31:3
an» d.

in

. 2b..
2.» wuﬂxﬁtu ...
A .33..»

. , , , :3.va a,
V . .1. . 1.. V
. . V V . 3.1.3: in
. . ‘ B3,... I- f!»
.3} . V . ,
.11! ‘ at: .
RT! 9 , .
. ‘ srh. a

. 2.3;.
“.3:

\ ‘!
'4": 3:33.
lrzsv. ’1.

 

 

IIIﬁiiuiiiiiiiiim

’L

This is to certify that the

dissertation entitled

Shame and Women's Self Orientation:
Perceptions of Dependency Needs and Connected-
ness in Bulimia Nervosa and in Recovery

presented by

Melissa Frisch McCreery

has been accepted towards fulﬁllment
of the requirements for

Ph -D- degree in W

 

   

Major professor

Date April 12, 1995

 

MSUi: an Afﬁrmative Action/Equal Opportunity Institution 0- 12771

 

 

LIBRARY
Mlchlgan State
Unlverslty

 

 

 

PLACE N RETURN BOX to remove thb checkout 1mm your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

NOV 1 8.1999
‘ Cliff:-

 

 

 

 

 

 

 

M

 

 

 

 

 

 
  

 

 

 

L__JL J
-E_-
| ll ll I

MSU to An Afﬂmettve WM Opportunity Inetnuion
Wiles-9.1

 

 

 

 
 
 

 

 

 

 

 

 

SHAME AND WOMENS' SELF ORIENTATION:
PERCEPTIONS OF DEPENDENCY NEEDS AND CONNECTEDNESS
IN BULIMIA NERVOSA AND IN RECOVERY
By

Melissa Fu'sch McCreery

A DISSERTATION

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

DOCPOR OF PHILOSOPHY

Department of Psychology
1995

ABSTRACT
SHAME AND WOMEN‘S SELF ORIENTATION:

PERCEPTIONS OF DEPENDENCY NEEDS AND CONNECTEDNESS IN BULIMIA
NERVOSA

AND IN RECOVERY
By
Melissa Frisch McCreery

Research suggests conﬂicts related to perceptions of autonomy and dependency are
important dynamics underlying bulimia. However, research in these areas is contradictory.
Additionally, bulimics' perceptions and ideals regarding dependency needs and self-
reliance have been insufﬁciently explored.

The study addressed two research questions: Do bulimics deﬁne themselves and
their ideals about relatedness differently than non-eating disordered women or recovered
bulimics, and, do bulimics view healthy dependency needs as signiﬁcantly more shameful
than non-mung disordered women and recovered bulimic women? Behaviorally-bulimic
(BB), behaviorally-recovered bulimic (BR) and non-eating disordered (NED) womens' real
and icbal self orientations were assessed using real and ideal responses to the Relationship
Self Inventory (RSI). Subjects' attributed levels of shame (using the Internalized Shame
Scale (188)) to audiotapes of women who expressed either dependency needs or self-
reliance. Subjects' own level of shame was assessed using the 188.

Subjects' own [88 scores showed a signiﬁcant declining trend from the BB to the
BRtotheNEDgroup. Correlationsbetweenreal RSI scalescoresandbetween RSI scores
and subjects' 188 scores showed signiﬁcant differences between groups. Bulimics' tended
to view concepts of autonomy, separateness, and interrelatedness as irreconcilable
opposites. The BR groups' responses showed less evidence of this tendency, supporting
arguments that an increased ability to integrate concepts of connection and individuation in
one's self deﬁnition is linked to recovery from bulimia. The BB group's ideal self was

signiﬁcantly higher on the ”Separate Self" RSI scale than the NED group. This runs
counter to theories that bulimics over-idsalize ”feminine" characteristics such as intimacy
and dependency (Bosldnd-Lodahl, 1976; Pettinati, et al., 1987) and supports theories
emphasizing over-idealizations of autonomy and self-reliance (i.e. Steiner-Adair, 1986).

The BB group attributed signiﬁcantly more shame to the woman expressing
depardency needs tlnn to the self-reliant woman and attributed signiﬁcantly more shame to
thedependencyneedsdepiction thandideitherthe BRorNEDgroups. Theﬁndingthat
the BR group's attribution of shame to the individual expressing dependency needs was
signiﬁcantly lower than the BB group suggests that a change in the paeeption of
depmdency needs is involved in the recovery process.

For Scott and for Cameron

iv

ACKNOWLEDGMENTS

This study was completed over a period of two years and over a distance of some
twothousandmiles.1twouldhavebeenmuchmoredifﬁcult, ifnotirnpossibletocomplete
without the signiﬁcant help and support of the following individuals, on whom I was
shamelessly dependent. ,

IwouldliketothankEllen Strommen form-chairing mydissertation comnritteeand
for her support, wisdom, and guidance throughout the project. Her knowledge and ideas
about relational development were integral to the development of the study. Her
enthusiasm, availability, and direction were most appreciated, especially during my
moments of panic and uncertainty.

IamalsobothindebtedandgrateﬁrltoBertKaron whom-chaired my dissertation
committee and who has been an advisor and supporter throughout my graduate experience.
Bert's clinical expertise, his sound judgment, and his view of the world (in general) and
psychology (speciﬁcally) have greatly inﬂuenced my thinking and my aspirations—far
beyond this research endeavor. Additionally, his generous loan of ofﬁce space made my
research magnitudes easier to complete.

Gersh Kaufman was extremely helpful in his role as ”resident shame expert." He
provided many resources and references and was a tireless editor ofboth my Master's
thesis and my dissertation. Thank you Gersh. Sharon Hoerr faced a challenge as the
”non-psychologist” mernba' of my committee. I greatly appreciated the input and fresh
perspective she provided me throughout this endeavor. She taught me much about the
physiological components of bulimia. Additionally, the organizational critiques and
suggestions sheprovided havemadethisabetterandmorecoherentwork. RickDeShon

V

probably has no idea how helpful he was to this study, providing statistical consultation
and expertise both in person and by phone at a moment's notice. Rick, I could not have
completed this dissertation without your help. Thank you, thank you, thank you!

My many research assistants were essential to the study. Thank you, all of you, for
your cheerfulness in the face of endless hours at the computer, your attention to detail, and
your tolerance of my sometimes incoherent instructions.

This research was supported by a Student Awards Grant (grant number: 132-
SAPl94-11) from the Michigan Health Que Education and Research Foundation, the
philanthropic afﬁliate of Blue Cross and Blue Shield of Michigan, and I wish to
acknowledge their generous support.

Finally, my deepest thanks go to my husband Scott and my son Cameron, who
havebeencar'riedalong on thisjourneyofmineand whobothmadealotofconcessions
and accomodations so that I could complete this adventure—all the while offering more love
and support than I could have expected. Thanks guys.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ x
LIST OF FIGURES .............................................................................. xi
RESEARCH RATIONALE AND RESEARCH QUESTIONS ............................. 1
INTRODUCTION AND REVIEW OF THE LITERATURE ............................... 4
Relational Needs and Women's Development ........................................ S
Shame, disconnection, and interpersonal failures .................................... 10
Affect Theory .................................................................... 13
Psychoanalytic Theories ........................................................ 15
Shame-based Disorders ......................................................... 17
Bulimia .................................................................................... 22
Dysfunctional Communication ................................................. 23
Denial of Needs and Feelings .................................................. 25
Perfectionism and Cultural Values ............................................ 27
Bulimics’ Perceptions of Their Disorder ..................................... 30
Physiological Considerations in Bulimia Nervosa .......................... 32
Research Objectives and Hypotheses. ................................................. 34
Overview of Design ...................................................................... 39
METHOD .......................................................................................... 41
Subjects .................................................................................... 41
Measures .................................................................................. 43
Eating Disorders Inventory (EDI) ............................................. 44
Internalized Shame Scale (188) ................................................ 44

vii

Relationship Self Inventory (RSI) ............................................. 46

Taped Vignettes .................................................................. 47

Procedure .................................................................................. 48
Ethical Issues ............................................................................. 51
Anonymity/Conﬁdentiality ..................................................... 51

Risks and Beneﬁts of the Study ............................................... 51

Consent Procedures ............................................................. 52

RESULTS .......................................................................................... 54
Demographic Data ........................................................................ 54

[SS ......................................................................................... 60

RSI ......................................................................................... 60
Responses to Vignettes .................................................................. 75
DISCUSSION ..................................................................................... 83
Demographics and descriptive information ............................................ 83

Individual Shame: Intemalized Shame Scale (188) scores - _ _ - 84
Self deﬁnition and ideals: Responses to the Relationship Self Inventory

 

(RSI) ....................................................................................... 85

Self orientation ................................................................... 85

Ideal self orientation ............................................................. 87

Shame and the Perception of Dependency Needs: Responses to Vignettes ....... 89

CONCLUSION ................................................................................... 95

APPENDIX A: Glossary ........................................................................ 97

APPENDIX B: Taped Vignettes ................................................................ 98
APPENDIX C: Pilot Testing .................................................................... 103
APPENDIX D: Measures ........................................................................ 104
APPENDIX E: Consent Forms and Information Forms ..................................... 122

viii

APPENDIX F: Prescreened Sample ........................................................... 126

APPENDIX G: Descriptive Information By Group .......................................... 128
APPENDIX H: Descriptions of Group Members ............................................ 129
REFERENCES .................................................................................... 133

ix

LIST OF TABLES

Table 1: DSM III-R Diagnostic Criteria for Bulimia Nervosa .................................... 4 2
Table 2: Eating Disorders Inventory: Drive For Thinness and Bulimia Subscale Items ...... 45
Table 3: Pilot Study: Subject Means ................................................................ 49
Table 4: EDI scale scoresandreported BMI scoresforthepre—screened sample .............. 55
Table 5: BMI and EDI descriptive statistics and bingeing and purging frequencies for

groups ..................................................................................... 57
Table 6: Psychotherapy sessions reported by groups and reasons for seeking

psychotherapy ............................................................................ 59
Table 7: ISS Scores ................................................................................... 61
Table 8: Means, standard deviations, and standard scores for RSI real and ideal scale

scores by group .......................................................................... 62
Table 9: RSI--"Real" and "Ideal" scale scores analyzed by group ............................... 69
Table 10: Intereorrelations of ”Real” RSI Scale Scores By Group .............................. 72
Table 11: RSI ”Real" Scale Score Intercorrelations for the Prescreening Sample ............. 74
Table 12: Correlations of R81 Real and Ideal Scale Scores with 188 Scores (By Group) ....76
Table 13: Ratings of Vignettes ....................................................................... 79

Table 14: Correlations of Eating Disorder Diagnostic Variables with 188 Score attributed to

the Interpersonal Needs Vignette: BB and BR Groups ............................. 82

LIST OF FIGURES

Figure 1: Physiological Vicious Cycle ......................................................... 35
Figure 2: Connected Self By Group (Standardized Scores) ................................. 64
Figure 3: Separate Self By Group (Standardized Scores) ................................... 65
Figure 4: Self and Other Care By Group (Standardized Scores) ............................ 66
Figure 5: Primacy of Other Care By Group (Standardized Scores) ........................ 67
Figure 6: RSI: Re‘al Responses By Group (Standardized Scores) ......................... 71
Figure 7: Shame Ratings Across Vignettes .................................................... 77
Figure 8: Ratings of Shame Across Vignettes By Group .................................... 78

xi

RESEARCH RATIONALE AND RESEARCH QUESTIONS

Bulimianervosaisadisorderwhichappearspredominantlyin females. Prevalence
estimates of bulimia and bulimic behavior range from 1-20% of high school and college
females, depending on the criteria used (Rand & Kuldau, 1992). Research indicates that at
least 90% of bulimics are female (Johnson, Lewis & Hagman, 1984). Reported recovery
ratesrangefrorn 29-71% overarangeof 14-72 months (Abralmn,Mira, & Llewellyn-
Jones, 1983; Keller, Herzog, Iavori, Bradburn, & Mahoney, 1992; Lacey, 1983; Mitchell,
Pyle, Hatsukami, Gogg, Glotter, & Harper, 1988; Pope, Judson, Jonas, & Yurgelun-
Todd, 1985; Swift, Kalin, Wamboldt, Kaslow, & Ritzholz, 1985). A recent long term
followup studyof30bulimicsrevealedthatonly 69% hadrecoveredafter3 t03.5 years
”despiternorethan six months oftreatmentin mostcases" (Kelleretal, 1992, p. 7).

Theempirical liwratureonbulimiaconsistentlyrepa'tsbulimics' difﬁcultieswith
interpersonal relationshipsandindeﬁningthemselvesinrelationtoothers (seeforexample,
Boskind-Lodahl, 1976; Dickstein, 1985; Garﬁnkel & Garner, 1983) and consistently
implicates conﬂicts or difﬁculties related to autonomy and dependency. Much research has
bemdevotedmdynamicsindrebuﬁnﬁc‘smtapasonalrehdmshipsandmissues
regarding the bulimic's actual levels of autonomy and dependency (Attic & Brooks-Gum,
1989; Bosldnd-Iodahl, 1976; Garﬁnkel & Garner, 1983; Humphrey, Apple, &
Kirschenbaum, 1986; Johnson & Berndt, 1983; Johnson 8: Maddi, 1986; Pettinati,
Franks, Wade, & Kogan, 1987; Steiger, Fraenkel, & Leichna', 1989; Strober &
Humphrey, 1987). However, surprisingly little research has focused on the bulimic's
paeeptionorideal conwptionofincrpersonal relationshipsand interpersonalneedsthe
meaning tlntrelationships hold forthebulimic,orchangesin thesepereeptionsaﬁer

1

recovery from bulimia nervosa (Teusch, 1988; McCreery, 1991). In the relative absence of
such research and of consistart ﬁndings in these areas, theories implicating the perceptions
and ideals ofbulimics have been advanced.

Theorists have posited that bulimics experience heightened identiﬁcation with
traditional feminine traits and ideals (Steiga', Fraenkel, & Leichner, 1989; Pettinati,
Franks, Wade, & Kogan, 1987). Others argue that the bulimic's self is characterized by
gendeidendtycmﬂictclnmdaizedbyideaﬁmﬁonofmascﬂheamibuesorauddr tobe
male (Rest, Neuhaus, & Florin, 1982) or by the shaming of “feminine" values of care and
connection (Steiner-Adair, 1986). Otters have theorized that the bulimic has developed a
comer-dependent ”false-selfideal" in ordatoavoidthe shame sheattributestoherreal
dependency and dependency needs (Jones, 1985).

Effecﬁvepsychodrerapeuﬁcheahnentrequhesanaccmateassesnnmtmd
understanding of the individual's real self, but also an accurate understanding of the
individual‘s perceptions, goals, ideals, and perceived shortcomings. For instance,
bulimics consistently display a heightened level of dependency and reveal strong neetB to
conformandgainapproval fromothers. Researchersandtheoristslmvesuggestedthat
treatment approaches should focus attention on strengthening and supporting autonomy and
separation andhaveaddressed dependencyasanegativeconstruct(forexample, Boskind-
Lodahl, 1976; Bornstein & Greerrberg, 1991; Rost, Neuhaus, & Florin, 1982). However,
if high levels ofactual dependency coexist with shameful perceptions of normal and
necessarydependencyneedsandwitharr idealwhichexaggerawsandrestrictivelygkrriﬁes
separaMesHashypothesizedinthepresent study),tharrsuchapproacheswouldneglect
impunmdynanucsmdwmddbemadcquateatbestandpowndaﬂyhnppropriatem
ineffective.

A morepreciseunderstanding ofthedymmics underlying bulinrianervosaand
recovayﬁombulimiacanleadtothedevelopmentand/autilizatim ofmoreeffective

treatment approaches. The present study sought to preliminarily investigate and clarify
paeeptions of depardurcy needs and the role intapa'sonal relationships play in the identity
ofboth fennlebulimicsandrecoveredbulimics with specialattention tofeelingsofsharrre.
The research questions speciﬁcally are : (1)do bulimics deﬁne themselves and their ideals
about relatedness drﬂ’eremly than non-eating disordered women or recovered bulimics and
(2) do bulimics view healthy dependency needs as shameﬁdusigm'ﬁcamly more shameﬂd
than non-eating disordered moron and recovered bulimic women view these needs.

Thestudyhastwocomponents. The ﬁrstphaseofthestudywasdirectedat
clarifying the bulimic's self orientation and examining differences in self orientation
between beluviorally bulinric, non-eating disorrbred, and behaviorally recovered bulimic
women. The present study examined behaviorally bulimic, non-eating disordered and
behaviorally recovered bulimic self orientations and reported ideal-self orientations through
these individuals' real and ideal responses to the Relationship Self Inventory (RSI).
Subjects' levels of internalized shame wae also assessed using the Internalizcd Shame
Scale (188).

The second component of the study examined behaviorally bulimic, behaviorally
recovered bulimic, and non-eating disordered womens' paceptions of incrpersonal or
dependency needs and self-reliance, speciﬁcally the shame attributed to these qualiﬁes. In
ordertoattempttodifferentiatetheshame bulimies' marrifestrelatedtobulirnic
symptomology from the shame they may attribute to the qualities ofcorrnection and self-
reliance,thestudyattemptedto separatedeeecomponartsbyinvestigating subjects'
attributions of shame to non-eating disordered women who reveal intapu'soual needs or
aredepictedasself-reliant. Diffaencesand similaritiesbetween groupswereexamined.

INTRODUCTION AND REVIEW OF THE LITERATURE

Research and theoretical contributions by Chodorow (1978), Gilligan (1982,
1986a, 1986b), and Miller (1986) (among others) have emphasized the signiﬁcant
meanings of relationships and connectedness for women's lives and have argued for the
importance ofattention to these meanings when attempting to understand the dynamics of
female development. The dynamics of connectedness and women's perceptions of
relationships should also be considered in efforts to conceptualize deviations in the
psychological development of women.

Bulimia nervosa is a disorder which appears predominately in females. Research
indicates that at least 90% of bulimics are female (Johnson, Lewis & Hagman, 1984). The
empirical literature on bulimia consistently reports bulimics' difﬁculties with interpersonal
relationshipsandin deﬁning themselvesinrelation toothers (seeforexample, Boskind—
Iodahl, 1976; Dickstein, 1985; Garﬁnkel & Garner, 1983). Much research has been
devoted to dynamics in the bulimic's interpersonal relationships and to issues regarding the
bulimic's levels of autonomy and dependency (Attic & Brooks-Gum, 1989; Boslcind-
Lodahl, 1976; Garﬁnkel & Garner, 1983; Humphrey, Apple, & Kirschenbaum, 1986;
Johnson & Berndt, 1983; Johnson & Maddi,. 1986; Pettinati, Franks, Wade, & Kogan,
1987; Steiger, Fraenkel, & Leichner, 1989; Strober & Humphrey, 1987). However,
surprisingly little research has focused on the bulimic's perception of interpersonal
relationships and interpersonal needs, the meaning that relationships hold for the bulimic,
or changes in these perceptions after recovery from bulimia nervosa (Teusch, 1988;
McCrea'y, 1991). The present study sought to preliminarily investigate perceptions of
inta'pa'sonal needs and the role interpersonal relationships play in the identity of both

4

female bulimics and recovered bulimics with special attention to feelings of shame. A
comparison was made with non-eating disordered females with the goal offurther
elucidating the dynamics which underlie bulimia nervosa and which mediate recovery.
Prior to outlining the present study, the literature concerning relational needs, women's
development,and shamewill bereviewed andpertainentresearchonbulimianervosawill
be discussed.

 

Traditionally, many of the major theories of identity development and personality
have conceptualized psychological growth as moves toward increased separation,
individuation, and self-reliance in which separation is viewed as a necessary precmsor for
the deveth of mature identity. Within such ﬁameworks, connection and dependence
have commonly been perceived as lack of individuation, as immature, or pathological (see
for example Erickson 1963; Freud; Mahler & Furer, 1968). Gilligan ( 1982, 1986a,
1986b) criticizes theories which emphasize developmental moves out of dependence as
failing to validate and grapple with the complexity of the construct. She argues that the
construct of dependence is viewed negatively in traditional developmental theories because
it is set up in a false dichotomy with independence. According to Gilligan, dependence is
really a construct with two polar opposites, independence and isolation; the emphasis on
independence has led to a neglect of the positive values and meanings of relationship and
connection. Increasingly theorists are recognizing that connection, dependence, and values
ofrelationshipareintegralcomponentsofemotional matmityandthatthedevelopmentof
the capacity for these characta'istics has been largely ignored in developmental theory.
Such realizations have led to a variety ofefforts to formulate more inclusive and complete
theories of human development (Franz & White, 1985; Berlin & Johnson, 1989; Gilligan,
1982, 1986a, 1986b; Stern, 1985; Miller, 1984, 1986; Chodorow, 1974, 1978).

Stern (1985) has reviewed the empirical literature on infant development and
concludes that theoretical models which posit development as a singular linear process of
separation and individuation are inaccurate. Contrary to the hypothesis that connection
results only from a failure at diffacntiation, he synthesizes compelling evidence that the
abilitytoconnectwithanotherisa learnedskill daivedfrom deliberate, activeeffortsorl
thepartofboth infant andmother; thattheability toconnectisaprowssofpsychic growth,
self-differentiation, and afﬁrmation which begins to develop almost from birth. Stern
(1985) argues that development involves not only moves towards individuation but also
towards relationship. One learns how to be with another, how to share one's self
experience and how to be in social relationship with others, begimring in infancy.
Experiences of being with anotha are seen as active acts of integration, through learning to
be in relationship a sense of self as individuated or agentic develops. For example, as one
learns to be with anotha', and one learns how one impacts on another, one builds skills at
differentiating from theother. The individual sees the selfdeﬁned in the context of
relationship (Stern, 1985).

Nancy Chodorow (1974, 1978) was one of the ﬁrst to afﬁrm the values of
connectedness and relationship in a theory which examines genda' differences with regard
to these constructs. She argues that male and female identity formation is necessarily a
different process because women are largely responsible for early child care (Chodorow,
1974). Feminity tends to be deﬁned tluough connection and relationship while for males,
separation and individuation are critical to gender identity.

Gender identity formation traditionally occurs in the context of an ongoing
relationship with the mother. Girls, in their female identification, experience themselves as
like their mothers and as more continuous than discontinuous with her. The little girl learns
about he own identity through a process of connection and relationship; to be a girl means
to be like mother (Chodorow, 1978).

Identity development for boys, by contrast, is a process of differentiation from the
mother. The little boy learns that he is male, or ”not female.” Development for males
involves "more empathic individuation and a more defensive ﬁrming of experienced ego
boundaries” (Chodorow, 1978, pp. 166). Chodorow argues that these sex differences in
identity formation lead females to ”emerge from this period with a basis for 'empathy' built
intotheirprimarydeﬁnition ofselfinawaythatboysdonot. . . .Girlsemergewitha
stronger basis for experiencing another's needs or feelings as one's own (or of thinking
that one is so experiencing another's needs and feelings). . . From very early then, because
they areparented by a person of the same gender . . . girls come to experience themselves
aslessdiﬁ'erentiatedthanboys,asmorccontinuous withandrelatedtotheextemalobject—
world” (Chodorow, 1978, pp. 167).

Other theorists have also argued that healthy identity development can occur within
thecontextofconnectionandrelatednesstoothers. Theseclinicians, theoristsand
researchers have emphasized relational themes in self development as healthy lathe than
pathological or immature. Gilligan (1982) has posited two pathways to self deﬁnition
along with corresponding moral ”voices” which develop out or the differing paths. In the
ﬁrst,which slemguesismmepredomimntmmaleabecausemalestaldtobesocialized
into roles which value separation and autonomy, the self is deﬁned in separation, and the
”justice voice,” emphasizing hierarchy, is dominant. Gilligan argues for a second mode of
self deﬁnition which occurs within a network of relationships in which the self is deﬁned
throughactivitiesofcormectionandcareforothers. This secondmode, which Gilligan
believes is more predominant in females, because women's socialization tends to
emphasize values ofconnectedness, relatedness and nurturance, leads to the development
of what she labels a ”care voice" emphasizing network and the maintenance of connection
ratha' than hierarchy in decision making.

A group of clinicians and researchers at the Stone Center at Wellesley College
including Miller (1984, 1986) and Surrey (1984) present a theoretical conceptualization of
the ”self-in-relation" in an attempt to capture and validate the developmental expeience
which is grounded in relationships. The self-in-relation is a theoretical understanding of
one mode of identity development in which

“the primary conceptualization of the self is relational, that is, the self is

organized and developed in the context of important relationships . . . The

notion of the self-in—relation makes an important shift in emphasis from

separation to relationship as the basis for self-experience and development.

Further, relationship is seen as the basic goal of development; i.e. the

deepening capacity for relationship and relational competence. The self-in-

relation model assumes that other aspects of self (eg. creativity, autonomy,

assertion) develop within this primary context . . . other aspects of self-
developrnent emerge in the context of relationship, and thee is no inheent

need to disconnect or to sacriﬁce relationship for self development.”

(Sm-rey, 1984, p. 2).

The self-inorelation model emphasizes growth and maturation within relationship, ”where
botha'aﬂpeopleinvolvedmeemnagedandchallelgedmmainmincmnecﬁmandm
foster, adapt and change with the growth of the other" (Surrey, 1984, p. 8).

'lheStone Certerwritersarguetlmttheself-in—relationis morclikelytodevelopin
women due to gender socialization practices, cultural patterns of hierarchical power
relations between women and men, and identiﬁcation processes in early childhood.
Through an ongoing collection of working papers, these theoreticians have examined the
dynamics of the self-in-relation as they relate to a wide variety of issues and have
undertaken a broad reaching theoretical investigation of the meanings of connection and
relationship to identity and maturity.

Developmental models which highlight the positive value of connection necessarily
complicate and enrich our understandings and theories of human development. Multiple
and diverse pathways towards psychological maturity appear to exist, culmimting in
diffeing processes of identity and difl‘e'elt world views. While many theorists highlight

gende' differences in developmental pathways and in the development of self, thee is no

evidence that such differences are "hard-wired." There are however factors, speciﬁcally,
traditional sex-role expectations and the inequality of males and females, which impact
ideltity development on all levels, from the most intimate, to the most institutionalized
cultural plane.

Societal norms, structures, and values clearly shape development differently for
females and males and lead to gender differences in developmental conﬂicts and goals. An
additional dynamic associated with this diffeential gender socialization involves the relative
value placed on traditional gender roles. Miller (1986) has convincingly argued that it is
impossible to understand the psychology of women without addressing this component.
While nurturance, cooperation, and interrelatedness are integral and necessary to the
culture, society extolls and rewards the “virtues” of autonomy and independelt
achievement (Gilligan, 1982; Miller, 1986; Steiner-Adair, 1986). Miller argues that the
role of connectedness in women's development and the degradation of concepts such as
depeldelce in culture and in many psychological theories is inseparably intertwined with
women's subordinate position within a hierarchical cultme.

