This is to certify that the

dissertation entitled

Splitting as a Characteristic of Perfectionism

presented by

David Brian Derr

has been accepted towards fulfillment
of the requirements for i

__J&L£L___m¥mem——Counseling Psychology i
l
1

W

ajor professor

Date z/Ziél

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

 

 

 

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

DATE DUE DATE DUE ' DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

SPLITTING AS A CHARACTERISTIC OF PERFECTIONISM
By

David Brian Derr

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Educational Psychology
and Special Education

2002

ABSTRACT
SPLITTING AS A CHARACTERISTIC OF PERFECTIONISM
BY

David B. Derr

Dichotomous thinking has long been associated with perfectionism without
empirical support. Splitting and dichotomous thinking are related constructs that
are believed to inﬂuence perfectionists cognitive and affective processes. The
purpose of the current study was to explore the relationship between
perfectionism and splitting. 264 male and female university students completed a
six part survey including the Almost Perfect Scale - Revised (APS—R), the Penn
State Worry Scale (PSWS), the Center of Epidemiologic Studies - Depression
scale (CES-D), and three versions of the Adjective Check List (ACL). Participants
completed ACLs describing their true, ideal, and projected self-perceptions. Raw
scores from each version of the ACL were converted into standard scores and
combined to create 8 personality scales. Splitting was estimated as the sum of
the eight scale scores consisting of the absolute difference of the scores between
(1.) the True-Self and Ideal-Self resulting in the dependent variable 'lntemal
Conﬂict Splitting” and (2.) the True-Self and the Projected-Self yielding the
dependent variable “Attribution Splitting.” Higher scores indicated a greater
tendency to split. A correlation matrix was created to test for a signiﬁcant
relationship between APS-R Discrepancy (maladaptive perfectionism [MP]),
APS-R High Standards (adaptive perfectionism [AP]), and the two forms of

splitting. A one way ANOVA, Tukey comparison, and KWANOVA were used to
detect order effects from six instrument combinations.

The results indicate that MP is signiﬁcantly associated with internal conﬂict
splitting, attribution splitting, anxiety, depression, and High Standards. AP is
signiﬁcantly negatively related to internal conﬂict splitting and Discrepancy, and
positively associated with anxiety. Anxiety and depression signiﬁcantly mediate
MP and internal conﬂict splitting, whereas only anxiety signiﬁcantly mediates MP
and attribution splitting. Results of an exploratory analysis between APS-R High
Standards and the Splitting types yielded a suppressor variable effect with
anxiety on AP and internal conﬂict splitting

The ﬁndings provide empirical support for the use of dichotomous thinking
by maladaptive perfectionists. The mediation effects of anxiety and depression
generally support historical research of perfectionism. An interesting ﬁnding is the
inverse relationship between adaptive perfectionism and internal conﬂict splitting.
The suppressor effect of anxiety on High Standards in relation to internal conﬂict
splitting was also a surprise. This study highlighted the effect of anxiety on MP
and AP and may suggest a greater focus on the affective aspects of

perfectionism for enhancing the effectiveness of treatment of perfectionism.

Copyright by
DAVID BRIAN DERR

2002

DEDICATION
This work is dedicated to my parents. To my mother, who instilled in me a
constant interest in learning all things new, and to my father, for passing along
his passion for creating. I would also would like to dedicate this dissertation to

perfectionists everywhere, who struggle to live a rich and full life without the ever-

present fear of failure.

ACKNOWLEDGMENTS

I would like to thank and acknowledge Ken Rice for his support, direction
and encouragement throughout this research. lam also grateful to Della
Faulkner for her much needed expertise in statistics, but mostly for her
friendship. Thank you to Matthew Goodman for his assistance in statistics.
Manchester, Inc. has my appreciation for its support of my work on this
dissertation.

I appreciate the support of my family and friends who encouraged and
supported me along this journey. I am especially grateful to my wonderful wife
Ellie, for her helpful collaboration, for sharing in my struggles and celebrating the
milestones, and for helping me to maintain my sanity during the course of this
project. Thank you to Dr. John Carter for suggesting to me that, “It’s only a
dissertation.” And ﬁnally, my appreciation goes to Drema, Aaron, Sepi and

Marcey, my fellow interns, whose feedback and support were invaluable.

vi

 

TABLE OF CONTENTS
LIST OF TABLES ............................................................................ v
LIST OF FIGURES ........................................................................... vi
CHAPTER 1
INTRODUCTION .............................................................................. 1
PROBLEM STATEMENT ................................................................... 8
PURPOSE ...................................................................................... 8
CHAPTER 2
A REVIEW OF RELATED LITERATURE ............................................... 10
Dimensions and Types of Perfectionism ............................................ 10
Perfectionism, Anxiety, and Depression ............................................ 15
Discrepancy Theory and Perfectionism ............................................. 18
Cognitive Implications of Perfectionism ............................................. 20
Splitting ...................................................................................... 22
Theories of Splitting .................................................................. 24
Elements of Splitting ................................................................. 27
Measurement of Splitting............ 29
Perfectionism and Splitting ............................................................. 31
Psychotherapy of Splits .................................................................. 32
Gestalt Therapy and the Origins of Splitting ................................... 34
Gestalt Therapy and the Mechanics of Splitting .............................. 34
Summary of Gestalt Therapy Theory on Perfectionism and Splitting... 38
RESEARCH QUESTIONS .................................................................. 40
HYPOTHESES ................................................................................ 42
CHAPTER 3
METHODOLOGY ............................................................................. 44
Participants ................................................................................. 44
Instruments ................................................................................. 45
Procedure ................................................................................... 52
Administration .............................................................................. 53
Analysis ...................................................................................... 54
Mediator Analysis ......................................................................... 56
CHAPTER 4
RESULTS ....................................................................................... 58
Preliminary Analysis ...................................................................... 58
Order Effect ................................................................................. 64

vii

TABLE OF CQN ! EQTS (antinggdl

Analysis ..................................................................................... 66
Primary Hypothesis .................................................................. 66
Secondary Hypothesis ............................................................... 67

lntemal Conﬂict Splitting ........................................................ 68

Attribution Splitting 70

Sobel Test of Mediation ......................................................... 71

Exploratory Analysis ..................................................................... 72

Summary of Findings ..................................................................... 75
CHAPTER 5

DISCUSSION .................................................................................. 76

Summary of Results ...................................................................... 76

Explanation for Findings ................................................................. 77
Maladaptive Perfectionism and Splitting ........................................ 77
Anxiety and Depression as Mediators ........................................... 79
Adaptive Perfectionism and Splitting ............................................. 80
Non-Signiﬁcant Variables ........................................................... 82

Comparison to Previous Studies ...................................................... 83

Implications ................................................................................. 86

Theoretical Implications ................................................................. 89

Study Limitations .......................................................................... 92
Issues Regarding lntemal Validity ................................................ 94
Issues Regarding External Validity ............................................... 95
Statistical Problems .................................................................. 96

Future Dimctions .......................................................................... 96

Summary of Results ...................................................................... 98

APPENDIX ...................................................................................... 100
REFERENCES ................................................................................. 109

viii

 

LIST OF TABLES

Table 1 . _ .
Adjective Check List (Abbrevrated) Description .......................................

Table 2
Descriptive Statistics .........................................................................

Table 3
Adjective Checklist True Self Comparison With Normative Population .........

Table 4
Correlation of Age, GPA, and Credits with Study Variables .......................

Table 5
Gender Means and Standard Deviations on Study Variables .....................

Table 6
Comparison of Class Level with Study Variables ....................................

Table 7
Means of Six Combinations of ACL Assessments Distributed to Participants

Table 8
Correlation Matrix of Perfectionism, Splitting, and Distress Variables ...........

Table 9
Mediation effect of anxiety and depression on maladaptive perfectionism

(Discrepancy): lntemal Conﬂict Splitting ................................................

Table 10
Mediation effect of anxiety and depression on maladaptive perfectionism

(Discrepancy):Attribution Splitting

Table 11
Mediation effect of anxiety and depression on maladaptive perfectionism

(High Standards): lntemal Conﬂict Splitting ..........................................

ix

48

60

61

62

63

54

65

67

69

71

74

 

LIST OF FIGURES

Figure 1
The relationship between discrepancy, splitting, and emotion

91

CHAPTER 1
“Have no fear of perfection. You'll never reach it.”
~Salvador Felipe Jacinto Dali~

“If a man should happen to reach perfection in this world, he would have to die
immediately to enjoy himself. ”

~Josh Billings~
“Striving to better, oft we mar what's well.
~William Shakespeare ~
Introduction
Perfectionism is a pervasive construct that has been blamed and heralded

for its effect on the world. There can be no question however, of its more
negative aspects. In studies from 1982 to the present, perfectionism has been
implicated in a wide range of psychological disturbances including eating
disorders (AxteIl & Newlon, 1993; Brouwers & Wiggum, 1993; Fairburn, Shafran
& Cooper, 1999; Slade, 1982), suicide and suicidal ideation (Hamilton &
Schweitzer, 2000; Hewitt, Newton, Flett & Callender, 1997; Hewitt, Flett &
Weber, 1994), social phobia (Antony, Purdon, & Swinson, 1998), anxiety (Alden,
Bieling, & Wallace, 1994; Blatt, 1995; Blatt, Quinlin, Pilkonis, & Shea, 1995; Flett,
Hewitt, & Dyck, 1989; Hill, McIntire, & Bacharach, 1997; Lynd-Stevenson &
Heame, 1999), worry (Juster, Heimberg, Frost, Holt, Mattia, & Faccenda, 1996,
Kawarmura, Hunt, Frost, & DiBartolo, 2001; Stcber & Joorman, 2001) and
obsessive-compulsive disorder (Broday, 1988; Frost, Steketee, Cohn, & Griess,

1994; Rheauma, Freeston, Dugas, Letarte & Ladouceur, 1995). Perfectionism

has been associated with depression (Blatt, 1995; Blatt et al., 1995; Hill et al.,
1997; Lynd-Stevenson & Heame, 1999), personality types such as Type A
behavior (Flett, Hewitt, Blankstein & Dynin, 1994), and personality disorders such
as obsessive compulsive personality disorder (Shafram & Mansell, 2001 ).
Perfectionism has been associated with procrastination, clinical symptoms of
psychopathology (Hill, Zrull, & Turiington, 1997), low self-esteem (Rice, Ashby, &
Slaney, 1998), and insomnia (Parker, 1997). Perfectionism has also been
implicated in a number of somatic diagnoses including migraine headaches
(Blatt, 1995; Brewerton & George, 1993), erectile dysfunction (Qualand, 1980),
Munchausen syndrome, irritable bowel syndrome, abdominal pain in children,
dysmorphOphobia, and ulcerative colitis (Pacht, 1984). More recent studies have
found that perfectionism plays a role in burnout of competitive tennis players
(Gould, Udry, Tuffy & Loehr, 1996), fatigue in career mothers (Michelson &
Burns, 1998), weariness in night shift workers (Magnusson, Nias, & White, 1996),
and increased severity of head pain in children and adolescents identiﬁed with
chronic headaches (Kowal & Pritchard, 1990).

Perfectionism has traditionally been labeled with a variety of attributes.
Hamachek (1978) described the “neurotic” and “normal” dimensions of
perfectionism. Perfectionism has been differentiated into “positive’ and “negative”
qualities (Slade & Owens, 1998) or “passive" and “active” (Lynd-Stevenson &
Heame, 1999). Other researchers have delineated perfectionistic tendencies into
“adaptive” beneﬁts and “maladaptive" deﬁcits (Slaney, Ashby, & Trippi, 1995;

Rice & DelIWO, 2002).

In their attempts to further explore and reﬁne their understanding of

maladaptive perfectionism, researchers have created a construct labeled

“Discrepancy” deﬁned as “...the perception that one consistently fails to meet the

high standards one has set for oneself” (Slaney, Rice, & Ashby, 2002, p. 69).
Slaney et al. found that discrepancy is positively associated with negative
psychological states including depression, anxiety, and distress. Positive
psychological states including achievement and self-esteem appear to be
negatively correlated with discrepancy.

lncongruency of internalized standards has also been the focus of
research in the area of self-discrepancy theory. Higgins (1987) and Higgins,
Tykocinsky, and Vookles (1990) propose that people are comprised of evaluative
psychic systems bounded by standards. Individuals evaluate themselves
according to self-imposed standards (e.g., evaluating a perception of self against
an idealized concept of self) or the introjected standards of others (e.g.,
evaluating a perceived self-concept against the injunctive evaluations of
signiﬁcant others). Several empirical studies appear to support the correlation
between discrepant self-standards and emotional distress (Hankin, Roberts, &
Gotlib, 1997; Higgins, Klein, & Strauman, 1985; Scott 8 O’Hara, 1993; Strauman,
1989)

Discrepancy, as deﬁned by Slaney et. al. (2001), appears to empirically
support the proposition by researchers and theorists that perfectionists may
strive to attain unattainable goals prompted by high personal standards. It is

thought that these unattainable goals manifest themselves as “shouldisms” in

 

perfectionists’ psyche, creating a pattern of divided or dichotomous thinking. For
example, perfectionists may earn a better than average score on an exam yet
feel disappointed for not having achieved their goal of earning a superior score.
These individuals tend to focus on what they “should” have earned and the
resulting disappointment at not having met their own high standards. The larger
the perceived gap or discrepancy created between actual achievement and
idealized goal, the greater the likelihood of internalized conﬂict.

Literature on self-discrepancy and perfectionism suggests that the greater
the difference between how individuals actually view themselves versus their
imagined ideal selves, the greater the propensity for emotional distress. Self-
discrepancy theory examines emotional ambiguity created by differences in the
perception of actual versus ideal selves. A study by Hankin et al. (1997) linked
the construct of perfectionism as described by Hewitt and Flett ( 1991a, 1991 b)
with self-discrepancy theory. Adolescents in the study were assessed on three
dimensions of perfectionism: self-oriented perfectionism (SOP) deﬁned as an
internal locus of high standards and motivation, socially prescribed perfectionism
(SPP) involving an external locus of high standards and expectations for oneself,
and other-oriented perfectionism (OOP) concerning unreasonable expectations
and high standards directed toward signiﬁcant relationships. The researchers
concluded that excessive internalized standards are associated with emotional
distress. Speciﬁcally, the study revealed that individuals with SOP and
actual/ideal discrepancies were uniquely prone to depressive symptoms.

Adolescents with SOP and actual/other discrepancies were given to symptoms of

 

anxiety. Participants with SPP were more likely to experience general emotional
distress than either anxiety or depression exclusively. These ﬁndings have also
been empirically supported in studies by Scott and O’Hara (1993) and Strauman
(1989). It is a thesis of this study that maladaptive perfectionism is associated
with negative emotions including anxiety and depression and discrepant states of
self-perception.

Splitting is a form of dichotomous thinking in which an individual identiﬁes
with one thought over another. Patterns of preferential splitting are seen as a root
of psychological difﬁculty. Individuals with perfectionistic tendencies may favor
the self-depreciating “’shoulds” created by expectations of high standards of
performance rather than ﬁnding satisfaction in simply doing well. The literature on
splitting indicates that individuals may utilize a defense mechanism of splitting
and dichotomous thinking in order to avoid, alleviate, and/or reduce emotional
distress created by discrepant views of self. Given this theoretical foundation,
investigation into the nature of splitting and its role in perpetuating dichotomous
thinking in perfectionists might offer further understanding into the mechanisms
that sustain perfectionism. Insight into these mechanisms may provide greater
understanding into psychotherapeutic interventions that may be administered to
diminish the effects of maladaptive perfectionism. It is therefore posited that
splitting is a psychological mechanism that is engaged when maladaptive
perfectionists experience dysphoria resulting from incongruent states of self-
perception.

Gestalt therapy is a psychotherapeutic orientation in which dichotomous

 

thinking and splitting are targeted as roots of pathology as well as vehicles for
change. Perls, Hefferline, and Goodman (1951, 1994) recognized the natural
creation of polarities within the individual as the inevitable process of
differentiation and Gestalt formation and destruction (the cycle of awareness and
satiation of needs). Without the smooth capitulation of the change process,
differentiation does not occur and rigid polarities may form. It is thought that
splitting has its origins in early development. Children are aware that they are
highly dependent upon their parents for primary needs of love and affection.
Some children may view their parental bond as fragile and the world as unsafe.
In order to survive, these children may feel that they must control themselves in
order to manipulate their environment. They may feel that an expression of anger
toward a parent may endanger their ability to obtain continued emotional support.
These children may decide to control and suppress their feelings in favor of
obtaining continued parental love and support. Dichotomous thinking or splitting
is the resulting intrapsychic mechanism that enables individuals to override their
emotions with rationalizations that are perceived necessary for survival. Splitting
may also have the effect of organizing feelings into polarities, such as “all good”
or “all bad”. This may enable the child to associate “all good” feelings with their
parents while diminishing “all bad” feelings that may cause anger and threaten
the child/parent bond. If the child feels that suppressing their feelings is
successful in getting what they require, they will continue to split as the need
arises. As the child matures, these polarities become patterns of behavior that

are reinforcing, thereby robbing the individual of the energy and resources

 

needed to engage in the demands and possibilities of the present (Fagan &
Shepherd, 1970).

Gestalt therapy offers a perspective in which both ends of a polar system
are accepted and preference of either is deemed appropriate in certain
situations. For example, Gestalt therapy views constructs such as love and hate
as valid emotions and appropriate in speciﬁc contexts. Healthy individuals are
seen as those who are able to integrate their experiences into a wide continuum
of polarities. Gestalt therapy offers a therapeutic approach in which to honor and
reify splits as a natural phenomenon rather than eradicate them as is the goal of
more cognitively oriented therapies. A further advantage of the Gestalt therapy
philos0phy of change is that it provides the practitioner with “experiments" or
activities that are speciﬁcally designed to ameliorate the adverse effects of
splitting.

A “two-chair” experiment is a therapeutic activity in which an individual is
instructed to role-play a polarity by enacting the two parts using two chairs facing
each other. Role-playing one of the ends of a polarity, the individual is instructed
to have a conversation with the opposite pole, as if it were actually in the facing
chair. The client is prompted to move back and forth between the roles as the
conversation develops. By encouraging the client to more fully experience the
oppositional poles, the opportunity for the individual to become more aware of
how the poles interact and struggle against each other increases. Catharsis may
occur when the individual is able to recognize how both poles may beneﬁt the

client, depending upon the context and situation in which the polarity is engaged.

 

It should be noted however, that Gestalt therapists believe that each person
experiences themselves and their environment in a way that is uniquely different
from others. Therefore, a Gestalt experiment that may be effective with one
individual may have no usefulness with another individual.

In total, Gestalt therapy may provide an appropriate theoretical ground
from which to better understand splitting mechanisms that drive perfectionism as
well as offer speciﬁc therapeutic methods for improving its maladaptive qualities.

Problem Statement

Most empirical research on perfectionism has focused on distilling
cognitive aspects that deﬁne the nature and dimensionality of the construct.
Perfectionism has been linked to three cognitive qualities of overgeneralization,
overly moralistic self-evaluation, and dichotomous thinking (Sorotzkin, 1985). To
date, no empirical investigation has validated these attributes. The literature
suggests that dichotomous thinking (splitting) is an especially important
characteristic to investigate. This construct is seen as a key mechanism to
pathology in many psychotherapeutic approaches. Empirical validation of a
correlation between perfectionism and dichotomous thinking may add to current
attempts to deﬁne the construct by increasing theoretical understanding of the
maladaptive mechanisms underlying perfectionism.

Purpose

The primary purpose of this study is to investigate the proposition by many

theorists and practitioners that individuals identiﬁed as “maladaptive”

perfectionists frequently engage in a deleterious pattern of dichotomous thinking

 

or splitting. If it can be shown that perfectionists have a tendency to dichotomize
or split their view of themselves (eg. “I am inferior (or average) and should
perform at a higher Ievel.”), and that splitting is a consistent cognitive/affective
pattern, then it follows that interventions aimed at integrating these splits could
potentially ameliorate poor self-concept, increase self-esteem, lower stress and
improve performance. Improvements in these areas of the psyche may serve to
reduce the harmful consequences of perfectionistic splitting.

A further purpose of this study is to investigate the underlying processes
by which maladaptive perfectionism and splitting are related. Past studies have
suggested a relationship between maladaptive perfectionism, affect, and splitting.
This study will explore the thesis that affect mediates maladaptive perfectionism
and splitting.

By virtue of it’s principal emphasis on phenomenological ﬁelds, polarities,
and their inﬂuence on cognitive/affective linkages, Gestalt therapy offers a
theoretical foundation that appears especially appropriate from which to examine
and understand the nature of dichotomous thinking and splitting. Identifying and
explicating the polarities of perfectionists may serve to illuminate underlying
intrapsychic schemes affecting these individuals’ ability to self-regulate affect and
cognition as well as determine ways in which to activate effective COpIng
mechanisms. Gestalt therapy provides a virtual elixir of theory and theory-derived
techniques from which to understand and treat splitting, and therefore splitting

associated with maladaptive perfectionism.

