Journal of Ctmsulting and Clinical I%ychology ~ t 1987 by the American Psycholo8ical Association, Inc.

1987, Vol. 55, No. 6, 889-897 0022-006X/87/$00.75

Cognitive-Behavioral Treatment of Women's

Body-Image Dissatisfaction
Jonathan W. Butters T h o m a s E Cash
Harlem Valley Psychiatric Center Old Dominion University

Body-image dissatisfaction is a problem that affectsa substantial minority ofwomen and cuts across
various diagnostic groups. Collegewomen with a significant level ofbody-image dissatisfaction were
randomly assigned to either a cognitive-behavioral treatment (CBT) program (n = 15) or to a wait-
ing-list control group (n = 16). The CBT program consisted of six structured, individual sessions
that applied cognitive-behavioral procedures to the problem of negative body image. At pretest,
posttest, and 7-week foUow-up,multiple aspects of body image and other areas of psychosocial func-
tioning were assessed. Relative to the control condition, the CBT program successfully improved
affectivebody image, weakened maladaptive body-image cognitions, and enhanced social self-esteem
and feelingsabout physical fitness and sexuality. Treatment effectswere largely maintained at follow-
up. After posttest, the control group received a 3-week treatment with immediate effects that gener-
ally replicated those obtained in the 6-wer program.

A substantial amount of research literature (cf. Cash, 1981,
1985; Hatfield & Sprecher, 1986) bas verified that one's objec-
tive level of physical attractiveness can influence one's psycho-
social experiences and deveiopment. The cumulative adverse
effects of unattractiveness may undermine one's mental health
(Archer & Cash, 1985; Cash, 1985; Farina et al., 1977; Farina,
Burns, Austad, Bugglin, & Fischer, 1986; O'Grady, 1982). How-
ever, individuals' self-perceptions of their physical appearance
bear only modest resemblance to reliable objective perceptions
of their physical attractiveness, and self-perceptions bave even
more salient implications for mental health (Cash, 1985; Noies,
Cash, & Winstead, 1985).
Among clinical and nonclinical samples, both men and
women who perceive themselves as physically unattractive or,
more generaUy, who report negative or dysphoric attitudes to-
ward their physical appearance (i.e., a negative body image)
typically bave been found to have poor self-esteem, social anxie-
ries and inhibitions, sexual difflculties, and a vulnerability to
depression (e.g., Archer & Cash, 1985; Cash, 1985; Cash, Cash,
& Butters, 1983; Cash, Winstead, & Janda, 1986; Lerner, Orlos,
& Knapp, 1976; Noies et al., 1985; Rosen & Ross, 1968; Secord
& Jourard, 1953). Moreover, persons who exhibit disordered
eating behaviors ofien have dysfunctional perceptions of their
physical appearance (Cash & Brown, in press; Garfinkel & Gar-
ner, 1982). The problem of body-image dissatisfaction is actu-
This study was completed by the first author under the supervision of
the second author in partial fulfillment of requirements for the doctor
of psychology degree from the Virginia Consortium for Professional
Psycbology.The autbors wish to thank Timothy Brown, Gary Cundiff,
Lisa Lee, and Julie Shelton for serving as therapists. A previous version
ofthis article was presented at the 1985 convention ofthe Association
for the Advancement of Behavior Therapy.
Correspondence concerning this article sbould be addressed to
Thomas E Cash, Department of Psychology,OId Dominion University,
Norfolk, Virginia 23508.
ally quite prevalent in the general population. A recent large-
sample national survey (Cash et al., 1986) of adolescents and
adults indicated that 31% of female respondents and 24% of
maie respondents reported negative attitudes about their physi-
cal appearance. For both sexes, dissatisfaction was strongest
with weight-sensitive body areas; for example, the midtorso
area elicited dissatisfaction for 57% of the women and 50% of
the men. This survey also round that a negative body image was
clearly associated with lower levels of self-reported psychosocial
well-being. Comparisons with earlier survey data (Berscheid,
Walster, & Bohrnstedt, 1973) have pointed to an increase in the
problem over the past decade (Cash et al., 1986).
Despite the growing prevalence of body-image problems and
their clinical significance, researchers have largely neglected to
develop and evaluate therapeutic interventions for the improve-
ment of a dysfunctional body image. A recent review (Butters,
1985) revealed that among the 18 studies evaluating a wide
range of treatments for the improvement of body image, no
study met ail three methodological criteria of (a) including a
control group, (b) randomly assigning subjects, and (c) collect-
ing follow-up data. Ofthe seven studies that did include a con-
trol group, rive round that the interventions failed to produce
significantly different effects on body-image indices. Moreover,
researchers have often relied on a single measure of body image
and have failed to recognize the multidimensional nature of the
body-image construct (Cash & Brown, in press; Keeton, Cash,
& Lewis, 1987; McCrea, Summerfield, & Rosen, 1982).
Accordingly, the present controlled investigation examined
the effectiveness of treatments designed to ameliorate negative
body image. Because a dysphoric body image is more closely
linked with appearance-related cognitions than with physical
realities (Cash, 1985; Cash et al., 1986), the problem seemed to
be a suitable application of cognitive-behavior therapy. Fur-
thermore, cognitive-behavioral techniques have demonstrated
therapeutic eflicacy for a variety of dysphoric conditions (Kaz-
din & Wilson, 1978; Kendall & Hollon, 1979; Miller & Berman,