Inherbmummdamhnlmmhmmmmwwmm
much of women's relational ability results ﬁom the relegation of the nurturing, caretaking
domain towomenandthelackofintegration ofrelational valuesin maleexpeience (Miller,
1986). Values of care and connection lead to the development and reﬁnement of important
relational skills which can be important strengths for females. However, the cultural
context in which such a relational stance is nurtured is problematic. Women's traditional
rolesarenotesteemedandrewarded by society, norarevalueswhich fosterrelationshipas
opposed to autonomy. Miller (1983, 1986) argues that the devaluation of activities of care
and relationship, and the relegation of the majority of such tasks to a subordinate group,
place constraints on the carries ofthese values and activities. The focus on fostering
intepersonalconnection maynotbereciprocaland islesslikelytotakeplacewithina

10

context of mutuality and shared empathy. Without mutuality, activites of care and
connection easily become one-sided experiences of caretalcing. Because caretaking and
nurturing are embedded within a hierarchy of power (at the low end), these activities
become something one person does for another, a chore, and the tasks, as well as the
resulting psychological characteistics, are trivialized and not highly valued, either by the
individual, or by society (Miller, 1986).

This lack of validation can be a signiﬁcant factor in women's development and life
expeience. Thedevaluation of'whatoneis' orofone'smostceltral values isthe
equivalelt of shaming. The experience of shame appears to develop as an important
dynamic in some women's perceptions of interpersonal needs and of their own identity.

 

While some have focused on validating and explicating the role of relationships in
human development, equally important are both an exploration of how dependence and
connectedness are inter-personally and institutionally devalued and the resultant impact of
this denigration and disregard.

Internal perceptions and judgments about the selfare linked to cultural expectations
and values. Such values are transmitted overtly and covertly in the most intimate of
relationships as well as at a broad cultural level. Connectedness, the desire to be in
relationship, and the capacity for intimacy are integral to the mature psychological
development of both males and females; their neglect and lack of validation in our culture as
wellasinclinical anddevelopmental theory hasbeen tothedetrimentofboth sexes.
However, in the face of this lack of validation, these capacities and values have been
encouraged differentially by gentle.

The widely cited Broverman study (Broverman et. al., 1972) elucidates the
dilemmathatthis situationcreates. In thestudy,clinicianswereaskedtoprovide the
psychological attributes of either a mature, healthy, socially compewnt "male”, "female" or

11

“adult.” While there emerged no signiﬁcant differences between standards for healthy
males and for healthy adults, the charactu'izations of a healthy female and a healthy adult
wee signiﬁcantly different. Healthy females were deﬁned in line with traditional gender
stereotypes and were assigned ”feminine" characteristics that were not deemed ”healthy" in
a mature “adult." The study provides empirical documentation ofthe identity paradox
women may face. To be ”female" may be judged differently than to be a mature ”adult".
Women's experience and socialimtion is different from men's experience. However,
becauseaccepted ”adult" nonnsappeartobegeneralizationsofmaleexperience, women's
authentic portrayal of themselves is still very likely to be judged as less than ideal.

Steiner-Adair (1986) argues that females are socialized towards values and
behaviasofcareandconnectednessandthenencolmteradilemmwhen facedwith the
reality that this culture does not value those traits, but extolls the "virtues" of autonomy and
independent achievement. Such social processes cream a developmental conﬂict for
females which is difﬁcult to resolve and which often leads to saa'iﬁces in self-esteem, self-
conﬁdence, and even the denial of one's own experience. This conﬂict is exacerbated by
the importance placed on sustaining relationship for many females, and the related fear or
discomfort which may be associated with separation from relationship or losing a
relationship by not accomodating to societal demands(Bernardez-Bonesatti, 1978; Gilligan,
1982; Jordan, 1990; Miller, 1986).

Gilligan (1989) has presented evidence that in adolescalce, many females begin to
loseconﬁdencein theirownidsasandpaceptions. Theybegin todoubttheirown
experience and values and become more hesitant about bringing their personal truths into
relationshipswithothers. Gilligan speaksofa "life-threatening splitbetween femaleand
adult” referring to the disparity between girls' own perspectives and perceptions and a
societalview which denigrates valuesofcareanddependence. Shebelieves such societal
tensions may throw young girls into saious conﬂict between their own values and those of

12

society (Gilligan, 1986b). Steiner-Adair (1986) labels this experience a “developmental
double-bind;" females are socialized to be one way and then learn that society places value
on something else.

If young women are unable to successfully negotiate this developmental dilemma, it
is possible that shameful perceptions of the self may develop (or be reinforced). Steiner-
Adair (1986) has argued that young females whoare unable to integrate theirown values of
relationship in formulating their ideals are at increased risk for eating disorders. Others
have also related shameful perceptions of the self and related conﬂicts over separation and
connectedness with eating disturbances (Kaufman, 1992, Wrrrmser, 1981, Gilligan
1986b). Theories of shame and shame-based disorders will be outlined and shame-based
conceptualizations of bulimia will be reviewed.

Although theorists differ in their understanding of the dynamics underlying shame,
their descriptions of the phenomonomlogieal expaience of conscious shame is fairly
consistent. 'lhe feeling of shame is an acutely painful affective expa‘ience involving
feelings of inadequacy, inferiority, and exposru'e (before the self or an other). Laing
(1960) has referred to shame as "an implosion of the self.” Lewis' description is
represemative:

"The body gestures and attitude include head bowed, eyes closed, body

crrrved in on itself, making the person as small as possible. At the same

time that it seeks to disappear, the self may be dealing with an excess of

autonomic stimulation, blushing or sweating or diffuse rage, experienced as

a ‘ﬂood' of sensations. Shame is thus regarded by adults as a primitive

reaction, in which body functions have gone out of control. It is regarded

as an irrational reaction for this reason also. . . Shame is a relatively

wordless state. The experience of shame often occurs in the form of

imagery, of looking or being looked at. Shame may also be pla ed out in

imagery of an internal auditory colloqu , in which the whole se f is

condemned by the 'other’" (Lewis, 1 1, p. 37).

Conceptualimtions of the dynamics of shame have developed from two major theoretieal
systems: Tomkins' affect theory, and applications and reformulations of psychoanalytic

theory.

13

W

Tomkins' (1963, 1987) theory of affect conceptualizes the affects as an innate
system, the primary motivational force in human beings, separate from the innate drives.
Affects are understood as a system of ampliﬁers which direct attention to the individual's
needs as indicated by physiological data inputs.

”. . . affects are sets of muscular, glandular, and skin receptor responses

located in the face (and also widely distributed throughout the body) that

generate sensory feedback to a system that ﬁnds them either inherently

"acceptable” or ”unacceptable." These organized sets of responses are

triggered at subcortieal centers where speciﬁc ”programs" for each distinct

affect are stored, programs that are innately endowed and have been

genetically inherited. They are capable, when activated, of simultaneously

capturing such widely distributed structures as the face, the heart, and the

endocrine glands and imposing on them a speciﬁc pattern of correlated

responses. Onedoesnot learn tobeafraidortocryortostartlemng' more

thanonelearnstofeelpainortogaspforair" (Tomkins, 1987p.1 7)

There are nine innate affects, interest-excitement, enjoyment-joy, surprise-startle, distress-
anguish, fear-terror, anger-rage, shame-humiliation, dissmell (the innate smell response to
bad odors), and disgust, (Tomkins, 1987).

Whileaffectisloeatedin subcortiealcentersin thebrain, theprimary siteofaction
of the affect system is the face. Fach innate affect is involved with groups of voluntary
muscles which are temporarily taken over by an affect as it emerges, creating a prototypical
facialresponse foreachofthenineaffects. Theshameresponseischaracterizedby
hanging the head, lowering or averting the eyes, and blushing. According to Tomkins
(1987), what is viewed and understood as facial display ofemotion is actually an "inward
feed” of information from the face to conscious awareness. Affect is primarily facial
behavior. As the developing individual becomes aware of these facial responses, she
becomes aware of her affects. Originally, psychological processes do not create affect.
Affect is innately activated by stimulation of speciﬁc receptors or the pattern of stimulation.
'I'hedensityofneural ﬁringalong with itsproﬁleova'timedetamineswhichaffectwillbe
innately triggered

14

Tomkins concepnralizesshameasanauxiliaryaffect, meaning thatitrequiresthe
presence of another affect, speciﬁcally interest or enjoyment. According to Tomkins
(1963, 1987), the incomplete reduction of interest or joy by some barrier activates shame.
Nathanson (1987) uses the term ”photo-shame” to describe the infantile experience of
shame. According to Nathanson, this proto-shame has no meaning, it is simply an innate
reaction to the rapid but partial reduction of positive affect. later, the infant “learns” to use
these innate facial expressions for voluntary expression as well. In addition, over time,
sharnebecornesassociated with input from interpersonal interactions,as lifeexperience
adds to the original physiologieal experience of shame.

Kaufman (1989, 1992) has expanded Tomkins' original formulation of shame and
provides a detailed explanation of the processes involved in the creation of a shame—based
identity. While classical Freudians posit libidinal and aggressive drives as the sources of
human motivation, and interpersonally-oriented theorists understand components of the
inta'personal relationship as the primary motivating force, affect theorists view affect as the
fundamental source of human motivation. Affect is viewed by both Kaufman ( 1989, 1992)
and Tomkins (1963, 1979, 1987) as distinct from drives and also from the need for
relationship. According to Kaufman ( 1989) it is affect which serves as the primary
motivator.

“It is affect that gives textru'e to experience, urgency to drives, satisfaction

to relationships, and motivating power to purposes envisioned in the future.

The affect system and the drive system are distinct, interrelated motivators.

They empower and direct both behavior and personality, but the drives must
borrow their power from affect. . .” (Kaufman, 1989 p. 61).

Affect is an ampliﬁer of all experience, including needs, drives, cognition, memory, or
even other affects (Tomkins 1963, 1987). When any of these is ampliﬁed by affect, that
affectean tbenbecomeattachedtotheneed,drive, cognition, memoryorexperience.
According to Kaufman (1989) individuals internalize their experience through imagery.
Scenes are internalized images that have become infused with affect. Scenes, imprinted

15

with affect, are stored in memory and become the foundations of personality (Kaufman,
1989).

When an affect, drive or interpersonal need is followed by shaming, shame scenes
are created. According to Kauﬁnan (1989), if a particular drive, affect, or need becomes
linked with shame, an internalized connection (shame-bind) to that affect, need, or drive
willbeestablished. Thecreationofshamebinds meansthatrecmrencesofthataffect, need
or drive will now spontaneously activate shame by reactivating the entire scene. Because
the shame-bound need, drive, or affect, is now experienced with shame, its expression will
be constricted, further restricting the expression of self.

Psychological magniﬁcation of scenes occurs when one affect-laden scene becomes
fused with a scene ampliﬁed by the identical affect (Kaufman, 1989; Tomkins, 1979);
when multiple affects about the same scene are combined; or through the combination of
multiple sources of shame about the same scene (Tomkins, 1987). Families of scenes are .
createdinthis way. Patternsofaction,calledscripts,arethencreatedasameansof
anticipating or controlling a magniﬁed group of scenes. In the case of shame-bound
scenes, scripts serve the defensive purpose of protecting the individual from experiencing
flu'ther shame. As additional shame binds are created, magniﬁcation takes place and
shame increases its power and control over the self.

2 I l . 1] .

A distinct group ofsharne theories is rooted in psychoanalytic theory. In these
particular theories, both physiological drives and interpersonal needs supercede the
importanceofaffcctasamotivating force. Freudconceptuallzed' sharneasareaction
formation against libidiml impulses and as a defense against cmiosity and self-exposure
(exhibitionism) (Freud, 1933; 1953). Other psychoanalytic theorists have greatly
augmentedtheshameliterature,conceptuallzlng’ ' thedynamicsofshameandaddressing
with greater speciﬁcity the developing context of shame, interpesonal experience.

16

Lewis (1971, 1987a, 1987c) understands shame as a state of self-devaluation, ”a
lapse from the ego-ideal” (Lewis, 1971, p. 37) which is experienced vicariously as the
negative evaluation by an other. According to Lewis, shame is a super-ego function; the
“affective-cognin've signal to the self that its basic affectional ties are threatened” (Lewis,
1987c p. 114). Shame is originally caused by a failure of a central attachment bond. It
neeessarily develops out of relationships with others. The development of shame requires
a relationship between the self and an other where one cares about the other's evaluation.

Wurmser's (1981) conceptualization of shame is similar. He believes that a failrne
to meet the standards of internalized objects results in shame. Wurmser (1981)
emplnsizes the power of early or arclmic internalized shame over later "realistic” or external
shame. Although our culture often equates shame with sexual exposure, he argues that
shamealsoinvolvesthebroaderexperienceofweaknessorfailrue. Tobeweakordirtyor
defective in one's own eyes is to be ashamed. To be ashamed, ultimately is to feel
unlovable.

”In asenseloveatitspeakmeans beingasfully acceptedasis humanly

possible in the wish for enriching self-expression and in the desire to be

gloriously and abidingly fascinated and impressed—and to have reciprocity

in this on uncounted levels of communication and attentiveness. Shame is

the defeat of such love . . ." (Wurmser, 1981, p. 166).

Wurmser posits that shame involves two modes of exposure. One is embarrassed
when one is revealed and also when one is caught viewing someone else's exhibitionism.
loordng and being looked at can both be shameful.

"Perceptual-expressive interaction is the zone cardinally important for the

development and the core of our identity. Only in seeing and being seen, in

hearing and being heard, can we match our self-concept with the concept

others have of us. The modes of attentive, curious grasping . . . and of

expressing oneself in nonverbal as well as verbal communication are the

arena where, in love and hatred, in mastery and defeat, our self is forged

and molded. If this interchange is blocked and warped, the core of the self-

concept is severely and permanently disturbed, twisted, deformed. . .The

consequence of such an interference is that expectations and reality never

seem to ﬁt: 'Ihe real (experienced) self of me never matches what 'they'
expect, nor do 'they‘ ever match what I expect” (Wurmser, 1981, p. 163).

17

Wurmser (1981) argues that much of severe psychopathology is based on often disguised
shameconﬂicts,andis setuptoundo,andatthesametimeperpetuate, the shametraumas
that have created a profound sense of unloveability.

Morrison (1984, 1987, 1989) has written about shame within a self psychology or
Kohutian framework. Morrison views shame as an affect of central importance which
reﬂects feelings of inferiority, defect, and failure of the self.

“Shame reﬂects decreased self-esteem—a manifestation of the selt’s sense of

failure with respect to goals and ideals, its deﬁcits with respect to early

insufﬁcient functions of its selfobjects” (Morrison, 1987 p. 289).

The phenomenological withdrawal experienced with shame is not only from external
objects; it is also a withdrawal from a negative or despairing self awareness. This self-
awareness is rooted in internalized “selfobjects” which reﬂect the empathic quality of early
relationships.

Although they differ in conceptions of how this occurs, both affect theorists and
psychoanalytictheoristsagreethatinterpa'sonal factorsareinwgraltodreﬁnkageofslrarne
to behavior and to identity. Where Kaufman discusses the binding of innate shame to
interpersonal needs, and the creation ofshame binds through the reactions ofothers,
Lewis, Wurmser, and Morrison, view shame as created and internalized within the context
of interpersonal relationships.

W

Shame is a powerful affect experienced as exposure before either selfor otha’s.
The association of shame with identity, with interpersonal expaience, with drives, or with
other affects can lead to inhibited expression in an effort to avoid the painful expaience of
shame. The personality can thus be profoundly affected. Shame theorists have posited
preliminary reconceptualizations of psychopathology, integrating the concept of shame with
the development of psychological disorders. Three theorists have speciﬁcally formulated
shame-based conceptualizations of bulimia nervosa. The theories of Kaufman, Wurmser,

18

and Iewis will be reviewed. They are notably different, reﬂecting their disparate
understandings of human motivation and development. Winnicott's related construct of a
false self will also be presented.

According to Kaufman (1989), repeatedassociation ofshame with interpersonal
needs, with hunger or sexual drives, or with other affects may lead to the development of
”shame syndromes” governed by central internalized shame scenes. These shame
syndromes are “constellations of affect, scene, and script” (Kaufman, 1989 p. 153). Thee
aredistinctshame syndromes, shapedbythenatureofthescmesandtheshame-binds,
which involve characteristic patterns of reproducing shame and further distorting the self.
The scripts or rules that an individual develops over time to predict, control, respond to,
and interpret a set of scenes magniﬁed by affect further solidiﬁes the individual's response
to these scenes (Kaufman, 1989; Tomkins, 1979, 1987). While Kaufman (1989) does not
believeallpsychopathologytobefoundedin shame, heargues that shamescenesand
scripts are central to the development of affective, narcissistic, borda'line, compulsive,
addictive, and eating disorders.

Kaufman (1989, 1992) argues that bulimia nervosa is a shame-based disorder.
According to Kaufman, both bingeing and prrrging are, in part, substitutions for more
shameful interpersonal needs. Bingeing on food is a substitute for interpersonal needs
which lave become bound with shame through repeated association. Bingeing on food
takestbeplaceoffulﬁllingtheneedforothers, which ispa'ceivedbythebulimicasacause
for shame. Purging isafutileand syrnbolicattempt by thebulimic toridherselfofthe
shame she feels, both for the bingeing behavior, and as a result of the unavoidable
’ experience of interpersonal needs.

Kaufman (1989) deﬁneatessevalintape'smalneedsﬂlefulﬁllmentofwhichare
necessary for the optimal cbvelopment of the individual: 1) the need for touching and
holding, 2) the need for identiﬁcation, the phenomenological experience of merging with

19

another, 3) the need to be in relationship with another, 4) the need for afﬁrmation, 5) the
need to nurture, 6) the need for power, and 7) the need for differentiation, embracing
separateness and autonomy. The ﬁrst four needs involve an aspect of submission to or
dependency on relationship. The need to nurture others and the need for power involve
some aspect of control over relationship, and the need for differentiation, encompassing
separateness and autonomy, indicates the need to be separate ﬁom a relationship.
According to Kaufman, the fulﬁllment of these needs is critical to the healthy development
oftbeindividual. Totheextentdlatanyoraﬂoftheseneedsarelinkedwiﬂltheexperience
of shame, optimal development is inhibited.

Bingeing, of course, does not adequately fulﬁll the individual's shame-bound needs
for others. The continued need for the other, combined with the secondary shame
associated with uncontrolled eating, serves to perpetuate and extend the bingeing behavior.
Shame is displaced ﬁom the self onto the act of bingeing. Purging, however, involves the
additional affect of disgust, which, like shame, is an auxiliary affect, according to Tomkins
(1987). Disgust becomes associated with the hunger drive, perhaps due to family and
cultural expectations of perfection, thinness and control over eating.

For Kaufman, the concept ofaffect magniﬁcation (Tomkins, 1963) is central to
understanding the binge-purge cycle. According to Tomkins (1963), affect magniﬁcation
is a proeess whereby an individual overwhelms herself with shame, bringing shame to
peak intensity. At this point, the affect is so intensiﬁed that it “erupts” or “explodes” and is
automatically reduced. Kaufman (1989) uses the descriptive metaphor of cleansing oneself
emotionally by bathing in shame; through the process of total humiliation and spending the
built up shame, the bulimic is puriﬁed or cleansed. Kaufman argues that the process of
purging not only lids the bulimic of food, but temporarily of shame as well. Bingeing
increases the build up of shame and then purging rapidly magniﬁes it. Shame and disgust
peak and then there occrn's a ”bursting effect” which leaves the bulimic feeling purged,

20

puriﬁed of shame. The shame, of course, is not eliminated entirely, and the cycle
eventually begins again. In addition, bulimia itself creates additional shame, leading to
increased isolation, which creates increased needs for interpersonal contact, further
penetrating the cycle»

Wurmser's ( 1981) very different understanding of the role of shame in
psychopathology and the involvement of slnme in disordered eating neva-the-less echoes
similar themes. Wurmser posits a shame syndrome as well, a continuum of neurotic to
psychotic behavior which includes varying degrees of four major symptoms,
depersonalization, eating disturbances (which include anorexic behavior as well as bulimic
binges), depression, and delusionally intense feelings of shameful exposure and rejection.
This shame syndrome originates in early conﬂicts over the desire for dependency and
symbiotic merging with the other and an intense desire for autonomy. In these
individuals, emotional intimacy has become equated with intrusiveness and loss of control;
the desire for autonomy, fueled by fears of total rejection, humiliation, or exposure
(”shame anxiety”) provides a safe haven, but results in painful isolation.

Wurmser (1981) maintains that orality and eating often play accrual role in the
shame-basedpersonality. Eating maybeusedasatietorealityinordertocounteract
overwhelming fears and wishes for symbiotic merging. The oral realm provides a concrete
arena for enacting the conﬂict betweal taking in and expelling; between allowing intrusion
(or intruding) and alternately nnintaining isolating control (or spitting out and rejecting the
other). Wurmser believes looking and eating can be tools for power and destruction. He
also argues that both are highly libidinized. Merging, through witnessing the other's
exposure, is frightening and according to Wurmser, the visual conﬂict is transferred to an
oral binge in an effort to regain power. From Wurmser’s perspective, eating binges are
shameﬁrlandarekeptsecretbecausetheyleadtosuong guilt feelings. Thisguiltisrelated
tothedestructivenessofone’soralimpulses, toshame, todisgust withtheoralgratiﬁcation

21

imelf,andwtheweahressdeemedmhaentmdledependencyonomlgraﬁﬁcaﬁon.
Bingeingisshameladenbutﬂmtshameisusedasadefenseagainstthemoresevereshame
over wishes for emotional intimacy and dependency. In this way bingeing behavior both

guardsagainstandperpetuatesshame.

”. . .eating merelyaddsto'theblackvomitinside me'andtothefecal
masses, 'to this through and through ﬁlthy nature of mine' thus swelling
ﬁn'ther the sense of shame. Eating is taking in from and of another and this
is intrusionubut a controlled one, one actively performed, not passively
suffered; yet in its symbolic equation with the other, it needs to be rejected .
. . beyond eating, an even more general dilemma appears. To be close
means to be intruded upon and swallowed up by the other—clearly and
insupportably a humiliating monument to one's weakness. Distance, on the
contrary, meansrejectionanddisdain: 'Iamtreatedliketheheapoftrashl
Eelasll)y am '«which once again is crushingly shameful” (Wurmser, 1981, p.

Lewis (1987a, 1987c) believes that neurotic symptomology or behavior is
ﬁemmﬂymeresldtofdecmsdomamenmtmmainmmmmlostaﬂ‘ecﬁmambmds.
Thefailmeofacentralattachmentbondresultsinshame. Thisshameaswellasthepainful
expaimceofbsingmlauachmmtbecausemehasnmbearabhmﬁvermmthemdards
ofan admiredintemalized image evokes rage, wlmt Lewiscalls ”shame-rage” or
”hurniliated-ﬁlry”. Sharne-basedrageisturnedagainsttheself,outoffearoflosingthe
valuedother. Lewis(1987d) believesthatbecauseofsocietalnorms,womenare
predisposedtoinﬁmlizehmniliatedfmyorshamerage. Inwritirlgaboutthegreater
ﬁequencyoqurressioninwomalshearticulatesafamiliarconﬂict:

'Ihebiologicalandculnnalexpectaﬁondlattheywﬂlbemoﬂrasmamit
appear natural that they should spend their lives devoted to othersnhusband
and children. But our society also scorns people who are not self-sufﬁcient
and independent of others. Women thus learn early that they should be
ashamed of the very set of qualities which are particularly theirs. Ironically,
at the same time, they are constantly threatened by the prospect that if they
are not affectionate enough and as close and loving to others as they ought
to be, they will have failed in their own and others' eyes. They are ashamed
of themselves if they are close to others and guilty and ashamed of
themselves if they are not. Within this profound conﬂict, the chances for
throttledhumiliatedfury aregreat. Any disturbancein theirrelationslripto
others” .can throw her into a state of unconscious fury at the way her self
hasbeentorn. Butatwhomis sheﬁnious—ha'selforthebelovedadmired

22

other with whom she is so close. This is the same confusion she faced

when ﬁrst she experienced rivalrous hatred of her mother. Then, also, it

was hard to separate the hatred of herself from the hatred of her ﬁrst

caretaker, in emulation of whom her self had been developed. In

adulthood, humiliated fury is deﬂected by women from the 'other' who is

its 'unjust' target, back upon the self” (Lewis, 1987d, p. 247).

Lewis (cited in Teusch, 1988) understands bulimia as one means of directing the
rage toward one's self. Bingeing on food becomes a means to direct the hostility against
one's self,incrdertoprotectothersfromtherage. Purgingactsasameansofcleansing or
removing the bad feelings. Bingeing and purging are selfrdestructive rageful acts which
alsoservetoarhancethe ”false self”,inordertomeetexternaldemandsthatwereatsome
point imposed by others. Attention to the ”false self" masks the rage, while at the same
time, it eases shame about the self (Lewis, 1987a).

Thereaction to shameistheimpulseorwish tohideand thedesiretoavoid
experiencing the affect (Kaufman, 1992. Wurmser). Winnicott's (1965) ”false self”
construct describes one means by which this may occur. Winnicottviewsthetrueselfas
the spontaneous selfthatexistsin the infant. Ignoring orreacting inappropriately to the
spontaneity of the true self is the equivalent of shaming (Morrison, 1987). According to
Winnicott,thefalse selfisan exaggerationofthepublic faceorimageoneextendstothe
external world in an effort to protect the true self. If the true self is sufﬁciently shamed,
thenthefalseselfcarr becomeoverdeveloped,andcanbecometheinter1ralizedsenseof
self, masking the true self (Winnicott, 1965).

Shametheoristsandothelshaveposited thatexcessive shame, speciﬁcally shame
related to conﬂicts over separation and connection, is the underlying basis for the
development of bulimia nervosa in women. The research on bulimic women and their
development is extensive. Several important areas of this research appear to reﬂect the
involvement of shame and conﬂicts over separation and connectedness in the development

of bulimia nervosa. The research on bulimic communication patterns and on the bulimic's

23

approachtoandperceptionofintapersonalneedsandoffeelingswillhereviewedand the
research investigating the role of cultlnal values in the development of eating disorders will
be presented.

12 ﬂ . I Q . .