 

CHAPTER 2
A Review of Related Literature
This review ﬁrst examines perfectionism as deﬁned by typological,
theoretical, and characterological aspects. Three distortions of thinking that
plague perfectionists are explored, with a particular focus on dichotomous
thinking. Dichotomous thinking and a related construct, splitting, are described
and deﬁned. The relevance of self-discrepancy theory is examined in relation to
perfectionism and splitting. Instruments used in assessing splitting are reviewed
with a particular focus on the Adjective Check List (ACL). Splitting and its
theoretical relevance to perfectionism are described. Gestalt therapy and its
functional explanation of how splitting occurs within organismic ﬁelds is
illustrated. The role of polarities and the intrapsychic organization and dynamics
of Perls’ “Topdog” and “Underdog” system of a split phenomenological ﬁeld are
explored. A linkage is made between the Gestalt explanation of splitting and its
inﬂuence on the perfectionist’s drive to maintain validation and to ﬁnd external
support for a fragile self-concept. Finally, it is proposed that an attribution split
appears to have particular relevance explaining the maladaptive functioning of
perfectionists.
Dimensions and Types of Perfectionism
Prior to eariy 1970, perfectionism has largely been deﬁned by
characteristics found in anecdotal stories, case studies, and linkages to
traditional theory. Some of the more commonly cited characteristics of

perfectionists include being tenaciously driven by high standards (Burns, 1980;

10

 

Burns 8. Beck, 1978; Halgin & Leahy, 1989; Hamachek, 1978; Hollander, 1978;
Mahoney & Amoff, 1979; Pacht, 1984), excessive standards accompanied by
overly critical self-judgement (Frost, Marten, Lahart, & Rosenblate, 1990), a fear
of loss of external support (Burns, 1980; Hamachek, 1978), a heightened sense
of shame and guilt (Hamachek, 1978, Sorotzkin, 1985), a need to hide or
otherwise avert attention from mistakes (Hamachek, 1978), self-depreciating
behaviors (Hamachek, 1978), and excessive precision, orderliness, and
organization (Frost et al., 1990, Hollander, 1978). Burns (1980) also included
impaired health, intense sensitivity to the judgments of others, difﬁculty in
determining the difference between adequate effort versus overindulgence, and a
view that others are more capable of managing life affairs. Several cognitive
implications are also posited including all or nothing thinking (Beck, 1976; Burns,
1980; Burns & Beck, 1978; Halgin & Leahy, 1989; Mahoney & Amoff, 1979), a
tendency to overgeneralize setbacks into never-ending failures (Beck 1976;
Burns, 1980; Burns & Beck, 1978), and a rigid adherence to moral imperatives
(Beck, 1976; Burns 1980; Mahoney & Amoff, 1979; Weisinger & Lobsenz, 1981 ).
Because it is has been correlated with so many psychological
disturbances, perfectionism’s nature and characteristics appear somewhat
ambiguous and difﬁcult to deﬁne. Hollender (1978) deﬁned perfectionism as a
mode of behavior in which individuals demand a higher standard of performance
than is required in a given situation. Early deﬁnitions of perfectionism also
focused on an unremitting drive to achieve unattainable goals (Burns, 1980;

Halgin & Leahy, 1989). More recent deﬁnitions have stressed the pursuit of high

11

standards accompanied by unrelenting self-criticism (Frost & Marten, 1990; Hill,
Zrull, & Turlington, 1997). Hamachek (1978) differentiated perfectionism into
“normal” and “neurotic.” Normal perfectionism is viewed as a healthy aspect of
individuals who ﬁnd pleasure in the task of pursuing excellence and are not
negatively affected by its outcome. The self-esteem of normal perfectionists is
minimally affected when they do not live up to their expectations. Hamacheck
views normal perfectionism as a desirable and nonpathological component of a
self-actualizing tendency. In contrast, the neurotic perfectionist seeks to avoid
failure at all costs. The primary way in which this is accomplished is by
maintaining unrealistic and unattainable standards. High standards serve to
reinforce self-defeating cycles of perceived poor performance that lead to low
self-esteem and are exacerbated by each new challenge. The neurotic
perfectionist is typically externally driven to live up to the real or perceived .
standards of others. Burns (1980) described this type of perfectionism as a
striving to accomplish impossible goals. Neurotic perfectionism has also been
viewed as a destructive force that entrenches individuals in the destructive
pursuit of unrealizable goals leading to maladjustment and pathology (Pacht,
1984).

In recent years, attention has been given to deﬁning and measuring the
construct of perfectionism. Burns made an initial attempt at operationalizing
perfectionism with the Burns Perfectionism Scale (1980), which primarily
measures the degree to which perfectionists measure their performance against

stringent standards. Two self-report measures attempt to capture the

12

multidimensional qualities of perfectionism. The Multidimensional Perfectionism
Scale (Frost et al., 1990) focuses on six factors attributed to perfectionism:
concern over mistakes, doubts about actions, setting excessively high standards,
the perception of inordinately high parental expectations, the perception of
extreme parental criticism, and a high preference for order and organization. The
Multidimensional Perfectionism Scale (Hewitt & Flett, 1991a) deﬁnes
perfectionism into three facets: other-oriented in which the perfectionist demands
that others meet distorted standards of perfectionism; self-oriented in which
unrealistic standards of perfectionism are self-imposed; and socially prescribed in
which the perfectionist assumes that others maintain exaggerated expectations
that must be attained in order to win approval and acceptance. Shafran and
Mansell (2001) have noted that the two multidimensional scales are similar, but
not entirely alike. The personal standards scale (Frost et al., 1990) is most
closely similar to the self-oriented subscale. The concern over mistakes, parental
expectations, and parental criticism subscales of the Frost measure are similar to
socially prescribed perfectionism. Criticisms of these scales include the
subjectivity of self-reporting used by the instruments to assess perfectionism, the
retrospective versus state measure of perfectionism, and the inclusion of
dimensions that are related to perfectionism but not characteristic of the more
classical qualities of perfectionism (Shafran and Mansell, 2001). These authors
suggest that the Frost et al. (1990) subscales of personal standards and concern

over mistakes combined with the Hewitt & Flett (1991) self-oriented perfectionism

13

 

subscale are closer to traditional deﬁnitions of perfectionism than other
subscales.

A more recent instrument devised in identifying perfectionism is the
Almost Perfect Scale (APS; Slaney 8 Johnson, 1992). The APS and its
subscales have been used in subsequent research to extend its application and
toward reﬁning the construct of perfectionism. One apparent outcome of these
studies was the substantiation of the construct into negative and positive qualities
characterized as either “adaptive” or “maladaptive” (Slaney et al., 1999). Studies
using the APS have also supported the dimensional constructs of perfectionism
deﬁned by Hamachek (1978) as “normal” and “neurotic“ (Slaney, Suddarth, Rice,
Ashby, & Mobley, 1998). Research ﬁndings clearly link the holding of high
standards and a sense of orderliness with perfectionism. Neither of these
qualities either alone or together however, appears directly problematic for the
majority of perfectionists.

Research using the APS led to the development of the Almost Perfect
Scale-Revised (APS-R; Slaney, Rice, Mobley, Rippi, & Ashby, 2001). The
authors of the instrument explored the negative or maladaptive effects of
perfectionism and added a new subscale labeled "Discrepancy.” Discrepancy is
deﬁned as “...the perception that one consistently fails to meet the high
standards one has set for oneself” (Slaney et al., 2002, p. 69). Current studies
with the APS-R appear to validate the effects of this concept on measures of
negative psychological states such as depression, anxiety, and distress.

Discrepancy also appears to be negatively correlated with achievement and self-

14

esteem. Although these studies offer implications for use in clinical practice, no
speciﬁc adaptation of the ﬁndings has as yet been integrated into existing
psychotherapeutic frameworks for use in reducing or eliminating the maladaptive
effects of perfectionism.

Other researchers have conducted studies afﬁrming the dualistic and
multidimensional qualities of perfectionism. More recently, Slade and Owens
(1998) differentiated perfectionism into positive and negative forms of cognitions
and behaviors based on Ieaming theory. Positive perfectionism entails the
achievement of high-level goals to obtain positive reinforcement and
consequences. Fear of failure or negative reinforcement drives negative
perfectionism, deﬁned as striving to achieve high-level goals in an effort to avoid
unpleasant or negative consequences. This dual process model of perfectionism
is based on Skinner’s view of behavioral conditioning from which the authors
derive therapeutic implications for treatment.

It should be noted that in the remainder of this review, the use of the word
"perfectionism” is a shortened use of the term “maladaptive perfectionism" as
deﬁned above.

Perfectionism, Anxiety, and Depression

Perfectionism has been linked to anxiety (Alden et al., 1994; Flett et al.,
1989) and depression (Blatt, 1995; Blatt et al., 1995; Hill et al., 1997; Lynd-
Stevenson & Heame, 1999). Maladaptive perfectionism has been empirically
linked to trait anxiety in non-clinical samples (Flett, Hewitt, Endler, & Tassone,

1995) and psychiatric populations (Antony, Purdon, Huta, & Swinson, 1998).

15

After controlling for depression, maladaptive perfectionism has been signiﬁcantly
associated with social anxiety, trait anxiety, and worry (Juster, Heimberg, Frost,
Holt, Mattie, & Faccenda, 1996, Kawarmura, Hunt, Frost, & DiBartolo, 2001;
Stober & Joorman, 2001). Social anxiety has been linked to perfectionism in
clinical (Blankstein, Flett, Hewitt, & Eng, 1993; Hewitt & Flett, 1991) and non-
clinical (Antony et al., 1998; Juster, et al., 1996) samples. Obsessive-compulsive
disorder has also been identiﬁed with elevated levels of perfectionism in clinical
samples (Antony et at., 1998; Frost & Steketee, 1997) and non-clinical samples
(Frost et al., 1990). Perfectionism appears to be related to the cognitive, but not
somatic factors related to anxiety (Enns & Cox, 1999; Juster et al., 1996;
Kawarmura et al., 2001, Stober & Joorman, 2001).

A collection of studies has explored ﬁner distinctions between
perfectionism, anxiety, and depression. In a study by Kawamura et al. (2001),
maladaptive perfectionism was found to be linked to anxiety after controlling for
depression. Depression was positively related to perfectionism after controlling
for anxiety, but to a lesser degree than the anxiety/perfectionism relationship.
Further, depression was found to have a signiﬁcantly greater correlation with
maladaptive than adaptive perfectionism. Norman, Davies, Nicholson, Cortese
and Malla (1998) found similar results in their exploration of maladaptive
evaluation concerns and positive striving.

Worry and depression are signiﬁcantly related to perfectionism. Stbber
and Joorman (2001) found that worry exhibited a substantial relationship with

procrastination and perfectionism. Their study indicates that worry and

16

depression are strongly related to concern for mistakes and doubts about one’s
actions. However, perfectionistic excessive standards are not related to worry.
The authors suggest that high-worriers may not possess excessively high
standards as compared to low-worriers and may even lower their standards while
under duress. The researchers also found that while there is extensive overlap
and signiﬁcant relationships between worry, anxiety, and depression, worry was
found to be a signiﬁcant component of both anxiety and depression. Further, the
study found that perfectionism is more related to how much individuals worry
versus how they feel about their worrying.

Several studies offer evidence that certain dimensions of perfectionism
are differentially related to perfectionism. Socially prescribed perfectionists
appear more maladaptive in nature than self-oriented perfectionists (Einstein,
Lovibond, & Gaston, 2000; Flett, Hewitt, Blankstein, & Gray, 1998; Frost,
Heimberg, Holt, Mattia, & Nebauer, 1993; Norton, Buhr, Cox, Norton, & Walker,
2000). For example, these studies also indicate that anxiety and depression
appear to be signiﬁcantly correlated with socially prescribed perfectionism but not
self-oriented perfectionism. In a study of passive versus active perfectionism and
depression, passive perfectionists were signiﬁcantly more likely than their active
counterparts to become depressed (Lynd-Stevenson & Heame, 1999). Passive
perfectionism was thought to contain a feature of hopelessness that increases
susceptibility to depression. Additionally, perfectionism was found to represent a

vulnerability factor increasing the chance of depression during times of stress.

17

 

 

A study of evaluative concerns and personal standards perfectionism
found that hassles, avoidant coping, and perceived social support mediate
evaluative concerns perfectionism (ECP) and distress (Dunkley, Blankstein,
Halsall, Williams & Winkworth, 2000). In fact, these factors accounted for nearly
100% of the relationship between ECP and distress. The researchers also
discovered that these perfectionists self-generate stress as well as engage
defensive and dysfunctional ways of c0ping resulting in increased stress. This
ﬁnding is consistent with Flett, Hewitt, and colleagues (Flett, Hewitt, & DeRosa,
1996; Flett, Hewitt, Blankstein, Solnk, & Van Brunshot, 1996) who found that
socially prescribed perfectionists react to stressful situations with hopelessness
and helplessness as evidenced by the use of defense mechanisms such as
splitting and denial. Dunkley et al. (2000) suggest that evaluative concerns
perfectionists may have lower self-efﬁcacy resulting from an inability to
adequately cope with stressful situations or to the satisfaction of others.

Discrepancy Theory and Perfectionism

Self-discrepancy theory is a social psychological model that purports that
anxiety and depression result from discrepancies in the self-concept. Higgins
(1987) argues that the self consists of three domains that interact to inﬂuence
individuals’ emotional experience: the actual self is comprised of attributes that a
person perceives they actually possess; the ideal self consists of attributes an
individual wishes to possess; and the ought self embodies characteristics that an
individual feels obligated to possess. According to self-discrepancy theory,

individuals evaluate their actual selves from the standpoint of either their ideal

18

and/or ought selves. Additionally, it is believed that individuals may also evaluate
themselves from the perspective of signiﬁcant others such as parents or
spouses. A primary hypothesis of discrepancy theory states that . .the greater
the magnitude and accessibility of a particular type of self-discrepancy
possessed by an individual (e.g., between the actual and ideal and/or ought
selves), the more the individual will suffer the kind of discomfort associated with
that type of self discrepancy” (Higgins, 1987, pp. 335-336). Empirical
investigation revealed that persons who experience discrepancies between their
actual and ideal selves are vulnerable to depression, whereas individuals
experiencing discrepancies between their actual and ought selves are more
prone to symptoms of anxiety (Scott & O’Hara, 1993; Strauman, 1989).

Hankin, Roberts, and Gotlib (1997) drew upon self-discrepancy theory and
Hewitt and Flett’s (1991a, 1991b) multidimensional model of perfectionism in an
effort to explore the relationship between self-standards and emotional distress in
adolescents. Hankin et al. conceptually related self-oriented perfectionism with
actual/ideal self discrepancies in that both require the individual to create and
adhere to self-standards. Socially-prescribed perfectionism was associated with
actual/ought discrepancies in that the individual creates self-standards based on
the real or perceived perceptions of others. Findings of the study support the
notion that excessive self-standards are associated with certain forms of
emotional distress in adolescents. After controlling for anxious symptoms, self-
oriented perfectionism and actual/ideal discrepancies were associated with

depressive symptoms. This ﬁnding was contradicted in a study by Einstein et al.,

19

(2000). The authors suggest that self-oriented perfectionists may anticipate
meeting their own high standards, thereby eliminating distress outcomes.

Alternately, actual/ought discrepancies were found to be related to
anxious symptoms after controlling for depression indicators. It was found that
socially-prescribed perfectionism was correlated with general emotional distress
(not speciﬁcally to depression or anxiety). Einstein et al. (2000) produced similar
results. Hankin et al. delineate clear conceptual differences between a
multidimensional model of perfectionism and self-discrepancy. For example,
although self-discrepancy theory examines emotional agitation created by
differences in the perception of selves, perfectionism focuses on the magnitude
of individuals’ self-standards, regardless of their perceived ability to attain these
standards.

Similarly, the APS-R identiﬁes individuals’ frustration of achieving
unattainable self-standards (labeled “Discrepancy”) as a deﬁning feature of
perfectionism. Self-discrepancy theory appears to encompass broader
dimensions of self that extend beyond the personality characteristics of self-
standards. These dimensions embrace the intensity, quality, and contextual
variability of emotional distress associated with discrepancy in self-concept
(Higgins, Klien, & Strauman, 1985).

Cognitive Implications of Perfectionism

From a cognitive behavioral perspective, perfectionists are viewed as

having three distortions of thinking that threaten their ability to function effectively

(Burns, 1980; Moore & Barrow, 1986; Sorotzkin, 1985). Overly moralistic self-

20

 

 

evaluation is a mechanism by which perfectionists compare themselves to a
world of absolutes. Black-or-white, wrong-or-right, all-or-nothing thinking
characterizes this distortion of thought. Perfectionists are prone to the tyranny of
what Karen Homey (1937) describes as “should” systems in which every
disappointment and failure is personalized. Instead of viewing failure with
compassion, the perfectionist harshly criticizes himself/herself and perpetuates a
cycle of self-blame and hatred.

Overgeneralization is a common mental distortion occurring when
perfectionists make dogmatic conclusions about their inability to perform based
on a single incident. Failing to meet a desired goal often predisposes
perfectionists into thinking that the negative event will inﬁnitely perpetuate itself.
Similar to the tendency to be overly moral in their behavior, perfectionists tend to
embed themselves in a pit of shouldisms that restrict their ability to ﬁnd self-
satisfaction and limit a system of unrealistic demands on the self.

Finally, dichotomous thinking limits the worldview of the perfectionist by
imposing a judgment system that polarizes events into absolutes. Corsini and
Wedding (1989, p. 592) deﬁne dichotomous thinking as “Categorizing
experiences or people in black-and-white or extreme terms only (e.g., all good
vs. all bad) with no middle ground.” Perfectionists view their efforts as all-or-
nothing, black-or-white, good-or-bad, wrong-or-right. When combined with critical
moralistic self-evaluation, dichotomous thinking contrives to rob perfectionists of
their ability to self-regulate a worldview that encourages patterns of thinking that

Burns describes as “. . .naive and self-defeating” (p. 38). Mahoney and Arnoff

21

(1979) and Barrow and Moore (1983) describe a “saint or sinner” syndrome in
which perfectionists are thwarted in their efforts to achieve personal goals by a
psyche that is prone to harsh judgments. When perfectionists attempt to desist
from cigarette smoking, they make rigid internalized rules of conduct - either they
smoke or do not smoke. This is the period of sainthood founded upon and
bounded by a moralistic judgment that perfectionists are good if they discontinue
smoking and bad if they resume. If perfectionistic smokers lapse in their goal and
smoke a cigarette, sainthood ends and a period of “sin” follows, characterized by
guilt and self-deprecation (Burns, 1980). Thus, the highly judgmental distortions
of the perfectionist become reinforced by the ensuing feelings of low self-esteem
and deﬂated self-worth. The perfectionist can be viewed as an individual stuck on
a merry-go-round of intrapsychic conﬂict which becomes ever more difﬁcult to
exit with every revolution.
Splitting

Related to dichotomous thinking is "splitting." Splitting is a term that has
created much semantic confusion in describing structural and functional states.
In a review of the history of the term, Pruyser (1975) indicated that the numerous
meanings attributed to splitting have served to make the term holographic in
nature. He advocates that the term be abandoned in favor of less confusing
descriptors. Marmar and Horowitz (1986) attempted to elucidate the complex
and contradictory ways in which the term has been used to describe the etiology
of personality disorders. Splitting has been described as a metapsychological

construct used to describe clients with dissociative disorders, schizophrenia and

22

schizoid disorders, narcissistic and most commonly borderline personality
disorders.

In his work with therapeutic process of change, Greenberg ( 1979) deﬁned
a split as “...a verbal performance pattern in which a client reports a division of
the self-process into two partial aspects of the self or tendencies” (p. 317). He
has identiﬁed three types of splits in his work. The Conﬂict Split is characterized
as a division of two parts of the self in opposition and experienced as a struggle
between the conﬂicting polarities. Greenberg, Rice, and Elliot (1993) state that a
“linguistic juxtaposition indicator” (p. 188) (eg. but, yet, if/why) serves to
differentiate opposing “I“ states. Greenberg and Sarkissian (1984) posit that the
client generally experiences “...a conﬂict between standards and values on the
one hand, and organismic emotional reactions and needs or wants on the other"
(Greenberg et al., 1993, p. 188). When this inner dialogue is processed,
Greenberg states that these conﬂicts result in an intrapsychic struggle between
“shoulds” and “wants.” For example, a perfectionist might intone, “I should have
achieved an A on the test, but I just didn’t study hard enough.“

The Subject/Object split occurs when one part of the self (the subject, I)
does something to the other part of the self (object). The person typically
expresses frustration with a behavior in which they are engaged. “I am critical of
myself” or “I tend to edit my paper as I write,” are examples.

Greenberg et al. (1993) also posit a form of splitting that appears similar to
Hewitt and Flett’s “socially prescribed” perfectionism and particularly relevant to

maladaptive perfectionism. An attribution split occurs when the individual

23

experiences a split “...as though it were originating in someone else” (p. 189). It
may indicate projected injunctions and judgments of others that are in conﬂict
with personal desires and wants. For example, a young woman expressing an
interest in marrying her boyfriend may state, “My mother thinks that I should get a
college education so I won’t have to be entirely dependent on my spouse for
ﬁnancial support.” While this projection may indeed be an accurate assumption
of parental concern, it actually reﬂects a conﬂict between personal desires and
the anticipated judgments of others. Individuals who engage in this style of
splitting attribute criticism and coercive efforts that deny personal desires to be
caused by the injunctions of others. From a developmental perspective, this type
of split is seen as originating from a person’s manifestation of negative
evaluations or expectations of parents, teachers, or signiﬁcant others. In the
process of accepting external judgments about themselves, individuals become
self-critical and self-pressuring and continue to expect this from others.
Theories of Splitting

Breuer and Freud (1893) were the ﬁrst to reference the term in describing
a “splitting of conscience" in which to illustrate dichotomous states of personality
or consciousness exhibited in hysterical patients. Freud initially expressed
splitting as the result of repressed and unconscious memories that denied
synthesis. Splitting was viewed as a psychic mechanism that enabled individuals
to avoid “. . .intolerable mental conditions” caused by contradictory interpretations
of the world. Freud (1916) later discussed the dualistic nature of a splitting ego,

“The disavowal is always supplemented by an acknowledgment; two

24

 

contradictory and independent attitudes always arise and result in the situation of
there being a splitting of the ego,” (pg. 204). The splitting ego was described as a
psychic division allowing the ego to sustain two mutually contradictory psychic
conﬁgurations. From this perspective, Freud posited that splitting was closely
associated with dissociation resulting in amnesia of core memories and their
associated affect.

