, M., Kuhlman, D. M., & Camac, C. (1988). What lies beyond E and
N? Factor analysis of scales believed to measure basic dimensions of

personality. Journal of Personality and Social Psychology, 141, 96-107.

Zuroff, D. C. (1986). Was Gordon Allport a trait theorist? Journal of Personalig

 

and Social Psychology, 5_1, 993-1000.