Much of the research involving the families of bulimics has focused on
cornmrmication patterns within the family. Families of bulimics show several
dysfunctional featrues fairly consistently. Humphrey and her colleagues (1986) compared
the intapersonal behaviors of 16 hulirrric families to non-bulimic family controls in a
problem solving role play situation. Researchers were able to blindly differentiate bulimic
families ﬁom non-bulimic family controls based on family communication patterns
(Humphrey, Apple, & Kirschenbaum, 1986). Through the use of complex observation—
rating systems, they found that pararts of bulimics had a tendency to use ”double-hind”
communications which presented contradictory directives. Bulimics' responses to self-
report measures have revealed indirect family communication styles (Johnson & Flach,
1985). Bulimics and their mothers have both indicated that their families approach conﬂict
indirectly, and that conﬂict tends to be elevated in these families (Attie & Brooks-Gum,
1989; Johnson & Flach, 1985; Stroher & Humphrey, 1987). These families have been
described as more disparaging and hostile (Humpln'ey et al., 1986; Strober & Humphrey,
1987), more walled off, less cohesive, disengaged and at the same time more enmeshed
(Humphrey et al., 1986; Johnson & Flach, 1985; Stroher & Humphrey, 1987), less
helpful or supportive (Humphrey et al., 1986; Johnson & Flach, 1985), less nurturing or
trusting (Humphrey et al., 1986; Strober & Humphrey, 1987), and less expressive
(Johnson & Flach, 1985). Humphrey and Stern (1988) argue impressively for the
importance of an ”integrative” analysis of the dynamics involved in bulimia. They present
a comprehensive theoretical conceptualimtion which stresses both individual intrapsychic
dynamics and the dynamics at the level of the family system.

24

Shame theorists also argue that shame is an inter-generational phenomenon; the
sense of shame about shame and the tendency, both at the level of the individual and the
level ofculture todeny, coveruporavoid shame leadstoits perpetuation inboth
individuals and families (Fossum & Mason, 1986; Kaufman, 1989, 1992; Wrrrmser,
1981). Fossum and Mason (1986) have proposed a set of characterisitics and rules which
theybelievecharacwrizefamiliesdorninatedhy shame. Dysfurlctionalcopinginthese
farniliesresults from therepeateddenialoftheshame. Through lackofdircct
acknowledgment, the shame is perpetuated. Fossum and Mason believe that family scripts
andrulesaredevelopedwhichreﬂectthe shameinthesefamiliesandthestrongneedsto
avoid and deny it. According to Fossum and Mason (1986) the script of a shame-based
family demands rigid control over all behavior and interaction, perfectionism~more aptly
deﬁned asperfectadherencetoaveryvaguelydeﬁnedexMralimage—and the useof
blame to cover shame over instances of lack of control or imperfect outcome. Other ”rules”
includethedenialoffeelingsthatareregativeorthatsignalaneedfornurturanceorneed
for an other; the use of unreliability, incompleteness and lack of resolution to avoid facing
isnlesﬂratnﬁghtamrseshameambmahommlldngabmnbehavimmmisshmneﬁﬂ; and
the use ofdenial or disqualiﬁcation to reframe and thus deny any occurrences of shameful
or abusive or compulsive behavior (Fossum & Mason, 1986).

The studiespresented supporttheadherenceofbulimicfamiliestosuch ”shame
scripts”. Bulimic families appear to lack the skills or ability to communicate honestly and
directly. Itcanbepositedthattheparentsin thesefamiliesaresufferingfrorntheirown'
shame. This shame, and fears of acknowledging it, leads to severely dysfunctional
communication ploys, invoked as a means of protection from painful affect. These
communication tactics may havebeenlearnedin theirown childhoodandwouldappearto
be a primary method by which shame is perpetuated intergalerationally. Shame-based

25

families fail to provide experiences which allow their members to learn and pracu'ce
assertive behavior and effective coping skills.

Bulimics certainlyappeartobelackingin theseareas. CattanachandRodin (1988)
reviewed the literature on the role of psychosocial stress and bulimia. They found that
while the stressors these women report are relatively normative, bulimic women tend to use
passive, and less effective strategies for dealing with stress. They suggest that bingeing
and purging eventually become the primary coping mechanisms for these women when
theyareconfrontedwith stress,asaway ofmanaging feelings,orwhen theenvironment
seems chaotic and beyond their control.

D . l [H I l E 1'

Difﬁculties in handling the conﬂicting needs of autonomy and dependence have
beendiscussedmtheshameﬁtaanneasameesmﬁonofshame(Fossum&Mason,
1986; Kaufman, 1989; Wurmser, 1981). Fossum and Mason (1986) believe that placing
an exaggaated priority on independence coupled with devalrring or denying needs for
nurturance and help (because neediness is viewed as shameful) leads to the inhibition of a
mature self. They argue that individuals or families who over-value autonomy never learn
tocreatebalancebetween theneedstobeindividualanddiﬁerentiatedandtheneedtohein
relationship with others. When the need to be independent is overly stressed, the
development of the self is stunted because of the continual need to deny natural (but shame-
bormd) needs for dependency on and relationship with otha' human beings.

Bulimic women appear to have great difﬁculty dealing with issues surrounding
autonomy and identity. Bulimics are reported to have an external locus of control and to
display a related sense of personal ineffectiveness (Dickstein, 1985; Johnson & Maddi,
1986). They are described as feeling helpless and somewhat out of control in relation to
their bodily experiences (Johnson & Maddi, 1986). Bulimics have been reported to display

26

strong needs to conform and gain approval ﬁom others and to be very sensitive to rejection
(Boskind-Lodahl, 1976; Garﬁnkel & Garner, 1983).

Bulimic families offer little support for autonomy (Attic & Brooks-Gum, 1989).
Family communication research emphasizes the lack of supportiveness or nurturance and
failure to encoruage self-sufﬁcient, assertive behavior in these families. These dynamics
can certainly be linked with the bulimic's feelings of ineffectiveness, need for approval,
and overall difﬁculties in coping with stressful situations. Johnson and Flach (1985) report
thathulinric fanriliestendtohavehigh standm’dsofpa-fonnancehutatthesametimeplace
a low emphasis on social and intellectual activities that might serve to foster that
achievement. Pafectionism is expected, while at the same time the family does not support
independent, assertive, or expressive behaviors. In addition, such a double-bind leads to a
no-win shame situation in which the individual is shamed for being dependent and yet is
leftashamedofherinability tobeindependentbecause shelacksthe skillsandsupportin
this endeavor.

Referring to Winnicott's ( 1965) false self construct, Jones ( 1985) theorized that the
bulinric's shameoverherneedforothersissointensethat shecreatesanexaggeratedfalse
self, a false self which emphasizes pseudo-independence and pseudo-achievement. The
false self, instead ofthe true self, is internalized and the submagence ofthe true selfis
posited to lead to the bulimic's feelings of emptiness, ineffectiveness, unrealness and
shame (Johnson & Maddi, 1986; Jones, 1985). In this way shame cycles or spirals,
leading toadaptations thatonlyincreaseandfurtherperpetuate shame.

The research literature on bulimia reﬂects these individuals' difﬁculties with
interpasonal relationships and in deﬁning themselves in relation to other pe0ple. Bulimics
suffer ﬁ'om disrupted social relationships and increased isolation (Johnson & Berndt,
1983). Bulimics lave been reported to display signiﬁcantly greater fears of intimacy than
non-bulimics (Pruitt, Kappius & Gorman, 1992). These women reportedly have great

27

difﬁculty dealing with time spent alone (Cullari & Redmon cited in Cattanach & Rodin,
1988) and display strong needs to conform and gain social approval (Boskind-Lodahl,
1976; Garﬁnkel & Gama, 1983).

E f: . . I C l l I! l

Bulimic families emphasize pafectionistic standards of bdlavior and achievement
(Attie & Brooks-Gum, 1989) and bulinrics tend to be perfectionists with high expectations
for themselves (Boskind-White and White, 1983; Garﬁnkel & Gama, 1983). In addition,
thesewornen display strong needstoconforrnandtogainapproval fromothers (Boskind-
Lodahl, 1976; Garﬁnkel & Gama, 1983). Such needs for approval may lead to behavioral
and even personality changes aimed at garnering positive evaluations ﬁom others.

Ithasbeenpositedthatlackofsupportforand shaming ofthetrueselfcan drive the
true self underground and encourage the development of a false self, built around external
ideals (Winnicott, 1965). The characteristics of the false self are related to those qualities
onewishestopresenttotheenvironment, thefalseselfisamashapuhlic facetbatone
believesismorelikelytogainsocialapprovalthanthetrueself. Characteristicsofthefalse
selfmayhereﬂectedinthevaluesofthefamilyandthecultrrre. Cultural attitudesabout
weight, body, and appearance, interpersonal needs and gender roles can result in shaming
on an interpesonal or societal level.

Theoristsarguetbatcharlgesin cultural idealsregarding the fenralebody have ledto
increasedbodyshameandanincreaseineatingdisorders. shrdiesrevealthatoverthelast
few decades, the ”ideal woman” has become slimmer; even Playboy centerfolds have
becomethinnerand moreangularoverthelastZOyears. MissAmericacontestants show
declining weight as well (Gama, Garﬁnkel, Schwartz, & Thompson, 1980). Silverstein,
Padue, Peterson, and Kelly (1986) provide convincing evidence that the media promotes
andperpetuates standardsofthinness forwomen. Astheideal body becomes thinnerand

28

lighter, statistics reveal that young women are growing heavia, further widening the
”shame gap” between cultural ideals and reality (Gama & Garﬁnkel, 1980).

Theolists have also posited that cultural values and expectations regarding gender
roles must be considered in developing an understanding of bulimia and of the bulimic's
conﬂict regarding autonomy and dependence. Some research has found evidence of
increased adherence to traditional female gender roles among bulinrics, with the traditional
role characterized by ”dependence and passivity” (Boskind-White & White, 1986; Steiger,
Fraenkel, & Ieichner, 1989; Pettinati, Franks, Wade, & Kogan, 1987). Silverstein,
Perdue, Wolf, and Pizzolo (1988) reported that eating disorders appeared to heparticularly
prevalent among women who reported that their parents held negative attitudes toward
female achievement.

Insufﬁcient research has focused on elucidating the perceptions and ideals of the
bulinrics themselves. Pettinati, Franks, Wade, and Kogan (1987) had 37 eating disordered
patients complete the Bern Sex-role Inventory twice, with self and ideal-self ratings. They
reputed that this group rated their ideal selves signiﬁcantly higha on feminine ratings and
concluded that eating disordered women over-idealized feminine traits. Paxton and
Sculthorpe (1991) assessed attitudes about sex role characteristics in a slightly different
manner, arguing that any relationship between sex role charactaistics and disordaed eating
wmﬂdbeobscmedifadiscﬁmimﬁmwasnmwdebaweenposiﬁveandnegaﬁveuaits.
In their study, the researchers differentiated between positive and negative masculine and
feminine characteristics. For example, ”gentle” was considered a positive feminine
characteristic while ”weak” was deﬁned as a negative feminine characteristic. The authors
found that the more eating disordered the individual, the few positive masculine
attributes she attributed to ha self and the more negative feminine characteristics were
attribumd. Although the authors investigated ideal-self perceptions as well, and report a

29

disclepencybetweenrealandideal, thedataprovideddonotprovideaclearpictureofthe
ideal characteristics reported by the subjects.

Rost, Neuhaus, and Florin (1982) report that bulimic women scored signiﬁcantly
higher than non-bulimic women on a scale of ”sex-role fatalism.” Silverstein, Carpman,
Palick, and Perdue (1990) report that women who exhibited genda identity conﬂict,
(deﬁned by drawing an androgynous ﬁgure on the Draw-a-person Test or by reporting
wishing theyhadbeenbommale) weremorelikely than otherwomentoreport frequent
bingeing or purging. They hypothesize that bulimia may be related to women's struggles
to deﬁne themselves in areas historically associated with male achievement.

Steiner-Adair (1986) theorizes that eating disorders are the result ofa cultural
overemphasis on autonomy which is rmhealthy and unrealistic, and a culture-wide shaming
of females. She argues that females are acculturated to view themselves in relationship
with othersandyetareshamedforthesevalues; insteadtheyaretaughttovaluethetraits -
for which male children are generally socialized, namely, independence and autonomy. In
other words, women are taught to be one thing and then told to be something else. \Vrthin
a culture which values ”male” tendencies, females shame themselves and are continually
shamed by others.

Using clinical interviews and diagnostic measures with a group of 32 adolescents,
Steiner-Adair (1986) was able to almost perfectly differentiate a subgroup of females who
medmﬂedixxdaedeaﬁngmngeonmeFaﬁngAtﬁmdesTemanobjecﬁvesdf-repm
instrumentdesignedtoassessabroadrangeofeating disorderedbehavior. This subgroup
identiﬁed cultural ideals of autonomy and success in deﬁning a ”superwoman” and did not
separatesociehlideals from theirownvaluesindescnhingwhattheybelievedtheideal
womantobe. Theyappearedtoundastand needing orinterdependencewithothersas
sharheful. Females who were able to recognize the ”superwoman” image and the emphasis
onautonomyasaproductofculture, hutwhoincludedthevalueofintadependenceintheir

30

own goals, did not score as eating disordered. Steiner-Adair’s (1986) important discovery
that women with disordaed eating could be distinguished from a larger group, solely on
the basis of their depiction of the ideal woman, maits further exploration.
Thevmiedreseuchmculnnalvaluesasdleyrelatembuﬁmiamﬂectsﬂledynanﬁc
of shame at sevaal different levels. Thae is some evidence that bulimia is more prevalent
among certain cultural groups, namely women ﬁ'om middle or upper-class families
(Shisslak, Crago, Neal, & Swain, 1987), suggesting that groups which espouse certain
values (high achievement, thinness, perfection and autonomy) and shame others may be at
increased risk. Preliminary evidence indicates that the strength of one's ties to the
“mainstream,” Caucasian American culture is related to one's risk for developing an eating
disorda (Purnariega, 1986), strengthening the evidence for a cultural link to this disorder.

 

While theorists are continually reformulating their understanding of the etiology of
bulimia, draehasbeenlitﬂedirectinvesﬁgaﬁonintothebulimic'sownpacepﬁonofha
disorder. Preliminary work in this area strikingly supports etiological theories involving
shameespedaﬂyasitrelatestointerpasonalneeeds.

Teusch (1988) interviewed 40 bulimic women in an attempt to understand how they
make sense of their symptoms. Subjects most often chose shame and guilt, over
depression, positive feelings, anxiety, or anger to describe their affective experience of
bulimia. One hrmdred percent of the sample attributed factors about themselves to the
development of their bulimia. Family factors were mentioned by 50 percent of the group in
this regard. Parental emphasis on food, weight, and diet was a prominent theme, but
within thiscontextitwasthelackofnurtmanceand connection witbtlreirparentsthatthese
women felt was problematic. Approximately one half of the women felt that their
”interpersonal beliefs” had contributed to the developmmt of bulimia, and 82 percalt
mentioned speciﬁc interpasonal experiences when discussing the development of their

31

eating disorder. When these women discussed their interpersonal beliefs, Teusch reports
tint feelings of emotional isolation and disconnection wee prevalent, as were negative (or
shaming) interpe'sonal experiences.

The motives given for bingeing and purging revealed conscious attempts by these
womer to satisfy needs independently of others and to cope with feelings of shame, rage,
and anxiety that result from the continued repression of wishes and needs and also from
emotional isolation. Needs for nrrrturance and concomitant inabilities to ask for or receive
nurturance were reported.

Though it did not set out to investigate eithe' shame or bulimics’ paceptions of
interpersonal needs, this study clearly supports theirrelevance to bulimia nervosa. These
huhmicwonrenreportedconﬂictsovaneedsfordepeldenceonothasan inabilityto
directly express feelings involving nruturance or neediness, a disruption of family
relationships, issues involving food and body, and interse personal shame about the self,
factorswhichhavebeenreportedelsewbereaswell. ItisimportanttonotethatTeusch
(1988) found no relationship between these womens' degree of insightfulness and
treatment history, making less likely theargumentthat thesewornen hadhadtheir
”motivations” explained to them in therapy.

In an earlier study, McCreery (1991) compared bulimics and non-bulimics on the
dimensions of shame, and real and ideal levels of ”emotional reliance on Me pe'son”
and autonomy. Bulimics reported a signiﬁcantly higher level of shame than non-bulimics.
Theyalsoreportedsigniﬁcantly moreemotionalrelianceonanotheperson. While thee
wasnotasigniﬁcantdifferencebetweengloupsin tbeideal level ofemotional relianceon
another person, the bulimic group reported a signiﬁcantly higher ideal level of autonomy.

This study further investigated bulimic and non-bulimic perceptions of inte'pesonal
needs involving dependency, speciﬁcally shameful paceptions. Subjects listened to three
short audiotaped ”interviews.” Fach interview depicted a conﬁdent, healthy, well

32

functioning female college student. The tapes differed in the main character‘s approach to
interpersonal needs. One was autonomous and self-reliant; one displayed and was
accepting ofinterpersonal needs; and the third served as a control—her stance toward
inmrpe'soual needs was not clearly deﬁned

Subjects wereaskedtocompleteameasureofshameas they thought thecharacter
would respond. Strikingly, both groups attributed significantly more shame to the
individual who displayed interpersonal needs, with the difference between groups being
one of extreme. The bulimic group attributed signiﬁeantly highe levels of shame to the
inte-personal needs characteization, peceiving this individual as experielcing above
wage levels of shame. The study lends emphasis to the importance of considering
shame, paru'cularly shame related to perceptions of dependency and autonomy, in the
dynamies of bulimia nervosa.

Shame, especially shame related to interpersonal needs, clearly appears to be woven
throughout the bulimic experience. Research consistertly supports the shame-based nature
ofbulimia,boththe shamerootedintheindividualandherfamilyandtheculnn'al shame
which works to enforoe societal ideals by shaping the standards of individuals and families.
Shame theory is a valuable addition to our knowledge of the developmelt of bulimia

mom and appears to offer a comprehersive and accurate understanding of the dynamics
involved in this disorder. While preliminary conceptualimtions of bulimia as a shame—
baseddisorde'appeartomakesense, ﬁn'themsearehmustseektoclarify and document
the relationship between the two.

 

Bulimia nervosa involves both physieal and psychologieal symptoms. In treating
the disorder one must be cognizant of possible physical sequalae; in one study of eating
disordered females, twenty-two pewnt of the bulimics required hospitalization for medieal
reasons (Palla & Litt, 1988). Menstrual irregularities, especially in bulimics with a history

33

of anorexia, are frequently noted (I-Ierzog & Copeland, 1985). Gastric dilation and rupture
may result from binge eating (Hemog & Copeland, 1985). Bulimies who vomit or abuse
laxatives or diuretics are at signiﬁcant risk of hypokalemia (abnormally low potassium
levels) which predisposes them towards eardiac arrythmias and relal damage (Agras, 1987;
I-Ierzog & Copeland, 1985; Palla & Litt, 1988). The repeated use of Ipacec to induce
vomitting can lead to Ipacec poisoning and, in rarecases, to fatal myocardial dysfunction
(Adler, Walinsky, Krall, & Cho, 1980). Othe' possible conrplications related to vomitting
include dental cavities and emmel eosion, swelling of the parotid glands, and esophageal
tearing and bleeding (Agras, 1987; Herzog & Copeland, 1985; Palla & Litt, 1988).

Additionally, bulimic behaviors of bingeing and purging trigger physiological
sequalae which ean potertially impact both psychological functioning and eating behaviors.
Seveedieﬁngandweightlosaalthough mostoftendiscussedinrelationtoanorexia
ne'vosa, may impact the functioning of some bulimics and may exacerbate bulimia. Keys
et. a1. (1950) described the psychological changes which occurred in response to starvation
in a group of male volunteers. When subjects' weight dropped below 85-90% of what
their ave'age weight should be, researchers noted intense preoccupations with food,
episodes of binge eating, obsessive drinking and behavior, and an inability to recognize
satiation.

Two important physiologieal reactions to bulimia are believed to fuel the cycle of
bulinﬁcbehaviorbyexacebaﬁngdisregrﬂaﬁoninfoodintakeandstorage: (1)
hypeinsulinisrn and (2) hypokalemia. When the bulimic binges, insulin is released from
thepancreas. Purging leavesthebulimic withnofoodinhersystembutwith elevated
levels of insulin. Insan arrouses the appetite ever when the stomach is full (Haskew &
Adams, 1989). Insulin also works to promote the movement of glucose into cells for
storage as fat, leading, potentially, to a slower metabolism (and an increased tendelcy to
gain weight) and to depleted glucose levels (Haskew & Adams, 1989). Vomiting, laxative,

34

and dimetic abuse all result in depleted levels of potassium (hypokalemia). Decreased
potassium levels and low blood sugar may also lead to increased appetite-triggering
another binge (Potes-Park & Bokram, 1989). Thee is some evidence that bulimics
eventually develop hypeinsulisrn, secreting insulin when they see, smell or think about
food (Haskew & Adams, 1989). This hyperinsulinism further interferes with food intake
regulation and may also impact energy level and mood Haskew & Adams, 1989). The
cycle hypodesizedtoresultﬁomandbeexacebatedbybulimiaisiﬂusuatedin ﬁgure 1.

Signiﬁeant research attention has centeed on better elucichting the
pathophysiologieal mechanisms which play a role in (and may in fact exacerbate) bulimia
nervosa. One major focus has beer on examining the eldocrinologic changes (and resultant
physiological mechanisms) brought about by bingeing, purging, and dieting behaviors (see
for example McBride, Anderson, Khart, Sunday & Halmi, 1991; Pirke, Friess, Kellner,
Krieg, & Fichter, 1994; Pugliese'& Lifshitz, 1985; Weltzin, et. al., 1991). While a
discussion of this research is beyond the scope of the present study, a much more
comprehensive discussion of the physiological issues and dynamics involved in bulimia
nervosa is presented by Pirke and his colleagues (Pirke & Vandereycken, 1988).

 

Researches have established that bulimic women often come from families who are
less accepting of interpersonal needs and who emphasize perfectionistic standards.
Bulimics consistently display a heightened level of dependency and reveal strong needs to
conform and gain approval from othes. The research reviewed suggests the involvement
of gender identity issues in bulimia, speciﬁeally difﬁculties related to autonomy and
depenMcy. However, the research in these areas is far from deﬁnitive and is at times
contradictory. In addition, the perceptions of bulimics themselves, speciﬁcally, their
pe'ceptions and ideal conceptions regarding interpersonal or depeldency needs and the
need to be autonomous or self-reliant have been insufﬁciertly explored. Effective

35

Vomiting or
/' Laxative Abuse \
. Decrease in potassium
Binge and blood sugar

Excess Insulin Over-stimulation of /
/ \ the feeding center

Fat deposition Low Blood Sugar

Figure 1: Physiological Vicious Cycle (Potes-Park & Bokram, 1989)

36

psychothe'apeutic treatment requires an accurate assessment and understanding of the
individual's real self, but also an accurate unde'standing of the individual's goals, ideals,
and perceived shortcomings.

Thepresent study has twocomponents. The ﬁrstphaseofthe study isdirectedat
clarifying the bulimic's self oriertation. Theorists have posited that bulimics experience
heightened identiﬁcation with traditional feminine traits and ideals (Boskind-White &
White, 1986; Steiger, Fraenkel, & Leichne', 1989; Pettinati, Franks, Wade, & Kogan,
1987). Othes have argued that the bulimic's self is characterized by gende- identity
cmﬂiachmacteizedbyidedimﬁmofmasculmeamibummawishtobenmlemost,
Neuhaus, & Florin, 1982) or by the shaming of ”feminine" values of eare and connection
(Steiner-Adair, 1986). Others have theorized that the bulimic has developed a counte-
dependent or pseudo-autonomous ”false-self ideal" in order to avoid the shame she
attributes to her rm] dependency and dependency needs (Jones, 1985; McCreery, 1991).
The present study examines bulimics' self orientation and bulimics' ideal self orientation
through their real and ideal responses to the Relationship Self Inventory (RSI).

The second component of the study investigated bulimic, recovered bulimic, and
non-eating disordered individuals' peeeptions of interpersonal needs and self reliance,
speciﬁmlly the shame these individuals may link to these qualities. It is important to
diffeentiatethe shamebulimiesmanifestrelatedtothdrbulimic symptomologyandtotheir
feelings about themselves from the shame they may attribute to the qualities of connection
andself-reliance. Therefore,thestudywillattempttoseparatetheeecomponentsby '
investigating bulimic, recovered bulimic, and non-eating disordered individuals'
attributions of shame (ISS scores) to non-eating disordeed womer who reveal
interpersonal needs or are depicted as autonomous and self-reliant. In addition, diffeenees
and similarities between groups will be examined.

37

It is hypothesized that the bulimic is indeed conﬂicted regarding dependency needs
and autonomy and attributes signiﬁeant shame to he experience of interpersonal needs.
While the bulimic characteristically displays heighteled levels of dependency, it is argued
that her feelings of shame exteld beyond he sometimes pathological dependence. It is
hypothesized that the bulimic ﬁnds all interpersonal needs shameful, even wher the
intepesomlneedsmemnnalandnmpathologieﬂandmedisplayedinamneaﬁng
disordered, conﬁdent, and successful woman. It is suggested that the bulimic's intense
shame ove interpesonal needs and her need to attempt to avoid these shame feelings, leads
totheattemptedderialofherown depeldency needsandtoattemptstogratifythem
indirectly by bingeing or ”taking in" food, a substitute for nurturance. Purging represents
an undoing, or a cleansing. The eltire bulimic cycle serves as a concrete manifestation of
the bulimic's conﬂict related to interpesonal needs. Therefore, it is We hypothesized
that successful recovery from bulimia nevosa involves not simply a change in eating
pattens,butachangein thepeoeption ofdependencyneedsandan increasedacceptance
of personal needs and desires for connection with others.

This argumert runs cormter to the proposition that the bulimic ave-idealism
traditionally feminine characteristics such as dependency. It is posited that the bulimic's
mtenseshmneofmterpesmalmdepeldencyneedsleadshermconsmctacounte-
dependent or pseudo-autonomous "false self ideal." While her own intense dependercy
needs preclude her ﬁom achieving this pseudo-autonomy, this ideal should be evident in
her concepualization of the ideal female.

Because this hypothesized formulation links shameful perceptions of dependency
needs to bulimia, it follows tint recovered bulinrics should approximate non-bulimics in
their perceptions of dependency needs and in their approximation of the ideal female. It
would be expected that recovered bulimics and non-eating disordered females, as opposed

38

to bulimics, would place less emphasis on separation and autonomy and would be more
accepting of connection and relationship in their conception of the ideal.