Mahler (1968, 1971) observed that children who have unsatisfactory
relationships with their mothers in the ﬁrst two years of life have a propensity for
splitting. She suggested that the child’s realization that he/she is not omnipotent
combined with a mother who is emotionally unavailable serves to create a hostile
dependence on the mother. The splitting serves as a defense mechanism by
which the child may attempt to reconcile conﬂicting feelings about the world and
thereby protect a fragile and developing good self-image against an inadequate
and bad maternal introject. Good and bad mother images are given internal
separation and the child's aggression is turned inward against the self. Mahler
believes that the splitting mechanism becomes the basis for responding
habitually to stress with negative mood swings, as well as for the relentless
pursuit of perfection in self and/or others (Marmar & Horowitz, 1986).

Kemberg’s (1975) metapsychological analysis of splitting is recognized as
the seminal work in object relations for understanding the phenomenon in
borderline personalities. According to Kemberg, splitting results from the normal
developmental process of infants as they attempt to manage and order chaotic

and ovenrvhelming environmental stimuli. Contradictory feelings of self and

25

 

objects are categorized within the ego as either good or bad. For example, an
infant may segregate conﬂicting feelings aroused by good and bad experiences
of its mother in order to protect a fragile ego. Splitting acts as a protective
measure that enables the organizing ego and superego to avoid anxiety that
might otherwise fragment the development of a cohesive psychic structure. St.
Clair (1986) describes early splitting as, "...the maturational inability to
synthesize incompatible experiences into a whole,” (p. 10). Splitting arms the
ego with a defensive tool that separates unwanted aspects of the self or
organizes threatening objects into more manageable aspects (Grotstein, 1981).
In adulthood, the same defense that allowed the infant to manage contradictory
feelings may become a pathological system of splitting used to avoid painful
ambivalence. “This defensive division of the ego, in which what was at ﬁrst a
simple defect in integration is then used actively for other purposes, is in essence
the mechanism of splitting” (Kemberg, 1975, p. 25).

For the purpose of this study, I will use the following deﬁnition of spjittﬂrg
(from Corsinsi & Wedding,1989, p. 599): “A situation in which a person splits off
part of self as a polar opposite. When aware of one pole, the person is oblivious
to the other. For example, an individual may split into competent and
incompetent selves and vacillate between these roles. A split is one form of a
dichotomy.”

To further elaborate on the deﬁning characteristics of splitting, Corsini and
Wedding deﬁne a dichotomy as “A split in which the ﬁeld is experienced as

comprising competing and unrelated forces that cannot be meaningfully

26

xi

integrated into a whole” (p. 592). The term “ﬁeld" originated from Gestalt
psychology and refers to an organized perceptual domain in which fragments of
a perceptual whole are interconnected and immediately responsive to each
other. No individual part is unaffected by activity elsewhere in the ﬁeld. For
example, an emotional malady such as stress may also inﬂuence chronic back
pain. Conversely, a back injury may promote agitation or worry.
Elements of Splitting
Marmar and Horowitz (1986) conclude that the work of object relations
theorists Kemberg (1975), Lichtenberg and Slapp (1973), Volkan (1976),
Horowitz (1977), and Grotstein (1981) most clearly represents a uniﬁed construct
of splitting, especially in relation to understanding borderline p0pulations. The
following tenants are central to this collective conceptualization of splitting:
1. Mental representations of self and others are separated into part
rather than whole images.
2. Objects are viewed as either all good or all bad rather than
possessing both attributes simultaneously.
3. Highly affective and emotional feeling states that are
in opposition are kept separate.
4. Individuals who exhibit splitting behaviors are unable to access
memories of previous positive feelings toward a frustrating
object which might otherwise mitigate their reaction to

frustration.

27

(Elements of splitting continued)

5. Individuals with a tendency to split their affect may overreact
to situations that engage emotional arousal due to an inability
to integrate mixed experiences over time.

6. The inability of the individual to ﬁrmly repress or deny the felt
ambivalence of contradictory states forces the individual to
identify with only one side of the ambivalence at a time. This
selective identiﬁcation of ambivalent states severely limits the
capacity of the individual to realistically integrate experience.

In summary, the theoretical frame used to describe splitting inﬂuences its
deﬁning features as a construct. The literature posits that splitting is
characterized as an intrapsychic defense utilized by emotionally vulnerable
individuals in anticipation of the consequences of negative affect. Splitting Is a
psychic mechanism by which individuals divert emotionally agitating experiences
between the ego and superego, it protects a fragile self-concept, and enables
persons to manage conﬂicting and contradictory perceptions of self or injunctions
of others. As described previously, recent research in self-discrepancy theory
buttresses the notion that incongruities between self-concept and ideal and/or
ought perspective of self generate emotional distress. Splitting would appear to
be a logical defense used to avert emotional distress created by discrepancies of
self-concept. Individuals who carry highly incongruent perceptions of their actual
versus ideal selves, such as perfectionists, may be prone to splitting in an effort

to manage intrapsychic disturbance. Measurement of discrepancy between self-

28

concept and ideal and/or ought self therefore, may provide evidence and an
estimate of the extent of splitting utilized by perfectionists.
Measurement of Splitting

Empirical literature indicates the development of two instruments that
measure individuals capacity to split. The Splitting Index (SI: Gould, Prentice, &
Ainslie, 1996) and the Splitting Scale (SPS: Gerson, 1984) were based on an
object-relations conceptualization of splitting. Test items of both instruments
appear to focus primarily on measuring the lability of affect and the identiﬁcation
of pathology such as borderline and narcissistic personality disorders,
dimensions of splitting which are not the focus of this study. Furthermore, the
psychometric properties of the instruments (i.e. reliability coefﬁcients) have
raised serious concerns about their utility in research.

Fortunately, more psychometrically sound measures are available to
measure splitting that are more aligned with the conceptual framework used in
the current study. As discussed, measures of disparity between varying
perceptions of self-concept may also provide an indication of splitting. It is argued
that instruments with the capacity to measure disparities in self-concept, such as
perceived self and ideal self, would give substantive support for identifying the
splitting mechanism. The Adjective Check List is a widely used standardized
assessment of personal saliency and employed in a number of studies to explore
self/ideal-self congruence. The instrument is comprised of a constellation of
adjectives that are delineated into 37 scales. The ACL produces a personality

proﬁle detailing the self-concept of respondents. Initial factor analysis of the

29

scales revealed intercorrelations that grouped the 37 scales into six factors. The
factor descriptions in order of loadings include Potency, Assertiveness,
Sociability, Individuality, Dissatisfaction, and Constriction. Individuals ranking
high in the Dissatisfaction factor are identiﬁed as anxious, introspective, and self-
critical. . .characteristics common to maladaptive perfectionists. More recent item
analysis of the ACL yielded ﬁve unipolar personality dimensions including
Aggressive/Dominant, Neurotic, Conscientious, Detached/Introverted, and
Surgent/Extraverted (Strack & Lorr, 1990). Recent studies indicate that the ACL
correlates highly with the NEO Five Factor Inventory of personality types
(FormyDuval, Deborah, Williams, Patterson, & Fogle, 1995; Piedmont, McCrae,
& Costa, 1991).

Several studies have focused on the use of the ACL to distinguish
between self and ideal-self (e.g. Gough, Fioravanti & Lazzari, 1979, 1983;
Gough, Lazzari & Fioravanti, 1978; Graves 8. Shearer, 1971; Lazzari & Gough,
1980; Small & Batlis, 1978). The literature appears to provide very good
empirical support concerning the ACL’s ability to discriminate between subjects’
description of their “real” selves versus their “ideal” selves. .

Of particular relevance is a study by Gough et al. (1983). The authors
found that subjects with large self/ideal-self discrepancies tended to be anxiety-
prone, riddled with self-doubt, and lacking in interpersonal skills. Subjects with
minimal discrepancy were characterized as conﬁdent, having good coping ability,
and socially balanced. Discrepancy between the real and ideal-self ACL proﬁles

was derived from summing the absolute differences between each pair of

30

standard scores. The results of the study were based on the assessment of
1,484 subjects on all 37 scales of the instrument and the authors discovered
eight scales yielding nearly the same discrepancy values as the index for the
overall instrument. Four scales exhibiting subject scores with higher ideal-self
scores than real-self scores as described in The Adjective Check List Manual
(1983) include Self-Conﬁdence, Ideal Self, Creative Personality, and Adult. Four
scales in which all self scores were higher than ideal-self scores include
Succorance, Unfavorable, Abasement, and Adapted Child. (See Table 1 in
Chapter 3 for scale descriptions.)

As indicated by the authors (Gough et al. 1983), “An abbreviated index,
based on just eight scales had correlations .90, .90, .86, and .81, respectively,
with the total index over all 37 scales In samples of American men and women
and Italian men and women” (p. 1219). Additionally, the abbreviated index and
total index of discrepancy yielded nearly identical results when contrasted with
the non-ACL test scores and ratings of observers. It would appear that
assessment of discrepancy in self-concept (splitting) could be efﬁciently
measured by the abbreviated ACL with nearly the equal effectiveness of the 37
scales of the ACL.

Perfectionism and Splitting

Many writers site dichotomous thinking as a dominant cognitive style of
perfectionists (Burns, 1980; Halgin & Leahy 1989; Hamachek, 1978; Moore &
Barrow, 1986; Pacht, 1984; Sorotzkin, 1985). When perfectionists engage in

destructive tendencies and behaviors, the polarization of opposing parts creates

31

an lntemal conﬂict most often in which an overly active system of unconscious
self-commands is required to alleviate the conﬂict. In order to cope with the
conﬂicted demands of the self, the individual forms an alliance with one part of
the self and suppresses the needs and feelings of the other part. The primary
quality of the cognitive style is all-or-nothing thinking. A “B” letter grade is
unacceptable to perfectionistic students who feel that they must earn an “."A
These cognitive commands are expressed as a series of “shoulds” such as “I
should be better person, I should not get angry, I should have done it differently, I
should have know better, I should have worked harder” (Pacht, 1984, p. 387).
These “shouldisms” are indicative of individuals who experience a disparity
between intended goals and goals that are actually achieved.
Psychotherapy of Splits

Splits are therapeutically managed in a variety of methods depending on
the orientation of the therapist. In Rational Emotive Behavior Therapy (R.E.B.T.),
splits are labeled as inappropriate thinking and an attempt is made by the
therapist to challenge and subjugate splits through a focus on cognition. The goal
of psychoanalysis is to locate the original source of the splitting pattern and
eliminate its inﬂuence upon a client’s personality through increased awareness
and reconciliation of past activating events.

Gestalt therapy offers a perspective in which splits are viewed as
disturbances in a healthy functioning phenomenological ﬁeld. Dichotomous
thinking creates an intrapsychic structure in which the mind is divided into an

assortment of competing forces or independent parts. Organismic self-regulation,

32

a state that is considered vital to the ability of the mind to integrate paradoxical
truths and embrace diversity, is severely restricted. The therapeutic goal is to
increase the present awareness of the split into its polarities allowing for
integration of the differences leading toward a balanced organismic ﬁeld.
Polarities are viewed as opposites that complement or explicate each other
(Corsini & Wedding, 1989). Gestalt therapy views polarities of an organismic ﬁeld
similar to the poles of an electrical ﬁeld in which both negative and positive poles
comprise the entire ﬁeld in opposition. The establishment of psychic poles
creates an internal continuum from which individuals can evaluate, categorize,
and make sense of their experiences of the world. Additionally, polarization
furnishes people with a structure from which to distill and reﬁne complex
phenomenon into discrete and predictable parts. This intrapsychic structure
provides a rapid and easy system from which to interpret experiences and
respond appropriately. While this structure may be experienced as safe, it may
also produce rigid dichotomous constructs that limit awareness and constrict a
person’s ability to integrate new and challenging experiences. For example, an
individual may feel safe in their ability to predict experience, but the consequence

of holding to rigid dichotomies may result in intellectualizing that limits a person’s

 

ability to appreciate new experiences that may promote growth and maturation.
In Gestalt therapy, fear Is seen as the primary motivating force behind an

individual’s attempt to control various aspects of the environment or self. By

creating rigid polarizations, individuals enact a two-value system that enables

them to generate simpliﬁed constructs and labels. These adopted labels are used

33

 

to give meaning to new experiences as if they represented the original
experiences from which they were created to categorize. In the process, a
person obtains an obscured sense of control and power built upon labels that
convey a false perception of their actual range of experience. Individuals who
believe that all things fall neatly into speciﬁc categories exemplify a reliance on
labels and use of them for control. When experiences fail to ﬁt neatly into
established categories, the safety of the rigidiﬁed labels is threatened resulting in
dissociation and extreme anxiety and angst. Individuals with a two-value system
may cling ever tighter to their labels in an effort to ward off future threats of
dissociation and reduce the potential for anxiety (Korb, Gorell, & Van De Reit,
1989)
Gestalt Therapy and the Origins of Splitting

Gestalt therapists believe that children accept ideals and behavior without
scrutiny (known as introjections). As a result, children indoctrinate themselves
with societal values and beliefs that may be in conﬂict with their own wants and
needs. The outcome is an internalized morality largely shaped by an externally
demanding environment rather than a healthy and balanced morality that
integrates compatibility between lntemal drives and external demands. Guilt
and/or shame typically result when individuals defer to their wants rather than the
“shoulds” imposed by the environment.
Gestalt Therapy and the Mechanics of Splitting

Theoretical literature may indicate that perfectionists exert a tremendous

amount of energy in maintaining the split between personal “wants” and societal

34

 

“shoulds,” speciﬁcally in the direction of following societal percepts (Burns, 1980; i
Pacht, 1984; Sorotzkin, 1985). They continually seek out ways in which to
maintain or improve an imagined ideal self-concept. Paris (1969) stated, “Many
people dedicate their lives to actualize a concept of what they should be like,
rather than to actualize themselves. . .This is again the curse of the ideal. The
curse that you should not be what you are” (p. 19). An example of one way in
which this actualizing process occurs is characteristic of dieters. These
individuals often become demoralized by environmentally imposed injunctions
about maintaining socially accepted levels of body images as opposed to
accepting want or craving to enjoy food or appear overweight. An inability to live
up to perceived internalized societal standards becomes an overwhelming
afﬁrmation of personal inadequacy rather than a disappointment or annoyance.
The result is often a circular system of attempts and failures to cease socially
inappropriate and/or medically harmful behavior, sabotaged by lntemal wants
and preferences.

The perfectionist on the other hand, will perhaps be successful more often
in his/her attempts to achieve a selected goal. However, the success of an
achievement is at best ﬂeeting and momentary to the hypercritical perfectionist
who is forever striving to achieve an ideal concept of self; an unattainable ideal
that is a socially prescribed mirage of the perfect person. Perfectionists are
driven by a need to maintain high personal standards in their efforts to preserve
ongoing validation from others. Failure to uphold superior performance in nearly

any capacity of living threatens the fragile self-concept of the perfectionist. It is

35

 

imperative that perfectionists adhere to what they perceive as the upper limits of
valued performance and behavior. These limits are reinforced by lntemal
“shoulds” dictating acceptable levels of performance. However, due to an
overidentiﬁcation with generalized societal values, perfectionists rarely ﬁnd
satisfaction with their achievements. They must always look to the next task to
receive validation. The perfectionist constantly seeks to embrace an unattainable
level of validation and therefore, acceptance.

Greenberg et al. (1993) describe this internal conﬂict in terms of conﬂicting
schemes. One polarity is based on emotional schemes representing biologically
adaptive emotions and needs. The opposing polarity consists of negative self-
judgments and introjected standards founded upon social measures that impose
themselves upon the individual as “shouldisms.” Conﬂict is created when internal
or external cues activate the opposing schemes. Typically, these oppositional
schemes create incompatible thoughts, feelings, or behaviors against desires
with either or both extending beyond the individual’s awareness.

Gestalt therapists believe that “shouldisms” sabotage healthy functioning.
Shouldlsms represent the self-coercive statements of socially introjected
standards or negative self-evaluations that condemn, coerce, or criticize the
emotional aspect of the self, similar to Hewitt and Flett’s (1991a) “self-oriented"
and “socially-prescribed “ perfectionism. This conﬂict has been classically
portrayed by Perls’ metaphor of “topdog vs. underdog.” Korb et al. (1989)
describe the topdog as “...the demander of perfection, the manifestation of a set

of introjected ‘shoulds’ and ‘shouldn’ts’ “(p. 63). The underdog is characterized as

36

the victim aspect of the psyche that protects its desires and interests by resisting
the topdog through hostility, passive-aggression, obsessiveness, and
compulsiveness. Perfectionists may overidentify with the coercive and cajoling
demands of the perfectionistic t0pdog in an effort to maintain a sense of internal
control over their emotions. The underdog may comply with the commands of the
dominant topdog, but at a cost. This “split” in preference of the topdog serves to
increase the resistance of the underdog. Feelings of validation and satisfaction
are lost, replaced by anxiety, depression, and angst. These feelings may assert
themselves through internalized anger, obsessive thoughts, or compulsive
pursuits to fulﬁll an emotional gap created by wants and needs that are highly
conﬂicted and polarized. Perfectionists must ﬁll this emotional gap by constantly
looking to others to provide validation of personal adequacy. It was for this
reason that Paris believed in a paradoxical theory of change. The more people
push to become what they are not (i.e., consort with their topdog) the more
resistance is established against intended change (as inﬂuenced by the resistant
underdog) and therefore the more they thwart their ability to achieve intended
change.

“Shouldisms” may also portray enacted introjected standards as
exempliﬁed by attribution splits. An individual may internalize real or perceived
evaluations of others that create the belief that they are contemptuous or
disgusting and therefore weak or inferior. Shame and embarrassment are the
hallmarks of all individuals who view themselves as inferior. Interestingly, shame

is often self-perpetuating and founded upon faulty projections in others that are

37

maintained by internalized splits of self-contempt and disgust. “T he ashamed
state can be seen as the attribution of disgust/contempt to others or as imagining
being contemptible/disgusting in the eyes of the other" (Greenberg et al., 1993, p.
191 ). Perfectionists may be particularly susceptible to this form of splitting. In
their attempts to validate their self-worth in the eyes of others, they may be prone
to overcompensating their performance on tasks, especially tasks that are under
the scrutiny of others. Additionally, the perfectionist who attempts to plug the dike
of his/her shame by overacheivement may create even greater intemallzed
shame and anger by feeling embarrassed that they are compelled to seek
external validation. This type of perfectionist may ﬁnd himself/herself in a
dilemma: how can a person be perfect if they are dependent upon the validation
of others? Pirot (1986) describes a pattern of irrational thinking in which
perfectionists focus on themselves to compensate for feelings of inferiority. This
drive for validation may push perfectionists to unwarranted educational
accomplishments such as the attainment of multiple degrees (which may have no
functional relationship) or unprecedented vocational achievements such as rapid
corporate advancement.
Summary of Gestalt Therapy Theory on Perfectionism and Splitting

Gestalt theory appears to provide theoretical support for the belief that
perfectionism is linked to intrapsychic splitting and dichotomous thinking. Gestalt
therapists believe that children are conditioned from birth to accept the
introjections of societal values and beliefs that create an internalized morality that

may favor perceived external demands over lntemal desires. lntemal conﬂict

38

results from the incompatibility of competing demands on the psyche. It is
believed that individuals are prone to “splitting” these oppositional demands into
a two-system intrapsychic structure. New experiences that challenge the psyche
are divided into preconstructed categories or polarities that enable individuals to
understand and give meaning to new phenomenological experience. These
categories or labels bring order and control to external experiences that create
confusion and threaten intrapsychic stability. Psychologically healthy individuals
are able to use the two-system structure to draw a balance between their lntemal
drives and external demands. Individuals with fragile self-concepts, such as
perfectionists, may ﬁnd an overreliance on the splitting mechanism. The
foundation for splitting appears to be the discrepancy between an individual’s
perceived sense of self and idealized self-image. The greater the disparity
between these two images, the greater the propensity to split. Splitting allows
these individuals to reduce the extreme anxiety producing effects of dissociation
that results when internalized labels are threatened.

Perfectionists may respond to intrapsychic threats by overidentifying with
one polarity. Greenberg et al. (1993) describe this polarity as a scheme
comprised of negative self-judgments and introjected standards. Perls described
“shouldisms” that form the demands of the “topdog” or internal critic in contrast to
the “underdog” the emotional “victim” side of the polarity. These internalized
demands may manifest themselves as exceptional drives for achievement in the
perfectionist. In order to avoid dissociation and divert anxiety, the topdog asserts

itself via self-coercive statements and demands in an effort to demean and

39

 

criticize the underdog. This serves to give the individual a false sense of control
over their emotions. Subverting feelings of anxiety and depression enhances the
resistance of the emotionally bound underdog who retaliates via internalized
hostility, obsessiveness, compulsiveness. or passive-aggression. In
perfectionists, the underdog may submit to the demands of the topdog, but at the
price of a continual need for external validation as the individual alienates the
emotional validation that the underdog is capable of providing.

A form of splitting particularly relevant to perfectionism is the attribution
split characterized by an inability of individuals to differentiate negative self-
evaluations and expectations of the self from their anticipation of others negative
evaluations and expectations. A typical indicator of this type of split is when
clients report that they are unable to fulﬁll a desire because of the disapproval of
others (Greenberg et al., 1993). This type of split appears relevant to
perfectionism such that perfectionists may be driven toward achievement
behaviors in an effort to validate the injunctions of others.

Currently, no empirical research exists supporting the previous arguments.
Although a number of articles give theoretical support for the link between
perfectionism, dichotomous thinking, and splitting, none provide quantitative or
qualitative data clearly elucidating these relationships.

Research Questions
1. Are dichotomous thinking and splitting characteristic of maladaptive
perfectionists? The literature supports the thesis that dichotomous thinking

and splitting are associated with maladaptive perfectionism. The literature

40

also indicates that “shouldisms” are characteristic of both maladaptive
perfectionism and splitting. Thus, maladaptive perfectionists would likely be
characterized by a signiﬁcant disparity between their “True” or actual image of
themselves and an idealized image (deﬁned by Greenberg [1993] as an
“Internal Conﬂict Splitting” style) as compared to adaptive or non-
perfectionists.

. Do maladaptive perfectionists appear to introject the external injunctions of
others? A key characteristic of attribution splitting is an individual’s
achievement behavior that is motivated by the need to ﬁnd validation through
the concordance of others’ perceived opinions of the individual and their
sitUation. By capturing the extent to which individuals introject the judgments
and injunctions of others, a split between the True-Self and lntrojected-Self
may be elucidated. The literature supports the notion that maladaptive
perfectionists may utilize an attribution splitting defense to gain acceptance,
favor, and possibly admiration from others.

. Does affect mediate the degree of splitting experienced by maladaptive
perfectionists? Baron and Kenny (1986) state that “...a given variable may be
said to function as a mediator to the extent that it accounts for the relation
between the predictor and the criterion” (p. 1176). Anxiety and depression
have been linked to perfectionism and splitting. Theories on splitting would
seem to support the mediator role of affect between perfectionism and
splitting. Individuals with high degrees of maladaptive perfectionism may be

more susceptible to intense feelings of anxiety and depression, prompting

41

equal levels of splitting in order to compensate or reduce the perceived
impact of negative affect. Said differently, splitting is expected to diminish in
individuals with maladaptive perfectionism when affect is controlled.
Hypotheses
Primary Hypothesis
Based on the foregoing theoretical assertions, the following hypothesis was
proposed for the present study: More maladaptive aspects of perfectionism were
expected to be signiﬁcantly and positively associated with splitting/discrepancies.
The discrepancy among maladaptive perfectionists was thought to indicate
particular styles of splitting.
Secondary Hypotheses
Affect was thought to mediate the perfectionism - splitting relationship:

1. It was posited that depression mediates perfectionism - internal conﬂict
splitting relationships as indicated by a positive correlation between
depression (CES-D) and splitting (ACL) measures. It was believed that
depression serves as a mediator between maladaptive perfectionism and
internal conﬂict splitting (discrepancy between actual and ideal self-
concepts). For example, the more maladaptive characteristics a
perfectionist has, the more vulnerable the individual is to experiencing
depression. lntemal conﬂict splitting is viewed as an internal mechanism
or defense that is an outcome of increased depression. As depression

increases, internal conﬂict splitting will increase.

42

2. It was also believed that anxiety mediates perfectionism - attribution
splitting relationships as indicated by a positive correlation between
anxiety (PSWQ) and splitting (ACL) measures. In the same way that
depression mediates internal conﬂict splitting, it was believed that anxiety
serves as a mediator between maladaptive perfectionism and attribution
splitting (discrepancy between actual and projected self-concepts). For
example, the greater degree to which individuals have maladaptive
perfectionism, the more vulnerable the individual is to experiencing
anxiety. Attribution splitting was viewed as an lntemal mechanism or
defense that is an outcome of increased anxiety. Increases in anxiety

result in responsive increases in attribution splitting.

43

CHAPTER 3
Methodology

The purpose of this study was to explore the relationship between
perfectionism and splitting. This chapter details the description of participants,
instruments, procedures, and analysis.
Participants

The sample included 263 students in which the majority of students were
drawn from the University of Oregon. . There were 91 men and 172 women with
a mean age of 22 years, ranging in age from 18 to 51, and a mean self-reported
GPA of 3.10 (SD = 0.5). Most of the sample was undergraduates and included 53
freshmen, 60 sophomores, 82 juniors, 51 seniors and 9 ﬁfth year students. No
information regarding race or ethnicity was obtained. This sample, according to
Cohen (1992), is large enough to detect a medium effect size to be used in a
multiple regression with an alpha of .05.

Participants for the study were recruited by two separate methods. First,
40% of the participants were recruited via advertising and direct contact. The
other 60% of the participants were solicited from Greek Life organizations as a
component of special outreach presentations on behalf of the University
Counseling Center at the University of Oregon. Each Greek Life organization
received credit toward activity requirements required by the university for their

participation. All respondents were entered into a drawing for a cash prize.

44

Instruments

Almost Perfect Scale-Revised (APS-R, Slaney et al., 2001). This 23-item
self-report measure assesses “adaptive” and “maladaptive” qualities of
perfectionism via three personality characteristics. These traits, High Standards
(e.g., “I try to do my best at everything I do.”), Order (e.g., "I an orderly person”),
and Discrepancy (e.g., “I’m hardly ever satisﬁed with my performance”),
comprise the three subscales of the instrument. The Discrepancy Scale
measures maladaptive qualities of the construct. For each item, respondents are
required to complete a Likert scale self-rating ranging from 1 “strongly disagree”
through 7 “strongly agree.”

According to Slaney et al. (2001), The High Standards subscale (7-items)
has an internal consistency (Cronbach’s alpha coefﬁcient) of .85. Order (4 items)
is reported to have an internal consistency of .86. The Discrepancy subscale (12-
items) has an internal consistency of .92. A conﬁrmatory factor analysis indicated
moderate correlations between Standards and Order (r = .42), and negligible r’s
between Standards and Discrepancy (r = -.12), and between Order and
Discrepancy (r = -.03).

Ashby and Rice (in press) found further evidence of the convergent and
discriminant validity of the APS-R. The results of a conﬁrmatory factor analysis
conducted on a study of the APS-R and Dysfunctional Attitudes Scale (DAS:
Weissman (1979) revealed a signiﬁcant correlation (r =.61) between the APS-R
Discrepancy Scale and the DAS Perfectionism subscale. Correlations between

the DAS Perfectionism and APS-R High Standards subscale (r =.09) and Order

45

subscale (r =.08) were found not signiﬁcant. Further conﬁrmatory factor analysis
offered evidence that self-esteem and High Standards were signiﬁcantly and
positively related (r = .25). Discrepancy and DAS Perfectionism, on the other
hand, where found to be signiﬁcant negative predictors of self-esteem (rs = -.31
and -.41, respectively). Ashby and Rice concluded that factors associated with
maladaptive perfectionism are closely linked with low-self esteem whereas High
Standards appears related to the adaptive qualities of perfectionism. The
signiﬁcant associations found in this study indicate that each factor of
perfectionism accounts for a unique variation in self-esteem after controlling for
the effects of other predictors.

The Adjective Check List (ACL, Gough, H. 8 Heilbrun, A., 1983). The
Adjective Check List is a self-report personality inventory comprised of 300
familiar adjectives and adjectival phrases used to elicit personal attributes.
Respondents are asked to identify all adjectives that they feel are self-
descriptive. Participants marking fewer than 10 items or more than 250 were
rejected from the study as suggested by the ACL Manual. Analysis of the items
yields 37 standard scales. Scale scores on the ACL may be used to gain an
overall personality proﬁle of subjects, or provide more speciﬁc detail by
examining individual scales or their combinations. Each scale is comprised of
indicative (favorable attribute) and contraindicative (unfavorable attribute) items.
Scale scores are the result of the sum of indicative and contraindicative items
converted into standard scores and are differentiated by gender and total number

of adjectives checked.

46

The overall reliability for the ACL is good as reported by Gough and
Heilbum (1983). The median alpha coefﬁcients for 591 males were .76 with a
range of .56 to .95. The median alpha coefﬁcient for 588 females was .75 ranging
from a low of .53 to a high of .94. Test-retest correlations for the instrument yield
a median of .65 for male subjects and .77 for female college students over a six-
month period.

The overall robust quality of the internal consistency and validity of the
ACL is supported by the discovery of an abbreviated version of the instrument. In
their study of the implications of ideal-self congruence, Gough et al. (1983)
administered the ACL to 490 American men, 216 American women, 600 Italian
men, and 178 Italian women. Congruence between scores on self and ideal-self
portraits was assessed by summing the absolute differences between each pair
of standard scores on the 37 scales yielding an index of congruence (labeled D-
T). To increase the efﬁciency of measurement, the investigators discovered a
subset of eight scales possessing nearly equal overall validity as the full 37-scale
instrument. Scales were selected on the basis of the four highest and lowest
mean scores for all scales from the sample. The scales with the highest means
included: Self-Conﬁdence (34-items), Ideal Self (46-items), Creative Personality
(30-items), and Adult (22-items). The scales with the lowest mean scores
included: Succorance (32-items), Unfavorable (75-items), Abasement (47-items),
and Adapted Child (44-items). (See Table 1 for a more detailed description of

each scale.)

47

Table 1

Adjective Check List (Abbreviated) Description

 

Self-Conﬁdence

Ideal Self

Creative Personality

Adult

Succorance

Unfavorable

High-scorers are initiators, conﬁdent in their ability to
achieve goals. Low-scorers have difﬁculty in
mobilizing their resources and taking action and are
viewed as shy, inhibited, and withdrawn.

High-scorers appear to be characterized by
interpersonal effectiveness and narcissistic ego
inﬂation. Low-scorers typically exhibit poor morale,
feel defeated by life, ﬁnd goal-setting difﬁcult, are
kind, modest and considerate.

High-scorers typically are venturesome, aesthetically
reactive, clever, and quick to respond. Low-scorers
are more subdued, less expressive, more
conservative, and less inclined to take action in
complex or ill-deﬁned situations.

High-scorers are characterized as productive,

work centered, reliable, ambitious, lacking in
spontaneity, jollity, and the ability to enjoy respite
and tranquility. High-scorers are also uncomfortable
expressing affection, love, and tenderness. Low-
scorers are viewed as more relaxed and responsive
but less effective in coping with work demands.

High-scorers feel inadequate in coping with stress
and crisis, avoid confrontation, and tend to retreat into
fantasy. They view others as stronger, more effective,
and their support is solicited. Low-scorers are
independent, relatively unbothered by self self-doubt
and equivocation, and effective in setting and
attaining goals.

High-scorers are characterized as disbelieving,
pessimistic about the future, changeable, headstrong,
and quick to take offense. The good fortune of others
is seen as undeserved, unearned, and unfair. Low
scorers are more dependable, more tactful, less
judgmental, and less easily offended.

48

(Table 1 continued)

Abasement High-scorers ask for little submit to the wishes and
demands of others, and avoid conﬂict at all costs.
Their interpersonal world is viewed with worry and
foreboding, and others are seen as stronger, more
effective, and more deserving. Low-scorers are
assertively self-conﬁdent and respond quickly; they
insist on obtaining what they judge to be their just
rewards.

Adapted Child High-scorers lack independence, feel unsure about
coping with the demands of adult life, fear and avoid
direct confrontation, and are easily disorganized by
stress and trauma. Low-scorers are autonomous and
effective, but inconsiderate. Personal feelings are
suppressed and others’ feelings are ignored. Low-
scorers strive for power, success, and tangible
accomplishments.

 

The sum of the absolute differences of the eight scales is a measure of
congruence labeled D-1. The correlation between D-T and D-1 among the four
samples was r =.90 for American men, r =.90 for American women, r =.86 for
Italian men, and r =.81 for Italian women. Alpha coefﬁcients for inter-item
reliability were .83 and .86 (American men and women, respectively) and .82 and
.83 (Italian men and women, respectively). The sample means and standard
deviations for the American and Italian samples on the abbreviated index were
signiﬁcantly different from each other.

As further evidence of concurrent and discriminant validity, Gough et al.,
(1983) found signiﬁcant correlations between the ACL, the California
Psychological Inventory (CPI), and the Minnesota Multiphasic Personality

Inventory (MMPI). The CPI is a metric of healthy self and personality while the

49

 

MMPI is an indicator of pathology. As expected, the congruence indexes (D-1
and D-T) negatively correlated with the majority of the CPI scales due to the
unfavorable or negative implications of differences between the self and ideal-
self. Positive correlations were obtained for the majority of correlations between
D-1 and the MMPI. As expected, participants with elevated levels of scores on
the D-1 (greater differences between self-views) exhibited higher levels of
anxiety, self-doubt, and social withdrawal.

The authors of the ACL admit that scoring of the instruments’ 37 scales
can be arduous, and when scored without the aid of a computer-scoring
program, the probability of errors increases. Similar to the procedure followed in
the Gough et al. study (1983), the abbreviated eight-scale version of the ACL,
which closely approximates the implications of the full measure, was used to
increase the efﬁciency and accuracy of scoring. It was also decided that a
composite score of the eight scales would be used to represent a global tally of
participants’ propensity to split. This was done in order to minimize the potential
complicating effects of examining and interpreting eight different aspects of
personality within and between each study participant.

Center for Epidemiologic Studies Depression Scale (CES-D, Radloff, LS,
1977). The CES-D was originally intended as a measure of depression. More
recent ﬁndings provide evidence that the instrument should be used as a broader
Indicator of distress (Blaney, 1986; Hesselbrock, Hesselbrock, Tennen, Meyer, &
Workman, 1983; Santor 8 Coyne, 1997). More speciﬁcally, Devins and Orrne

(1985) suggest that CES-D scores should be interpreted as indicators of distress

50

 

that accompany depression rather than a deﬁnitive gauge of clinical depression.
This instrument was selected over others because it focuses on symptomatology
rather than diagnostic criteria. The components include depressed mood,
feelings of guilt and worthlessness, feelings of helplessness and hopelessness,
psychomotor retardation, loss of appetite, and sleep disturbance. The instrument
includes 20 items scored on a four-point scale. The twenty items are summed
and higher scores indicate greater depression. The standard cutoff score used
as a baseline for detecting depressive symptoms is 16 or higher (Breslau, 1985;
Husaini, Neff, Harrington, Hughe, 8 Stone, 1980; Radloff, 1977; Weissman,
Sholomskas, Pottenger, Prusoff, 8 Locke, 1977). Conﬁrrnatory and exploratory
factor analyses have supported single factor and second-order factor models for
the instrument (Radloff, 1977; Sheehan, F iﬁeld, Reisine, 8 Tennen, 1995). The
CES-D appears to have strong internal consistency (Cronbach’s coefﬁcient alpha
= .90) for the general population (Breslau, 1985; Corcoran 8 Fischer, 1987;
Radloff, 1977). The instrument also demonstrates a strong relationship (r =.87)
with the Beck Depression Inventory (Santor, Zuroff, Ramsay, Cervantes, 8
Palacios, 1995).

Penn State Worry Questionnaire (PSWQ, Meyer, T. J., Miller, M. L., Metzger,
R. L., 8 Borkovec, T. D., 1990). The Penn State Worry Questionnaire is a 16 item
instrument measuring the tendency, intensity, and general nature of individuals to
worry. It is a measure of pathological worry in which both the degree of worry and
the intrusiveness of worry are reported by respondents (Stober 8 Joormann,

2001). Respondents are required to rate their perception of concern (worry)

51

regarding statements for each item on a scale from 1 (“very typical of me”) to 5
(“not at all typical of me”). Scores are summed for the 16 items. Higher scores
indicate higher degrees of worry. While no ofﬁcial cutoff score has been
determined as an indicator of anxiety, 47.65 (SD = 12.99) was found as an
estimate of anxiety in a large college sample (Molina 8 Borkovic, 1994).

In their initial development of the instrument, Meyer et al. (1990) found
that the PSWQ is a reliable and valid measure of worry. lntemal consistency from
initial use of the instrument was reported as high (or = .94). College samples also
support high lntemal consistency (.92 in Davey, 1993; .91 , .88, .91 in Ladouceur,
Freeston, Dumont, Letarte, Rheauma, Thibodeau 8 Gagnon (1992); and .94, .95,
and .91 in Meyer et al., 1990). Test-retest reliability also appears stable across
varying periods of time. Four week interval testing using the PSWQ yielded an r
= .86 (Ladouceur et. al, 1992) while a multi—week study found a range of
correlations from a low of r =.74 (4 week) to a high of r =.92 (8-10 week) (Meyer
et al., 1990). The PSWQ demonstrates strong concurrent validity and correlates
well with the State Trait Anxiety Inventory (.64 [trait] and .49 [state]), the
Cognitive Somatic Anxiety Questionnaire (.69), Self-Handicapping Scale (.33)
and others.

Procedure

The primary source of participants was drawn from students attending the
University of Oregon. The questionnaires were administered individually and in
several groups ranging from 20 to 80 participants. Several data collection

sessions were advertised via ﬂyers and word of mouth communication. Students

52

were offered prizes and to have their name entered into a grand prize drawing for
a cash reward for their participation. Each participant completed a written
informed-consent form, a brief demographic questionnaire, and a packet of self-
report questionnaires that included the APS - R, CES-D, PSWQ, and three
different versions of the ACL. In order to determine possible order effects caused
by the sequence in which the assessments were distributed, participants
received one of six surveys in which the order of the instruments was varied.
Participants took 30 to 50 minutes to complete the assessment battery. When
possible, the surveys were reviewed upon completion to make certain
participants had ﬁnished all assessments as instructed. At the completion of the
data collection process, a participant’s name was randomly drawn and awarded
the cash grand prize.
Administration

Participants completed three ACL’s. In one version of the ACL,
participants were instructed to complete the instrument as is typically required
(labeled “True Self” [T S]). In a separate ACL administration, participants were
asked to complete the assessment as they would ideally like to see themselves
(labeled “Ideal Self” [IS]). In yet a different administration, respondents were
asked to complete the form as to how they imagine a signiﬁcant caretaker (e.g.,
mother, father, grandparent, or other) might characterize them (labeled

“Projected Sell” [PS]).

53

Analysis

Splitting scores were obtained in the following manner: participants
completed each of the three versions of the ACL, (b) each of the eight scales is
comprised of a pre-selected combination of adjectives, and a raw score for each
scale is obtained by counting the number of adjectives selected for each
particular scale, (c) conversion of the raw to standard scores is determined by
subgroups deﬁned by the gender of the participant and the total number of
adjectives selected for each scale, (d) the standard scores for the eight scales
for each version of the ACL are summed to yield a composite number that
represents a splitting index for the participant on each dimension of self-concept
(i.e. actual, ideal, and projected), (e) composite scores for the three self-concepts
were compared (actual versus ideal and actual versus projected), (f) the absolute
difference between the comparisons represents the numerical representation of
lntemal conﬂict splitting (actual versus ideal) and attribution splitting (actual
versus projected).

It should be noted that directionality of splitting was not considered as a
part of this study because it is not known which self-concept is used as a point of
reference. For example, it is not known whether a person uses his/her actual
self-perception as a reference for creating an ideal self-perception, or visa versa.
Therefore, it was felt that an absolute value of Splitting variance between self-

concepts offered a more global representation of an individual’s propensity to

split.

54

In order to examine the relationships between the dependent variables
and participant demographic variables, the means of the splitting variables were
compared across gender, class, age, gpa, and credits. A correlation matrix was
also used to examine associations between the dependent variables and
independent measures of discrepancy, standards, anxiety, and depression. To
answer the primary hypothesis (maladaptive perfectionism is associated with
splitting), a correlation matrix was created to test for a signiﬁcant relationship
between APS Discrepancy (a measure of maladaptive perfectionism) and the
dependent variables.

To detect possible order effects, six surveys were distributed, each with a
different sequence of assessments. For example, one survey included
assessments in the following order: APS-R, ACL (Actual), CES-D, ACL (Ideal),
PSWQ, ACL (Projected). Another survey included assessments presented in the
following order: CES-D, ACL (Ideal), PSWQ, ACL (Projected), APS-R, ACL
(Actual). Surveys were randomly distributed to participants. A one way ANOVA
was conducted on the dependent variables (lntemal Conﬂict and Attribution
Splits) to determine if any of the six orderings inﬂuenced the responses of
participants. A Tukey comparison was used to examine between group
vanaﬁons.

Because order effects were discovered to inﬂuence the pattern of answers
on the ACL scores and as a consequence, the dependent variables, order effects
were held constant in the correlational operations in order to eliminate their bias

on this analysis. This was accomplished by creating ﬁve dummy variables from

55

 

the six types of surveys distributed using the sixth order (PIA order) as a
reference value for the other ﬁve orders. The dummy variables were partialed out
of the correlational analysis to control for order effect.

Mediator Analysis

Kenny (2001) describes mediator variables as having an intervening effect
between predictor and outcome variables. Mediators are mechanisms that serve
to more fully explain variable relationships. These variables play an intermediary
role in a chain reaction process in which a predictor variable inﬂuences a
mediator variable, which in turn inﬂuences an outcome variable (Holmbeck,
1997). Holmbeck states that four criteria must be in order to validate the
existence of a mediator variable: “(1.) the predictor must be signiﬁcantly
correlated with the hypothesized mediator, (2.) the predictor must be signiﬁcantly
associated with the dependent measure, (3.) the mediator must be signiﬁcantly
related to the dependent variable, (4.) the impact of the predictor on the
dependent measure must be less after controlling for the mediator,” (p. 602).

To answer the second hypothesis (affect mediates perfectionism and
splitting), three multiple regression analyses were conducted for each
independent, mediator, and dependent variable combination. In the ﬁrst
regression, the association between perfectionism and depression/anxiety is
examined. In previous studies, depression and anxiety have been implicated as
affective characteristics of perfectionists and are identiﬁed as mediator variables
in this study. In the second regression, the association between perfectionism

and splitting is examined. In the ﬁnal regression, perfectionism and

56

depression/anxiety are used as predictors of splitting. If a true mediation effect
exists, the association between perfectionism and splitting is less when
depression/anxiety is present. The degree to which depression/anxiety reduces
the association between perfectionism and splitting is an indicator of the potency
of affect on splitting. As in the correlational analysis, order effects were used as
independent variables in each regression to control for order effect. All
signiﬁcance tests were performed with an alpha value of .05 with the anticipation

of medium effect sizes.