The following hypotheses follow from this formulation:
Me
1. The behaviorally-bulimic group will report signiﬁeantly highe' levels of inte'nalized
shame (as measured by responses on the 188) than either the behaviorally recovered or
non-eating disordered groups.
2. There will be no signiﬁcant differences between the behaviorally recoveed-bulimic and
non-eating disordered groups on the 188.
3. It is hypothesized, that unde' the "ideal" response set, the behaviorally bulimic group
will score signiﬁeantly higheronthe Separate Selfscaleofthe RSI thaneitherofthe two
other groups.
4. It is predicted that thee will be no signiﬁcant difference between behaviorally recovered
andnon-eatingdisorderedgroupsontheirscoresontheldeal RSI.
5. It is hypothesized that thee will be no differences between behaviorally recovered and
non-eating disordered scores on the "real” RSI.
6. The behaviorally bulimic group's “real" RSI responses are difﬁcult to predict. Because
research consistently reports bulimics' pathologieal levels of dependence, it is likely that
thesewornenwill showelevatedConnectedSelfandespecially PrimacyofOtherCare
scores. However, their hypothesized shame and conﬂictual feelings related to dependency
needs appear to result in an emphasis on self-reliance and isolation. It is hypothesized tint
bulimics are extrerrlely conﬂicted in their expressions of connection and separation and that
theirpatte'nofscoresontheRSIwillreﬂectthisconﬂictandwilldiﬁ‘erfromthescoresof
theothetwogroups. The SeparateSeifandConnected Selfarenegatively correlatedin the
normative female sample (r = -.23) and the Separate Self and Primacy of Other scales show
a very low correlation (r = .09) (Pearson et al., 1991) . It is hypothesized that the

39

behaviorally bulimic group's when of scores will not approximate this relationship;
bulimicswill score mosthighly on the PrimacyofOthersandthe Separate Selfscales,
reﬂecting their high levels of depeldence, their extrene discomfort with this dependence
and their attempts to avert it. While the exploratory nature of this research precludes a
conﬁdent hypothesisrelatedtotherelationship between the Separate SelfandConnected
Self scores for the bulimic sample, it is quite possible that the negative correlation obtained
in thenormativefemalesample maynotbereplicated.
MM

In a previous study (McCreery, 1991), a group of women reporting bulimic
behaviors attributed signiﬁcantly highe levels of shame to a non-eating disordeed woman
who displayed dependency needs than to a woman depicted as self-reliant. Additionally the
bulimic-type group attributed a signiﬁcantly higher level of shame to the woman displaying
dependency needs than did a group of non-eating disordered women. It is hypothesized '
thattheseresultswill bereplicatedin thepreserrt study. Threeadditional hypotheses
follow from this assumption:
7. The behaviorally bulimic group will attribute signiﬁcantly higher levels of shame to the
woman expressing dependency needs than to the woman expressing self-reliance.
8. The behaviorally bulimic group will attribute signiﬁcantly higher levels of shame to the
woman expressing dependercy needs than will the non-eating disordeed or the
behaviorally recoveed-bulimic groups. '
9. The level of shame attributed to interpersonal needs across groups will be positively
correlated with the individuals' level of bingeing and purging.

Q . [D .

The study is a two step design. In the ﬁrst phase, 680 female undergraduates
completed the lnternalrzed' Shame Scale (ISS), two subscales of the Eating Disordes
Inventory (EDI), a demographic questionnaire, a questionnaire about eating habits and

40

behaviors, and two versions of the Relationship Self Inventory (RSI), one version
requesting them toanswerasthey wouldrespond, andoneasthey wouldrespondifthey
wee their ideal self. Subjects for the second phase of the experimert were selected from
thissubjectpoolonthebasisoftheirresponsestotheeating habitsquestionnaireandthe
EDI scales. Twenty-three behaviorally bulimic (BB), twenty-ﬁve non-eating disordered
(NED), and thirty behaviorally recovered-bulimic women (BR) wee identiﬁed.

Inmesecondphaseofdreexpeimelgmreeaudiotapedchmacteizaﬁonswee
presentedtoeachofthoseidentiﬁed subjectswhoagreedtoreturn. Onetapedepictsa
woman who is autonomous and nonreliant; one depicts a woman who displays and is
accepting of interpersonal needs for touching and holding, afﬁrmation, identiﬁcation, and
theneedtobeinrelationshipwith another; andthethirddepictsacontrol situationin which
neithe style is clearly discernable. Subjects listened to each tape and then completed the
188 as they believed the woman they heard would respond. Group responses to each
vignette wee compared both within and across groups.

METHOD

Subjects

It was deemed important that this study not simply investigate diffeences in beliefs
and peeeptions betweel a clinical group (i.e. bulimics identiﬁed by their participation in
some treatmert group) and nonclinical populations, or that any signiﬁcant differelces
between groups merely reﬂect the treatment ideology of a giver eating disordes treatment.
Additionally, this study sought to avoid examining a restricted sample of bulimics by using
a clinical group or by advertising for “bulimics willing to participate in psychological
research,” thereby restricting the generalizability of any ﬁndingsl. In order to avoid these
limitations, the undergraduate psychology subject pool at a large midwesten univesity
was used as the subject source for all subjects. Subjects signed up to participate in a study
on "Female Personality." Although this method narrowed the gerealizeability of the data
to college students, the sampling of subjecm in this manner more closely approximates a
random sampledlanmaybeachieveddrroughdlesoﬁcitaﬁmofsubjectsinnmredhect
marines. Subjects wereinformedatthetimethey signed up forpartipation thatthey might
be recontacted and asked to complete a second phase of the expeiment.

Six hundred-eighty undergraduate females participating in the subject pool (as an
option toearrrextraclasscredit) wee screeredin theﬁrstphaseoftheexpeimerrt.
Subjectsweechosenforinclusioninthe secondphaseofthestudybasedontheir
responses to a structured self-report instrnmert which was a modiﬁed vesion of the Eating

 

1 Research suggests that studies identified as ”eating disorders research”
discourage the participation of some eating disordered subjects (Beglin &
Fairburn, 1992). It is not known if these non-participants represent a distinct
subgroup different from other bulimics.

41

42

Table 1: DSM III-R Diagnostic Criteria for Bulimia Nervosa

A. Recmrent episodes of binge eating (rapid consumption of a large amount of food in a
discrete peiod of time).

B. A feeling of lack of control over eating behavior during the eating binges.

C. The person regularly ergages in either self-induced vomiting, use of laxatives or
diuretics; strict dieting or fasting, or vigorous exe'cise in order to prevent weight gain.

D. A minimum aveage of two binge eating episodes a week for at least three months
E. Persistent ove'concern with body shape and weight.

43

Disorders Invertory Symptom Checklist (EDI-SC). The DSM III-R diagnostic criteria for
bulimia ne'vosa are listed in Table 1. Behavioral criteia~recurrent binge eating and
purging we a three month peiod (criteria A, C, and D)--were assessed directly by
subject's self report. Subjects who reported recurrent episodes of binge eating (at least two
episodes a week for at least three months), who ergaged in either vomiting, laxative or
diuretic use, fasting or vigorous exercise1 on a regular basis (twice a montir or more) were
included in the ”behaviorally bulimic group" (BB). Subjects who reported that they met
thesebeiravioralcriteiainthepastbuthadnotmetthiscriteriain tirelastfourmonthswere
included in the ”behaviorally recoveed-bulimic group" (BR). Two of the DSM III-R
diagnostic criteria for bulimia nervosa are more subjective to assess: ”A feeling of lack of
control over eating behavior during the eating binges” (criterion B) and "Persistent
overconcern with body shape and weight” (criteion E). Scores from two subscales of the
Eating Disorde's Inventory, ”Drive for Thinness” and "Bulimia," were used to assess the
seveity of these dynamics. However, these scores were not used to group subjects.
Subjects who reported no history of bingeing or purging behaviors and who reported an
ideal weight of no more than ﬁve pounds below or above their presert weight were scored
in random order. The ﬁrst 25 females whose scale scores on both ”Drive for Thinness"
and 'Bulimia"rankedbelowthe50thpercertile(scoresoftwoorlessandzeo
respectively) and who agreed to return for fur-the participation comprised the non-eating
disordered group (NED). '

Memes
Thecompleteassessmentbaue'ﬁbothprescreeningandthemeasmesusedas

responses to the vignettes) is included as Appendix D.

 

1Vigorous exercise was included as a criterion only when the subject reported
that exercise was engaged in specifically to ”burn off or 'get rid of' large
quantities of food you ate (binges)."

44

E . 11' I 1 ED E

The EDI is a self-rating scale designed to assess the psychological characteristics
relevant to anorexia nevosa and bulimia nervosa. Scores on the subscales of the EDI have
been found to be predictive of clinician's ratings and diagnoses (Garner, Olmstead &
Polivy, 1983). The two scales used in this study wee ”Drive for Thinness" (DT), an
indicator of "concen with dieting, preoccupation with weight, and entrenchment in the
extreme pursuit of thinness" (Garner, et al., 1983), and ”Bulimia" (B), which ”indicates
the tendency toward episodes of mrcontrollable overeating (Bingeing) and may be followed
by the impulse to engage in self-induced vomiting" (Garner et al., 1983). The individual
items are listed in Table 2. Further validity and reliability data are available (Game et al.,
1983).

Subjects also completed a self-report questionnaire designed to assess the presence
and history of bulimic behaviors. This questionnaire consisted of some diagnostic items
from the EDI symptom checklist (EDI-SC) (Garner, 1990) as well as additional items
created by the author.

W

The ISS consists of 30 items which subjects rate on a ﬁve point scale. It is
designed to measure the level of inte'nalized shame.

”lntemalized shame, as it is deﬁned operationally by a high score' on the

ISS, essentially results from the frequent triggering of shame in

circumstances or situations that intensify or magnify the shame feelings,

with a corresponding diminishment of sustained experiences of interest or

enjoyment . . . The constellation of feelings triggered by shame are those

associated with incompetence, inferiority, defectiveness, unworthiness,

threats of exposure, emptiness, alienation, and self-contempt, among

others. . . The items are couched in language that reﬂects a high degree of

negative affect intensity, speciﬁcally associated with cognitions about the

self, reﬂective of the feelings noted above. Thus the shame items on the

ISS are a sample of the most internally consistent statements that tap into

this central sense of incompetence or infeiori that represents the core of
the shame experience." (Cook, 1993, p. 18-] ).

45

Table 2: Eating Disorders Invertory: Drive For Thinness and Bulimia Subscale Items

>199“?pr

$9999.“?

I eat sweets and carbohydrates witirout feeling nervous
I think about dieting.

I feel extremely guilty afte' overeating.

I am terriﬁed of gaining weight.

I exaggerate or magnify the importance of weight.

1 am preoccupied with the desire to be thinner.
Iflgainapound,lworrythatlwillkeepgaining.

Ieat when I am upset

1 stuff myself with food.

I have gone on eating binges where I felt that I could not stop.

I think about bingeing (overeating).

I eat moderately in front of others and stuff myself when they're gone.
I have the thought of trying to vomit in orde to lose weight.

I eat or drink in secrecy.

Items are scored on a six point Likert-type scale of "always,” ”usually," ”often,”
”sometimes,” ”rarely,” or ”never.” In scoring, responses are weighted from zero to three,
with tirree being the strongest or most symptomatic response. The three choices opposite
indirectiontothesymptomaticresponsearescoredaszero.

46

The ISS has two scales, self- esteem and internalized shame. Alpha reliability in a non—
clinical college sample was reported to be .94 for the shame scale and .88 for the self-
esteem scale. Test-retest reliability coefﬁcients at a sever week inte'val were .84 for the
shame scale and .69 for self esteem. The six self-esteem items balance the direction in
which items are scored to reduce the possibility of response set bias. Furthe' validation
data is available (Cook, 1993).

BI° l° SIEI [BSD

The Relationship Self Inventory (RSI) was designed to assess self orientation as
discussed by Gilligan and the Stone Certer group (among other), diffeentiating
individuals who deﬁne themselves in separation and those who deﬁne themselves in
connection (Pearson, et al., 1991). While tirese two self orientations certainly overlap and
coexist, the RSI is designed to assess the centrality of these different means of self
deﬁnitiontotheorganintionofthe self.

A self-report instrument, the RSI is made up of four scales, Connected Self (CS),
in which relations with others are most central to one's self deﬁnition; Separate Self (SS),
in which independence, separation, autonomy and justice are central for self deﬁnition; and
two scales assessing diffeerrt manifestations of CS, Primacy of Other Care (POC), in
which caring for the needs of others, frequertly at one's own expense, is a core self-theme;
and Self and Other Care (SOC), in which care of the self is integrated with care of othe's.
The authors report internal consistencies of .77 for Separate Self, .76 for Connected Self,
.68forPrimacyofOtherCare, and .78 forSelfandOtherCareinfemalepopulations
(Pearson, et al., 1991).

Thefourscaleshavedemonstratedextenal validityandappeartomeasure
meaningful and distinct constructs (Pearson, et al., 1991). The CS scale shows signiﬁcant
low to modeate positive correlations with sociability (r = .36), nurturance (r = .17), and
communion (r = .17) in female populations. The SS scale shows signiﬁcant low

47

correlations with autonomy (r = .24), and low negative correlations with nurtruance (r = -
.23), sociability (r = -.21), and communion (r = -.15). The constructs of Connected and
Sqrarate Self appear to be related to but distinct from nurturance, autonomy, agency,
communion, and sociability. Further information on the development and validation of
the RSI is available (Pearson, et al., 1991).
MM

Subjects selected for participation heard three audiotaped vignettes presented as
“portions of interviews with female college students.” In actuality, the “interviews” wee
written by the expeimenter and recorded by three graduate students. The three
mte'vieweeseachpresentadiffeertatﬁmdemwmdsmtepermnalmlaﬁmships Onetape
depicts a woman who is autonomous and self-reliant. The woman in the second tape
displays an interpersonal style characteized by mutual dependence on others and
interpersonal needs as conceptualized by Kaufman (1989), namely, the need for touching
and holding, the need for identiﬁcation, the need to be in relationsirip, and the need for
afﬁrmation. The third interviewee seves as a control; here expressions of autonomy are
‘balancedbyexpressionsofintepersomlneeds. Thethreewomendepictedontire
audiotapes all present themselves as happy with their lives. All report conﬁdence in their
academic life and satisfaction with their intepersonal relationships. All three womer report
that they have a boyfriend. Transcripts of the vignettes are included as Apperdix B.

Apﬂmstudywasconmreedmmdertodetermmewhctherdrctapesmhahlypretmt
the hypothesized values towards relationships and whether raters reliably assess the tapes
as differing along these hypothesized dimensions. Twenty-one undegraduate females
participating in the Psychology Research Pool served as subjects in the pilot study.
Subjectsﬁsmnedtoeachtapeandﬂrelcompletedaquesﬂernaireconsisﬁng ofeightitems
requiring a “true” or “false” response (see Appendix C). The eight itens (completed for
eachtape) assess thepresenceorabsenceofautonomyand inte'personal needs ,

48

speciﬁcally, the need for touching and holding, the need for identiﬁcation, the need to be in
relationship, and the need for afﬁrmation.

Individual items wee scored eithe' one or zero. A total score ofeight indicated the
deﬁnite preselce of the interpersonal needs for relationship, for idertiﬁcation, for touching
and holding, and for afﬁrmation. A cumulative score of zero indicated the absence of these
needs in that particular vignette. Statistics of the subjects’ ratings of the three tapes are
presented in Table 3. .

All subjects rated the tape depicting interpersonal needs with a score ofsix or above
(86% rated it with the most extreme score of eight). All rating scores of the tape depicting
the absence of intepersonal needs wee two or less (86% of the ratings wee either zero or
one). In addition, the majority of subjects (90.5%) rated all vignettes in the order
hypothesized—the tape depicting interpersonal needs received the highest score, followed
by the control tape, with the tape depicting the absence of intepersonal needs scoring
lowest overall. The two pilot study subjects who did not show this pattern scored the two
extremetapesinthedesiredduectionbutgavetheconu‘oltapeanoveall scoreiderticalto
the tape depicting no interpesonal needs (in neitirer case was this score zero). Finally, a
repeated measures analysis ofvariarrce revealed a signiﬁcant difference in the ratings
between tapes (F = 216.81, p<.001).

Emeline
In the ﬁrst phase of the experiment, 680 female undergraduates completed the
Internalized Shame Scale (ISS). This group also completed the Relationship Self Invertory
(RSI) two times, from two diffeent pespectives; ﬁrst, as they perceive themselves, and
second, as they would respond if they were their ”ideal” self. Denographic information
was collected from subjects including age, marital status, parents' marital status, estimated
family income, religion, ethnicity, history of psychotherapy or treatment for mting

49

Table 3: Pilot Study: Subject Means

Tape 1 Tape 2 Tape 3
(Self-Reliant) (Expresses Interpersonal Needs) (Control)
N 21 21 21
Man 0.667 7.810 4.000
Minimum 0.000 6.000 1.000
Maximum 2.000 8.000 7.000
Standard 0.730 0.5 12 1.612

50

disordes, and the use of psychotropic medications. Subjects additionally completed a
questionaire which uses DSM III-R criteria to diagnose bulimia nevosa including portions
of the Eating Disorde's Invertory Symptom Checklist (EDI-SC). All subjects wee asked
about past eating disorders and past eating behavior and information on weight and eating
style was collected. Finally, all subjects completed two scales of the Eating Disorde's
Inventory (EDI), ”Drive for Thinness" (or) and ”Bulimia” (B) 1.

Subjects selected forparticipationinthesecondphaseoftire study (basedon their
classiﬁcation as behaviorally bulimic, non-eating disordeed, or behaviorally recoveed-
bulirnic) wee contacted and offered additional course credit to return. They wee informed
tint additional participation was voluntary. Returning subjects completed the second phase
of the experiment in an individual setting to protect conﬁdentiality. The expeimert was
administered by undegraduate research assistants. Subjects wee seated at a desk and
provided with headphones and a tape recorder. They were giver the following
instructions:

”Todaylamgoingtoaskyou tolisten to several veryshorttapes. The

tapes contain portions of inte'views with female college students. Please

listen carefully toeach tape. Trytoform an ideaofwhatyou tirinktlre

woman you are listeling to is like. After listening to a tape I will ask you to

respond to a short questionnaire in the way you think the woman you heard

would respond. The answers to the questions are not necessarily in the

tapes. What we are interested in are your opinions about the woman, the

impressions that are formed ﬁom the short tape you hear of he.”

After answering any questions, subjects weeprovided with one of the audiotaped
interviews. The tapes were presented in random order.

After listening to each tape, the subject was asked to complete a copy of the
Intemlized Shame Scale (ISS). The instructions to the scale wee alteed slightly. Instead

of responding to the scale in tems of themselves, subjects wee asked to assess how well

 

1Materials pertaining to the diagnosis of eating disorders were provided at the
end of the screening battery in order to avoid the effects of any secondary
shame (or other response set) on the other measures.

51

each staternentcharacterrzed' theindividualonthetape. Afterﬁnishingtirescale,the
procedrue was repeated with the remaining tapes.

Upon completion of the experiment, participants were debriefed. They wee
informed that the initial screening was used to identify women with a range of attitudes
about their bodies and with diffeent eating habits. Subjects were provided with referral
information related to any concerns they might have about their own eating or body issues.

E l . l I
E . 59 ﬁ 1 . 1'

Several steps were taker to ensure subjects' conﬁdertiality and anonymity.
Subjects wee assigned a code number by the primary experimenter and the subjects'
responses were identiﬁed only by that code. Subjects' names were stored in a secure place
separate from the code-identifed data and accessible only to the primary expeimerter. This
information was destroyed afte data was collected.

Subjects who wee selected for participation in the second phase of the experimert
were contacted by the primary experimenter only. No other persons had access to names
or phone numbers of the subjects. Research assistants who participated in the second
phase of data collection were not informed ofthe criteria used to select returning subjects.
Additionally, research assistants did not have access to the subject's previous responses.
The subject's responses in the second phase of the expeimert wee idertiﬁed only by code
number.

Whimsy

Students were asked to report on their eating behaviors, on possible eating
disorders, and on their history of psychological and psychiatric treatment. Procedures
describedtoprotecttheconﬁdentialityandarronymity oftheparticipanthelptominimizethe
discomfortthis may havecaused. Theexperimentwasrurr by undergraduateresearch
assistants giver speciﬁc training in the importance of conﬁdentiality. These research

52

assistants wee not informed of the criteia for subject selection. The subjects themselves
wee informed in the consert agreenert of their right to discontinue participation at any
time and subjects selected for continued participation wee informed (verbally and in
writing) that their continued participation was completely voluntary.

Bereﬁts to the individual participants in this study were limited to the course credit
they received, the experierce they gained ﬁom participating, and any personal insights they
may have gained from completing the experimert. It is possible, that answeing such
detailed questions about one's eating pattens might heighten an individual's awareness
about eating problems or disordes. At no time wee subjects' eating behaviors labeled for
them in any way. However, a handout listing resources for individuals who feel
dissatisﬁed with their eating behaviors was provided to participants in the ﬁnal phase of the
expeinrent. Subjects were also informed in the consent agreement that the expeimerter or
he advisor was available to discuss any concerns related to the experimental procedure or
content. No such contacts were made.

The potential beneﬁts of this study are primarily to the ﬁeld of psychology and to
society. The purpose of the study was to elucidate the perceptions, values, and self
orientation of bulimics and to gain information on changes in these areas that may or may
not occur with recovery from bulimia nevosa. An improved understanding of differences
between bulimics, non-bulimics, and recovered bulimics, as well as an understanding of
any unrealistic or inaccurate perceptions and ideals bulinrics may have about interpesonal
relationships will aid in improving the focus and effectiveress of the treatmert of bulimia
nervosa.

W

Consent was obtained from subjects in both phases of the experiment; the
screening stage (Phase I), and the formal study itself (Phase 11). Subjects were given
corrsert forms after being preserted with a brief oral introduction to the research. After

53

their participation in each portion of the study they were given an informational sheet
outlining the purpose of that phase of the expeiment. This form included the names of

individuals they might contact for further information. Consent forms and informational
sheets are included as Apperdix E.

RESULTS

W

Six hundred eighty undergraduate females were screered for participation in the
study. Subjects ranged in age from 16—24 years ( mean = 18.64, SD = 1.07). The sample
was 84.1 percent Caucasian (n = 572), 7.9 pecent Aﬁican Ameican (n = 54), 1.8 percent
Latin American (n = 12), 0.4 percent Native Ameican (n = 3), and 4.1 pecent Asian (n =
28). Elever individuals (1.6 pecent) identiﬁed themselves as ”other." Six hundred
seventy-four (99.0 percent) of the individuals screened wee unmarried (one was divorced,
one was widowed). Five subjects (0.7%) were currently married (one subject did not
respond to the item on marital status).

Means and standard deviations of EDI subscale scores for the pre-screened sample
are reported in Table 4. Quetelet's Body Mass Index (BMI), a method of standardizing
body weight across heights, was calculated for all subjects' reported real and ideal weights
as kglm(m) (BMI and IDEAL BMI). The diffeence between real and ideal BMIs was also
calculated for each subject (REAL-IDEAL BMI). Summaries of this data are also preserted
in Table 4.

Two hundred-eight of the individuals screened (30.6 percent) report that they have
been in some type of psychotherapy or counseling (mean number of sessions = 28.15, SD
= 56.62) and 3.7 percent ( n = 25) report having received some type of psychotherapy or
counseling for an eating disorde. Further descriptive information on the prescreened
sample is presented in Apperdix F.

From the original subject pool, 23 subjects (3.38 percent) met the criteria for
inclusion in the behaviorally bulimic group (BB); 30 subjects (4.41 percent) met the criteria

54

55

Table4: EDI scalescoresandreportedBMI scoresforthepescreened sample

Variable

BULIMIA (EDI)
DPT (EDI)

BMI

IDEAL BMI
REAL-IDEAL BMI

Mean

2.16
7.75
21.82
19.98
1.86

SD

3.42
7.10
3.19
1.83
2.20

Minimum

0.00
0.00
16.18
15.08
-4.61

Maximum

21.00
21.00
45.48
31.00
26.26

N

679
679
677
667
667

56

for inclusion in the behaviorally recovered bulimic group (BR). The ﬁrst 25 subjects who
met the criteria for inclusion in the non-eating disordered group (NED) were also selected
for participation in the experiment. Subject ages ranged from 17-20 years (mean = 18.48)
in the BB group, 18-22 years in the BR group (mean = 18.67), and 17-22 years in the
NED group (mean = 18.56). A one-way analysis of variance (alpha = .05) showed no
signiﬁcant age difference between groups (F = .231 (2, 75), p = 0.794). Group means
and standard deviations for reported BMI, Real-Ideal BMI, Bulimia and Drive for Thinness
(DFI') scale scores, and bingeing and purging behaviors of the BB and BR groups are
reported in Table 5.

A one-way analysis of variance revealed a signiﬁcant difference between groups in
BMI (F (2, 74) = 10.046, p < .001), Real - Ideal BMI (F (2, 74) = 10.932, p < .001), but
not reported Ideal BMI (F (2, 74) = 2.30, p = .107). A Scheffe post hoc comparison of
BMI scores across groups (alplm = .05) indicated that the NED group's mean BMI was
signiﬁcantly lower than that of the BB group.

0f the subjects selected to return for participation in the expaiment (N = 78), 38
(48.7 percent) report that they have been in some type of psychotherapy or counseling;
60.9 percent of subjects in the BB group (n = 14), 53.3 percent of subjects in the BR
group (n =16), and 32 percent of subjects in the NED group (n = 8). A Chi square test of
signiﬁcance indicated no signiﬁcant difference between groups on this variable (Chi square
= 2.73, df = 2, p = 0.25). However, there was a much wider range of within group
variance among the BB and BR groups in the number of psychotherapy sessions reported.
These ﬁgures as well as the reasons group members reported for seeking psychotherapy
‘ae summarized in Table 6. Two BB subjects reported that they were currently taking
antidepressant medication. One member of the BR group also reported current use of an
antidepressant and one BR subject was taking Ritalin. None of the NED subjects reported
psychotropic medication use.

57

Table 5: BMI and EDI descriptive statistics and bingeing and purging frequencies for
groups

Behaviorally

 

Bulimic (BB)
Mean 21.65 2.5 9.26 18.13
' ' 18.11 0.3 0.00 4.00
Maximum 29.80 7.6 20.00 21.00
22.00 2.00 23.00 23.00
SD 2.58 2.03 5.81 3.63
Behaviorally
Recovered (BR)
Mean 22.58 2.78 4.70 11.98
Minimum 16.69 0.51 0.00 0.00
Maximum 37.97 17.33 21.00 21.00
11 30.00 30.00 30.00 30.00
SD 4.06 3.05 5.54 6.55
Non-Eating
Disordered (NED)
19.05 0.15 0.00 0.20
Minimum 16.18 -0.94 0.00 0.00
Maximum 21.29 0.94 0.00 2.00
n 25 .00 25.00 25 .00 25.00
SD 1.17 0.51 0.00 0.50
Group Average Average Avaage Avaage Average Average

Binges Diuretic use Diet Pill use laxative use Vomiting Exercise
(“WW

 

Behaviorally

Bulimic (BB)

Mean 13.57 0.35 4.87 3.04 9.96 1557.17
Minimum 8.00 0.00 0.00 0.00 0.00 0.00
Maximum 28.00 8.00 35.00 25 .00 60.00 10800.00
n 23.00 23.00 23.00 23.00 23.00 23.00
SD 6.12 1.67 1 1.10 7.59 15.44 2393.66
Behaviorally

Recovered (BR)

Mean 1.35 0.00 3.97 0.00 0.80 315.08
Minimum 0.00 0.00 0.00 0.00 0.00 0.00
Maximum 4.50 0.00 60.00 0.00 10.00 3600.00
n 30.00 30.00 29.00 30.00 30.00 30.00
SD 1 .57 0.00 12.62 0.00 l .97 691.47

58

Table 5 (cont'd)

* Exercise is measured in minutes and reﬂects minutes per month exercised to burn off or
"get rid of“ large quantities of food eaten or a "binge."