57

CHAPTER 4
Results

This chapter details the results of the study. A preliminary analysis
provides descriptive information regarding the four assessments used in the
analysis. The procedure used to detect assessment order effects is described
along with the results of this assessment. Finally, the major analyses are
presented.

Preliminary Analyses

Instrument ranges, means, standard deviations, and reliability estimates
are summarized in Table 2. Mean scores on the APS-R were 20.39 on Order,
40.68 for High Standards, and 44.56 for Discrepancy. These results are
comparable to other university samples using the APS-R. Reliability of the three
subscales for the APS-R ranged from .86 to .94.

The mean number of adjectives endorsed by participants on the ACL was
highest for ideal-self (M = 91.92, SD = 35.18) followed by true-self (M = 83.67,
SD = 39.13). The standard deviation for all categories was similar with a low of
33.80 (projected-self) to a high of 39.18 (actual self). An item analysis of the ideal
self-concept measure indicates that a minimum of 200 (76%) or more of the
participants endorsed 28 popular items. This is in contrast to the actual and
projected measures in which participants selected two popular items. Sixteen
participants were dropped from the study due to incomplete surveys or not

meeting the cut-off for ACL validity requirements.

Participants generally appeared to worry at a similar level to the general

58

population with a mean of 50.66 (SD = 15.07). In their comparison study of
anxious and nonanxious groups, Molina and Borkovic (1994) reported a mean of
47.65 (SD = 12.99) for the general population of their sample comprised largely
of college students. Depression for the study population appeared equal to the
typical cutoff score of 16 (Breslau, 1985; Husaini et al., 1980; Radloff, 1977;
Weissman et al., 1977). Fifty six percent of the sample scored at or below the
cut-off score of 16. Twenty ﬁve percent of the sample scored within in the 16 to
26 range (one standard deviation above the mean) or in the mild depressive
area. Twelve percent of the sample scored in the 26 to 36 range (two standard
deviations above the mean) classiﬁed as moderately depressed. Approximately
four percent of the sample exhibited high depression scoring above 36. These

higher scores have the effect of raising the overall sample mean.

59

 

Table 2

Descriptive Statistics

 

MEASURE Scale Scale
Range — Range -
Possible Actual M SD

 

Perfectionism
(Almost Perfect Scale —Revised)

Order
Standards
Discrepancy
Self-Concept
(Adjective Check List)
ACL (TRUE SELF)
ACL (Ideal Self)
ACL (Projected Self)
Wow
(Penn State Worry
Questionnaire)

PSWQ

Depression

(Center for Epidemiologic
Studies — Depression Scale)

CES-D

4-28
7-49
1 2-84

1 0-250
1 0-250
1 0-250

1 6-80

0-60

4-28
1 1-49

1 3-84

1 3-236
1 3-224

12-213

1 6-80

0-51

20.38
40.68

44.56

83.67
91.92

77.01

50.66

15.99

5.50
7.00

16.76

39.13
35.18
35.80

15.07

10.09

.88
.86

.94

.94

.89

 

60

Table 3 exhibits a comparison of means and standard deviations between
the men and women in the current study and as reported in the ACL manual
(1983). Means and standard deviations are reported for the “true self” only
because no previous studies have recorded these statistics for the measurement
of ideal or projected self-concepts. A t-test comparing genders of the current
study to the normative sample indicated signiﬁcant differences between the male
samples t(2, 5498) = 3.17, p < .001, and between the female samples I (2, 4424)
= 3.93, p < .001. The effect sizes of the samples were .33 for males and .31 for
females. Given the wide demographic variability between the samples,
heterogeneity of variance was assumed in conducting the t-test. See Table 3 for
comparison of gender means and standard deviations.

Table 3

Adjective Checklist True Self Comparison With Normative Population

 

ACL MALES MALES FEMALES FEMALES

MEASURE
OF SELF- (STUDY) (MANUAL) (STUDY) (MANUAL)

CONCEPT
M SD M SD M SD M SD

ACL (True
Self) 78.90 43.55 93.40 36.36 86.17 36.50 97.37 34.64
Table 4 exhibits the relationship between age, grade point average (GPA),
and the number of academic credits (Credits) claimed by participants in the term
in which they completed the survey. The table also displays correlations between
these variables, the perfectionism variables (Order, High Standards, and
Discrepancy), depression, anxiety, and the splitting variables. Age and GPA were

signiﬁcantly associated with Credits. High Standards was signiﬁcantly related to

61

GPA.
Table 4

Correlation of Age, GPA, and Credits with Study Variables

 

 

AGE GPA CREDITS

AGE - .12 -.13*
GPA - - .13*
CREDITS - - -
ORDER .05 .01 .06
HIGH STANDARDS -.02 .21“ .07
DISCREPACY .06 .05 -.02
DEPRESSION .04 .00 .03
ANXIETY .08 .11 .12
INTERNAL CONFLICT SPLITTING -.00 .08 .03
ATTRIBUTION SPLITTING .08 .04 .04

 

* p < 0.05 ,** p < 0.01
Table 5 is a comparison of gender means and standard deviations with

study variables. T-tests indicate that levels of anxiety were signiﬁcantly different

between men and women.

62

Table 5

Gender Means and Standard Deviations on Study Variables

 

 

_E_N .VMQM_EN
M SD M SD t

ORDER 20.19 4.88 20.49 5.83 .44
HIGH STANDARDS 40.17 7.43 40.95 6.78 .83
DISCREPANCY 42.32 16.01 45.73 17.08 1.60
DEPRESSION 14.01 9.03 16.98 10.50 2.38
ANXIETY 45.07 13.28 53.59 15.16 4.68*
INTERNAL CONFLICT 70.53 46.68 87.60 55.64 2.63
SPLITTING

ATTRIBUTION SPLITTING 67.50 46.20 71.83 39.05 .76

 

* t> 3.291 at p = .000, df= 262

Table 6 compares class level means and standard deviations with each of

the variables in the study. Anxiety was the only variable found to be signiﬁcantly

different between class levels of participants. Participants past their fourth year in

undergraduate education and graduate students represented less than six

percent of the total sample and were not included in the tab

63

Table 6

Comparison of Class Level with Study Variables

 

 

 

FRESH. SOPH. JUNIOR SENIOR

M SD M SD M SD M SD F
ORDER 20.00 5.19 19.98 6.06 20.92 5.24 20.24 6.11 .53
HIGH 39.08 8.08 40.92 6.69 41.46 6.20 40.92 6.20 .84
STANDARDS
DISCREPANCY 45.25 15.01 47.14 14.80 45.27 19.79 42.84 14.90 1.55
DEPRESSION 15.55 8.79 15.75 8.63 17.27 11.47 15.86 10.69 .73
ANXIETY 47.38 13.98 51.51 13.81 51.44 16.01 54.98 14.87 3.43“
INTERNAL 73.42 54.02 88.41 53.47 88.21 55.81 82.57 52.42 1.14
CONFLICT
SPLITTING
ATTRIBUTION 62.47 37.85 71.20 43.80 77.07 45.07 70.06 39.41 1.26
SPLITTING

 

* p < .005, df= 262

Order Effect

In an attempt to detect order effects, each participant was given one of six
different surveys, each with a different order of assessments. The ordering of
assessments should not be confused with the variable measure of Order, a
subscale associated with the Almost Perfect Scale-Revised. A Tukey HSD and
Kruskal-Wallis H test were used to examine the inﬂuence of test order on
participants. The ﬁndings of the Tukey HSD revealed no signiﬁcant differences
between groups for the internal conﬂict variable. The order analysis showed a
signiﬁcant order effect for the attribution splitting results. The API order group
showed signiﬁcantly higher than the PIA and IPA in the attribution splitting

scores. Table 7 describes the means, standard deviations between the

64

dependent variables and the associated order groups.

Table 7

Means of Six Combinations of ACL Assessments Distributed to Participants

 

 

 

 

INTERNAL
SEQUENCE CONFLICT ATTRIBUTION
m SPLITTING SPLITTING
M SD M SD
AIP 6.25 4.90 5.90 4.96
IPA 5.00 4.41 4.123 4.38
PAI 5.30 4.42 4.46 4.25
IAP 4.51 4.21 4.32 4.10
API 4.77 3.90 7.02 5.37
PIA 4.63 4.05 3.933 3.35

 

Note: A = Actual, I = Ideal, P = Projected;
3 denotes signiﬁcantly lower means than 7.02; **p < .001.

In order to further examine the inﬂuence of assessment order on
participants’ responses, a Kruskal-Wallis H test (KWANOVA) was used to
analyze survey order and participant response independence. The KWANOVA
detects differences in location and is the nonparametric equivalent of ANOVA.
The KWANOVA results indicate that a signiﬁcant order effect was detected in the

attribution splitting variable, F (5,256) = 3.32, X2 = 13.89, p < .001, but not the

lntemal conﬂict variable, F (5,256) = 1.03, x2 = 5.27, p > .01.

65

Analysis

Primary hypothesis: Maladaptive perfectionism is signiﬁcantly and
positively associated with splitting tendencies.

Previous studies have made use of cluster analysis as a method for
identifying maladaptive perfectionists (i.e., Parker, 1997; Rice 8 Mirzadeh, 2000).
However, in the current study, the results of this procedure did not converge on
an interpretable solution of maladaptive, adaptive, and non-perfectionist clusters.
It was decided to pursue an approach using continuous variables allowing the
researcher to study the dimensional characteristics of perfectionism rather than a
cluster focus.

The Discrepancy and High Standards subscales were signiﬁcantly
correlated (r = .23, p < .000). Using an analytic technique created by Blatt et al.
(1996), residualized versions of discrepancy and high standards were created in
order to remove shared variance between the variables. Blatt et al. described
the new variables as “pure” because the inﬂuence of one variable on the other
was effectively removed, retaining a cleaner measure of each subscale. In the
current study, controlling for High Standards created a pure measure of
Discrepancy, while a pure measure of High Standards was generated by
partialling out the effects of Discrepancy. Pure Discrepancy and pure High
Standards each correlated .97 with their respective original subscale. Pure
measures of Discrepancy (maladaptive perfectionism) and High Standards
(adaptive perfectionism) were used in all study analyses. A positive and

signiﬁcant association between Discrepancy and both internal conﬂict and

66

attribution splitting supported the ﬁrst hypothesis. Also of interest in Table 5 are
signiﬁcant negative correlations between High Standards and lntemal conﬂict
splitting, and between High Standards and Discrepancy. High Standards was
positively and signiﬁcantly associated with anxiety. Depression was signiﬁcantly
and positively associated with anxiety in addition to internal conﬂict splitting,
attribution splitting, and discrepancy. Anxiety was signiﬁcantly linked to all the
other variables. The results of the correlations between the dependent and
independent variables are detailed in Table 8.

Table 8

Correlation Matrix of Perfectionism, Splitting, and Distress Variables

 

 

INTERNAL ATTRIB. HIGH DISCR. DEPR. ANX.
CONFLICT SPLIT. STAND.
SPLIT.

INTERNAL -- .53“ -.19* .45“ .51“ .37"
CONFLICT
SPLITTING
ATTRIBUTION -- -.01 .30“ .38“ .25“
SPLITTING
HIGH -- -.27** -.09 .16“
STANDARDS
DISCREPANCY -- .56" .52“
DEPRESSION -- .56“

 

Note: *p<.01, **p<.001. Order effect controlled for in analysis. Residualized values of Discrepancy
and High Standards were used in the analysis.

Secondary hypothesis: Depression and anxiety mediate the maladaptive
perfectionism/splitting association.

As outlined by Holmbeck (1997), three regressions were completed for
each model: (Step 1) anxiety and depression were regressed on Discrepancy,

(Step 2) lntemal or attribution splitting were regressed on Discrepancy and

67

(Step 3) lntemal conﬂict and attribution splitting were regressed on both anxiety
and depression. Differences in the standardized beta coefﬁcients were compared
across regressions. The mediation effect was determined by a decrease in the
beta coefﬁcient between the independent and dependent variables when either
anxiety or depression was included in the analysis. Order effects and gender
were controlled in each regression.

lntemal Conﬂict Splitting

In Step 1, Discrepancy and anxiety were signiﬁcantly associated, R2 =
.36, F(7, 254) = 20.43, p < .001 as were Discrepancy and depression, R2 = .33,
F(7, 251) = 17.67, p < .001. In Step 2, Discrepancy and lntemal conﬂict splitting
were signiﬁcantly associated after controlling for order effects, R2 = .24, F(7, 254)
= 12.78, p < .001.

In Step 3, Discrepancy and depression (and anxiety) combined to account
for signiﬁcant variation in splitting. Upon comparing the standardized beta values
between Step 1 and Step 3 of the analysis, it was discovered that both anxiety
and depression satisﬁed the required criteria as mediators (the standardized beta
value between Discrepancy and internal conﬂict splitting was reduced with the
addition of anxiety [AB = .11] and depression [AB = .221). Anxiety appears to
signiﬁcantly mediate Discrepancy and lntemal conﬂict splitting. Nevertheless,
when mediated by anxiety, Discrepancy and internal conﬂict splitting were
remained signiﬁcantly associated, R2 = .28, F(8, 253) = 12.53, p < .001.
Depression also seems to signiﬁcantly mediate Discrepancy and internal conﬂict

splitting. When mediated by depression, Discrepancy and internal conﬂict

68

splitting remained signiﬁcantly associated, R2 = .36, F(8, 250) = 17.27, p < .000.
The standardized beta results for each step of the analyses are summarized in
Table 9. All standardized betas were signiﬁcant at p < .001.

Table 9

Mediation effect of anxiety and depression on maladaptive perfectionism
(Discrepancy): lntemal Conﬂict Splitting

 

CRITERIQN PREDICTOR UNSTANDARDIZED STANDARDIZED

 

 

COEFFICIENTS COEFFICIENTS
(3 SE3 (3 t
Step Anxiety Discrepancy 7.46 .76 .49“ 9.80
1
Step Internal Discrepancy 22.91 2.90 .45“ 7.90
2 Conﬂict
Splitting
Step lntemal Discrepancy 17.88 3.35 .34" 5.33
3 Conﬂict
Splitting
lntemal Anxiety .67 .24 .19* 2.86
Conﬂict
Splitting
Step Depression Discrepancy 5.63 .53 .56“ 10.70
1
Step lntemal Discrepancy 22.91 2.90 .45“ 7.90
2 Conﬂict
Splitting
Step lntemal Discrepancy 12.17 3.31 .23” 3 .68
3 Conﬂict
Splitting
Internal Depression 1 .97 .33 .37“ 6.00
Conﬂict
Splitting
* p <. 01, **p<.001, df= 262
69

Attribution Splitting

In Step 1, Discrepancy and anxiety were signiﬁcantly associated, R2 = .36,
F(7, 254) = 20.43, p < .001, as were Discrepancy and depression, R2 = .33, F(7,
251) = 17.67, p < .001. In Step 2, Discrepancy and attribution splitting were
signiﬁcantly associated after controlling for order effects, R2 = .11, F(7, 254) =
4.26, p < .001.

In Step 3, Discrepancy and anxiety (but not depression) combined to
account for signiﬁcant variation in splitting. A comparison of the standardized
beta values between Step 1 and Step 3 of the analysis offered evidence that
anxiety satisﬁed the required criteria as a mediator (the standardized beta value
between Discrepancy and attribution splitting was reduced with the addition of
anxiety [AB = .071). Anxiety appears to signiﬁcantly mediate Discrepancy and
attribution splitting. When mediated by anxiety, Discrepancy and attribution
splitting remained signiﬁcantly associated, R2 = .17, F(8,250) = 6.30, p < .000.
The standardized beta results for each step of the analyses are summarized in

Table 10. All standardized betas were signiﬁcant at p < .001.

70

Table 10

Mediation effect of anxiety and depression on maladaptive perfectionism
(Discrepancy): Attribution Splitting

 

 

 

 

RITERI N PR DI TOR UNSTANDARDIZED STANDARDIZED
COEFFI IENTS COEFFICIENTS
13 SE3 (3 t

Step Anxiety Discrepancy 7.46 .76 .49“ 9.80
1
Step Attribution Discrepancy 8.43 1.71 .30“ 4.94
2 Splitting
Step Attribution Discrepancy 6.63 2.00 .23“ 3.32
3 Splitting

Attribution Anxiety .24 .14 .13 1 .72

Splitting
Step Depression Discrepancy 5.63 .53 .56" 10.70
1
Step Attribution Discrepancy 8.43 1.71 .30” 4.94
2 Splitting
Step Attribution Discrepancy 3.72 2.01 .13 1.85
3 Splitting

Attribution Depression .86 .20 .30“ 4.30

Splitting
**p<.001, df = 262

Sobel Test of Mediation

A Sobel test (1982) was performed as a test of mediator signiﬁcance. The
Goodman (l) (1960) version of this test calculates the critical ratio as a test of
whether the indirect effect of maladaptive perfectionism on splitting via the
mediator is signiﬁcantly greater than zero. This test provides a sample-based
estimate versus the more conservative approximation of the Sobel. As Baron and

Kenny (1986) suggest, the Goodman (I) was selected because it does not make

71

the assumption that the product of the standard errors is insigniﬁcant. The
Goodman (I) test indicates that the mediation effect created by the addition of
either anxiety or depression to discrepancy is signiﬁcant for both types of
splitting. Unstandardized beta coefﬁcients and standard errors were used to
calculate all Goodman (I) test statistics.

The Goodman (I) test indicates that the effect between Discrepancy
(maladaptive perfectionism) and lntemal conﬂict was signiﬁcantly mediated by
anxiety (G (I) = 4.67, p=. 000) and depression (G (I) = 3.47, p=.001). The effect
between Discrepancy and attribution splitting was also signiﬁcantly mediated by
anxiety (G (I) = 3.13, p = .002), but not depression (G (I) = 1.81, p =. 07).
Exploratory Analysis

If maladaptive perfectionism embraces the negative aspects of
perfectionism, then it stands to reason that adaptive perfectionism encompasses
its more beneﬁcial (opposite) qualities. It was reasoned that if maladaptive
perfectionists utilized splitting mechanisms, that adaptive perfectionists would
either use splitting in a more productive way or not at all. If it could be found that
adaptive perfectionists react differently or oppositely to splitting than their
maladaptive counterparts, then these ﬁndings may provide further insight into the
maladaptive perfectionism] splitting dynamic. Therefore, High Standards, a
deﬁning quality of adaptive perfectionism, was examined for its association with
splitting. As with maladaptive perfectionism and splitting, a series of steps was
used to examine multiple regression outcomes in this analysis. Order of the ACL

sequence was once again controlled in each regression.

72

Step 1 analyses indicated that High Standards was signiﬁcantly
associated with anxiety, R2 = .14, F(7, 254) = 6.01, p < .000. High Standards
was not associated with depression, R2 = .03, F(7, 251) = 1.25, p < .278.
Because of the lack of signiﬁcant ﬁndings between High Standards, depression,
and attribution splitting, these variables were not examined further in this study.

In Step 2, High Standards was signiﬁcantly associated with internal
conﬂict, R2 = .12, F(7, 254) = 4.61, p < .000. High Standards was not associated
with attribution splitting, R2 = .02, F(7, 254) = .71, p < .663. Due to a lack of
signiﬁcant effects between High Standards and attribution splitting, no further
analyses were conducted on these variables.

In Step 3, anxiety was found to be signiﬁcantly associated with High
Standards and internal conﬂict splitting, R2 = .26, F(8, 253) = 11.33, p < .000. An
examination of the standardized beta coefﬁcients between Step One and Step
Three for each of the variable combinations revealed that only anxiety met the
required criteria for mediation (the standardized beta value between High
Standards and lntemal conﬂict splitting was reduced with the addition of anxiety
[AB = -.06]). The Goodman (I) Test for signiﬁcance indicates that High Standards
and anxiety were signiﬁcantly associated with lntemal conﬂict splitting (G (I) =
2.25, p = .023). The standardized beta results for each step of the analyses are

summarized in Table 11. All standardized betas were signiﬁcant at p < .001.

73

Table 11

Mediation effect of anxiety and depression on adaptive perfectionism (High
Standards): lntemal Conﬂict Splitting

 

 

 

RIT RION P DI T R UNSTANDARDIZED STANDARDIZED
C EFFI IENT COEFFICIENTS
I} SE: B I

Step Anxiety High 2.32 .89 .15" 2.63
1 Standards
Step lntemal High -9.98 3.18 -.19* 3.14
2 Conﬂict Standards

Splitting
Step lntemal High -13.42 2.94 -.25" -4.56
3 Conﬂict Standards

Splitting

lntemal Anxiety 1.48 .21 .42“ 7.20

Conﬂict

Splitting
Step Depression High -.96 .63 -.10 -1.52
1 Standards
Step lntemal High -9.98 3.18 -.19* -3. 14
2 Conﬂict Standards

Splitting
Step lntemal High -7.34 2.78 -.14* -2.64
3 Conﬂict Standards

Splitting

lntemal Depression 1 .97 .33 .37“ 6.00

Conﬂict

Splittirg

 

* p <. 01, **p<.001, df= 262

Interestingly, the standardized beta value between High Standards and

lntemal conﬂict splitting is negative and signiﬁcant, suggesting an inverse

relationship between these two variables. Anxiety appears to augment this

relationship. The negative and signiﬁcant association between High Standards

and lntemal conﬂict splitting contraindicate mediation, in which a reduction in a

74

positive relationship between the independent and dependent variable is
expected. An indirect effects model may further explain these ﬁndings.
Summary of Findings

The results of this analysis provide evidence that maladaptive
perfectionism is signiﬁcantly associated with splitting. Internal conﬂict and
attribution splitting are related to maladaptive perfectionism. Anxiety and
depression signiﬁcantly mediate the maladaptive perfectionism/splitting
relationship. The results of the exploratory analysis provide further information
regarding the dynamics of perfectionism/splitting associations. High Standards
appear to be signiﬁcantly and inversely related to splitting. Anxiety is an
intervening variable that appears to signiﬁcantly enhance the inverse relationship
between High Standards and lntemal conﬂict splitting. Mediators are variables
that help to explain the inﬂuence of predictors on outcomes. Because anxiety
decreases High Standards association with lntemal conﬂict splitting, it is not
considered a mediator. It would appear that when adaptive perfectionists become
anxious, their ability to set and perhaps achieve high personal standards might
be a factor, reducing their preference for internal conﬂict splitting. These ﬁndings

will be discussed in greater detail in Chapter 5.

75

CHAPTER 5
Discussion
Summary of Results

The primary purpose of this study was to ﬁnd empirical evidence
supporting theoretical and anecdotal propositions that perfectionists utilize a
psychological splitting mechanism. The main hypotheses predicted that
maladaptive perfectionists would exhibit splitting, exhibited as incongruity in self-
concept. Splitting styles were conceptualized as “Internal Conﬂict Splitting”
(actual versus ideal self-concept) or “ Attribution Splitting” (actual versus
projected self-concept). A secondary thesis of the study was that anxiety and/or
depression are mitigating factors that mediate maladaptive perfectionism and
splitting. Speciﬁcally, it was posited that depression mediates perfectionism and
lntemal conﬂict splitting whereas anxiety mediates perfectionism and attribution
splitting. An exploratory analysis examined the association between adaptive
perfectionism (deﬁned as high standards) and both types of splitting as well as
the roles of anxiety and depression in this relationship.

The study provided evidence that maladaptive perfectionism is
signiﬁcantly correlated with both lntemal conﬂict and attribution splitting. Anxiety
and depression appear to be signiﬁcantly related to maladaptive perfectionism.
Anxiety proved to signiﬁcantly mediate the effects of maladaptive perfectionism
on both lntemal conﬂict and attribution splitting. Depression, however, only
appears to mediate maladaptive perfectionism and lntemal conﬂict splitting. The

Goodman (I) tests provided empirical validation for these ﬁndings.

76

It was discovered that adaptive perfectionism was inversely associated
with lntemal conﬂict splitting and not attribution splitting. Further, anxiety was
found to increase the inverse relationship between adaptive perfectionism and
internal conﬂict splitting. Depression was not identiﬁed as a mediator of adaptive
perfectionism and either form of splitting.

Explanation for Findings

Mala daptive Perfectionism and Splitting

A deﬁning feature of maladaptive perfectionism is a concern over failing to
meet self-deﬁned goals or standards or the perceived expectations of others.
This would suggest that individuals with maladaptive perfectionism are sensitive
to gaps between what they hope to accomplish and what they actually achieve.
Splitting, as deﬁned by this study, is the recognition of gaps in self-concept, and
therefore an indication of an intrapsychic tendency to polarize differences in self-
concept. In order to eliminate a highly anxiety provoking disparity between real
and ideal self-concept, it is the thesis of this study that maladaptive perfectionists
react by dichotomizing their thinking. Perfectionists believe that there is only one
right outcome and one way to achieve that outcome. Anything less than
perfectionism is considered a failure. Perfectionist behavior serves to control
events. The ﬁndings of this study appear to support the assumption that
maladaptive perfectionists are sensitive to a divided self-concept. In other words,
the more affected a person is by maladaptive perfectionism, the greater the
likelihood that they will split their thinking. In order to reconcile these differences,

maladaptive perfectionists may engage a splitting mechanism in which the gap is

77

effectively eliminated in order to maintain equilibrium of self-concept. Qualities of
maladaptive perfectionism such as fear of failure, self—doubt, relational problems,
and anxiety, appear to activate splitting.

A further ﬁnding of the study is the relationship of maladaptive
perfectionism to different forms of spitting. The results indicate that maladaptive
perfectionists are prone to internal conﬂict and attribution splitting. Internal
conﬂict splitting is operationally deﬁned as the magnitude of differences between
the actual and ideal selves. In addition, Gestalt Therapy further distinguishes
lntemal conﬂict spitting as the difference between intemallzed standards and
values versus emotional reactions, needs and wants. The ﬁndings of the study
support the notion that the drive to achieve an ideal self-image may be driven by
fear and doubt. Maladaptive perfectionists also appear to attribute their drive to
excel to others' expectations. This type of splitting is deﬁned operationally as the
difference between the actual and projected selves and theoretically as the
difference between the accepted injunctions and perceived standards of others in
conﬂict with personal desires and wants. This ﬁnding would explain eating
disordered individuals, for example, who strive to conform to a model of
perfection that is perceived as a demand of the individual by others. The crucial
factor in motivating this type of thinking is the belief that a person is not
acceptable to others unless they are able to meet the perceived requirements of

others.

78

Anxiety and Depression as Mediators
The study indicates that anxiety and depression play mediating roles in the

perfectionism/splitting relationship. Emotions play a key role between
perfectionism and splitting. Anxiety is associated with both types of splitting,
while depression appears to mediate only lntemal conﬂict splitting. The
signiﬁcant role of negative emotions supports the idea that maladaptive
perfectionists are sensitive and reactive to gaps or experiences that challenge
the way in which they perceive themselves. These same individuals appear
vigilant regarding how others view and value them as well. Anxiety and
depression are signiﬁcant affective agents that serve to enhance
perfectionism/splitting dynamics. Said another way, if these agents are removed,
the relationship between maladaptive perfectionism and splitting is less potent.

When maladaptive perfectionists seek to achieve a lofty goal, anxiety
and/or depression result from an insecure ego and fragile self-esteem. Splitting
may be engaged as an avoidant coping mechanism in order to reduce distress
and feelings of vulnerability and shame. For example, maladaptive perfectionists
may experience stress resulting from a fear of failing to meet an expected
standard of performance (their own or the perceived expectations of others). The
source of this fear may result in a perceived threat to a fragile sense of pride and
the way in which maladaptive perfectionists view themselves. This threat may
create an imbalance between the actual versus ideal self-image. In order to
reduce distress, the perfectionist utilizes increasing degrees of splitting to cope

with the emotional disorientation created by the perceived gap in self-concept.

79

The resulting “black or white“ view of the world serves to eliminate the
indecipherable and threatening “gray” of incongruent self-concepts. An internal
state of limbo created by different and possibly conﬂicting self-concepts may
challenge an individual’s beliefs about themselves and understanding of others.
Contradictory information leads to feelings of confusion and doubt about one’s
beliefs and perceptions. Splitting enables maladaptive perfectionists with the
ability to simplify their thinking. This process helps to reduce feelings of fear and
confusion. By channeling their thoughts through a “black and white" ﬁlter,
perfectionists are able to regain an lntemal sense of stability and equilibrium
thereby reducing confusion and self-doubt.
Adaptive Perfectionism and Splitting

In contrast to maladaptive perfectionism, adaptive perfectionism appears
to have the opposite effect on splitting. Interestingly, the results indicate that
adaptive perfectionism (as deﬁned by the presence of high standards) is
negatively related to lntemal conﬂict splitting The literal interpretation of this
ﬁnding indicates that individuals who maintain high standards are less likely to
split. It should be noted that the inﬂuence of maladaptive perfectionistic
characteristics were removed from the APS-R variable of High Standards in the
analysis. Fear of failure, self-doubt, and other negative attributes of perfectionism
were extracted from the High Standards variable. Therefore, it may be concluded
that high standards in and of themselves are not necessarily detrimental to

perfectionists. In fact, high standards may be a factor of resilience that helps to

80

prevent high achievers from splitting. It could be that a side beneﬁt of being an
adaptive perfectionist is better mental health.

Additionally, the study found that when anxiety is removed from the
perfectionism/splitting relationship, perfectionism is even less related to internal
conﬂict splitting. Conger (1974) deﬁned a suppressor variable as one that
increases the predictive validity of another variable (or set of variables) by Its
inclusion in a regression equation. Classic suppressor variables carry no
signiﬁcant relationship with criterion variables and serve to “clean” irrelevant
variance from predictor variables, subsequently improving the predictive power of
the predictor. In this study, anxiety has a signiﬁcant relationship with internal
conﬂict splitting, and further “puriﬁes” High Standards after partialling out
Discrepancy effects, improving its predictive power with splitting. Since there is
an inverse relationship between adaptive perfectionism (High Standards) and
lntemal conﬂict splitting, anxiety increases the polarity of this relationship (i.e.
adaptive perfectionism becomes more negatively correlated with splitting). This
association exists despite anxiety’s signiﬁcant correlation with internal conﬂict
splitting.

The result of this ﬁnding suggests that the presence of anxiety in adaptive
perfectionists may impede their ability to resist splitting behavior. In other words,
anxiety may create a vulnerability factor for adaptive perfectionists. It could be
that adaptive perfectionists establish lofty goals with less fear of failure than
maladaptive perfectionists, allowing them to experience greater purpose in their

efforts and a reduced need to split. Perhaps these individuals feel that they have

81

 

 

more reasonable expectations of themselves than do their maladaptive
counterparts. Adaptive perfectionists may also feel less threatened, wounded, or
perhaps shamed, when they do not achieve their goals, allowing for more
effective maintenance of a polarized self-concept. Not only do maladaptive
perfectionists have difﬁculty appreciating their accomplishments, but also they
may ascribe a different meaning to their achievements than do adaptive
perfectionists.

Non-Signiﬁcant Variables

Several non-signiﬁcant associations were discovered. Depression does
not appear to mediate maladaptive perfectionism and attribution splitting, despite
being individually associated with these variables. This ﬁnding suggests that
maladaptive perfectionists are more prone to experience anxiety than depression
when they internalize others’ expectations of their performance. These results
support previous research regarding anxiety’s stronger link to perfectionism than
depression (Kawamura, et al., 2001; Norman et al. 1998).

Depression was not correlated with high standards (adaptive
perfectionism) or attribution splitting. It was negatively associated with internal
conﬂict splitting. These ﬁndings suggest that depression has a less potent
relationship with adaptive perfectionism than anxiety. And although anxiety was
found to negatively mediate adaptive perfectionism and internal conﬂict splitting,
it did not mediate this type of perfectionism and attribution splitting. These
ﬁndings suggest that affect does not play a signiﬁcant mediating or suppressing

role between adaptive perfectionism and attribution splitting.

82

Comparison to Previous Studies

Currently, no study has been found that speciﬁcally duplicates the focus or
method of the current research on splitting and perfectionism. However, several
studies are similar to portions of the current study.

The current study supports eariier research linking perfectionism with
anxiety and depression (Alden et al., 1994; Blatt, 1995; Blatt et al., 1995; Flett et
al., 1989; Hill et al., 1997; Lynd-Stevenson 8 Heame, 1999). As in previous
studies, maladaptive perfectionism was associated with worry in the current
study (Juster et al., 1996, Kawarmura et al., 2001; Stbber 8 Joorman, 2001).
Kawamura, et al. (2001) discovered that depression was not as strongly
associated with perfectionism than anxiety, a ﬁnding supported by the current
research. Alternately, the ﬁndings of Stbber and Joorman (2001) that wony was
not related to high standards, was not supported by the current research.
However, it should be noted that the characteristics of Discrepancy were
partialed out from Standards for the purpose of creating a more “pure” version of
the Standards variable for analysis. Wony, and similar characteristics, may have
been removed from the Standards variable as a result.

Studies into perfectionism indicate that individuals prone to negative
emotional states such as depression, anxiety, and anguish are also likely to carry
discrepant self-images. Likewise, achievement, self-esteem and other positive
psychological states are negatively correlated with Discrepancy (Hewitt 8 Flett,
1991 a, 1991b; Slaney, et al., 1998). Past research into self-discrepancy purports

that depression is related to actual/ideal discrepancy while anxiety is related to

83

actual/ought (in which the ought-self is identiﬁed as “projected” in the current
study) self-image (Hankin et al., 1997; Higgins et al., 1985; Scott 8 O’Hara, 1993;
Strauman, 1989). The emotional/seIf-image relationships of these studies are in
contrast to a study by the authors of the Adjective Check List in which

participants were generally found to be more anxious when their actual and ideal
self-concepts are incongruent (Gough et al., 1983). The ﬁndings of the current
study appear to support the Gough et al. ﬁndings. The differences between the
ﬁndings may be attributed, at least in part, to how the researchers

operationalized differences in self-concept.

One difference between the current and previous research was the
method for determining differences in self-concept. In the other studies, self-
discrepancy was typically measured by requiring participants to create lists of
descriptive adjectives associated with each self-concept (actual/ideaI/ought) and
then comparing the similarities and differences of the lists rather than responding
to a predetermined list of adjectives (The Selves Questionnaire, Higgins et. al.,
1985). Measures of depression and anxiety were correlated with these self-
generated lists in order to examine various discrepancies. The researchers felt
that the use of an adjective checklist of adjectives would constrain the
spontaneity of participants’ ability to maximize the likelihood that the listed
attributes would be important to the respondent. A limitation of this methodology
as compared to the current study is the lack of generalizability of these studies.
By offering participants a controlled list of descriptors, all respondents are

accessing their splitting capacity from closely deﬁned adjectival parameters. An

84

added beneﬁt of adjective checklists is that they allow for cross-group
comparisons, which past studies do not.

Another difference between the current study and a related study (Hankin
et al., 1997) is the difference of perfectionism measures. The Hankin et al. study
utilized the Multidimensional Perfectionism Scale (MPS: Hewitt et al., 1991) as a
means of measuring perfectionism. The scales of the MPS measure three
different types of orientation to perfectionism (self, other, and socially
prescribed). The Almost Perfect Scale focuses on the character qualities of
order, standards, and a person’s tendency to self-deprecate. In other words, the
MPS largely differentiates perfectionists by their perceived locus of control
(external versus lntemal) while the APS-R offers the facility to categorize
perfectionists into good or bad groupings. While the actual and ideal selves of the
multidimensional scales are metrics similar to those used in the current study, the
ought scale focuses on an individual’s sense of obligation to others’ injunctions
rather than the projection itself — the metric used in this research. The Hankin et
al. study attempted to explore the role of excessive standards in depression and
anxiety during adolescence in contrast to the current study attempted to
delineate among “good" and “bad" qualities of perfectionism and how these
qualities affect or are affected by emotionality and splitting. The incongruent
ﬁndings among the current and past studies may be conceptual differences
between how perfectionism is operationalized.

An additional complication that appears consistent with both the past and

current research may be the attempt to integrate two similar but different theories

85

 

of perfectionism (as deﬁned by the MP8 and APS) with self-discrepancy. Both
theories made use of measures of self-standards, but did not necessarily target
the same constructs. Self-discrepancy theory posits that differences between
self-concepts will result in emotional distress. Perfectionism, as deﬁned by the
APS-R, suggests levels of distress resulting from unmet standards, but not
necessarily as the result of incongruent self-concepts, a feature of the
Multidimensional Perfectionism Scale. The differences of ﬁndings between past
and current studies in this area suggest that there may be signiﬁcant differences
between these theories that should be explored more fully. For example, ﬁndings
of the current study provide evidence that distress is a by—product of maladaptive
perfectionism that may lead to splitting (perceived differences in self-concept). It
may be useful to explore the affect of maladaptive perfectionists before and after
an episode in which they have made an unsuccessful attempt at attaining a goal.
A study of this kind may elucidate the dynamics of splitting and self-concept
further.
Implications

This study has shed light on the relationship between perfectionism,
affect, and splitting. Additionally, this paper is helpful in identifying differences
between maladaptive versus adaptive perfectionists. A product of this study was
a clearer understanding of how these variables interact, leading to useful clues in
creating interventions that may alleviate the deleterious effects of maladaptive
perfectionism. For example, it will be important for practitioners to recognize the

connection between adverse perfectionism and splitting, and the role of affect in

86

mediating these two characteristics. The results of this study may offer mental
health practitioners clues in assisting maladaptive perfectionists to recognize how
maladaptive traits may trigger a system of splitting. And further, that these traits
may be enhanced by anxiety and/or depression, depending upon the splitting
style of the perfectionist. Anxiety, for example, appears to be a negative element
common to both maladaptive and adaptive perfectionists. It was discovered that
anxiety plays a role in mediating maladaptive perfectionism and splitting as well
as suppressing adaptive perfectionism and internal conﬂict splitting. It may be
useful for clinicians to educate clients in methods of stress relief and problem
solving that could reduce the adverse effects of anxiety. Depression was
discovered to be a somewhat less potent mediator of perfectionism and splitting.
Practitioners and researchers may ﬁnd more value in attending to the level of
anxiety that clients or study participants are experiencing and give less attention
to depressive affect.

Additionally, it may be useful for practitioners to help their perfectionistic
clients explore their cognitive styles. Clinicians attempting to ameliorate the ill
effects of perfectionism may ﬁnd it useful to conduct a thorough exploration of the
maladaptive behaviors resulting from perfectionist/splitting dynamics. It would be
helpful to know whether an all-or-nothing thinking style is a common occurrence
for the client. The frequency of dichotomous thinking may reveal the extent to
which a client utilizes a splitting mechanism to support and maintain a
perfectionist pattern. Clinical interventions aimed at increasing awareness and

appreciation of conﬂicting internal polarities may effect a reduction in

87

perfectionism. Gestalt Therapy offers a therapeutic framework in which polarities
may be explored, thereby reducing maladaptive patterns of splitting that plague
perfectionists. For example, if perfectionists could be taught to recognize patterns
of self-recrimination when they fail to meet their goals, they may learn how to
better cope with disappointment, thereby averting future maladaptive habits of
perfectionism. A therapeutic experiment in which the perfectionist is encouraged
to verbalize their thoughts immediately following a failed attempt to achieve a
goal may elucidate lntemal messages that reinforce behavioral patterns that
support and maintain cycles of shame and blame. Exploring the connection
between dichotomous thinking and the meaning perfectionists give to the pursuit
of their goals, accomplishments, and failures may enlighten the practitioner
regarding the perfectionism/splitting relationship.

A further implication is to assessment considerations. Given the results of
this study and the important mediating roles of affect, it seems important that
levels of anxiety and depression should be assessed when using perfectionism
instruments. Practitioners and researchers may omit a crucial determinant of the
adverse effects of perfectionism affecting respondents if anxiety and depression
are not assessed when detecting perfectionism. When used in combination with
the APS-R, for example, an anxiety survey could provide information as to the
current extent of maladaptive perfectionism experienced by the respondent. The
assessment of anxiety may also prove useful to adaptive perfectionists who may

be thwarted in their ability to achieve their goals by a high level of distress.

88

As indicated earlier by Pruyser (1975), splitting is a difﬁcult concept to
clearly deﬁne. The current study examined splitting as a difference between self-
concepts. The study provides evidence that splitting is associated with
perfectionism, anxiety, and depression. It appears to be positively related to
maladaptive perfectionism and negatively associated with adaptive
perfectionism. These ﬁndings appear to support the theoretical context in which
splitting was deﬁned for this study. However, as noted in the literature review, the
underlying mechanisms and process by which splitting operates is dependent
upon the theoretical framework in which splitting is deﬁned. In the current study,
Gestalt Therapy is used to explain how and why splitting and perfectionism are
related. Empirical evidence resulting from this study offers support for the
conceptualization of Gestalt polarities and the process of splitting. Consideration
of other theoretical frameworks may offer further insight that may add to
understanding the nature of splitting and more accurate ways in which it may be
operationalized for future research.

Theoretical Implications to Maladaptive Perfectionism

The current study provides support that maladaptive perfectionists are at
risk of distress. These individuals may have character qualities that make them
more vulnerable to psychological ambiguity, distress, and the resulting
dysfunctional uses of mental mechanisms and defenses than other groups.
Because of their inability to tolerate cognitive dissonance, it is theorized that
maladaptive perfectionists tend to polarize their thinking to reduce distress

resulting from incongruent self-concepts. Gestalt Theory suggests that

89

maladaptive perfectionists will intellectually gravitate toward one pole over
another in order to eliminate differences in self-concept. Because maladaptive
perfectionists fear failure and the resulting feelings of inadequacy, it seems
plausible that these individuals would habitually gravitate toward an ideal self-
image. Maladaptive perfectionists are constantly running from the threat of a
divided self-concept and in pursuit of an ideal that can never be realized, leaving
these individuals forever fettered to personal dissatisfaction.

According to Greenberg (1993), internal conﬂict arises out of a difference
between intemallzed standards and values versus organismic emotional
reactions and needs or wants which are in opposition. Attribution splitting occurs
when a person’s wants/desires are in conﬂict with the projected injunctions and
judgments of others. In accepting judgments about themselves, these individuals
become self-critical and self-pressuring. Because of maladaptive perfectionists'
tendency to form rigid categories of experience, it is likely that these individuals
have a greater likelihood to experience extremes in self-concept. Maladaptive
perfectionists most likely have a strong reaction (both in degrees of affect and
splitting) to shifts in self-concepts. As described in terms of Gestalt Therapy, the
maladaptive perfectionist selects the top-dog, a self-concept that is sensitive to
socially deﬁned values. This is at the expense of the individual’s own emotional
needs and wants. The splitting process serves to partition thought, thereby
reducing the anxiety/depression that would otherwise result from experiences
that conﬂict with established perceptions of self-concept. Figure 1 illustrates the

relationships between discrepancy, internal conﬂict splitting, and emotion.

90

Figure 1

The relationship between discrepancy, splitting, and emotion

 

 

 

 

 

Anxiety or
Depression
(Mediators)
/ °<>\
00
Maladaptive Internal
Perfectionism Conﬂict or

Attribution
Splitting

 

 

 

 

The ﬁndings of the study support this theoretical postulate. Across all
attributes of maladaptive perfectionism, as anxiety and depression increase,
splitting increases between actual/ideal or actual/projected self-images.

Similarly, as anxiety and depression decrease, discrepancy decreases between
the self-images.