59

Table 6: Psychotherapy sessions reported by groups and reasons for seeln'ng
psychotherapy

 

Wm: SD Mann—n:
BB 70.85 120.79 12.00 13
BR 45.29 39.63 38.75 12
NED 5.70 3.03 5.00 5

I"Notall subjects whoreportedthat they had been in therapyreported the numberof
sessions

Reasons for seeking therapy Group
(chose as many as applied)
BB BR NED Total

Problems with

AlQhOl n = 1 n = 0 n = 1 ﬂ = 2
4.3% 0.0% 4.0%

may 11 = 4 n = 3 n = 2 n = 9
1 .4% 1 .0% 8.0%

Depression n = 7 n = 9 n = 1 n = 17
30.4% 30.0% 4 0%

Drugs 11 = 1 n = 1 n = 0 n = 2
4.3% 3 3% 0.0%

FatingDisorda n=7 n=8 n=0 n=15
30 4% 26.7% 0 0%

FarnilyProblems n=8 n=9 n=5 n=22
34.8% 30.0% 2 0%

Otha' n = 2 n = 3 n = l n = 6
8.7% 1 .0% 4.0%

SleepProblems n=3 n=0 n=O n=3
13.0% 0.0% 0.0%

60

Furtha descriptive information is included as Appendix G. Summaries of subject
responses relevent to placement in the BB or BR group are provided in Appendix H.

Of the 23 BB subjects identiﬁed, 22 returned and completed the experiment. Of the
30 BR subjects identiﬁed, 26 returned and participated in the vignette phase of the
experiment. Twenty-ﬁve NED subjects completed the experiment.

_I§§

Hypotheses related to subjects' score on the ISS were tested using a one-way
analysis of variance. Means and standard deviations of each group's ISS scores are
presented in Table 7

The difference between groups was signiﬁcant (F (2, 75) = 12.11, p < .0001).
Planned comparisons found that the BB group reported signiﬁcantly higher ISS scores than
the averaged responses of the BR and NED groups (T (30.4) = 3.116, p < .004).
Hypothesis one, that the BB group would report signiﬁcantly higher ISS scores than
either the BR or NED groups was supported

Hypothesis two, that the difference between the BR and NED groups' ISS
scores would not be signiﬁcant was tested using a planned comparison. The difference
between scores was found to be signiﬁcant (T (47.5) = 4.273, p < .001) and the
hypothesis was not supported.

RSI
Means and standard deviations of RSI real and ideal scale score responses for each

grouparepresentedinTable 8. Group means foreach scalewere standardizedusing Z
transformations in order to correct for unequal scale lengths (which resulted in unequal
maximum scores for each scale) and allow for comparisons across scales. Transformed

mean scores are presented in Table 8 and graphically in ﬁgures 2—5.

Table 7: ISS Scores

BB Group
BR Group
NED Group

Mean
52.39
43.72
23.13

61

Standard Deviation
26.80
22.34
12.81

Table 8: Means, standard deviations, and standard scores for RSI real and ideal scale

scoresby group

BB Group

W
Real

Ideal
W

Real

Ideal
W
W

Real

Ideal

W

@1631?)
Real

Ideal
BR Group

WES.)

Real

Ideal
W

Real

Ideal
W
W

Real

Ideal
Selfandﬂther

mm
Real

Ideal
NED Group

WW
Real

Ideal
W

Real

Ideal
W
W

Real

Ideal

Mean

51.91
53.39

49.82
50.93

46.88
43.65

64.95
70.20

50.60
54.92

47.68
49.87

45.69
42.90

66.07
73.57

51.80
54.68

44.81
41.48

44.10
45.48

62

Standard Deviation Mean Z Score

6.33
6.56

11.24
11.34

9.55
11.01

7.13

5.48
4.86

8.73
11.98

-0.22
-0.02

0.12
0.08

-0.35
-0.53

-0. 15
0.06

63

Table 8 (cont'd)
NPD Group (cont'd):
Mean Standard Deviation Mean Z Score
mm:
mm
Real 66 48 5.77 0.02

Ideal 70212 6.48 -0.03

64

 

 

 

 

 

 

 

 

 

 

 

 

 

BB étoup BRGroup

I: Zsoore(CSIDEAL)
- Zecore(CSREAL)

Figure 2: Connected Self By Group (Standardized Scores)

 

65

 

 

 

 

 

 

 

 

 

BB Group BRGroup

[:3 Zscore(SSlDEAL)
- Zscorc(SSRL)

Figure 3: Separate Self By Group (Standardized Scores)

 

 

NEDGroup

 

66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BB Group BR Group

C3 Zscore(SOCIDEAL)
- Zecore(SOCREAL)

Figure 4: Self and Other Care By Group (Standardized Scores)

 

67

 

 

 

 

 

 

 

 

 

 

 

 

 

BBGroup BRGroup NEDGroup

E3 Zscore(POCIDEAL)
- Zscore(POCREAL)

Figure 5: Primacy of Other Care By Group (Standardized Scores)

 

68

A repeated-measures analysis of variance was conducted on the original scores with
group as thebetween factorvariable and RSI scale andreallideal conditionaswithin
subject factors. The diffeence between scale scores was signiﬁcant (F (4, 71) = 6097.48,
p < .ml) (as would be expected). The ANOVA revealed no signiﬁcant diffeence between
groups on RSI scale scores (F (8, 144) = 1.68, p < .107). However, the interaction
between group and real or ideal condition was signiﬁcant (F (8, 144) = 2.26, p < .026).
Finally, there was a signiﬁcant difference between the real and ideal response conditions
across groups (F (4, 71) = 24.07, p < .001).

Becausethedifferencebetweelrealandidealconditionswas foundtobe
signiﬁcant, the data was broken down into these two conditions for further analysis.
Multivariate analyses of variance were conducted betweel groups for both real and ideal
conditions separately. These analyses are summarized in Table 9.

Under the ”ideal" condition, a signiﬁcant difference was found betweel groups in
the Separate Self scores only (F (2, 75) = 5.79, p < .005). In order to We examine
group differences on the Separate Self ”ideal“ condition variable, a one-way analysis of
variancewasconductedwiththisparticular scaleandconditiOn asthedependentvariable.
Planned contrasts indicated that the aveaged responses of the BB and BR groups were
signiﬁcantly highe than those of the NED group (F (75) = 3.40, p < .001) but that there
was no signiﬁcant diffeence between "ideal” Separate Selfresponses of the BB and BR
groups (T (75) = 0.36, p < .722). These results offer only partial support for hypothesis
three: that the BB group would score signiﬁcantly higher on the Separate Self "ideal”
scalethaneithertheBRorNEDgroups (astherewasnodifferencebetweentheBB and
3 BR groups). Hypothes’s four, that there would be no signiﬁcant difference between
BRandNEDgroupsontheirscoresonthe "ideal” RSIwasnot supported.

”Real” RSI responses wee also analyzed using a multivariate analysis of variance.
Hypothesis five predicted that thee would be no signiﬁcant differences between BR and

69

Table 9: RSI—"Real” and ”Ideal" scale scores analyzed by group

Real RSI:
Variable

CSReal
SSReal
SOCReal
POCReal

Ideal RSI:
Variable
CS Ideal
SS Ideal

SOC Ideal
POC Ideal

FrrorMS

28.93268
82.65758
36.44677
57.70067

ErrorMS

27.94613
1 16.17145
36.38729
87.28210

CS = Connected Self
SS = Separate Self
SOC = Self-Othe' Care

POC=PrirnacyofOtherCare

.49771
1.40399
.39629
.56396

.59693
5.79053
2.94800

.53641

Signiﬁcance of F

.610
.252
.674
.571

Signiﬁcance of F

.553
.005
.059
.587

70

NED scores on the ”real” RSI. No signiﬁcant group diffeences wee found in any of the
RSI scales completed under the ”real” condition supporting the hypothesis. However, the
nonsigniﬁcant group effect does not support the component of hypothesis six: that the
BB group's ”real” RSI responses would differ signiﬁcantly from the other two groups.
Power analyses revealed small effect sizes for all four "real” scale score variables (ranging
from 0011-0037). The combination of small effect size and the size of the groups
examined led to a less powerful test than would ideally be desired (power ranged from
0.291 on the Separate Self ”real" variable to 0.11 on the SOC ”real” variable). Once again,
scale scores were standardized using Z transformations in order to correct for unequal scale
lengths and allow for direct comparisons across scales. Figure 6 depicts the three groups'
standardized RSI scale scores under the ”real” response set. As predicted in hypothesis
six, the BB group scored most highly on the Primacy of Others and the Separate Self
scales.

Intercorrelations of RSI ”real" scale scores for each group are presented in table 10.
Table 11 depicts intercor'relations of real RSI scale scores for the prescreening sample.
Hypothesis six additionally predicted that the normative negative correlation (-.2 3)
betweel theSeparate SelfandConnectedSelfscoreswouldnotbeobtainedinthe BB
group due to the group's hypothesized conﬂicts related to dependency needs and self-
reliance. This hypothesis was not supported. Both the BB and BR groups' Separate Self
(real) and Connected Self (real) scores wee signiﬁcantly negatively correlated (-.560 and -
.576 respectively).

Fisher's Z transformations were performed on RSI "real'scale score correlations
foreachgroupand fortheprescreening samplein ordertotest for diffeelcesbetween
groups and between groups and the prescreening sample. The correlation between the
Connected Self and Separate Self scales was signiﬁcantly diffeent in the BB and NED
groups (p < .047) and in the BR and NED groups (p < .024). Correlations between the

71

 

 

 

 

 

 

 

 

 

.5:

A:

 

Real Responses By Group (Standardized Scores)

RSI:

Figure 6

72

Table 10: Intercorrelations of “Real" RSI Scale Scores By Group

GROUP: BB group
- - Correlation Coefficients - -
CSREAL POCREAL SOCRBAL SSRL
CSREAL .5577 -.1705 -.5603
( 22) ( 22) ( 22)
P- .007 P- .448 P- .007
POCREAL -.6877 —.2970
( 22) ( 23)
P- .000 P- .169
SOCRBAL .4890
( 22)
P- .021
SSRL

(Coefficient / (Cases) / 2-tailed Significance)

" . ” is printed if a coefficient cannot be computed

GROUP: BR group
- - Correlation Coefficients - -

CSREAL POCREAL SOCREAL SSRL
CSREAL .5431 .0423 -.5758
( 30) ( 30) ( 30)
P- .002 P- .824 P- .001
POCREAL —.3296 -.3253
( 30) ( 30)
P= .075 P- .079
SOCREAL .2110
( 30)
P= .263

SSRL

(Coefficient / (Cases) / 2-tailed Significance)

” . " is printed if a coefficient cannot be computed

Table 10 (cont'd)

- - Correlation Coefficients

GROUP: NED group
CSREAL POCREAL
CSREAL .3594
( 25)
P8 .078
POCREAL
SOCREAL
SSRL

SOCREAL

(
PI:

(
Pa

.2009
25)
.336

.2340
25)
.260

SSRL

-.0144
( 25)
P= .946

-.O300
( 25)
P- .887

.5161
( 25)
P- .008

(Coefficient / (Cases) / 2-tailed Significance)

" . ” is printed if a coefficient cannot be computed

74

Table 11: RSI “Real” Scale Score Intecorrelations for the Prescreening Sample

- - Correlation Coefficients - -

CSREAL POCREAL SOCREAL SSRL
CSREAL .3969 .2499 -.2270
( 679) ( 679) ( 679)
P- .000 P- .000 P= .000
POCREAL -.1350 -.1373
( 679) ( 680)
P- .000 Pt .000
SOCREAL .3523
( 679)
P- .000
SSRL

(Coefficient / (Cases) / 2-tailed Significance)

' . ” is printed if a coefficient cannot be computed

75

Primacy of Others Care and Self and Other Care scales varied signiﬁcantly as well. The
negative correlation between these scales in the BB group was signiﬁcantly grants in
magnimdethanthenegativeeonelation betweenthesesealesin theprescreening sample (p
< .002) or in the NED group (p < .05). No other signiﬁcant differences were detected.
However, an examination of the correlations between Connected Self and Self and Other
Care scales showed a notable trend. The correlation between the two scales was negative
in the BB group (r = -.1705), was almost nonexistent in the BR group (r = .0423) and was
positive in the NED and prescreening groups (r = .2009 and r = .2499 respectively). The
differenceinthemagrﬁmdeoftheoorrelaﬁonbetwemtheprescreening sampleandtheBB
group barely escaped signiﬁcance (p < .064).

Finally, correlations between RSI real and ideal scale scores and subjects' ISS
scores were examined by group. These correlations are presented in Table 12.

Fisher‘s Z transformations were performed on the correlations between RSI scale scores
and ISS scores in order to test for differences between groups. The correlation between
real PrimacyofOtha'CareandISS scorewas signiﬁcantly differentin the BBandNED
groups (p < .002). The difference between BB and NED groups in the magnitude ofthe
correlation between ideal Connected Self and ISS score narrowly missed statistical
signiﬁcance (p < .055).

Wilmette:

Figure 7 displays the mean ratings of shame across the three vignettes. Ratings of
shameby groupacrossvignettesaredepictedinFigure 8. All threegroupsattributeda
higherlevelofshametothewoman whoertpresseddependencyneedsthantothe
individual who expressed self reliance. Means and standard deviations both across and
within groups for each vignette are presented in Table 13.

76

Table 12: Correlations of RSI Real and Ideal Seale Scores with ISS Scores (By Group)

BB BR NED
Connected Self (real) .2785 -.01 81 -.2572
Connected Self (ideal) .1314 -.0531 -.4283*
Separate Self (real) .0155 .1732 .2262
Separate Self (ideal) .1301 .2280 .3384
Primacy of Otha' Care (real) .6270“ .3094 -. 1817
Primacy of Other Care (ideal) .2 850 .0731 -.2271
SelfandOtherCare(rml) -.3058 .1918 .1106
Self and Other Care (ideal) .0818 .3460 .2099

*Indieates signiﬁeance at alpha = .05

77

 

 

 

 

 

 

 

Mean

10

 

 

 

 

 

 

autism mm

Figure 7: Shame Ratings Across Vignettes

78

 

 

.aa'p-p
.BRM
.Dmm

 

 

 

104

 

Figure 8: Ratings of Shame Across Vignettes By Group

79

Table 13: Ratings of Vignettes
AmssAllﬁmms:
Self reliant individual

Individual who expessed dependency needs
Control

W

BB :
Self reliant individual
Individual who expressed dependency needs
Control
BR Group:
Self reliant individual
Individual who expressed dependency needs
Control
NED Group:
Self reliant individual
Iggvidual who expressed dependency needs
trol

Mean

17.88
35.97
20.90

20.61
45.09
20.66

17.04
36.38
23.16

16.36
27.52
18.76

SD

15.31
18.63
15.72

SD

20.02
20.22
15.77

8.40
17.72
16.59

16.41
14.32
15.06

80

In order to investigate hypothesis seven, that the BB group would attribute
signiﬁcantly higher levels of shame to the woman who expressed interpersonal needs than
tothewoman expressing selfreliancearepeated measureanalysisofvariancewasrun for
the BB group separately. The difference in attributions of shame across vignettes for the
BB group was signiﬁcant (F(2, 42) = 13.60, p < .0001). A planned comparison revealed
a signiﬁcant difference between the BB group's rating of the vignettes depicting
interpersonal needs and selfreliance (F (1, 21) = 16.07, p < .001) supporting the
hypothesis.

Thedatawerealsoanalyzedtoexaminedifferenwsin the ISS scoresacross
vignettesandbetweengroups. Arepeatedmeasmesanalysisofvarianceindieatedthatthe
group effect fell just short of the .05 level of signiﬁeance ( F (2, 70) = 2.98, p < .057).
The difference in the level of shame scores across vignettes was signiﬁcant (F (2, 140) =
33.79, p < .0001) The interaction between group and shame ratings was not signiﬁcant (F.
(4, 140) = 2.24, p = .068).

In order to investigate hypothesis eight, that the BB group differed signiﬁcantly
from both the BR and NED groups in their attribution of shame to the wonmn who
expressed dependency needs, a one-way analysis of variance was conducted comparing the
three groups' responses to that particular vignette. The ANOVA indieated a signiﬁcant
difference between groups in their ISS ratings for the dependency needs vignette (F (2, 70)
= 5.93, p = .0042). A planned comparison revealed that the BB group's attribution of
slnme to this vignette was signiﬁcantly higher than the averaged responses ofthe BR and
NED groups (T = 2.949, p < .004), as hypothesized.

The ﬁnal hypothesis, hypothesis nine, involved the relationship between level of
bingeing and purging and the magnitude of the ISS score assigned to the woman who
expressed dependency needs. It was hypothesized that the level of shame (ISS score)

attributed to the vignette depicting dependency needs across groups would be positively

81

correlated witln the individuals' levels of bingeing and purging. Because the NED group
was speciﬁcally based on the absence of bingeing and purging, the data from this group
was excluded from examimtion. The BB and BR groups wee combined and the
correlations between the dependency needs ISS, current binge frequency, and current
purging frequencies (vomiting, laxative use for weight loss, diet pill use, diuretic use, and
exe'cise to work off a binge) as well as "worst ever" levels of bingeing arnd purging were
examined Anove'all purgingindexwasalsocreatedforeach individualconsistingoftlne
sum of all current purging activities reported other tlnn exercise (the sum of vomiting
episodes, laxatives taken, diet pills taken, arnd diuretics taken). None of the bingeing or
purging variables wee signiﬁcantly correlated with the dependency needs ISS score.
Hypothesisninewasnotsupported. ThedataobtainedfromtheBBandBRgroupswas
also examined for signiﬁcant correlations between the self-reliance vignette ISS score and
the bingeing and purging variables. Again, there wee no signiﬁcant correlations. The BB
and BR groups' data was examined individually as well. Within the BB group, the ISS
score attributed to the dependerncy needs vignette was signiﬁcantly negatively correlated
with magnitude of current diet pill usage (r = -.4518, p = .035), however, only ﬁve
individualsin thegroupreportedanycurrentuseofdietpills. Thecorrelation between the
interpe'sonal needs ISS variable and ”worst eve'" diet pill use was also highly signiﬁcant (r
= -.8156, p < .004). Within the BR group, although there was no apparent correlation
between reported current levels of bingeing and purging and the dependency needs ISS
score, thee wee signiﬁeantrelationships between the dependency needs ISS score and the
reported ”worst ever” level of bingeing (r = .4294, p = .029) as well as between the
dependency needs ISS score and the reported ”worst ever" level of vomiting (r = .3992, p
= .053), and "worst ever" use of laxatives to purge (r = .7328, p < .001). BB and BR
correlations of these eating disorde' diagrnostic variables with the dependency needs ISS
score are presented for BB and BR groups in Table 14.

82

Table 14: Correlations of Eating Disorder Diagnostic Variables with ISS Score attributed to
the Interpersonal Needs Vignette: BB and BR Groups

BB BR

Bulirrnia (EDI) -.4084 .1084
Drive For Thinness (EDI) -.0267 -.3287
Current binge frequency -. 1865 —.02 10
Oment level of overall purging -.35 30 .0057
Current vomiting frequency -. 1612 .2330
Current laxative use frequency -.0337 .
Current dinnetic frequency .2642 .
Current diet pill frequency -.4518* -.0385
Current amount of exercise

to work off a binge -. 1786 .0227
"Worst eve'” binge frequency -.35 80 .4294“
'Worst ever vomiting frequency —. 1790 .3992"
I'Worst ever" diet pill frequency -.8156* .4459
"Worst eve" laxative use frequency .1088 .7328"

* Indicates statistical signiﬁcance (alpha = .05)

DISCUSSION

 

Thee is no evidence that this sample of behaviorally bulinnic (BB) and behaviorally
recoveed bulimic (BR) subjects diffeed in prevalence or on demographic variables from
similarly aged samples of women who report that they meet (or have met) the diagnostic
criteria forbulimiarnervosa. Oftheoriginal subjectpool, 3.38 pecentmetthecriteria for
inclusion in the behaviorally bulimic group (BB), cornsistent with available prevalence data
for this age group (Neuman & Mitchell, 1986; Rand & Kuldau, 1992). No known
prevalence data is available on the pecentage of college-age worrnen who report previous
bulimia nervosa. Although the BR subjects no longe met diagrnostic criteria for bulimia
nervosa, subjects in this group did report some bulimic tendencies and behaviors. Ideally,
this study would have irncluded ornly ”pefectly recovered” individuals for inclusiorn in the
BR group, that is, subjects who reported no evidence of eating disordered cognitions or
behaviors. Howeve, given the low prevalence estimates for bulimia ne'vosa, the relatively
low pecentage of individuals who report recovey (Keller et al., 1992), the young age of
the sample popnrlatiorn, and the limits of data collectiorn capabilities, more lenient criteia
wee employed. This may have limited the degree of differences between the BB and BR
groups (anddegreeofsinﬁlmitiesbetweendneBRandnon—eaﬁngdisordeedeD)
groups).

The ﬁnding that the non-eating disordeed (NED group) reported signiﬁcantly
lowe body mass indexes (BMIs) than the BB group is notable. The NED group's mean
BMI (19.05) fell below the 15th pecentile for women ( at age eighteen) while botln the BB
and BR group's reported BMIs fell solidly within the normal range. This ﬁnding is
cornsistent with previous research suggesting that college women who report satisfaction

83

84

with theirbodyweightandshapeandwhodonotdietarethinnethanaveage (Mortenson,
I-loer, & Game, 1993), altlnough it shouldbenoted that BMIs weebasedon subjects'
selfreport and weenotobjective measures. It isapossibility thattheNEDsubgroup has
beenlessvuheabbmthedevebpmentofsymptonmmlatedmweightbodysizeandw
eating disorders because they arephysiologically predisposed to be thin.

 

Subjects' scores on the Intenalized Shame Scale (ISS) reveal important diffeences
betweengroups. AshypoﬂnesizedtheBBgroupreportedsigniﬁcantly highe levels of
intenalizedshamethantheBRorNEDgroups. ThemeanscorefortheBBgroup(52.39)
faﬂsabwethe85thpeoenﬁlefeﬂnefemalenm-chnicalmrmaﬁvesamplempemdby
Cook (1993). The as group's ISS score also falls above the cutoff point (a score of 50)
identiﬁed by Cook as indicative of ”painful, possibly problematical levels of internalized
shame” Thisﬁndingisnotsnnprising. Itprovidesempirlcalsupportforcommonclinical
observations and replicates previous research (McCreey, 1991).

ThehypothesisﬂnattheewonddbenosigrﬁﬁcantdiffeencebemtheBRand
NEDgroups'ISSscoreswasnotsupported. TheBRgrouphadsigniﬁcantlyhighe
shamescoresthantheNEDgroup. IfISSscoresarerelatedinanywaytofeelingsof
shameaboutcatingdisordeedbehavior,orifeating disordeedbehaviorandcogrnitionsane
the result of underlying shame, this ﬁnding would be somewhatexpected, given that the
BRgroupcontinuestorqnortsomebulimichenciesandbehaviors. 'Ihetrendof
decliningISSscoresfrom BBtoBRtoNEDgroups supportsatheoryofarelationship
between tlne intensity of bulimic pathology arnd feelings of internalized shame. However, it
does not explain the cause of the shame. Is the shame solely caused by the individual's
feelingsaboutheeatingdisordeedbehaviorsﬁe.shamethatshebingesorpurges)orisit
related to feelings, ideas orpeceptions which may nmdelie, pepetuate, or inﬂuence
bulimic behaviors? Responses to the Relationship Self Inventory (RSI) and to the

85

audiotapedvignettessupportthetheorythattheroleofshameinbulimianervosaappearsto
benneemanasecondeyracﬁmmbuﬁmicsympmmsandappeeswbemteminedwith
peceptiorns of dependency needs and self-reliance.

 

Ornepnrrposeofthestudywastoexplorewhetlnerornotbulimicsdeﬁnetlnemselves
and their ideals about relatedrness differently tlnarn recoveed bulirrnics or non-eating
disordeed individuals. Responsestothe Relationship Self Inventory (RSI),bothreal and
ideaLshowedfewdiffeenwsbetweengroups. Howeve,theoveallpatternofscores
supportthehypothesis tlnatbulirrnics senseof tlnenselves--both inrelation toandapartfrom
others-is more conflrcted' than non-eating drsordered' or recoveed bulimic women .

SIE . .

Although no signiﬁcant diffeences wee found between groups on ”real" RSI scale
scores, an examination of each groups' pattens of scores reveal important diffeences in
therelationslnipbetweenscales. Itwashypothesizedthatbulinnicshavebothshameand
cmfhcnmlfedingsrelamdmdependencyneedswhichrendtmanenmhasismself-
relianceandisolation. RealRSIscoresweeexpectedtoreﬂectconﬂictsinthisgroup's
expressionsofcornrnectiornandseparation. Thediffeencesinmagnitudeweenot
statistically signiﬁcant; however, as hypothesized, the BB group's highest scores wee on
theSeparateSelfandthePrimacyofOthesscales. 'lIneSeparateSelfsmleenrphasizes
separation, independence and autonomy as a means of self deﬁnition. The core tlneme of
dnePrimcyofOthesscaleisceeofotheeﬁequentlyatone'sownexpense. Careof
otheshaspriorityovercareofselfinthetypeofconnectedness'whichsubjectserndorseon
-thisscale. ThePﬁmacyodenessmlewasdesigrnedtoreﬂectanimmannedevelopmental
phaseofcornnectedness. 'Ihismodeoftheconrnectedselfistheoreticallymoresubjectto
pmblemaﬁcwwomesthmthatmoderepresentedbydneCmneeedSelfeSelfmethe
Care Scale (Pearson etal, 1991).

86

The BB group's lowest scale score was on Self and Other cee, the scale which
assesses mature intedependence, in which care of the self is integated witln care of othes.
A tentative hypothesis that the normative negative correlation between the Connected Self
and SeparateSelfscaleswouldnotbereplicatedin theBBgroupbecauseofconﬂicted
feelingsrelatedtothesetwoconstructswasnotsupported. Infact,the BBandBRgroups'
Separate SelfandConnected Selfscores showedextremenegativecorrelations, incontrast
to the nonsigrniﬁcant correlation found in the NED group and tlne modeate correlations
found in the original normative group. The correlatiorns between Connected Self and
Separate Self wee signiﬁcarntly different between both the BB and NED groups and the
BR arnd NED groups. Although this outcornne is different than what was hypothesized, it
supports the theory that bulimics peoeive concepts of autonomy, separateness, and
interelatednessasirreconcilablepolaropposites. 'llnisappearstobetruefe'the BRgroup
as well.