From a Gestalt Therapy perspective, this study is valuable because it
supports the use of gestalt interventions that could help to diminish Splitting,
reduce anxiety, and perhaps even reduce maladaptive perfectionism. For
example, maladaptive perfectionists may be educated to recognize the emotional

effect of a discrepant self-image and to understand the internal mechanism of

91

splitting inﬂuencing their feelings and behavior. A gestalt therapist might
encourage a maladaptive perfectionist to increase their awareness of their
polarizing through the use of an intervention, called an “experiment" in gestalt
terminology. The use of an experiment, such as a two-chair technique, might
serve to integrate the highly developed polarities of perfectionists. By opening a
dialogue between the split polarities, increased awareness and insight into the
internal process of splitting may be identiﬁed. Once this process is identiﬁed, it
would enable perfectionists to observe their own pattern of splitting, and then
make more constructive choices regarding their thoughts and behaviors than
previously available. This intervention would hopefully reduce perfectionists’
typical reaction to failure, thereby reducing anxiety and gaps in self-image, and
perhaps even maladaptive perfectionism itself.

Study Limitations
An attempt was made in the study to capture a broad cross-section of university
students with varying class associations, cultural diversity, and gender. The
demographic data collected from the respondents did not require race or ethnic
identiﬁcation. The majority of sampling was conducted at the University of
Oregon, which reports an ethnic population comprised of 13.1% non-
White/Hispanic students. It is assumed that a portion of this ethnic audience is
reﬂected in the sample. Class afﬁliation in the undergraduate sample was
relatively evenly distributed. Graduate students and undergraduates in their ﬁfth

year of study represented 6% of the sample.

92

A factor that may have unduly inﬂuenced the results of the study was the
gender sampling in which females outnumbered males by a margin of two to one.
Women were shown to have a slightly greater propensity to split than men. This
difference may have had the effect of inﬂuencing the splitting comparisons,
creating a larger splitting effect than would be found in a more gender-balanced
population.

Data collection procedures were carefully followed throughout the study.
However, several collections were made in large group settings of fraternity and
sorority organizations in which the presence of student peers may have unduly
inﬂuenced the responses of the participants. Peer pressure may have especially
impacted groups of young males in fraternity settings in which denial is a
common characteristic of male members. When possible, attempts were made
to separate the participants so as to reduce the temptation to interact with fellow
respondents.

A difﬁculty in conducting a survey that requires respondents to complete
several instruments is the possibility of order effects. In order to evaluate and
provide a measure of control for this issue, six possible arrangements of
instrument were randomly distributed to respondents. Order effect was found to
confound the responses of a portion of the sample. A potential problem
associated with the distribution of differing orders of instruments is in participants’
ability to gauge different self-concepts. Self-images are typically based on an
individual’s actual perception of self. ldeal and projected self-concepts are

examined against the context of an individuals’ actual self-concept. Those

93

participants who were not directed to evaluate their actual self-concept before the
other self-concepts may have been at a disadvantage in accurately explicating
the characteristics of their ideal and projected self-concepts.

A methodological concern was the manner in which the Adjective Check
List was used as a measure of splitting. The current study followed a method of
exploring differences between self-concepts as developed by the authors of the
ACL (Gough et al., 1983). However, unlike the 1983 study, a composite score of
the eight scales was used as a global representation of splitting propensity rather
than an examination of each of the eight scales individually. Also, no distinction
was made as to the positive or negative nature of personality as represented by
each scale as it may relate to splitting. The choice of limiting the methodology in
this manner was a deliberate attempt to focus more on the extent of splitting and
its association to perfectionism rather than investigating the complexities of the
relationships between perfectionism and personality. Consideration of the
personality factors affecting or inﬂuencing the relationship between perfectionism
and splitting may offer greater detail into the mechanics behind this association.
Issues Regarding lntemal Validity

No distinction has been made regarding causation between the independent

and dependent variables. Given the nature of this descriptive study, it is not fully
known whether perfectionism, anxiety, and/or depression cause or inﬂuence
splitting or visa versa. Despite the discovery of a relationship in which anxiety
and depression serve to mediate perfectionism and splitting, it is difﬁcult to

determine whether reverse or reciprocal causation affected the study. For

94

example, it is not known whether perfectionism results from individuals with the
propensity to split.

As discussed earlier, it was discovered that perfectionism is a signiﬁcant
characteristic of splitting. However, other factors may signiﬁcantly inﬂuence
perfectionistic, emotional, or splitting behavior of the sample. Some of these
factors may include intelligence, emotional sensitivity, age and/or developmental
life-stage inﬂuence, hardiness, degree of optimistic or pessimistic outlook,
emotional tolerance, and temperament.

Issues Regarding Extemal Validity

Although the study draws from a relatively large university student
population, the ethnic balance of the University of Oregon may be not be
representative of similarly sized universities, thereby limiting the generalizability
of the research against more diversiﬁed populations. Cultural dynamics that may
be unique to the northwestern United States, Oregon, and the University of
Oregon may also limit generalizability. The large number of female respondents
may have inﬂuenced the results of this study, possibly compromising its
generalizability to university student populations.

A second area of concern is the operationalization of splitting. Although a
sound theoretical case was presented linking self-concept discrepancy with
splitting, the use of a measure that captures gaps in self-image may not have
adequately represented more traditional deﬁnitions of splitting as outlined by
earlier in this study by Marmar and Horowitz (1986). It could be that self-concept

discrepancy is not an indication of a person’s preference for one polar identity

95

over another. Although this study supports the proposition that affect and self-
discrepancy are related, splitting may not be inﬂuenced by root emotions but by
habitually formed cognitions or physiological factors not measured in this
research.
Statistical Problems

The ﬁrst attempt to analyze the Almost Perfect Scale was through a
cluster analysis process, a procedure commonly used to deﬁne groups of
maladaptive, adaptive, and non-perfectionists. This analysis did not yield three
groups that were cleariy deﬁnable and was therefore abandoned. However,
cluster techniques may often present difﬁculties in identifying meaningful
groupings of psychological characteristics. It was decided that a correlation
analysis would be used to explicate signiﬁcant relationships among the variables
and maintained for an examination of the APS continuous variables consistent
with the other measures. The decision to view the data in this manner enabled
the researcher to examine the dimensions of perfectionism rather than
categorizing the sample by speciﬁc characteristics. This method of deﬁning
perfectionism as either continuous or categorical variables however, remains in
debate.

Future Directions

This study attempted to quantitatively link perfectionism with dichotomous

thinking, a characteristic identiﬁed by researchers as a distortion of thinking

utilized by perfectionists. Examination of other cognitive distortions attributed to

96

perfectionists, such as overly moralistic self-evaluation and overgeneralization,
may also provide valuable insight into the minds of perfectionists.

An opportunity exists to utilize the breadth of personality scales of the
Adjective Check List. The current study derived scores of self-discrepancy from a
composite of eight scales of the Adjective Check List. An additional investigation
could include a composite of all 37 ACL scales to attain the self-concept
discrepancy values for internal conﬂict and attribution splitting. This analysis
would extend help in more accurately identifying discrepancies between self-
concepts.

One of the challenges of the study was in attempting to capture the
essence of the splitting mechanism through self-discrepancy. It may be
informative to repeat the study adding an additional measure of splitting (e.g. The
Splitting Index [Gould et al., 1996] or Splitting Scale [Gerson, 1984]). The focus
of these scales is on the lability of affect and may provide added insight into
splitting behavior in perfectionists.

Stress is a factor that has been associated with perfectionism (Chang,
2000; Chang & Rand, 2000). The researchers found that stress fully mediates
the inﬂuence of perfectionism on positive outcomes, while only partially mediating
perfectionism on negative outcomes, such as worry and negative affect.

Similarly, Dunkley et al. (2000) found that hassles, avoidant coping, and lack of
perceived social support mediate evaluative concerns perfectionism and distress.
In order to capture the inﬂuence of stress (as measured by anxiety and

depression) on perfectionists across a speciﬁc time period, it may be useful to

97

repeat the current study at different time frames during a university semester with
the same participants. Because it is believed that maladaptive perfectionists are
inﬂuenced by external validation, it may prove informative to measure this factor
in relation to self-image variables. A study of this nature may serve to more
clearly deﬁne the relationship between self-image, splitting, and perfectionism.

Self-reporting of perfectionism and splitting behavior lack observable
veriﬁcation. It may prove useful to ﬁnd participants who have been clinically
identiﬁed with perfectionism and splitting behaviors to repeat this study so as to
conﬁrm or dispute the ﬁndings of the current research.

Given the ﬁndings regarding adaptive perfectionism and splitting, it may
be useful to assess high performers on the basis of their propensity to split. A
study of this nature would serve to further validate the ﬁnding that high
standards, absent of their maladaptive qualities, are beneﬁcial. Athletes, for
example, may possess expect high standards of performance from themselves,
but may not exhibit splitting.

Finally, the development of clinical interventions in which to reduce
splitting, anxiety, depression and maladaptive perfectionism may be created and
explored as a result of this study. Gestalt Therapy appears to provide a
framework from which interventions may be effectively created.

Summary

It was the intent of this study to advance perfectionism research. This

study has sought to further deﬁne perfectionism and its relationship with other

variables. Speciﬁcally, the study has provided evidence supporting the

98

interaction of bad perfectionism with self-concept discrepancy deﬁned as
splitting. Further, this research has explored the relationships of perfectionism
with anxiety and depression, their internal dynamics, and subsequent effect on
the psyche. Hopefully, insight gained from this study will assist future researchers
as they continue to deﬁne and reﬁne our understanding of perfectionism.

Anne Lamott (1995) nicely summarized how perfectionists may constrict
their lives: “I think perfectionism is based on the obsessive belief that if you run
carefully enough, hitting each stepping-stone just right, you won't have to die.
The truth is that you will die anyway and that a lot of people who aren't even
looking at their feet are going to do a whole lot better than you, and have a lot

more fun while they're doing it.” (p. 28).

99

 

APPENDIX

Research Instruments

100

 

Appendix

Almost Perfect Scale - Revised

The following items are designed to measure certain attitudes people have toward themselves,
their performance, and toward others. It is important that your answers be true and accurate for
you. In the space next to the statement, please enter a number from “1” (strongly agree) to “7"
(strongly disagree) to describe your degree of agreement with each item.

 

 

 

 

 

 

 

 

 

STRONGLY AGREE SLIGHTLY NEUTRAL SLIGHTLY DISAGREE STRONGLY
AGREE AGREE DISAGREE DISAGREE
1 2 3 4 5 6 7
___1. l have high standards for my performance at work or at school.
__2. I am an orderly person.
_3. I often feel frustrated because I can't meet my goals.
_4, Neatness is important to me.
__5, If you don't expect much out of yourself you will never succeed.
_6. My best just never seems to be good enough for me.
_7. I think things should be put away in their place.
_8. l have high expectations for myself.
_9 I rarely live up to my high standards.
_10, I like to always be organized and disciplined.
‘1 1. Doing my best never seems to be enough.
~12, I set very high standards for myself.
__13, I am never satisﬁed with my accomplishments.
__14. I expect the best from myself.
‘15, I often worry about not measuring up to my own expectations.
_16, My performance rarely measures up to my standards.
_17, I am not satisﬁed even when I know I have done my best.
__18. I am seldom able to meet my own high standards for performance.
_19, I try to do my best at everything I do.
_20_ I am hardly ever satisﬁed with my performance.
_21, I hardly ever feel that what I’ve done is good enough.
__22. l have a strong need to strive for excellence.
23, I often feel disappointment after completing a task because I know I
‘ could have done better.
24. Using the scale above, please rate the degree to which you agree that

you are perfectionistic.

101

 

 

 

II:

Using the scale below, circle the number which best describes h

Center for Epidemiologic Study — Depression

DURING THE PAST WEEK.

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

20.

1.

2.

3.

4.
During the past week:

i was bothered by things that usually
don’t bother me ...............................

I did not feel like eating: my appetite was
poor ..............................................

I did not feel that I could not shake off
the blues even with help from my family
and friends ......................................

I felt that l was just as good as other
people ............................................

I had trouble keeping my mind on what I
was doing .....................................

I felt depressed ................................

I felt that everything I did was an
effort ..............................................

I felt hopeful about the future ..............
I thought my life had been a failure.......
I felt fearful ......................................
My sleep was restless .......................
l was happy .....................................
I talked less than usual ......................
I felt lonely ......................................
People were unfriendly ......................
i enjoyed life ....................................
I had crying spells .............................
I felt sad .........................................
I felt that people disliked me ................

I could not get 'going' ........................

= Rarely or none of the time (less than 1 day)
= Some or a little of the time (1-2 days)

= Occasionally or a moderate amount of time (3-4 days)
= Most or all of the time (5.7 days)

Rarely or
none of
the time

102

Some or
a little of
the time

NNNNNN

Occasionally or a
moderate
amount of the
time

3

ow often you felt or behaved this way —

Most or
all of
the time

Penn State Worry Scale

Please read the following statements and decide how closely each relates to you.
In the space next to the statement, enter a number from “1 ” (very typical of me)
to “5” (not at all typical of me).

 

 

 

Very typical of Not typical at
me all of me
1 2 3 4 5

 

 

 

 

 

1.

10.

11.

12.
13.
14.
15.

16.

If I do not have enough time to do everything, I do not worry about it.
My worries overwhelm me.

I do not tend to worry about things.

Many situations make me worry.

I know I should not worry about things, but I just cannot help it.
When I am under pressure I worry a lot.

I am always worrying about something.

I ﬁnd it easy to dismiss worrisome thoughts.

As soon as I ﬁnish one task, I start to worry about everything else I
have to do.

I never worry about anything.

When there is nothing more I can do about a concern, I do not worry
about it any more.

I have been a worrier all my life.

I notice that l have been worrying about things.
Once I start worrying, I cannot stop.

I worry all the time.

I worry about projects until they are all done.

103

 

 

The Adjective Check List

DIRECTIONS: This assessment contains a list of adjectives. Please read them quickly
and put an x on the space beside each one you would consider to be self-descriptive.
Do not worry about duplications, contradictions, and so forth. Work quickly and do not
spend too much time on any one adjective. Try to be frank, and check those adjectives
which describe you as you really are, not as you would like to be.

1 . _ Absent-minded 21. _ Bitter 41 . _ Conﬁdent

2. _ Active 22. _ Blustery 42. _ Confused

3. _ Adaptable 23. _ Boastful 43. _ Conscientious
4. __ Adventurous 24. _ Bossy 44. __ Conservative
5. _ Affected 25. _ Calm 45. _ Considerate
6. __ Affectionate 26. _ Capable 46. _ Contented

7. _ Aggressive 27. _ Careless 47. _ Conventional
8. _ Alert 28. _ Cautious 48. _ Cool

9. __ Aloof 29. _ Changeable 49. _ Cooperative
10. _ Ambitious 30. _ Charming 50. _ Courageous
1 1 . _ Anxious 21 . _ Cheerful 51. _ Cowardly

1 2. _ Apathetic 32. _ Civilized 52. _ Cruel

13. _ Appreciative 33. _ Clear-thinking 53. _ Curious

14. __ Argumentative 34. _ Clever 54. _ Cynical

15. __ Arrogant 35. _ Course 55. _ Daring

16. _ Artistic 36. _ Cold 56. _ Deceitful

17. _ Assertive 37. _ Commonplace 57. _ Defensive
18. _ Attractive 38. __ Complaining 58. _ Deliberate

1 9. _ Autocratic 39. _ Complicated 59. _ Demanding
20. Awkward 40. Conceited 60. _ Dependable

104

61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.

80.

Dependent
Despondent
Determined
Digniﬁed
Discreet
Disorderly
Dissatisﬁed
Distractible
Distrustful
Dominant
Dreamy
Dull
Easy-going
Effeminate
Efﬁcient
Egotistical
Emotional
Energetic
Enterprising

Enthusiastic

88.
89.
90.

91.

Evasive
Excitable
Fair-minded
Fault-ﬁnding
Fearful
Feminine
Fickle
Flirtatious
Foolish
Forceful
Foresighted
Forgetful
Forgiving
Formal
Frank
Friendly
Frivolous
Fussy
Generous

Gentle

105

101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.

120.

Gloomy
Good-looking
Good-natured
Greedy
Handsome
Hard-headed
Hard-hearted
Hasty
Headstrong
Healthy
Helpful
High-strung
Honest
Hostile
Humorous
Hurried
ldealistic
Imaginative
Immature

Impatient

 

121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.

133.

134.
135.
136.
137.

138.

139.

140.

Impulsive
Independent
Indifferent
Individualistic
lndustrious
Infantile
Informal
Ingenious
Inhibited
Initiative
Insightful
Intelligent

Interests
narrow

Interests wide
lntolerant
lnvenﬁve
Irresponsible

Irritable

Jolly

Kind

141.
142.

143.

145.
146.
147.
148.
149.
150.
151.
152.

153.

154.
155.
156.
157.

158.

159.
160.

Lazy
Leisurely
Logical
Loud
Loyal
Mannerly
Masculine
Mature
Meek
Methodical
Mild
Mischievous

Moderate

Modest
Moody

Nagging

Natural

Nagging

Natural

Nervous

106

161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.

173.

174.
175.
176.
177.

178.

179.

180.

Obnoxious
Opinionated
Opportunistic
Optimistic
Organized
Original
Outgoing
Outspoken
Painstaking
Patient
Peaceable
Peculiar

Persevering

Persistent
Pessimistic
Planful
Pleasant

Pleasure-
seeking

Poised

Polished

 

 

181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.

196.
197.
198.
199.
200.

Practical
Praising
Precise
Prejudiced
Preoccupied
Progressive
Prudish
Quarrelsome
Queer
Quick
Quiet
Quitting
Rational
Rattlebrained
Realistic
Reasonable
Rebellious
Reckless
Reﬂective

Relaxed

201 .
202.
203.
204.
205.

207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.

Reliable
Resentful
Reserved
Resourceful
Responsible
Restless
Retiring

Rigid

Robust

Rude
Sarcastic
Self-centered
Self-conﬁdent
Self—controlled
Self-denying
Self-pitying
Self-punishing
Self-seeking
Selﬁsh

Sensitive

107

221 .
222.
223.
224.
225.

226.
227.
228.
229.
230.
231 .

232.

233.

234.
235.
236.
237.
238.
239.

240.

Sentimental
Serious
Severe
Sexy
Shallow
Sharp-witted
Shiftless
Show-off
Shrewd
Shy
Silent
Simple
Sincere
Slipshod
Slow
Sly
Smug
Snobbish
Sociable

Soft-hearted

 

241 .
242.
243.
244.
245.
246.
247.
248.
249.
250.
251 .
252.
253.
254.

255.

256.
257.
258.
259.
260.

Sophisticated

Spendthrift

Spineless

Spontaneous

Spunky

Stable
Steady

Stern

Stingy

Stolid

Strong
Stubborn
Submissive
Suggestible
Sulky
Superstitious
Suspicious
Sympathetic
Tactful

Tactless

261 .
262.
263.
264.
265.
266.
267.
268.
269.
270.
271 .
272.
273.
274.
275.
276.
277.
278.
279.
280.

Talkative
Temperamental
Tense
Thankless
Thorough
Thoughtful
Thrifty

Timid

Tolerant
Touchy

Tough

Trusting
Unaffected
Unambitious
Unassuming
Unconventional
Undependable
Understanding
Unemotional

Unexcitable

108

281 .
282.
283.

284.

285.
286.
287.
288.
289.
290.
291 .
292.
293.
294.
295.
296.
297.
298.
299.
300.

Unfriendly
Uninhibited
Unintelligent
Unkind
Unrealistic
Unscrupulous
Unselﬁsh
Unstable
Vindictive
Versatile
Warm
Wary
Weak
Whiny
Wholesome
Wise
Withdrawn
Witty
Worrying
Zany

 

REFERENCES

Alden, L. E., Bieling, P. J., & Wallace, S. T. (1994). Perfectionism in an
interpersonal context: A self-regulation analysis of dysphoria and
social anxiety. Cognitive Therapy and Research, 18, 297-316.

Antony, M. M., Purdon, C. L., Huta, V., & Swinson, R. P. (1998). Dimensions of
Perfectionism across the anxiety disorders. Behaviour Research and
Therapy, 36, 1143—1154.

Ashby, J., & Rice, K. (in press). Perfectionism, dysfunctional attitudes, and self—
esteem: A stmctural equations analysis. Journal of Counseling
Development.

Axtell, A., & Newlon, B. J. (1993). An analysis of Adlerian life themes of
Bulimic women. Individual Psychology: Journal of Adlerian Theory,
Research & Practice, 49, 58-67.

Baron, R. M., Kenny, David A. (1986). The moderator-mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51(6), 1173-1182.

Barrow, J. C., & Moore, C. A. (1983). Group interventions with perfectionist
thinking. The Personnel and Guidance Journal, 61, 612-615.

Beck, A. R. (1976). Cognitive therapy and the emotional disorders. New York:
lntemational Universities Press.

Blaney, P. (1986). Affect and memory: A review. Psychological Bulletin, 99,
229-246.

Blankstein, K. R., Flett, G. L., Hewitt, P. L., & Eng. A. (1993). Dimensions of
perfectionism and irrational fears: An examination with the Fear Survey
Schedule. Personality and Individual Differences, 15, 323-328.

Blatt, S. J. (1995). The destructiveness of perfectionism. American Psychologist,
50(12), 1003-1020.

Blatt, S. J., Quinlan, D. M., Pilkonis, P., & Shea, T. (1995). The effects of need
for approval and perfectionism on the brief treatment of depression.
Journal of Consulting and Clinical Psychology, 50, 113—124.

Breslau, N. (1985). Depressive symptoms, major depression, and generalized
anxiety: A comparison on self-reports on CES-D and results from
diagnostic interviews. Psychiatry Research, 15, 219-229.

 

109

 

Breuer, J., & Freud, S. (1893-1895). Studies in hysteria. In J. Strachey (Ed.), The
complete psychological works of Sigmund Freud. (Standard ed., Vol. 2, ).
London: Hogarth.