Anexaminationofthecorrelationsbetwcen PrimacyofOtherCareandSelfand
Otlne Care furthe suggests the possibility that bulimics have much greater difﬁculty
integrating concepts of individuation and self care with ideas of connectedrness and
intedependency. The scales are signiﬁcantly more negatively correlated in the BB group (r
= -.688) than in the NED group (r = -.234). The correlation in the BR group, perhaps
better labeled the "recover'ng " group, falls between the othe two groups. Correlations
between the Connected Self and Self and Other Care scales lend additiornal support to this
theeetical formulation. TheConnectedSelfand SelfandOtlnerCarescalesare
cornceptually linked. The Connected Self scale reﬂects a mode of self deﬁnition in which
relationswithothesarecentral. 'I'heSelfandOtheCarescalewascrcatedtodescribedne
most developmentally mature form of a connected self orientation, in which the self is
chariyirntegratedandincludedinthosewhoarecaredfor; selfandotheareunderstoodas
equally deseving of care. In the BB group, scores on these two scales are negatively

87

correlated. In the BR group, they are virtually uncorrelated, and in both the NED and
prescreening groups the two scales show modeate positive correlations.

An examination of conelations between real RSI scale scores and subjects' own
ISS scores fur-the elucidates the diffeences in the way the groups deﬁne themselves
relativetorelationships withotlners. 'llnecorrelationbetweenintemalized shamearndthe
Primacy of Other Care score in the BB population is striking (r = 0.627, p < .001)
indicating a signiﬁcant relationship between feelings of shame about the self and a mode of
self deﬁnition focused on the care of others. The same correlation in the NED group is not
sigrniﬁcantly diffeent from zeo, offeing evidence that feelings of sharnne about the self are
much less strongly linked with feelings of dependency. Once again, the correlation in the
BRgroupis moremodeate, and does notreach signiﬁcance (r= .3094, p < .096),
offeing support for the hypothesis that with recovey, feelings of shame and feelings of
dependency become less linked.

The BBgroupappearstohavemoredifﬁcultythan theothegroupsdeﬁning
themselves in a way which integrates both concepts of individuation and connection. They
appear to have difﬁculty diffeentiating developmental concepts of mutual intedependence
fromunilatealdependency. WhiletheBRgroup didnotlookidenticaltotheNEDgroup
in their real RSI respornses, their scores slnow less evidence of difﬁculties with concepts of
separateness and connection. This supports the theory that the resolution of this conﬂict is
linked to recovey from bulimia nevosa in some way.

I I I If . .

Under the "ideal" response condition, the groups diffeed signiﬁcantly only on the
Separate Self scale. The BB and BR groups deﬁned their ideal self as sigrniﬁcantly highe
on the Separate Self dimension than the NED group. It was hypothesized that the BB
groupwould scoresigrniﬁcantly higheonthe SeparateSelfideal than theNEDgroup,
reﬂecting a "false self idal" developed to counte feelings of shame related to intepesornal

88

ordependency needs. Itwasadditionally predictedthatthe BRandNEDgroupswould
have scale score proﬁles that wee similar to each otlner, placing less emphasis on ideals of
separation and autonomy. However, the ideal responses of the BR group resembled the
BBgroupandnottheNEDgroupin tlnisrespect.

The ﬁnding that the BB group expressed signiﬁcantly highe Separate Self ideals
tlmn the NED group is important. This result runs directly counter to theories that bulimic
women ove-idmlize traditional feminine characteistics such as intimacy and dependency
(Boskind-Lodahl, 1976; Pettinati, Franks, Wade, & Kogan, 1987) and lends support to
tlneorieswhicln enphasizecenﬂictsrelatedtocornnectedness/separationoran
overidmlization of autonomy and self-reliance. This ﬁnding receives some corroborating
snrpport from earlie research in which bulirrnic women professed sigrniﬁcantly higher ideals
of autonomy than non-bulimics (McCreery, 1991).

Furtherlongitudinal research wouldbenecessarytoteaseoutwhetheornotthe
magnitudeoftheideal Separate Selfscoreforthe BRgroupwould declinewitln continued
recovey. Again, the level of subdiagrnostic bulimic behaviors and cogrnitions in this group
indicate that the BR group has not completely resolved the issues, feelings, or conﬂicts that
led to their bulimia. An additional possibility exists. Even if dependency/autonomy
cornﬂicts are key dynamics in the developrnnent of bulinnia nervosa, it is possible that
recovey from bulirnnia does not involve a complete resolution of these conﬂicts. It may
be, that for some individuals, recovey occurs as the individual learns to toleate these
conﬂicts without focusing on bingeing and purging behaviors and issues of weight and
appearance. However, the present study presents evidence that the conﬂict is at least
amelieatedifnotfullyresolved. Interoorrelationson manyoftheRSI scalescoresand
correlations ofRSI real scale scores with subjects' ISS scores show a tendency to modeate
and/or movein thedirectionoftheNEDgroup’sscoresintheBRgroup. Anexampleis
thecorrelation between subjects' ISS scoresandtheirscoreson theideal Connected Self

89

scale. In the NED group these variables are signiﬁcarntly negatively correlated. As the
ideal Connected Self score rises, ISS scores decline. The correlation is small but positive
(although not statistically signiﬁcant) in the BB group (r = .1314). A rise in ideal
ConnectedSelfscoresislinkedwith increasedshame. IntheBRgroupthevariablesare
virtually lmcorrelatcd (r = -.053 l).

 

Responses to the audiotaped vignettes provide additiornal informatiorn about the
possible meaning of the groups' diffeing real and ideal RSI proﬁles. This data offes
support for the hypothesis that bulirrnies view healthy dependency needs as shameful,
signiﬁcantly more shameful than recoveed bulimics or non-eating disordeed women.

The BB gmup attributed a sigrniﬁcantly highe level of shame to the wornarn who
expressed dependency needsthantothewomarn depictedasself-reliarnt. Infactaccording
to the BB group, using non-clinical norms for ferrnales aged 17—63 (Cook, 1993), the
wmnmwhoexpresseddepmdemyneedsfalhmﬂeﬂmpeoenﬁleformtenalindshame
The individual depicted as self-reliant was classiﬁed in the 21st peoeltile by the BB group.
While all three groups attributed highe ISS scores to the womarn who expressed
dependency neeth, the BB group's attributiorn of shame to the dependency needs vignette
was signﬁcantly highe than the BR or NED groups' attributions of shame. As would be
expected, hypothesizing that theBRgroupisin the midstofarecoveryprocess , the BR
group's attribution of shame for this vignette fell between the BB and NED group's scores.

TheﬁndingdmtdneBRgroupratedthewomanwhoexplessedintepesonalneeds
as lessasharnedtharn the BBgroupis important tonote. Thepresentresearch replicatesan
eariier study in which bulimic women attributed signiﬁcantly mote shame to the individual
in tine intepersonal needs vignette than did non-bulimic women (McCree'y, 1991). That
worrnen who no large nneet the diagnostic criteia for bulimia attribute sigrniﬁcantly lowe
levels of shame to an individual expressing intepersonal needs than do women who are

90

currently bulinnic suggests that a change in the peoeption of dependency needs is involved
in the recovery process. This hypothesis has somewhat more validity given that botln BB
and BR subjects wee identiﬁed from a general population of college students and wee not
speciﬁcally recruited forastudyorn "eating disorders," creating less likelihoodthatornlya
cetain subgroup of recovered individuals volunteeed for inclusion in the study.

It shouldbenotedthatnoneofthegroups seemedtodiffeentiatetlnecontrol
vignette from the self-reliant vignette. While it may not be possible to create a truly neutral
depictiorn, the self reliant vigrnette and the control vigrnette wee clearly diffeent The
woman in the self reliant vignette professed autonomy and non-reliance on othes and the
woman in the control vignette balanced expressions of autonomy with expressions of
interpesonal rneed and interdependence. Previonrs validity testing demonstrated a
diffeencebetween thevignettes. Thelackofdiscrimination between theselfreliantand
control vignettes in attributions of shame raises an inteesting possibility. All three groups
peoeived the individuals in the neutral arnd self-reliant vigrnettes as equals in terms of their
level ofshame. Allthreegroupsratedthewonnan whoexpresseddeperndencyneedsas
havingahigherlevel ofshamethan tlneotlnetwo women. ltappearsthatall subjects wee
much more sensitive to tire absense of a strong value of self-reliance (as in the dependency
needsvigrnette) than they weretotlneabsenseofavalueofconrnectedrness (in the self-reliant
vignette). This may, in part, reﬂect an age appropriate developmental corncen. Yonmg
college women are typically working through issues related to a signiﬁcant physical and
psychological separation from their family of origin. This lack of sensitivity to the absence
of values of connectedness and intedependence may also reﬂect'the lack of emphasis on
and validation of these concepts in the mainstream culture.

Ananempttoclmifyﬂnerelaﬁonslﬁpbetweenﬂnelevelofshameatuibutedto
dependency needs arnd the magnitude of bingeing and purging behaviors yielded inteesting
ﬁndings. It was hypothesized that the level of shame attributed to the worrnan who

91

expressed dependency needs would be positively correlated with subjects' levels of
bingeing and purging. Shame over dependency needs was hypothesized to lead to the
denial of dependency rneeds and attempts to gratify them indirectly by bingeing. Purging
washypothesizedtorepresentan undoing ofthebingeoraresmganceofdenialof
interpesonal need

No such relationship emeged in the data. Although thee was some evidence of a
statistically signiﬁcant relationship between the severity of bingeing and purging symptoms
andtheshameattributedtoinwrpesonalneedsintheBRgroup,thiswasonlyapparentin
the BR's report of their worst level of bingeing, vomiting, and laxative use—not their
presentlevel. Evenmoreglaringwasthelackofany suchpatternofcorrelationsintheBB
group (statistically sigrniﬁcant or not). In fact, in the BB group, the seveity of bingeing
and vomiting, amount of laxatives used to purge, amount of diet pills used, and the amount
of exercise presently used to “burn off a binge," as well as ”worst ever” levels of vomiting,
bingeing, and diet pill use and Drive for Thinness and Bulirrnia EDI scores wee all
negatively correlated with the shame score for the intepesonal needs vignette. (It needs
tobe su'essedthatonly twooftlnesecerelationkthectxrelationsoftheintepesonal needs
vigrnette with current diet pill use and witln worst diet pill use wee statistically signiﬁcant).

One explarnatiorn for this ﬁnding is consistent with the theoretical hypothesis that the
bulimic is engaging in bulimic behaviors in order to defend against the deepe internalized
shame she feels about he own dependency needs (Kaufman, 1989; Wurmse, 1981). An
errorinthedevelopmentofthehypothesesofthis studyappearsrelatedtoalackof
attention to the bulimic's need to defend against her feelings of shame. According to both
Kaufman and Wurmse, tlne bulinnic not only defends against he dependency needs
(through denial and by using food as a substitute for emotional intimacy), she also defends
against the intense feelings of shame she experiences related to those needs (Kaufmarn,
1989; Wurmser, 1981).

92

Based on this tlneoretical formulation, the BB group can be viewed as using bulimic
behaviors in a defensive manneruas a substitute for dependency needs which are associated
with shame too painful to allow one's self to expeience. To the extent that the bulimic is
actively bingeing and purging, she is defending heself against the shame she associates
with intepesonal needs and is likely to deny the shamefulness ofintepesonal needs—in
heselfor in anothe individual. Speciﬁcally, the data for the BB group indicate that as
bingeing and purging incrmse, the shame bulinnics attribute to the intepesonal needs
vigrnettedecrease. Ahypothesizedexplanation deseving furtheresearchattention istlnat
bulimia seves to help the bulimic avoid her affective expeience (shame) of intepesonal
needs. This would suggest that, wee the bulimic group not bingeing and purging, the
level of shame attributed to the woman who expressed dependency needs nnight have been
even higher.

Tlnetrendofnegativecorrelationsbetween rrneasuresofeating disordeandtlne
slumeattributedtointepersonal needsisnotas stronglyorclearly depictedin the BR
group. The BR group shows three statistically insigniﬁcant negative correlations between
thelevelofslmmetheyattributedtotheintepersonalneedsvignetteandtheircurrentlevel
of bingeing, current level of diet pill usage, and their Drive for Thinness score on the EDI.
Howeve two of these negative correlations (the correlation of the vigrnette with current
bingeing and with current diet pill usage) are extremely close to zeo (-.02 and -.04
respectively). On all measures of current bingeing and purging (including the use of
exercise to purge), on measures of "worst eve" levels of bingeing and purging, and on the
EDI Bulimia scale, the correlation with the intepesonal need shame score was highe arnd
more positive for the BR group than for the BB group (current level of laxative and diuretic
use wee excluded because none of tlne recoveed bulirrnics reported current use). Not only
does the BR group attribute less shame to the interpersonal needs characterization, they do

93

not appear to use bingeing, purging, and eating disordeed cogrnitions to modulate their
peceptiorn ofshamein the same way the BB group appearsto.

This ﬁnding is preliminary arnd reﬂects only statistical trends noted in the data
(although it should be noted that because of the effect of group size on statistical
sigrniﬁcance, some of the "nonsigniﬁcant" trends wee quite large-in the BB group the
correlationbetween shameon tlneintepersonalneedsvigrnetteand the Bulimia scalewas
-0.41 (p < .059) and between the same vignette and level of purging was -0.35 (p <
.107)). The data suggest, however, that bulinnic symptoms-bingeing, purging, the
tendency towards bingeing, impulses to purge, and extreme concern with dieting, weight,
and the pursuit of thinness-~seve a defensive purpose for the BB group that they do not
seve for the BR group. It may be that recovery from bulimia must involve not ornly an
increasedacceptanceofdependelcyneedsanddecreasedlevels ofsharneattributedto those
needs, but also an increased ability to tolerate and acknowledge affect (a common clinical
observation), or at least, feelings of shame.

The results of the present study suggest that bulimics deﬁne themselves and their
ideals about relatedness somewhat diffeently than non-eating disordeed women or
recoveed bulirrnics. Additionally, the study provides evidence that behaviorally bulimic
women appear to peoeive nonpathological dependency needs as shamefulnsigniﬁcantly
more shameful than non-eating disordeed women and recovered bulinnic women view
these needs. The ﬁndings of the present study are however, somewhat preliminary.
Additionally, the study only begins to examine the role of shame and of perceptions of
dependency needs in bulimia and in recovery. Furthe' research aimed at replicating and
further exploring differences between bulimics and recoveed bulirrnics, as well as
lorngitudinal studieswhich detail therecovery prowss,areessentialinordertobetterclarify ,
and deﬁne the dynamics underlying recovey. longitudinal research which seeks to begin
examining these variables before eating disordeed behavior develops could also provide

94

important data about the meaning arnd motivations undelying bulimia nevosa. This
research might also furtlner elucidate the relatiornship between bulimics' peoeption of
dependency needs and seveity of bulimic symptoms.

The present study has seveal linnitations which should be considered in developing
future research. The population studied consisted exclusively of college students who may
have somewhat diffeent peceptions and ideals and who may represent a distinct (though
still signiﬁcant) subgroup of eating disordeed women. Because of a desire not to bias the
sample by recruiting speciﬁcally for eating disordeed individuals, it was difﬁcult to recruit
large sample sizes. Important trends or diffeences between groups may have gone
undetected. In this same vein, the BR group was not "pefectly recoveed” It would be
bothinteesting andimportanttoexaminethe samevariablesinagroupofwomenwho
showed long termandcompleterecovey frombulimia.

Finally, subjects were assigned to groups based on their fairly arnonymous, written .
self reports of behaviors. While this mehodology has some advantages, it must be noted
that the study lacks ml objective measln'es of eating disordeed behaviors, history, body
mass index, and other diagnostic variables. Findings based on these more speciﬁc
clmacteistics must be viewed tentatively and intepreted with this in mind.

CONCLUSION

The study sought to answe two major research questions:

1. Do bulinnics deﬁne tlnemselves and their ideals about relatedrness diffeently than non-
eating disordeed or recoveed bulimic women?

2. Do bulinnics view healthy dependency needs as shameful, signiﬁcantly more slnrneful
than non-eating disordeed arnd recovered bulimic women peceive these needs?

The results of the study support positive answe's to both questions. The
behaviorally bulimic group (BB) reported both clinically arnd statistically elevated levels of
inte'nalized shame. The signiﬁcant trend of declining ISS scores from behaviorally bulimic
(BB) to behaviorally recoveed bulimic (BR) to the non-eating disordeed (NED) group
supports a relationship between the intensity of bulimic pathology and feelings of
internalized shame. Responses to the Relationship Self Inventory (RSI) and to the
audiotaped vignettes support the theory that the role of shame in bulimia nevosa appears to
bemorethanasecondaryreactiontobulinnic symptomsandappearstobeintetwinedwith
peceptiorns of dependency needs and self-reliance.

Intercorrelations between real RSI scale scores and correlations between RSI scale
scores and subjects' ISS scores provide evidence supporting the theory that bulinnics tend
to peceive concepts of autonomy, separateness, and interelatedrness as irreconcilable polar
opposites. They appear to have difﬁculty deﬁning themselves in a way which irntegrates
both concepts of individuation and connectiorn. The responses of the BR group show some
of these same patterns. Both the BB and BR groups real RSI scores revealed negative
correlations between the Connected Self scale (CS) and the Separate Self scale (SS) that
wee signiﬁcantly highe than the NED group. Botln the BB and BR groups deﬁned tlneir

95

96

ideal self as signiﬁcantly higher on the Separate Self dimension than the NED group. The
BR group did not resemble the NED group as closely as hypothesized. Howeve, BR real
RSI responses show less extreme evidence of the difﬁculties with concepts of separation
and connection found in the BB group. This supports the theory that an increased ability to
deﬁne the self in a way which integrates concepts of cornnection and individuation is linked
to recovey from bulimia nevosa.

Group responses to the audiotaped vignettes provide compelling evidence that
bulinnics view dependency rneeds as shameful, signiﬁcantly more shameful than recoveed
bulimics or non-eating disordered individuals. The ﬁnding that the BR group attributed
signiﬁcantly lower levels of shame to an individual expressing interpersonal needs than did
the BB group suggests that a change in the perceptiorn of dependency needs is involved in
the recovey process. Additionally, trends in the data suggest that in the BB group, the
level of eating disordeed diagnostic responses is negatively correlated with the shame
attributed to dependency needs. This suggests that bulimia itself may be a means of
defending against painful feelings of slnarne. This relationship is not apparent in the BR
group. Not only does the BR group attribute less shame to the interpersonal needs
characterimtion, they do not appear to use bingeing, purging, and eating disordeed
cogrnitionstomodulatetheirpeceptionofsharneintheway theBBgroupappearsto.

APPENDICES

97

APPENDIX A

91mm

WW mmismadeupofwbjectswhomﬂyrepm
recurrentepisodesofbingeeating(atleasttwoepisodcsaweekforatleastthreemonths)
and who use vomiting, laxative use, diuretic use, or extreme execise to ”get rid of" food

catenortoloseweightaminimum oftwotimespermonth.

 

t; Thisgroupismadelrpofsubjectswho
presently do not meet the criteia for inclusiorn in thebehaviorally bulimic group but who
reportthatinthepast(morethanfourmonthsago)theydidmeetthecriteia.

 

This group is nnade up ofthe ﬁrst twenty ﬁve
subjects (scored at random) who report no history of bingeing or paging belnaviors of any
kind, who report an ideal weight of no more than ﬁve pounds below their present weight,
and whose scores on both the "Drive for Thinness“ and the "Bulimia" scales of tire Eating
Disordes Inventory fall below the 50th pecentile for female college students (scores of
two or less and zeo respectively).

98

APPENDIX B

W

a) Marla

Interviewer : Okay, we can get started whenever you're ready.

Marla : Alright . . . let's see . . . what kind of person am 1? Well, I’ve always been
described as an individualist in my family . . . and I guess that's right. I think that the best
way to get something done is do it yourself. I‘m pretty independent minded . . .andl
enjoybeing on my own too. I’vealwaysbeen likethat . . likelremembeoncewhen I
wasalittlegir'lIgotlostinadepartmentstorebecauselleftmymomtoﬁndthetoy
department. And they asked me when they found me why I hadn’t just asked someone
howtogettheebutitseemedtomeatthetimelikeI’djustdoitonmyown. . .andl
guess now it’s important to me to do things at my ownpaceand the way I want them done.
I mearn, ultimately I’m the one I have to please. Right?

Imervr‘mr: So can you tell me how that plays into your relationships?

Marla: Yeahltlnnkmyboyﬁiendundetsnandsthatpattofme Ithirnkbecauseofthat
we're really compatible. We're botln really busy all the time, and our relationship is the
idealescapeflomallthat. . . Ithinkhe'stlnepefectboyfriend.

Interviewer: Can you tell me wlmt that mearns?

Marla: Well, I have someone to enjoy my free time with and to relax with. I nnean, we
both have our own friends and we each have jobs and we have our school stuff, so we
don'tlike,needeachotherandwedon'thassleeach otheralltlnetime... Ican'tinnagine
_ being like that, you krnow, like those women from the ﬁfties, who relied on their husbands
for everything . . . John and I have fun together. But I don't like, rely on him for stuff, we
can each takecareofourselves. We haveaprettygoodtimetogetheandthat'sideal for
me. We understand each othe too.

99

Interviewer: So how would your friends describe you?

Marla: They’d describe me as independent too I think. I mean I have a lot of friends, but I
like to spend most of my time by myself. I get my best ideas when I’m by myself . . . . I
liketotlninkthingsthrough when I’malone. Imean,lcantalktomy friends,andldotalk
tothem,butldon’tliketellthemeveylittlethingthatl’mdoingalltlnetime. Likelast
summe I tried out for the swim team and I didrn’t tell anybody in my house until I made the
ﬁnalcuts. lguessljustdidn’tfeel likelneededtheirsupportorencouragement. . . Ijust
ﬁgured,hey,iflmakeitlmakeit.

Interviewer: Do you think this independence of yours affects how you are in school?
Marla: Probably . . . yeah, like in class, I guess I‘m not one of these people who asks a
lot of questions. I mean, I don’t really go to professors a lot for help . . . You know, I like
to go off and try to ﬁgure things out on my own. It’s kind of a challenge. Ienjoy it.

b) Audrey

Interviewer : We can get started whenever you're ready.

Audrey : Okay.

Interviewer: I'll just start by asking you the real geneal question. What kind of a person
do you think that you are and how do you feel that affects how you function in school and
in your relationslnips?

Audrey : . . . that's a complicated question . . . umm, I'm not sure what you mean by all
that but, well, people are important to me. I‘m not like Suzy cheerleader or anything, but
my friends, you know, having good friends, that is important. I 'mean I like to be by
myself sometimes, but not all the time.

Interviewer: How would your friends describe you?

Audrey: Hopefnrllyasagoodfniend. . . umm . . . Ihaveasmallgroupofpeoplethatl
am pretty close with, people I have known since I was little. It's almost like we’re family.

100

. .Inneanthey lunoweverythingaboutmeandlknowallaboutthem. Likelastyear,when
my mom was in the hospital. . .I don't know what I would have done without them.
Having people there to comfort me. . . it was horrible. . .I don't krnow what I would have
donewithoutfriendstheretohug meandletmecryonthemandtokeepmecompany
throughit. ..... Myﬁiendsaresogreatlmeanlreally...lreallyrespectthem. And
theiropinion meansalottome. . . .Like myonefriend,shegraduatedlastyear,and she's
been helping me work on my resume. I'm graduating this spring and I’ll be looking for a
jobandit'sgreattohave someonewhoknowstheropes,whoisshowing mehowtodoit
right. She was the same major as me so she's been through it.

Interviewer: You said that you have a boyfriend?

Audrey: Yeah, Joe . . . We get along so well together. It's so nice having someone I care
about, you know, that much . . . someone I'm close to, who I can rely on and who
dependsonme. . .IfI'vehadalong day hemakesmedinnerandldothesamething for
him. And sometimes it feels so good just to be held . . . We help each other out, support
eachother,giveadvice. Likelreadhim therough draftsofmy Englishpape'sandhe
gives me feedback. I just love having someone like him. It's a lot of compromises
though, when you have two people with different goals and schedules and stuff. It's more
work than not having a boyfriend sometimes. But I think it's worth it. I'm really happy.
Interviewer: Andwhatdoyou thinkofschool? Doyou thinkthatthekindofpe'son you
areaffectshowyouarein school?

Audrey: Not really. I do ﬁne in school. I like parts of it (pause) . . .I like smaller classes
much better than those huge ones I had my freshman year. It's much easier to get
feedback, to ask questions and to make sure I'm understanding. I think school is easie in
thelasttwoyearsbecauseyougettoknowpeoplebetter. Ihavethisprofessorwhol‘ve
beenworking with, she'sgreat. 'lhiswomanisexactlywhatlwouldliketobelikewlnenl

101

ﬁnally get out ofhere and start working. And she's been giving me advice on classes and
instructors and things like that.

c) Leslie

Interviewer: Okay, let's get started with the general question. What kind of a person do
you think you are and how do you feel that affects how you function in school and in your
relationships?

Leslie : That's a confusing question . . . I guess I think I'm a happy person. I do well in
school. Ilikewhatl‘m studying. Ilikewritingpapersmorethanlliketaking tests. . .I
think it's the challenge of being creative. I'm not sure how what kind of person I am
affectshowldoinschool. ..ImeanIguessyoucouldsay I'mresponsible. Igetmy
work done and I turn it in on time, but I'm not a real pe'fectionist about school. I have a
lot of other interests as well. That’s what I enjoy about a big school. You can really get
lost here ifyou want to, likeyou can takeaclasswhee you neverhavetospeaktothe
teacher, or go to a football game and just lose yourself in the crowd . . . or, you know, you
can take advantage of opportunities to meet people and get involved with smaller groups. I
like having a choice. . .I mean, thee are plenty of opportunities to make friends bee, but
thereisalotofspaoetobealoneifthat’swhatlreallywantatthemoment.

Interviewer: So what are your friendships like?

Leslie : Well, I have friends that I would call really close and I have friends I krnow well
enough to do things with . . . you know?

Interviewer: Do you spend a lot of time with your friends, do you rely on them a lot?
Leslie : Well. . . it really depends. During the school year I don't see my friends at home
much. I’m going to parties and stuff on the weekends and hanging out at people's
apartments . . . I'm not sure if] rely on them . . . I mean for some things, sure. . .like

when my carbroke down last week and I had to call my friend to rescue me, or when we

102

take a class together and we study for the ﬁnal . . . but in some things I’m real
independent. I mean, I krnow a girl who won't go to the mall by herself, and I’m not like
that. I like to do some things alone.