Brewerton, T. D., & George, M. S. (1993). Is migraine related to eating
disorders? lntemational Journal of Eating Disorders, 14, 75-79.

Broday, S. F. (1988). Perfectionism and Millon basic personality patterns.
Psychological Reports, 63, 791 ~794.

Brouwers, M., & Wiggum, C. D. (1993). Bulimia and perfectionism: Developing
the courage to be imperfect. Joumal of Mental Health Counseling, 15,
141-149.

Burns, D. D. (1980). The perfectionist's script for self-defeat. Psychology
Today (November), 34—52.

Burns, D. D., & Beck, A. R. (1978). Cognitive behavior modiﬁcation of mood
disorders. In J. P. Foreyt & D. P. Rathjen (Eds), Cognitive Behavior
Therapy: Research and Application (pp. 109-134). New York: Plenum.

Chang, E. C. (2000). Perfectionism as a predictor of positive and negative
psychological outcomes: Examining a mediation model in older and
younger adults. Journal of Counseling Psychology, 47, 18-26.

Chang, E. C. & Rand, K. L. (2000). Perfectionism as a predictor of subsequent
adjustment: Evidence for a speciﬁc diathesis-stress mechanism among
college students. Joumal of Counseling Psychology, 4 7, 129-137.

Cohen, J. (1992). Quantitative methods in psychology. Psychological Bulletin,
112(1), 155-159.

Conger, A. J. (1974). A revised deﬁnition of suppressor variables: A guide to their
identiﬁcation and interpretation. Educational and Psychological
Measurement, 34, 35-46.

Corcoran, K. & Fischer, J. ( 1987). Measures for clinical practice: A source book.
New York: Free Press.

Corsini, R. J., & Wedding, D. (1989). Current psychotherapies. (4rth ed. ed.).
Itasca: F. E. Peacock Publishers, Inc.

Davey, G. C., (1993). A comparison of three worry questionnaires. Behaviour
Research & Therapy. 31 (1), 51-56.

110

 

Devins, G. & Orrne, C. (1985). The Center for Epidemiologic Studies depression
scale. In D. J. Keyser & R. C. Sweetland (Eds), Test Critiques (Vol. 2, pp.
144-160). Kansas City: Westport Publishers.

Dunkley, D. M., Blankstien, K. R., Halsall, J., Williams, M., & Winkworth, G.
(2000). The relation between perfectionism and distress: Hassles, coping,
and perceived social support as mediators and moderators. Journal of
Counseling Psychology, 47, 437-453.

Einstein, D. A., Lovibond, P. F., & Gaston, J. E. (2000). Relationship between
perfectionism and emotional symptoms in an adolescent sample.
Australian Joumal of Psychology. 52, 89-93.

Enns, M. W. & Cox, B. J. (1999). Perfectionism and depression: Symptom
severity in major depressive disorder. Behaviour Research and Therapy,
37, 783-794.

Fagan, J., 8. Shepherd, l. L. (Eds.). (1970). Gestalt therapy now: Theory,
techniques, applications. Palo Alto: Science 8. Behavior Books.

Fairburn, C. G. (1997). Eating disorders. In D. M. Clark, & C. G. Fairbum (Eds.),
Science and practice of cognitive behaviour therapy. Oxford University
Press.

Fairburn, C. G., Shafran, R., & Cooper, 2. (1999). A cognitive-behavioral
theory of anorexia nervosa. Behaviour Research and Therapy, 37,
1 -1 3.

Flett, G. L., Hewitt, P. L., & DeRosa, T. (1996). Dimensions of perfectionism,
psychological adjustment, and social skills. Personality and Individual
Differences, 20, 143-150.

Flett, G. L., Hewitt, P. L., Blankstein, K. R., & Dynin, C. B. (1994). Dimensions
of perfectionism and Type A behaviour. Personality and Individual
Differences, 16, 477-485.

Flett, G. L., Hewitt, P. L., Blankstein, K. R., & Gray, L. (1998). Psychological
distress and the frequency of perfectionistic thinking. Journal of
Personality and Social Psychology, 75, 1363-1381.

Flett, G. L., Hewitt, P. L., Edler, N. S., & Tassone, C. (1995). Perfectionism and

components of state and trait anxiety. Current Psychology:
Developmental-Leaming Personality-Social, 13, 119-126.

111

 

Flett, G. L., Hewitt, P. L., Blankstein, K. R., Solnik, M., & Van Brunschot, M.
(1996). Perfectionism, social problem-solving ability, and psychological
distress. Journal of Rational-Emotive and Cognitive-Behaviour Therapy,
14, 245-274.

FormyDuvaI, D. L., Williams, J., Patterson, D. J., & Fogle, E. E. (1995). A "big
ﬁve" scoring system for the item pool of the Adjective Check List. Journal
of Personality Assessment, 65(1), 59-76.

Freud, S. (1916). Introductory lectures on psychoanalysis. (Standard ed.). (Vol.
15). London: Hogarth.

Frost, R. O., Marten, P., Holt, C. S., Mattia, J. l., & Neubauer, A. L. (1993).
A comparison of two measures of perfectionism. Personality and
Individual Differences, 14, 119-126.

Frost, R. O., & Marten, P. A. (1990). Perfectionism and evaluative threat.
Cognitive Therapy and Research, 14(6), 559-572.

Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of
perfectionism. Cognitive Therapy and Research, 14(5), 449-468.

Frost, R. O. & Steketee, G. (1997). Perfectionism in obsessive-compulsive
disorder. Behaviour Research and Therapy, 35, 683-692.

Frost, R. O., Stekette, G., Cohn, L, & Griess, K. (1994). Personality traits in
subclinical and non-obsessive-compulsive volunteers and their parents.
Behavior Research and Therapy. 32, 47-56.

Gerson, M. (1984). Splitting: The development of a measure. Journal of Clinical
Psychology, 40, 157-162.

Goodman, LA. (1960). On the exact variance of products. Journal of the
American Statistical Association, 55, 708-713.

Gough, H. G., Fioravanti, M., & Lazzari, R. (1979). A cross-cultural unisex ideal
self scale for the Adjective Check List. Journal of Clinical Psychology, 35
(2). 314-319.

Gough, H. G., Fioravanti, M., & Lazzari, R. (1983). Some implications of self
versus ideal-self congnrence on the revised Adjective Check List. Journal
of Personality and Social Psychology. 44(6), 1214-1220.

Gough, H. G., & Heilbrun, A. B. (1983). The Adjective Check List Manual. (2nd
ed.). Palo Alto: Consulting Psychologists Press.

112

 

Gough, H. G., Lazzari, R., & Fioranvanti, M. (1978). Self versus ideal self: A
comparison of ﬁve Adjective Check List indices. Journal of Consulting and

Clinical Psychology, 46(5), 1085-1091.

Gould, J. R., Prentice, N. M., 8. Ainslie, R. C. (1996). The Splitting Index:
Construction of a scale measuring the defense mechanism of splitting.
Journal of Personality Assessment, 66(2), 414-430.

Gould, D., Udry, E., Tuffey, S., & Loehr, J. (1996). Burnout in competitive junior
tennis players: A quantitative psychological assessment. Sport
Psychologist, 10, 322-340.

Graves, W. H., & Shearer, R. A. (1971 ). Use of the Adjective Check List to elicit
description of self-ideal discrepancy. Perceptual and Motor Skills, 32(2),

781 -782.

Greenberg, L. S. (1979). Resolving splits: Use of the two chair technique.
Psychotherapy: Theory, Research and Practice, 16(3), 316-324.

Greenberg, L. 8., Rice, L. N., 8. Elliott, R. (1993). Facilitating emotional change:
The moment by moment process. New York: The Guilford Press.

Greenberg, L. S., & Sarkissian, M. G. (1984). Evaluation of counselor training in
Gestalt methods. Counselor-Education-and-Supervision., 23(4), 328-340.

Grotstein, J. S. (1981). Splitting and projective identification. New York: Jason
Aronson.

Halgin, R. P., 8. Leahy, P. M. (1989). Understanding and treating perfectionistic
college students. Journal of Counseling Psychology, 68, 222-225.

Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism.
Psychology, 15, 27-33.

Hamilton, T. K., 8. Schweitzer, R. D. (2000). The cost of being perfect:
Perfectionism and suicide ideation in university students. Australian
and New Zealand Journal of Psychiatry, 34, 829-835.

Hankin, B. L., Roberts, J., & Gotlib, I. H. (1997). Elevated self-standards and
emotional distress during adolescence: Emotional speciﬁcity and gender
differences. Cognitive Therapy and Research, 21(6), 663-679.

Hesselbrock, M., Hesselbrock, V., Tennen, H., Meyer, R., 8. Workman, K. (1983).

Methodological considerations in the assessment of depression in
alcoholics. Joumal of Consulting and Clinical Psychology. 51, 399-405.

113

Hewitt, P. L., 8. Flett, G. L. (1990). Perfectionism and depression: A
multidimensional analysis. Journal of Social Behavior and Personality, 5,
423-438.

Hewitt, P. L. 8. Flett, G. L. (1991a). Perfectionism in the self and social contexts:
Conceptualization, assessment and association with psychopathology.
Journal of Personality and Social Psychology, 60, 456-470.

Hewitt, P. L. 8. Flett, G. L. (1991b). Dimensions of perfectionism in unipolar
depression. Journal of Abnormal Psychology, 100, 98-101.

Hewitt, P. L. 8. Flett, G. L. (1991c). The Multidimensional Perfectionism Scale:
Reliability, validity, and psychometric properties in psychiatric samples.
Psychological Assessment, 3, 464-468.

Hewitt, P. L., Flett, G. L., 8. Weber, C. (1994). Dimensions of perfectionism and
suicide ideation. Cognitive Therapy and Research, 18, 439-460.

Hewitt, P. L., Newton, J., Flett, G. L., 8. Callender, L. (1997). Perfectionism and
suicide ideation in adolescent psychiatric patients. Journal of Abnormal
Child Psychology, 25, 95-101.

Higgins, E. T., Klein, R., 8. Strauman, T. (1985). Self-concept discrepancy theory:
A psychological model for distinguishing among different aspects of
depression and anxiety. Social Cognition, 3, 51-76.

Higgins, T. E. (1987). Self-discrepancy: A theory relating self and affect.
Psychological Review, 94(3), 319-340.

Hill, R. W., McIntire, K. 8. Bacharach, V. R. (1997). Perfectionism and the big ﬁve
factors. Joumal of Social Behavior and Personality, 12, 257-270.

Hill, R. W., Zrull, M. C., & Turlington, S. (1997). Perfectionism and interpersonal
problems. Journal of Personality Assessment, 69(1), 81 -1 03.

Hollander, M. H. (1978). Perfectionism: A neglected personality trait. Joumal of
Clinical Psychiatry, 39, 384.

Holmbeck, G. H. (1997). Toward terminological, conceptual, and statistical clarity
in the study of mediators and moderators: Examples from the child-clinical
and pediatric psychology Iiteratures. Journal of Consulting and Clinical
Psychology, 65(4), 599-610.

Homey, K. (1937). The neurotic personality of our time. New York: W. W. Norton
and Company, Inc.

114

 

Horowitz, M. J. (1977). Cognitive and interactive aspects of splitting. American
Journal of Psychiatry, 135, 549-533.

Husaini, B. A., Neff, J. A., Harrington, D., Hughe, P. & Stone, L. (1980).
Depression in rural communities: Validating the CES-D scale. Journal of
Community Psychology. 8, 20-27.

Juster, H. R., Heimberg, R. 6., Frost, R. O., Holt, C. S., Mattia, J. |., 8. Faccenda,
K. (1996). Social phobia and perfectionism. Personality and Individual
Differences, 21, 403-410.

Kawamura, K. Y., Hunt, S. L., Frost, R. O., 8. DiBartolo, P. M. (2001).
Perfectionism, anxiety, and depression: Are the relationships
independent? Cognitive Therapy and Research, 25, 291-301.

Kenny, D. A. (2001). Mediation, Retrieved December 1, 2001 from
[http://nw3.nai.netl~dakennylmediate.htm].

Kemberg, O. (1975). Bordertine conditions and pathological narcissism. New
York: Jason Aronson.

Korb, M. P., Gorrell, J ., 8. Van De Riet, V. (1989). Gestalt therapy: Practice and
theory. (2nd ed. ed.). New York: Pergamon Press.

Kowal, A., 8. Pritchard, D. W. (1990). Psychological characteristics of children
who suffer from headache: A research note. Journal of Child Psychology
and psychiatry and Allied Disciplines, 31, 637-649.

Lamott, A. (1994). Bird by bird: Some instructions on writing and life. New York:
Pantheon Books.

Lazzari, R., & Gough, H. G. (1980). Adjective Check List self and ideal self
correlates of MMPI proﬁles classiﬁed according to the Meehl-Dahlstrom
rules. Journal of Clinical Psychology, 36(4), 905-91 0.

Lichtenberg, J., & Slap, J. (1973). Notes on the concept of splitting and the
defense mechanism of splitting of representations. Journal of the
American Psychoanalytic Association, 21, 772-787.

Lynd-Stevenson, R. M., 8. Heame, C. M. (1999). Perfectionism and depressive l
affect: The pros and cons of being a perfectionist. Personality and
Individual Differences, 26, 549-562.

Magnusson, A. E., Nias, D. K. B., 8. White, P. D. (1996). Is perfectionism
associated with fatigue? Journal of Psychosomatic Research, 41,
377-383.

 

115

 

Mahler, M. (1968). On human symbiosis and the vicissitudes of individuation.
New York: lntemational Universities Press.

Mahler, M. ( 1971 ). A study of the separation-individuation process and its
possible application to borderiine phenomena in the psychoanalytic
situation. Psychoanalytic Study of the Child, 10, 195-212.

Mahoney, M. J., 8. Arnkoff, D. B. (1979). Self-management. In 0. F. Pomerleau 8.
J. P. Brady (Eds), Behavioral Medicine: Theory and Practice (pp. 75-96).
Baltimore: Williams and Wilkins.

Marmar, C. R., & Horowitz, M. J. (1986). Phenomenological analysis of splitting.
Psychotherapy 23(1), 21 -29.

Meyer, T. J., Miller, M. L., Metzger, R. L. & Borkovec, T. D. (1990). Development
and validation of the Penn State Worry Questionnaire. Behaviour
Research & Therapy, 28 (6), 487-495.

Mitchelson, J. K., & Bums, L. R. (1998). Career mothers and perfectionism:
Stress at work and at home. Personality and Individual Differences, 25,
477-485.

Molina, S. & Borkovec, T. J. (1994). The Penn State Worry Questionnaire:
Psychometric properties and associated characteristics. In C. L. Graham
8. F. Tailis (Eds.), Wonying: Perspectives on Theory, Assessment and
Treatment (pp.265-283). New York, NY: John Wiley 8. Sons.

Moore, C. A., & Barrow, J. C. (1986). Perfectionistic thinking in university
students: Implications for individual treatment. In J. E. Talley & W. J. K.
Rockwell (Eds), Counseling and Psychotherapy with College Students
(pp. 100-112). New York: Praeger.

Norman, R. M. 6., Davies, F., Nicholson, l. R., Cortese, L., 8. Malia, A. K. (1998).
The relationship of two aspects of perfectionism with symptoms in a
psychiatric outpatient population. Joumal of Clinical and Social

Psychology, 1 7, 50-68.

Norton, G. R., Buhr, K., Cox, B. J., Norton, P. J., & Walker, J. R. (2000). The role
of depressive versus anxiety-related cognitive factors in social anxiety.
Personality and Individual Differences, 28, 309-314.

Pacht, A. R. ( 1984 ). Reﬂections on perfectionism. American Psychologist, 39(4),
386-390.

116

 

Parker, W. D. (1997). An empirical typology of perfectionism in academically
talented children. American Educational Research Journal, 34(3), 545-
562.

Perls, F. (1969). Gestalt therapy verbatim. Lafayette: Real People Press.

Perls, F., Hefferline, R., & Goodman, P. (1951, 1994). Gestalt therapy:
Excitement and growth in the human personality. (2nd ed.). Highland, NY:
The Gestalt Journal Press, Inc.

Piedmont, R. L., McCrae, R. R., 8. Costa, P. T. (1991). Adjective Check List
scales and the ﬁve-factor model. Joumal of Personality and Social
Psychology, 60(4), 630-637.

Pirot, M. (1986). The pathological thought and dynamics of the perfectionist.
lndividualPsycho/ogy: Journal of Adlerian Theory, Research and Practice,
42( 1 ), 52-58.

Pruyser, P. W. ( 1975). What splits in ”splitting”? A scrutiny of the concept of
splitting in psychoanalysis and psychiatry. Bulletin of the Menninger Clinic,
31(1), 1-46.

Quadland, M. C. (1980). Private self-consciousness, attribution of responsibility,
and perfectionistic thinking in secondary erectile dysfunction. Journal of
Sexual and Marital Therapy 6, 47-65.

Radloff, J. (1977). The CES-D Scale: A self-report depression scale for research
in the general population. Applied Psychological Measurement, 1, 385-
401.

Rheaume, J., Freeston, M. H., Dugas, M. J., Letarte, H., & Ladouceur, R. (1995).
Perfectionism, responsibility, and obsessive-compulsive symptoms.
Behaviour Research and Therapy 33, 785-794.

Rice, K . G., Ashby, J. S., & Slaney, R. B. (1995). Self-esteem as a mediator
between perfectionism and depression: A structural equations analysis.
Joumal of Counseling Psychology 45, 304-314.

Rice, K. G., & Dellwo, J. P. (2002). Perfectionism and self-development:
Implications for college adjustment. Journal of Counseling and
De ve/opment, 80, 188-196.

Rice, K. G. & Mirzadeh, S. (2000). Perfectionism, attachment, and adjustment.
Jouma/ of Counseling Psychology, 47(2), 238-250.

117

 

Santor, D., Zuroff, D., Ramsay, J., Cervantes, P. 8. Palacios, J. (1995).

Examining scale discriminability in the EDI and CES-D as a function of
depressive severity. Psychological Assessment, 7, 131-139.

Santor, D., 8. Coyne, J. (1997). Shortening the CES-D to improve its ability to
detect cases of depression. American Psychologist, 9(3), 233-243.

Scott, L., & O'Hara, M. W. (1993). Self-discrepancies in clinically anxious and

depressed university students. Journal of Abnormal Psychology. 102, 282-
287.

Shafran, R., 8 Mansell, W. (2001). Perfectionism and psychopathology: A review
of research and treatment. Clinical Psychology Review, 21, 879-906.

Sheehan, T., Fiﬁeld, J., Reisine, S., & Tennen, H. (1995). The measurement

structure of the Center for Epidemiologic Studies Depression Scale.
Joumal of Personality Assessment, 64, 507-521.

Slade, P. D. (1982). Towards a functional analysis of anorexia nervosa and
bulimia nervosa. British Journal of Clinical Psychology, 21, 167—179.

Slade, P. D., & Owens, R. G. (1998). A dual process model of perfectionism
based on reinforcement theory. Behavior Modiﬁcation, 22(3), 372-390.

Slaney, R., 8., Rice, K., G., & Ashby, J., S. (2002). A programmatic approach to
measuring perfectionism: The Almost Perfect Scales. In F. Flett and P.
Hewitt ( Eds.) Perfectionism: Theory, Research, and Treatment, (pp. 63-
88), Washington DC, American Psychological Association.

Slaney, R. B., Ashby, J. S., & Trippi, J. (1995). Perfectionism: lts measurement
and career relevance. Joumal of Career Assessment, 3, 279-297.

Slaney, R. B., 8. Johnson, D. G. (1992). The Almost Perfect Scale. Unpublished
manuscript, The Pennsylvania State University.

Slaney, R. 8., Rice, K.G., Mobley, M., Trippi, J., Ashby, J. S. (2001). The

Revised Almost Perfect Scale. Measurement 8. Evaluation in Counseling
8 Development. 34(3), 130-145.

Small, A. C., 8. Batlis, N. C. (1978). Factor structure of Adjective Check List

under different instructional sets. Psychological Reports, 43(3 Pt 2), 1111-
1 1 14.

118

 

Sobel, M. E. (1982). Asymptotic conﬁdence intervals for indirect effects in
structural equation models. In S. Leinhardt (Ed.) Sociological
Methodology, 1982 (pp.290-312). Washington DC: American Sociologist
Association.

Sorotzkin, B. (1985). The quest for perfection: Avoiding guilt or avoiding shame?
Psychotherapy, 22(Fall), 564-571.

St. Clair, M. (1986). Object relations and self psychology. Monterey: Brooks/Cole
Publishing Company.

Stdber, J., 8: Joorman, J. (2001). Worry, procrastination, and perfectionism:
Differentiating amount of worry, pathological worry, anxiety, and
depression. Cognitive Therapy Research, 25, 49-60.

Strack, S., & Lorr, M. (1990). Item factor structure of the Personality Adjective
Check List. Journal of Personality Assessment, 55(1-2), 86-94.

Strauman, T. J. (1989). Self-discrepancies in clinical depression and social
phobia: Cognitive structures that underlie emotional disorders? Journal of
Abnormal Psychology. 98, 5-14.

Volken, V. D. (1976). Primitive lntemational object relations. New York:
International Universities Press.

Weisinger, H., & Lobsenz, N. (1981). Nobody's perfect. New York: Warner
Books.

Weissman, A. N. ( 1979). The Dysfunctional Attitude Scale: A validation study.
Dissertation Abstracts, 40(3-8), 1389-1390.

Weissman, M. M., Sholomskas, D., Pottenger, M., Pmsoff, B. A. 8. Locke, B. A.

(1977). Assessing depressive symptoms in ﬁve psychiatric populations: A
validation study. American Joumal of Epidemiology. 106, 203-213.

119