Interviewer: Are you involved in a relationship right now?

Leslie : mrrnhuh . . .

Interviewer: Can you tell me a little about that?

Leslie : Oh . . . okay. My boyfriend Kevin and I have been going out for a while now
and things are really good. I'm really happy. We get along vey well. We complement
each otlner. It’s so great having someone in your life with lots of the same goals and
interests. We both love to go eamping in the summer . . . we like the same music, have the
same taste in movies. He's a great support in some ways . . . but thee are just some
things males just don't seem to understand, you know? . . But that's okay. I think I'd go
crazy if we wee that compatible. My friends and relationships are important to me, but I
nwdmyownspaceandtimetoo.. .IguessIlikeandIneedtokeepsomethingsto
myself.

103

APPENDIX C

2.1.11.0 ..

Please rate the following statements as either true or false:

I.

999999!”

This individual needs relationships with others.

This individual is autonomous.

This individual is dependent on other people.

This individual does not need other people.

This individual displays a need for touching and/or holding.

This individual is independent of other people.

This individual does not show a need for other people's approval.

This individual showed a need to have someone slne can identify with or model herself
after.

Scoring:

"True" responses on items 1, 3, 5, 8 are scored one point.
“False” responses on items 2, 4, 6, 7 are scored one point.
Higher scores indieate the presence of interpersonal needs.

104

APPENDD( D

M

1. Demographic items

2. Items 1-60: Relationship Self Inventory ”Real"

3. Items 61-120: Relationship Self Inventory "Ideal”

4. Items 121-150: Internalized Shame Scale

5. Items 151-164: Eating Disorders Inventory Bulimia and Drive for Thinness Scales

105

l . Age at last birthday
2. MaritalStatus (check one)

 

 

snngle divorced, rennarried
___married separated
divorced, single widowed

__cohabitating (living with signiﬁeant other)

3. Marital status of Wmheck one)
single
married
divorced, single
divorced but at least one parent has remarried
widowed
cohabitating (not married but living together)

4. Religion (check one)

 

 

 

 

___Catholic __Protestant __Iewish
Other ( ) ___No religious afﬁliation
5. How regular are you in your religious observance?
__Attend regularly Never attend
__Attend occasionally Does not apply: no Attend rarely
religious afﬁliation
6. Primary ethnic or racial identiﬁeation (check one)
. Black/African-Ame'iean Asian
Native American White/Caucasian
Hispanic Other
7. Your family's estimated gross income for last year (check one)
__310,000-20,000 ___350,000-$60,000
$20,000—30,000 $60,000-$70,000
330,000-340,000 $70,000—$80,000
___540,000-50,000 __$80,000-$90,000
over $90,000

8. Number of people in your family _____
9. Number of people supported by your parents

Are you currently taking any medicatiorn prescribed by a physician yes

no

If yes, please list the medications you are taking and the dosages (if you know them)

 

 

Have you ever been had any counseling or

psychotherapy yes
no

106

if yes, was it:

individual treatment

family therapy
__couples therapy (marital therapy)
__tlmw 810“?

Please estimate the number of psychotheapy or counseling sessions you attended:
individual sessions

family therapy sessions

couples therapy sessions

n therapy group sessions

 

How old were you at the time?
Did you seek counseling for: (check all that apply)

family problems/problems with parents
___marital problems

depression

drug addiction/substance abuse
___problems with alcolnol

eating disorder
anxiety

sleep problems

school problems

other

 

 

 

 

 

 

107

For the following items, please read each statement CAREFULLY. Decide how much it
describes you. Using the following rating scale, circle the most appropriate response.

SCALE
Not at all like me Very much like me
1 2 3 4 5

1 2 3 4 5 l.Ioftentrytoactonthebeliefthatself-inteestisoneoftheworst
problemsfacingsociety.

l 2 3 4 5 2. A close friend is someone who will help you wherever you need
help and krnows that you will help if they need it.

1 2 3 4 5 3. Ieannotchoosetohelpsomeoneelseifitwillhindermyself-
development.

1 2 3 4 5 4. I wanttoberesponsible for myself.

1 2 3 4 5 5.Inmakingdecisions,leanneglectmyownvaluesinordertokeepa
relationship.

1 2 3 4 5 6. Iﬁndithardtosympathizewithpeoplewhosemisfortuneslbelieve
areduemainlytotheirshortcomings.

1 2 3 4 S 7.1trytocurbmyangerforfearofhurtingothe's.

1 2 3 4 5 8. Being unselﬁsh with othes is more important than making myself
happy.

1 2 3 4 5 9. Loving is like a contract: Ifits provisions aren't met, you wouldn't
lovetlrepersonanymore.

l 2 3 4 5 10. Inmyeve'ydaylifelamguidedbytlnenotionof"aneyeforaneye
andatoothforatooth.”

l 2 3 4 5 11.1wanttoleamtostandonmyowntwofeet.

l 2 3 4 5 12.Ibelievetlratorneofthemostimportantthingsﬂnatparentscanteach
their children is how tocoope'ate and livein harmony with others.

1 2 3 4 5 l3. Itrynottothinkaboutthefeelingsofotherswbentheeisa
principleatstake.

l 2 3 4 5 l4. Idon't oftendomuch for others unless tlneycandosome good for
melateron.

1 2 3 4 5 15. Activitiesofearethatlperformexpandbothmeandothers.

l 2 3 4 5 l6. Ifwhatlwanttodoupsetsotherpeople,ltrytothirnkagairntosee
iflreallywanttodoit.

l 2 3 4 5 l7. Idonotwantotlnerstoberesponsibleforme.

108

18. Iamguidedby theprincipleoftreating othe'saslwanttobe
treated.

19. I believe that I have to look out for myself and mine, and let others
shift for themselves.

20. Being unselﬁsh with othes is a way I make myself happy.

21. When a friend traps me with demands and negotiation has not
worked, I am likely to end the friendship.

22. I feel empty if I‘m not closely involved with someone else.

23. Sometimes I have to accept hurting someone else if I am to do the
things that are important in my own life.

24. In order to continue a relationslnip it has to let botln of us grow.

25. Ifeelthatmydevelopmenthasbeen shapedmoreby thepersonsl
care about than by what I do and accomplish.

26. People who don't work hard to accomplish respectable goals can't
expect me to help when they're in trouble.

27. Relationships are a central part of my identity.

28. I often keep quiet rather than hurt someone's feelings, even if it
means giving a false impression.

29. If someone offers to do something for me, I should accept the offer
even if I really want something else.

30. The worst thing that could happen in a friendship would be to have
my friend reject me.

31.1f1amreally surethatwhatlwanttodoisﬁghgldoitevenifit
upsetsotherpeople.

32. Beforelcanbe surelrmllycarefor someonelhavetokrnow my
true feelings.

33. Whatitallboilsdowntoisthattheonlype‘sonIcanrelyonis
myself.

34. Even though I am sensitive to others' feelings, I make decisions
baseduponwhatlfeel isbestfor me.

35. Even though it’s difﬁcult, I have learned to say no to others when I
need to take care of myself.

36. I like to see myself as inte'connected with a network of friends.

nun-third
NNNN
UUUU
5585

UIUIUIUI

109

37. Those about whom I care deeply are part of who I am.
38. I accept my obligations and expect others to do the same.

39. I believe that I must eare for myself because othes are not
responsible for me.

40. The people whom I admire are those who seem to be in close
personal relationships.

41. It is necessary fornretotakeresponsibility fortheeffect my actions
haveon others.

42. True responsibility involves making sure my needs are cared for as
well as the needs of others.

43. The feelings of otlners are not relevant when deciding what is right.

44. If someone asks me for a favor I have a responsibility to think
about whether or not I want to do the favor.

45. I make decisions based upon what I believe is best for are and
mine.

46. Once I‘ve worked out my position on some issue I stick to it.

47. I believe that in order to survive I must corncertrate more on taking
care of myself than on taking care of othes.

48. The best way to help someone is to do what they ask even if you
don't really want to do it.

49. Doing tlnings for others makes me happy.
50. All you really need to do to help someone is to love them.
51. I deseve the love of othes as much as they deserve my love.

52. You've got to look out for yourself or the demands of
circumstances and of othe people will eat you up.

53. Icannot affordto give attentiontotheopinions of othes when lam
ce'tainlamcorrect.

54. If someone does something for me, I reciprocate by doing
something for them.

55. Caring about other people is important to me.

56. If other people are going to sacriﬁce something they want for my
sake I want them to understand what they are doing.

110

1 2 3 4 5 57.WhenImakeadecidionit'simportanttousemyownvaluesto
maketherightdecision.

l 2 3 4 5 58.1n'ytoapproachrelationshipswiththesameorganizationand
efﬁciency as I approach my work.

1 2 3 4 5 59. IfIamtohelpanotherpe'sonitisimportanttometounde'stand
myownmotives.

1 2 3 4 S 60.11iketoacquiremanyacquaintancesandﬁiends.

111

Now respond to the following statements. You have seen the statements before, but THIS
T'IME PLEASE RESPOND AS IF YOU WERE EXACTLY AS YOU WISH. In other
words, n . , , , _ . r in . Pleaseanswereach
question very carefully ESE m2 NQI: SKQ AIS! ITEMS. IF AN ITEM SEEMS
HARD TO AN SWFR, CHOOSE THE ANSWER WHICH IS MOST APPROPRIATE.

SCALE
Not at all like me Very much like me
1 2 3 4 5

 

1 2 3 4 5 61. Ioftentrytoactonthebeliefthatself-inteestisoneoftheworst
problemsfacingsociety.

l 2 3 4 5 62. Acloseﬁiendissomeonewhowillhelpyouwhereveryouneed
helpandknowsthatyouwillhelpiftheyneedit.

l 2 3 4 5 63. Icannotchoosetohelpsomeoneelseifitwillhindermyself-
development.

1 2 3 4 5 64.1wanttoberesponsibleformyself.

l 2 3 4 5 65.1nmakingdecisions,lcanneglectmyownvaluesinordertokeep
arelationship.

1 2 3 4 5 66.Iﬁndithardtosympathizewithpeoplewhosemisfortunesl
believe are due nrainly to their shortcomings.

l 2 3 4 5 67.1trytocm'bmyangerforfearofhurtingothers.

l 2 3 4 5 g. Beingrmselﬁshwithothesismoreimportantthanmaldngmyself
PPY

1 2 3 4 5 69. Loving is like a contract: If its provisions aren't met, you wouldn't
love the person any more.

1 2 3 4 5 70. Inmyeverydaylifelamguidedbythenotionof 'aneyeforaneye
andatoothforatooth."

l 2 3 4 5 71.1wanttolearntostandonmyowntwofeet.

l 2 3 4 5 72.1be1ievethatoneofthemostimportanttlungsﬂratparentseanteach
theirchildrerishowtocoopeateandliveinharmonywithothe's.

1 2 3 4 5 73. In'Lnottothinkaboutthefeelingsofotheswhertheeisa

l 2 3 4 5 74. Idon'toftendomuchforothe’sunlesstheycandosomegoodfor
melateron.

1 2 3 4 5 75. Activitiesofcarethatlpe'formexpandbothmeandothes.

112

76. Ifwhatlwanttodoupsetsothepeople,ltrytothinkagaintosee
ifIreallywanttodoit.

77.1donotwantothestoberesponsibleforme.

78.1am guidedby theprincipleoftreating othesaslwanttobe
treated.

79. I believe that I have to look out for myself and mine, and let others
shift for themselves.

80. Being unselﬁsh with others is a way I make myself happy.

81. When a friend traps me with demands and negotiation has not
worked, I am likely to end the ﬁiendship.

82. I feel empty if I'm not closely involved with someone else.

83. Sometimes I have to accept hurting someone else ifIam to do the
things that are irnportarnt in my own life.

84. In order to continuea relationship it has to let both ofus grow.

85. Ifeeltlratmydevelopmenthasbeenshapedmorebythepersonsl
eareaboutthan by whatldoand accomplish.

86. People who don't work hard to accomplish respectable goals can't
expect me to help when they're in trouble.

87. Relationships are a central part of my identity.

88. I often keep quiet rather than hurt someone's feelings, even if it
means giving a false impression.

89. If someone offe's to do something for me, I should accept the offer
even if I really want something else.

90. The worst thing that could happen in a ﬁiendslrip would be to have
my friend reject rue.

91.1fIamreally surethatwhatlwanttodoisrighudoitevenifit
upsets other people.

92. Beforelcanbesurelreallycarefor someonel havetokrrow my
true feelings.

93. What it all boils down to is that the only peson I can rely on is
myself.

94. Even though I am sensitive to otlne's' feelings, I make decisions
based upon what I feel is best for me.

HHHH

HHHH
NNNN
“MUN
858:3
MUIUIM

NNNN

“NOD03

huh-BIB
UIUIUIUI

113

95. Even tlnough it's difﬁcult, I have learned to say no to others when I
need to take care of myself.

96. I like to see myself as inte'connected with a network of friends.
97. Thoseaboutwhom Icaredeeplyarepartofwholam.
98. I accept my obligations and expect othes to do the sarrne.

99. Ibelievethatlmustcare formyselfbecauseothesarenot
responsible for nne.

100. The people whom I admirearethose who seen tobe in close
personal relationships.

101. It is necessary for rrne to take responsibility for the effect my
actions have on others.

102. True responsibility involves making sure my needs are cared for
aswellastheneedsofothers.

103. The feelings of othes are not relevant when deciding what is
right.

104. If someone asks me for a favor I have a responsibility to think
aboutwhethe'ornotlwanttodothe favor.

105. I makedecisions based upon what I believe is best for me and
mac.

106. Once I've worked out my position on some issue I stick to it.

107.1believethatinorderto survivelmustcercentratemoreontaking
eare of myself than on taking care of othes.

108. The best way to help someone is to do what they ask even if you
don't really want to do it.

109. Doing things for otlners makes me happy.
110. Allyoureallyneedtodotohelpsomeoneistolovethem. ’
111. I deserve the love of othes as much as they deserve my love.

112. You've got to look out for yourself or the demands of
circumstances and of othe people will eat you up.

113. Icannotaffordtogiveattentiontotheopinionsofothe'swherI
amcertainlamcorrect.

114. If someone does something for me, I reciprocate by doing
sonnething for them.

114

115. Caring about other people is important to me.

1 16. If othe' people are going to sacriﬁce something they want for my
sake I want tlnem to undemnd what they are doing.

117. When I make a decidion it's important to use my own values to
make the right decision.

118. ln'ytoapproachrelationshipswiththesameorganizationand
efﬁciencyaslapproachmywork.

119. If I am to help another person it is important to me to undestand
my own motives.

120. I like to acquire many acquaintances and friends.

115

Below is a list of statements describing feelings or experiences that you may have from
time to time or that are familiar to you because you have had these feelings and expeiences
for a long time. Most of these statenents descrrbe feelings and expeiences that are
generally painful or negative in some way. Some people will seldom or never have had
many 0 these feelings. Everyone has had some of these feelings at some time, but if you
ﬁnd that these statements describe the way you feel a good deal of the time, it can be
painfuljustreadingthem. Trytobeashonestasyoucaninresponding.

Read each statenent earefully arnd circle the WW to tlre left of the item that indieates the
ﬁequency with which you ﬁnd yourself feeling or experiencing what is described in the
statement. Use the scale below.

DO NOT OMIT ANY I'I'EM.

SCALE:
luNEVER 2--SELDOM 3--SOMETIMES “FREQUENTLY
5--ALMOST ALWAYS

1 2 3 4 5 121. Ifeellikelamnever quite good enough.

1 2 3 4 5 122. I feel somehow left out.

1 2 3 4 5 123. Ithinktlnatpeoplelookdownonme.

1 2 3 4 5 124. Allinall,IaminclinedtofeelthatIamasuccess.

1 2 3 4 5 125. I scold myself and put myself down.

1 2 3 4 5 126. I feel insecure about others opinions of me.

1 2 3 4 5 127. Compared to other people, I feel like I somehow never
measure up. .

l 2 3 4 5 128. I see myself as being very small and insigniﬁcant.

1 2 3 4 5 129. I feel I have much to be proud of.

1 2 3 4 5 130. I feel intensely inadequate and full of self doubt.

I 2 3 4 S 131. Ifeelasiflarnsomehowdefectiveasaperson,liketheeis
something basically wrong with me.

1 2 3 4 5 132. WhenIcomparemyselftoothe‘sIamjustnotas
important.

1 2 3 4 5 133. I have an overpowering fear that my faults will be
revealed in front of othes.

1 2 3 4 5 134. IfeelIhaveanumberofgoodqualitiesf

l 2 3 4 5 135. Iseemyselfstrivingforpe'fectiononlytocontinually fall
short.

1 2 3 4 5 136. Ithinkothersareabletoseemydefects.

1 2 3 4 5 137. Icould beat myself over the head with aclub when!
makeamistake.

papa-spa

N
M
8

NNNN
UUUU

5855
UIUIUIUI

138.
139.
140.

141.

142.
143.

144.
145.

146.

147.
148.
149.
150.

116

On the whole, I am satisﬁed with myself.

I would like to shrink away when I make a mistake.

I replay painful events over and over in my mind until I
am overwhelmed.
Ifeellamapesonofwortlratleastonanequal plane
with others.

At times I feel like I will break into a thousand pieces.
I feel as if I have lost control over my body functions and
my feelings.

Sometimes I feel no bigger than a pea.

Attimeslfeel soexposed that] wishtheearth would
open up and swallow me.
I have this painful gap within me that I have not been
able to ﬁll. ‘

I feel empty and unfulﬁlled.

I take a positive attitude toward myself.

My loneliness is more like emptiness.

I always feel like thee is something missing.

117

Please provide the following inforrrration. There are no right or wrong answers so try hard
to be completely honest in we answers. RESULTS ARE COMPLETELY
CONFIDENTIAL Read each question carefully and circle the letter under the column

which applies to you. Please answer each question vey carefully.

A=ALWAYS
B=USUALLY
C=OFTEN
D=SOMETIMES
E=RARELY
=NEVER

151. I at sweets and carbohydrates without feeling nervous. A B C D E F
152.Ieatwhenlamupset. A B C D E F
153.1thinkaboutdieting A B C D E F
154. I stuff myself with food. A B C D E F
155. I feel extremely guilty after overeating A B C D E F
156. I have gone on eating binges wheel felt I could not stop. A B C D E F
157. I am terriﬁed of gaining weight. A B C D E F
158. I think about bingeing (overeating). A B C D E F
159. Iexaggerateor magnify tlneimportance of weight. A B C D E F
160. I sat moderately in front of othes and strnff myself when

they're gone A B C D E F
161.1ampreoccupiedwiththedesiretobethinner. A B C D E F
162. I have the thought of trying to vonrit in order to lose

weight. A B C D E F
163.IfIgainapound,Iworrythathillkeepgaining. A B C D E F
164.1eatordrinkinsecrecy. A B C D E F

Please answer the following questions by ﬁlling in the appropriate blank. Please answer as
honestly as You mn- Again. MW

165. Your present weight (in pounds)

 

 

 

 

 

166. Height (specify feet and/or inches)

167. Highest past weight (excluding pregnancies)
How long ago was this? months ago

168. Lowest past weight
How long ago was tlnis? months ago

 

169 What do you consider your ideal weight (in pounds)?

118

PLEASE ANSWER THE FOLLOWING QUESTIONS. ALL RESPONSES ARE
STRICTLY CONFIDENTIAL (you're almost ﬁnished).

A. Have you eye had an episode of eating an amount of food that othes would regard as
(a b rnge) __yes ___no
(IF NO PLEASE SKIP TO 'B' BELOW)

During thebstmmmmohowoftenhaveyw typicallyhadaneating binge?
a. I have not binged m the last 3 months

 

 

 

b. Monthly
I usually binge time(s) a month.

c. Weekly
I usually binge time(s) a week

d. Daily
I usually binge time(s) a day.

At the worst of times, what was your aveage number of hinges per week?
binges pe' week. When was this?

 

B. Haveyoumtriedtovomitafteeatinginordetogetridofthefoodeaten
yes no (IF NO PLEASEGO TO 'C').

Drning the W, how often have you typieally induced vonniting?
a. I have not vomited in the last 3 months

 

b. Monthly

I usually vomit time(s) a month.
c. Weekly

I usually vomit time(s) a week
d. Daily

1 usually vonrit time(s) a day.

 

At the m of tinres, what is the average numbe' of vomiting episodes pe week?
vomiting episodes per week. When was this?

C. Have you ever taken diet pills?
yes ___no

Ifyouhavetakendietpills, dnnin ngthelastﬂmmgnthghowoften
have you typieally taken diet pi.lls
a. Ihavenottakendietpillsinthelast3months

 

b. Monthly
1 usually take diet pills time(s) a montln.

c. Weekly
I usually take diet pills __time(s) a week

d. Daily
I usually take diet pills time(s) a day.

D. Have you ever used laxatives to control your weight or ”get rid of food"?
Yes ___no

 

Dming the W3, how often have you been taking laxatives for weight control?

119

a. Ihavenottakenlaxativesinthelast3months

 

b. Monthly

I usually take laxatives time(s) a month.
c. Weekly

I usually take laxatives time(s) a week
(1. Daily

1 usually take laxatives time(s) a day.

 

E. Have you are taken diuretics (water pills)o to control your weight?

If you have taken diuretics, during the W, how often
have you typically taker dinnetics?
a. I have not taken diuretics 1n the last 3 months.
b. Monthly
I usually take diuretics time(s) a month
0. Weekly
I usually take diuretics time(s) a week
(1. Daily
I usually take diuretics time(s) a day.

 

 

THAT'S ALLI!
THANK YOU VERY MUCH FOR YOUR PARTICIPATION!!!

120

Vignette Respornse Measure (Part 2)
The following statements or items ask for your opinion. There are no right responses. We
onf ' h 1 respond

 

to the items as you feel the woman you just heard would respond. Circle the numbe' to the
left of the item which indicates the frequency with which you believe the peson
experiences the described feeling.

Please complete all the items in the order provided. DO NOT OMIT ANY ITEM.
EXAMPLE:

For instance, the ﬁrst statement is "I feel like I am neve' quite good enough.” Is it your
opinion that the subject of the tape feels this way (l)"never" , (2)“seldom",

(3) "sometimes", (4)"frequently", or (5)”almost always”?

SCALE:

l--NEVER 2--SELDOM 3-SOMETIMES 4-FREQUENTLY
5--ALMOST ALWAYS

12345

D—5
O

I feel like I am never quite good enough.
1 . I feel somehow left out.
1 . I think that people look down on me.
1
. I scold myself and put myself down.

. I feel insecure about othes opinions of me.

NNNNNN
UUWUUM
555558
UIUIUIUIUIUI

2
3
4. Allinall,IaminclinedtofeelthatIamasuccess.
5
6
7

. Compared to other people, I feel like I somehow never
measure up.
8. I see myself as being very small and insigniﬁcant.
9. I feel I have much to be proud of.
10. I feel intensely inadequate and full of self doubt.

FIHHH
NNNN
MUM”
5885
UIUIUIUI

11.1feelasif1am somehow defectiveasapersorn, likethereis
something basieally wrong with me.

1 2 3 4 5 12. WhenIcompare myselftoothersIamjustnotas

important.

HHHH

NNNN
9303030)

5855
UIUIUIUI

13.

14.
15.

16.
17.

18.

19.

20.

21.

22.
23.

26.

27.
28.
29.

121

I have an overpowering fear that my faults will be revealed
in front of others.

I feel I have a number of good qualiﬁes.

I see myself striving for perfection only to continually fall
short.

Itlrinkotlnersareableto seemy defects.

1 could beat myself ove- the head with a club when I make
a mistake.

On the whole, I am satisﬁed with myself.

Iwouldliketoshrinkawaywhenlmakeannistake.

Ireplaypainful eventsove'and overin my mind untilI
am ove'whelmed.

Ifeellamapesonofworthatleastonanequalplanewith
others.

Attimeslfeellikelwill breakintoathousandpieses.

Ifeelas ifI have lostcontrolove' my body functionsand
my feelings.

. Sometimeslfeelnobiggethanapea.
25.

Attimesl feel soexposedtlratlwish theeartlrwouldopen
up and swallow me. I

I lmvethispainful gap within nnethatlhavenotbeen able
to ﬁll.

I feel empty and unfulﬁlled.

I take a positive attitude toward myself.

My loneliness is more like emptiness.

I always feel like thee is something missing.

122

APPENDD( E
Cmsetﬁmnsanﬂnfmationﬁonas
Consent Form I (Screening)
Michigan State Univesity
Department of Psychology

DEPARTMENTAL RESEARCH CONSENT FORM

1. Ihave freely consentedtotakepartinascientiﬁc study being cornductedby Melissa
McCreery under the supervision ofDr. Bertram Karon.

This research will require that I respond to some statements and answer some questions
about myself and about my feelings and expeiences

Participation in this expeiment usually takes approximately one hour. I urnderstand that
Imaybeaskedtoretmrratalatertimetoparticipateinanadditionalonehourexperiment
for additionalresearchcredit.

2. Thestudyhasbeenexplainedtomeandlunde'standtheexplanatimthathasbeengiver
and what my participation will involve.

3. Iunderstandthatlamfreetodiscontinuemyparticipationinthestudyatanytime
witlroutpenalty.
4. Iunderstandthattheresultsofthestudywillbetreatedinstrictconﬁdenceandthatl

will remain anonymous. Within these restrictions, results of the study will be made
available to me at my request.

5. I understand that my participation in the study does not guarantee any beneﬁcial results
to me.

6. I undestand that, at my request, I can receive additional explanation of the study after
my partrcnpatn‘ ' 'on is completed.

Signed:

 

Please print name

Date

123

Information Shwl (Screening)

Thank you for yom' participation. The purpose of this study was to examine diffeences in
a variety of variables such as age, family background, inteests and concerns, eating habits,
and sel ’

orientation. Yonn responses will be kept strictly conﬁdential and will not be
associatedwith yournameinanyway.
Itispossiblethatyouwillbecalledandaskedtoretnu'nforadditionalparticipationatalate
time. Youarenotrequiredtocontinueyourparticipationifyoudonotdesire. Ifyouare
called back and choose toparticipate you will earn additional credit for your time.

If ou have any questions about your participatiorn in this study or would like more
information, you may contact myself or Dr. Bertram Karon at 353-5258.

Melissa McCreery
Dr. Bertram Karon

124

Consent Form 1] (Vignettes)
Miclnigarn State Unive'sity
Departrrrent of Psychology
DEPARTMENTAL RESEARCH CONSENT FORM

1. IhavefreelycornsentedtotakepartinascientiﬁcsmdybeingconductedbyMelissa
McCreey under the supervision of Dr. Bertram Karon.

Thisresearchwillrequirethatlrespondtosomestatementsandanswesomequestions
aboutmyselfandaboutmy feelingsandexperiences. Iwillalsobelisteningtoaudiotapes
of inteviews and giving my opinions about what] thinkthepe'son I heard is like.

Participation in this expeirrnent usually takes approximately one hour.

2. The smdyhasbeenexplainedtomeandlundestandtheexplanationthathasbwn
given arnd what my participation will involve.

3. Inmdestandthatlamfreetodiscontinuemyparticipationinthesundyatanytime
witlroutpenalty.
4. lunderstandtlrattheresultsofthestudywillbetreatedinstrictconﬁdenceandthatl

will remain anonymous. Within these restrictions, results of the study will be made
available to me at my request.

5. I understand that my participation in the study does not guarantee any beneﬁcial results
to me.

6. I unde'stand that, at my request, I can receive additional explanation of the study after
my participation is completed.

Signed:

 

Please print name

Date.

125

Informational Form 1] (Vignettes)

Thank you for your participation. The expeiment you have just completed was a two part
study investigating females' perceptions of interpesonal relationships, especially the
aspects of autonomy and dependency. The purpose of this study is to investigate whether
or not females' eating habits and feelings about their bodies are related to their self
orientation, and to their feelings and peceptions about autonomy and needing other people.
An additional component of the study was designed to examine whethe any of these

dynamicsarealtecdbyrecove'y fromaneating disorde.
Participants for this stu were selected to represent a broad range of eating behaviors.
Participation in this does not mean that your eating behaviors are disordered If you

are concerned about your eating behaviors and attitudes towards food and your body, there
are resources available on eampus. A partial list is included with this form.

If you have any further questions about this study or would like to talk about issues that it
has raised, you may contact myself or Dr. Bertram Karon at the numbes indicated below.

Thank you again for your participation.
Melissa McCreery
Dr. Bertram Karon

 

 

 

$10-30 62 9.1
$30-50 156 23.0
$50-70 185 27.2
$70-90 131 19.3
over $90 120 17.6
don't krnow or missing 26 3.8
Total 680 100.0
Valid Cases: 662

Missing Cases: 18

8 1° .

Religion Frequency Percent
Catholic 28] 41.3
Jewish 34 5.0
Protestant 187 27.5
Othe 94 13.8
No Religious

Afﬁlatiorn 83 12.2
Missing 1 0.1
Total 680 100.0
Valid Cases: 679
W

Percentage of subjects who have been in therapy or counseling = 30.6 (n = 208)

127

Reasons endorsed for seeking Frequency
cournseling or psychotheapy

Problems with alcholrol ll
Anxiety 26
Depression 58
Drugs 9
Eating Disorde 25
Family Problems 128
Marital Problems 1
Other 59

(Subjects endorsed as many reasons as applied)

3

Nu—

ass—I
mO—N8QNUI

hinin'o'w'soia'w

N

 

128

 

 

 

APPENDIX G
Descriptive Information by Group
E I . I 1 .5 . 1 Ci .
Group African Asian Caueasian Latin Othe Row total
Auntie!) Amm'satr
BB n=1 n=1 n=20 n=1 n=0 n=23
4.3% 4.3% 87.0% 4.3% 0.0%
BR n=1 n=0 =28 n=0 n=l n=30
3.3% 0.0% 93.3% 0.0% 3.3%
NED n=2 n=3 n=19 n=l n=0 n=25
8.0% 12.0% 76.0% 4.0% 0.0%
Income (rn thousands of dollars)
91mm 10-30 30-50 50-70 70-90 We 90 Denim
BB n=5 n=2 n=7 n=3 n=5 n=0 1
22.7% 9.1% 31.8% 13.6% 22. 7% 0.0%
BR n=3 n=6 n=5 n=4 n=10 n=2
10.0% 20.0% 16.6% 13.4% 33.3% 6.7%
NE) n=0 n=7 n=8 n=7 n=2 n=0
0.0% 29.2% 33.3% 29.2% 8.3% 0.0%

(Two individuals, 1 BB and l NED did not respond)

E 1' . E E .
Religiorn

G C I 1' I . 1

BB n=10 n=3 n=4
43.5% 13.0%

BR n =10 n=3 n=6
33.3% 10.0%

NED n=7 n=l
29. 2% 4.2%

Pro 01 11 EEﬁl"

 

17. 4%

20.0%

n=10
41.7%

n=4
17.4%

n=4
13.3%

n = 3
12.5%

129
APPENDIX H

Descriptions of group members (BB and BR groups)

BB Group:

Age Current Vomiting Laxative Diet Pills Diuretics Exercise Length of
Bingeing Episodes Use" (monthly) (monthly) frequency“ exercise
(monthly) (monthly) (monthly) (monthly) period

(nninutes)

18.00 8.00 0 .00 24.00 0.00 8.00 15.00 60.00

19.00 8.00 5.00 0.00 0.00 0.00 2.50 90.00

18.00 24.00 2.00 0.00 0.00 0.00 25.00 50.00

18.00 20.00 10.00 0.00 0.00 0.00 12.00 120.00

18.00 16.00 0.00 0.00 0.00 0.00 30.00 90.00

20.00 16.00 4.00 0.00 0.00 0.00 20.00 40.00

18.00 20.00 20.00 16.00 30.00 0.00 100.00 60.00

18.00 8.00 3.00 0.00 0.00 0.00 90.00 120.00

18.00 8.00 2.00 0.00 0.00 0.00 8.00 52.50

18.00 8.00 12.00 0.00 35.00 0.00 5.00 45.00

17.00 8.00 4.00 0.00 30.00 0.00 30.00 60.00

19.00 16.00 3.00 0.00 0.00 0.00 0.00 0.00

18.00 20.00 0.00 5.00 2.00 0.00 8.00 30.00

18.00 8.00 1.00 0.00 0.00 0.00 31.00 45.00

19.00 12.00 1.00 0.00 0.00 0.00 60.00 35.00

18.00 20.00 10.00 25.00 0.00 0.00 25.00 45.00

20.00 16.00 48.00 0.00 0.00 0.00 16.00 120.00

18.00 8.00 2.00 0.00 0.00 0.00 3.00 90.00

20.00 28.00 25.00 0.00 0.00 0.00 0.00 0.00

19.00 8.00 1.00 0.00 15.00 0.00 30.00 37.50

18.00 8.00 12.00 0.00 0.00 0.00 12.00 45.00

18.00 12.00 60.00 0.00 0.00 0.00 12.00 45.00

20.00 12.00 4.00 0.00 0.00 0.00 0.00 0.00

*Laxative use refers to laxatives used for the purpose of weight control.
*Exercise refers to exercise for the purpose of burning oﬂ‘ or " getting rid of' large
quantities of food eaten (or binges).

130

BR Group:
Age Current Vomiting Laxative Diet Pills Diuretics Exercise Length of
Bingeing Episodes Use“ (monthly) (monthly) frequency“ exercise

(monthly) (monthly) (monthly) (monthly) period
18.00 4.00 1.00 0.00 0.00 0.00 5.00 30.00
18.00 4.50 1.00 0.00 0.00 0.00 2.00 120.00
18.00 2.00 1.00 0.00 0.00 0.00 13.00 60.00
18.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18.00 2.00 0.00 0.00 0.00 0.00 20.00 60.00
18.00 0.00 0.00 0.00 0.00 0.00 4.50 45.00
19.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18.00 4.00 1.00 0.00 30.00 0.00 11.50 90.00
18.00 0.00 0.00 0.00 60.00 0.00 3.00 120.00
18.00 4.00 4.00 0.00 0.00 0.00 8.00 25.00
18.00 0.00 0.00 0.00 0.00 0.00 5.00 60.00
19.00 3.00 0.00 0.00 0.00 0.00 1.00 35.00
20.00 2.00 3.00 0.00 4.00 0.00 0.00 0.00
19.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18.00 1.00 10.00 0.00 0.00 0.00 5.00 30.00
18.00 0.00 0.00 0.00 20.00 0.00 4.00 120.00
19.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20.00 1.00 1.00 0.00 0.00 0.00 2.00 120.00
19.00 0.00 0.00 0.00 0.00 0.00 3.00 60.00
18.00 1.00 1.00 0.00 1.00 0.00 1.00 60.00
22.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19.00 4.00 0.00 0.00 0.00 0.00 0.00 0.00
19.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19.00 0.00 0.00 0.00 . 0.00 0.00 0.00
18.00 2.00 0.00 0.00 0.00 0.00 2.00 30.00
19.00 0.00 0.00 0.00 0.00 0.00 60.00 60.00
19.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19.00 3.00 1.00 0.00 0.00 0.00 0.00 0.00
18.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00
19.00 2.00 0.00 0.00 0.00 0.00 3.00 60.00

*Laxative use refers to laxatives used for the purpose of weight control.

131

BR Group:

BMI Real - Bulimia DFT Number of
score Ideal BMI (EDI) (EDI) Therapy Sessions
22.85 .88 .00 6.00 .
28.07 4.95 21 .00 19.00 40.00
20.02 2.40 4.00 21 .00
16.69 .51 2.00 16.00
20.56 1.23 3.00 11.00 .
21.09 2.64 6.00 13.00 60.00
25.03 2.89 1.00 10.00 15.00
19.31 1.96 14.00 19.00
20.02 1.60 1.00 20.00
22.63 3.77 20.00 18.00
25.54 4.26 6.00 11.67
20.98 2.33 3.00 8.00
21.30 2.88 4.00 10.00
20.64 1.65 0.00 5.00 .
19.42 1.63 2.00 2.00 15.00
23.31 3.11 9.00 18.00 61.00
21.97 1.76 1.00 2.00 156.00
19.61 1.51 2.00 19.00 .
18.81 0.80 1.00 7.00 52.00
22.99 1.70 1.00 18.00
23.04 0.54 0.00 0.00 .
27.50 1.49 2.00 6.00 50.00
21.12 1.51 0.00 6.00
17.87 1.55 5.00 1.00 .
20.81 2.60 1.00 14.00 37.50
37.97 17.33 7.00 14.00 .
21.75 2.33 1.00 12.83 30.00
25.03 2.89 6.00 21.00 26.00
23.43 1.87 6.00 11.00 .
27.97 6.68 12.00 20.00 1.00

132

BB Group:

BMI Real - Bulimia DFT Number of

score Ideal BMI (EDI) (EDI) Therapy Sessions
20.97 1.32 10.00 16.00
22.42 2.40 7.00 18.00
20.21 0.88 12.00 19.00
20.21 1.76 18.00 18.00 .

. . 0.00 17.00 4.00
22.65 1.36 7.00 20.00 83.00
20.21 1.76 18.00 21.00 70.00
19.33 0.88 3.00 20.00 52.00
20.32 2.18 13.00 20.00 1.00
24.08 5.44 2.00 19.00 .
20.99 2.25 14.00 21.00 12.00
24.86 6.99 4.00 21.00 2.00
22.68 3.18 20.00 19.00 35.00
21.56 2.81 11.00 14.00
18.11 1.34 3.00 18.00
20.26 0.68 0.00 4.00 .
20.87 0.91 12.00 20.00 260.00
29.80 7.66 14.00 15.00 396.00
19.49 0.36 13.00 21.00 .
19.33 0.88 5.00 20.00 1.00
21.31 1.94 7.00 17.00 .
25.47 5.02 13.00 20.00 1.00
21.14 3.25 7.00 19.00 4.00

LISTOFREFERENCES

REFERENCES

Abraham, S. F. ,Mira, M., & Llewellyn-Jones, D. (1983). Bulimia: A study of

 

Agras, W. S. (1987). W New York: Pegamon Press.

Attic, 1. , & Brooks-Gunn, J. (1989) . Development of eating problems in adolescent
sirls= A longitudinal smdy- W15. 70-79.

Beglirr, S. J., & Fairburn, C. G. (1992). Women who choose not to participate in
sulrgeyson eating disorde's. . .. f - ,12, 113-
1

Berlin, S. , & Johnson, C. G. (1989) . Women and autonomy: Using Structural Analysis
of Social Behavior to ﬁnd autonomy witlnin connections. Psychiatry, 52, 79-95.

Be'nardez-Bonsatti, T. (1978). Women and Anger: Conﬂicts with Aggression in
Contemporary Women. . .. '
31(5), 215-219.

Boskind-Lodahl, M. (1976). Cinderella's stepsistes: A feminist perspective on anorexia
nervosa and bulima. WW 2, 342-356

Boskind-White, M. . & White. W- C. (1983) - W New
York: W.W. Norton and Company.

Boskind-Wlnite, M. ,& White, W. C. (1986). Bulimarexia: A historical-sociocultural

perspective. InK. D. Brownell&J. P. Foreyt(Eds.,) W
W (pp. 353- 366). New York. Basic Books.

Broverman,]. K. ,,Voge1 S. R. ,Broverman, D. M., Clarkson, F. K., & Rosenkrantz,

P. S. (1972). Sexrole stereotypes: Acurnentappraisal. Wages,
28, 59- 78.

 

 

Cattanach, L., & Rodin, J. (1988). Psychosocial components of the stress process in
bulimia. _ -- . _. _ . ,.Z,75-88

Chodorow, N. (1974) . "Family Structure and Feminine Personality." In M.Z. Rosaldo

& 1- IamPhene. (Eds) WW. Stanford: Stanford
University Press.

Chodorow, N. (1978) . Wm. Berkeley: Univesity of
California Press.

 

133

134

Cook, D. R. (1993). WWW Menomonie, WI: Channel
Press. (Available from author: Rt. 7, Box 270A, Menomonie, WI 54751.)

Dickstein, L. J. (1985). Anorexia nervosa and bulimia: A review of clinical nssues.
W in, 1086-1092.

Eikson, E. H. (1963) . W (2nd ed. ) . New York: W. W. Norton.

Fairburn, C. G. & Beglin, S. J. (1990). Studies of the epidemiology of bulimia nervosa.
WW 1.17.401-408

Fossum, M. A. , & Mason, M. J. (1986) . aningjhame. New York: Norton.

Franz, C. E, & White, K. M. (1985). Individuation and attachment in pesonali
development: Extending Erikson' s theory. MW, 53, 22 -.256

Freud, S. (1933). ~ . , - ' (J. StracheyEd.and
Trans.) New York: W. W. Norton and Co.

Freud,S. (1953). ‘ kit .1. r Im r _r our ‘-i- n 4;; rt
Strachey Ed. and Trans.) Volumes 8,9,11, 0. London: Hogarth Press.

Garﬁnkel, P. E., &Garner, D. M. (1983). The multidetermined nature of anorexia
nervosa. InP. L.Darby,P. EGarﬁnkel, D. M. Game, &D. V. Coscina(Eds.),
..- ' .. _. -- 11(ppH3-14) NewYork: Alan

 

  

  

Game, D. M., & Garﬁnkel, P. E. (1980). Socio-cultural factors m the development of
anorexia nervosa. We .19. 647-655

Garner, D. M, Garﬁnkel, P. E, Schwartz, D, & Thompson, M. (1980). Cultural
expectations of thinness in women. W 41, 483-491.

Garner. D. M. , & Olmstead, M. P. (1984) .W
Odessa: Psychological Assessment Resources.

Garner, D. M. ”Olmstead M. P., & Polivy, J. (1983). Development and validation of a
multidimensional eating disorde' inventoryzfolrsaggrexra nervom and bulimia.

Gilligan, C. (1982) . W. Cambridge, MA: Harvard University Press.

Gilligan, C. (1986a). Remapping the moral domain: New images of the self in
relationship. In T. C. Heller, M. Sosna, &. D. E. Wellbery (Eds. ), .
s." '1 91L! ' 1'! 1.: i.'l‘_l 1' l'.l 1'1n‘.‘.-t11">-’-:ll

thought (pp. 37-252) Stanford: Stan ord University Press.

Gilligan, C. (1986b). Exit-Voice dilemmas rn adolescent development. In Foxley, A.,
McPherson, M. S., & O’Donnell, G. (Eds),
. n.(pp 283- 3 ). Notre

 

 
  

Dame: A University of None Dame Press.

Gilligan, c. (1989). n. '

 

Troy, NY: Emma Willard Schoo1.

Haskew,P. ,&Adams,C.H. (1989) . .."
Food. Mission Hills, CA: Glencoe.

Herzog,D B Copeland,P M (1985) Eating Disorders. W
We all, 295'303

Humphrey, L. L. ,Apple, R. F. ,& Kirschenbaum, D. S. (1986). Differentiating
bulimic-anorexic from normal families using mta'personal and behavioral

 

 

 

observational systems.
195.
Humphrey, L. L., & Stern, S. (1988). Object Relations and the family system in bulimia:
A theoretieal integration. » _ f _ r _ . , 15, 337-350.
Johnson, C. & Berndt, D. J. (1983). Preliminary investi aﬁonm of bulimia and life
adjustment. W139

Johnson, C, & Flach, A. (1985). Family characteristics of 105 patients with bulimia.
Aminmalntﬂmhm. 1.41. 1321-1324.

Johnson, C. & .Maddi, K. L. (1986). The etiology of bulimia. Biopsychosocial
273 ,- _. . . . .. _u,_1_3,253-

 

Johnson, C. L. ,Iewis, C. & Hagman, J. (1984). The syndrome of bulimia: Reiview
and Synthesis We. 1. 247-273

Jones, D. M. (1985) . Bulimia: A false self identity. W, 1.3,
305-316.

Jordan, J. V. (1984) . Empathy and Self Boundaries. mm No.46
Wellesley, MA: Stone Cents Working Paper Series.

Jordan, J. V. (1990). Courage 1n connection: Conﬂict, compassion, creativity. mm
m, M. Wellesley, MA: Stone Center Working PaperSerr es.

Kaufman. G- (1992) W (3rd edition) Cambridge:
Schenkman Books.

 

Conference on Re-visiomng Knowledge and the Curriculum: Feminist
Perspectives, East Iansing, M].

Kaufman, G. (1989) . W. New York: Springer.

Keller,M. B. ,,Herzog D. B., Lavori, P. W. ,Bradbum,l. S. ,Mahoney,E M. (1992)
Th.e naturalistic history of bulimia nervosa: extraordinary high rates of chronicity,

1 3 6
relapse. recurrence. and psychosocial morbidity. latemaﬁonallmrrnalnﬂﬁating
mm. 12. 1-9.

Keys, A., Brozch, J., Henschel, A. ,Michelson, 0. ,,Taylor H. L. (1950).]:11;
WW Minneapolix University of Mmem m

laoey, H. (1983). Bulimia Nervosa, binge eating and psychogenic vomiting: a controlled
treatment study and long term outcome. MW, 286, 1609-1613.

Lewis, H. B. (1971) . W New York: International
Universities Press.

Iewis, H. B. (1987a). Introduction: Shame-The ”sleeper" in psychopathology .In H.
r . , . U)! (ppl-28). Hillsdale:

  

Lewis,H. B. (1987b). Shameanddepression. InH. B. Lewis(Ed.), Iheﬁoloof
Wm (pp 2960) Hillsdale: Iawmnce Erlbaum

Lewis, H. B. (1987c) . Shame and the Narcissistic Personality. In D. L. Nathanson
) .W (pp 93432). New York: Guilforcl

Lewis, H. B. (1987d) . WW (Rev. Ed.) . Hillsdale: Lawrence Erlbaum
Associates.

Mahler M S &Furer.M (1968) Lmﬂmnﬁxmmonmdthﬂieisrinm
W. NewYork: BasicBooks.

McBride,P. A., Anderson, G. M, Khart,V. D. ,Sunday,S. R. ,Halmi,K. A. (1991).
We mavensivity in eating disorders MW 21.

McCreery M F (1991) MW
._ ~. Unpublished master's thesis,

 

Michigan State Uversity,Fast Lansing,

Miller, J. B. (1982) . Women and Power. W No._1. Wellesley, MA:
Stone Center Working Paper Series.

Miller, J. B. (1983) . The construction of anger in women and men. Whom
No,_4,. Wellesley, MA: Stone Center Working Paper Series.

Miller, J. B. (1984) . The Development of Women's Sense of Self. mm
No._1_2. Wellesley, MA: Stone Center Working Paper Series.

Miller. J. B. (1986).Imrd.amholosy.of_ﬂomm (2nde¢)-Boston: Beacon
Press.

Mitchell, J. ,.Pyle,R L., Hatsukami, D. Goff, G. Glotter, D. ,&Harper, J. (1988).
Atwo twoto ﬁve $227-$165“ patients treated for bulimia. mmaﬁonalloumalof

137

  

Morrison, A. P. ( 1984). Working with shame in psychoanalytic treatment. Mom:
t. ' .‘Lv I I) ,32, 479-505.

Morrison, A. P. (1987) . The eye turned inward: Shame and the self. In D. L. Nathanson
(FA ) .IheMmEaeesntShame (PP 271-291). New Yorln Guilford

Mom'son. A P. (1989)S_hame:_1heiindetside_o£nardssism New Yorln' Analytic
Press

Mortenson,G. M. ,Hoerr,S. L. ,&Garner,D. M. (1993). Predictorsofbody
satisfactionincollegewomen. . _ - . 2.:12,..,-o123,
1037-1039.

Najjar, M. J. , & Rowland, M. (1987) . . .- - 1
' ' l l (DI-{HS Publication No. 87-1688).
Washington DC: National Carter for Health Statistics

Nathanson, D. L. (1987) . A timetable for shame. In D. I... Nathanson (Ed. ) , Thom
W (pp 1-63). New York: Guilford.

Palla, B. ,& Iatt, I F. (1988). Medieal Complications of eating disorders in adolescents.
Pediatrics 5 613-623

Paxton, S. J. , & Sculthorpe, A. (1991) . Disordered eating and sex role characteristics in
young women: Implications for sociocultural theories of disturbed eating. m
Boles, E. 587-598.

Pearson, J. L., Reinhart, M. A., Strommen, E A. ,Donelson, E, Barnes, C. ,Blank,
L. ,Cebollero,A. m. ,Cornwell,K., Kamptner, N. L. 1(1991). Mano
'-‘1I.‘-Ir-- av) ‘. mm «:I I . ‘11 [In II I II I II I Unpublished

 

 
   

   

Pettinati, H. M. , Franks, V. , Wade, J. H. , & Kogan, L. G. (1987) . Distinguishing the
role of eating disturbance from depression in the sex role self-perceptions of
anorexic and bulimic inpatients W as. 280-282.

Pirke, K. M. ,Friess, E, Kellner, M. B. ,Krieg,J. C., Fichter, M. M. (.1994).
ngmatostatininmtingdisorda's. _r‘ . .. _ _ . .—. g
l

Pirke. K. M., Vandereyckm. w. (an) (1988) .mmmmmmﬂma
Springer-Vedag.

Pope, H. G. ,,Judson J. 1., Jonas, J. M., & Yurgelun-Todd, D. (1985).
Antidepressent treatment of bulimia, a two year follow up study. Moi

W192! 5. 320

PoteI-Park. E & Bokram R L (1989) W (HNF 102)
Fast Iansing: Olin Health Center Health Education Savices, Michigan State
University.

 

138

Pn1itt,J. A. ,Kappius,R. E, Gorman, P. W. (1992). Bulimia and fear ofintimacy.
MW 48.472475

Pugliese, M. T., Lifshitz, F. (1985). Endocrine adaptations to undernutrition. _Qmjcal
Nutrition 4 84--53

Pumariega, A. J. (1986). Acculturation and eating attitudes 1n adolescent girls: A

 

Rand, C. S. ,& Kuldau, J. M. (1992). Epidemiology of bulimia and symptoms in a
general population: Sex, age, ,racg, and socioeconomic status. W
W, 1.1. 7-44

Rost, W. ,Neuhaus, M. ,& Florin, I. (1982). Bulimia nervosa: Sex role attitude, sex
role behavior, and sex role related locus of control 1n bulimarexic women. [cm

Wt: 25, 403-408

Shisslak, C. M. , Crago, M. ,Neal, M. E, & Swain, B. (1987). Primary prevention of
eating disordem WW :5 660-667

Silberstein, L. R. ,Striegel-Moore, R. H. ,& Rodin, J. (1987). Feeling fat: A woman's
sham. 11111 B Lewis (Ed).1heRQle_Q£’§lhame_in_Sxmntom.Ethatinn(pp
51-87). Hillsdale: Lawrence Erlbaum.

Silverstein, B. , Carpman, S. , Perlick, D. , & Perdue, L. (1990) . Nontraditional sex
role aspirations, gender identity conﬂict, and disordered eating among college
women. Moles, 23, 687-695.

Silverstein, B. ,Perdue, L., Peterson, B. ,Kelly, E (1986). The role of the massmedia
in promoting a thin standard of bodily attractiveness for women. $539125, 15,
5 19-5 32.

Silverstein, B. , Perdue, L. , Wolf, C. , & Pizzolo, C. (1988) . Bingeing, purging and
egﬁmates of parental attitudes regarding female achievement W 12, 723-
7 3.

Steiger, H. ,,Fraenkel L., & Leichner, P. P. (1989). Relationship of body-image
distortion to sex-role identiﬁcations, irrational cognitions, and body weight 1n

eating disordered females Won 4.2, 61-65

Steiner-Adair, C. (1986). The body politic: Normal female adolescent development and
the develovment of eating disorders WWW

thnanalxsis 1195-114

‘ Stern. D. N. (1985). W New York: Basic Books.

Striegel—Moore, R. H. ,Silberstein, L. R., & Rodin, J. (1986). Toward an
understanding of risk factors for bulimia. W 51, 246-263.

139

Strober, M., & Humphrey, L. L. (1987). Familial contributions to the etiology and
course of anorexia nervosa and bulimﬁ. mm

mm 25. 654-659

Surrey, J. L. (1984) . Self-in-Relation: A theory of women's development. m
m, No.13. Wellesley, MA: Stone Center Working Paper Series.

Swift,W. J, Kalin, N. H., Wamboldt, F. S, Kaslow, N., & Ritholz,M. (1985).1

Degressron 1n bulimia at two to ﬁve year follow-up. W16 1-
12

Teusch, R. (1988). Level of ego development and bulimics' comWons of their
‘ - - . , Z,

 

Tomkins, S. S. (1963) . -
York: Springer.

Tomkins, S. S. (1979). Script theory: Differential magniﬁcation of affects. In H. E
Howe and R A Dienstbier (Eds ) Won“
(pp. 201-236). Lincoln: University of Nebraska Press.

Tomkins, S. S. (1987) . Shame. In D. L. Nathanson (Ed. ) , W
(pp 133-161). New York: Guilford.

Weltzin, T. E, McConaha, C., McKee, M. ,Hsu,L. K., Perel, J. Ka,y,W. H. (1991)
.Circadian patterns of cortisol, prolactin, and growth hormonal secretion during

bingeing and vomiting in normal weight bulimic patients. W,
39, 37-48.

Winnicott,D. W. (1965). _ .
London: Hogarth Press.

Wurmser, L. (1981) . W. Baltimore: John Hopkins University Press.

 

 

111011an 57an UNIV. LIBRRRIES
illillllllWIINI“IIIHWIIINIWIN‘IHIIIWIH‘I
31293014172732