889
890 JONATHAN W. BUTTERS AND THOMAS E CASH
1983) known to be associated with a negative body image (e.g.,
depression, interpersonal anxieties, sexual dysfunctions, eating
disorders). Therefore, the present study describes the design and
evaluation o f two brief cognitive-behavioral treatment pro-
grains for body-image dissatisfaction a m o n g young women,
who are especially likely to report a negative body image (Ber-
scheid et al., 1973; Cash et al., 1986; Fallon & Rozin, 1985). A
6-week, individual cognitive-behavioral treatment was com-
pared with a waiting-list control condition whose members
were subsequently treated in an abbreviated 3-week program. A
7-week foUow-up o f the initial treatment was carried out. Focal
effects on multiple body-image parameters, including affective
and cognitive components, and generalization to other aspects
ofpsychosocial functioning were evaluated.
Method
Subjects
Sixteen undergraduate psychology classes containing 186 female stu-
dents were tested using the Body-self Relations Questionnaire (BSRQ;
Winstead & Cash, 1984) and the Symptom Checklist-90-Revised (SCL-
90-R; Derogatis, 1977). ~ A project description and a form to request
body-image therapy wr also included. Women who met ail of the fol-
Iowing inclusionary and exclusionary cr9 were telephoned: (a) a de-
sire to receive body-image therapy; (b) a score below the 25th percentile
ofthe BSRQ Appearance Evaluation subscale norms for college women
(Winstead & Cash, 1984); (c) T scores under 70 on both the Psychoti-
cism and Paranoid Ideation scales of the SCk-90-R (to exclude more
severely disturbed persons); (d) sr weight that was not more
than 25% above or below the median desirable weight for her height
(Metropolitan Life Insurance tables, 1983); (e) denials of a serious medi-
cal condition or of current receipt of counseling or psychotherapy. Of
the 186 women tested, 36 met ail eriteria, and from these 32 were ran-
domly selected.
The subjects' average age was 21.3 years (SI) = 5.0). Thirty subjects
were White, 1 was Black, and 1 was Asian. They were informed that
due to the program's popularity and the number of therapists available,
halfofthe participants would begin therapy in a week and the other half
would start midway through the semester. Subjects then drew lots to
be randomly assigned to one of the two conditions and to one of four
therapists. AU but one subject participated for the study's duration. This
subject, assigned to the experimental group, was dropped from the
study after one session and was referred to a therapist due to reports of
suicidal ideation.
Procedure
Four clinical psychology doctoral students (two men and two women)
served as trained therapists in the study. Therapists did not conduct th•
research assessments and were blind to subjects' data. Each therapist
was assignr to treat 4 subjects in the experimental gronp and 4 subjects
in the treated control group.
Individuat treatment consistr of structured l-hr weekly sessions.
Each session began with the therapist's 10-15-min meeting with the
client to establish rapport, to review any homework 8 the prior week,
and to introduce the session's topics. The client was then escorted to a
soundproof room where ail auditory stimulation could be controlled by
the therapist from an adjacent room. While comfortably reclined in a
chair, the client received an approximate 30-min intervention that var-
ied in focus from session to session. Finally, the subject returned to the
consulting room where the therapist reviewed the highlights of the inter-
vention, elicited from the client what was most personally meaningful,
and assigned a homework task.
Experimental treatment group. In Session 1, clients received infor-
mation concerning the problem ofbody-image dissatisfaction (e.g., pos-
sible causes, prevalence, effects on personal happiness) and the underly-
ing rationale for the treatment program. In the remainder of the session,
audiotaped instructions taught clients progressive muscle relaxation
plus the use of autogenic suggestions and pleasant imagery (Goldfried
& Davison, 1976). The importance of relaxation for managing body-
image dysphoria was emphasized. Each client was given a copy of the
tape and was assigned to practice relaxation daily during the week.
In Session 2, clients constructed a personalized hierarchy of rive body
parts or areas, ranging from those with which the client was moderately
satisfied to those with which she was least satisfied. Following a detailed
script, the therapist then guided the client through imaginal systematic
desensitization (Goldfried & Davison, 1976; Wolpe, 1973) for her anxi-
ety concerning body areas and the body as a whole. The idea of desensi-
tization as a method for taking control (Goldfried & Davison, 1976)
of body-image feelings was emphasized, and daily practice of imaginal
desensitization was assigne&
Session 3 stressed applying the techniques to real-life situations. The
therapist guided in vivo desensitization in which the client Iooked at
herself in a full-length mirror in the soundproof room. The assigned
homework was the practice of in vivo desensitization privately at home,
clothed and unclothed, in front o f a mirror. Also, the client was intro-
duced to the concept of automatic thoughts (Beck, 1976) and was asked
to monitor and record her automatic thoughts as she Iooked at her body
in the mirror.
In Session 4, the focus was on understanding the foie that thinidng
plays in maintaining body-image dissatisfaction. The A-B-C (anteced-
ents--beliefs--emotional consequences) model of Ellis (1962) was ar-
ticulated, and the concepts of irrational beliefs (Eliis, 1962) and cogni-
tive errors (Beck, 1976) were presented on audiotape. Several cognitive
errors regarding appearance were detailed, and clients identified those
that were personally applicable. The therapist taught the client rational
counterarguments to combat her cognitive errors concerning appear-
ance. Homework required the subject to write several counterargu-
ments for the beliefs covered in the session. In addition, the subject was
taught and assigned Beck's (1976) triple-column technique to use
whenever she felt dissatisfied with some aspect ofher appearance: (a) to
monitor negative body-image affect in specific situations, (b) to identify
the automatic thoughts and cognitive errors that occur, and (c) to dis-
pute or correct the thoughts with counterarguments and to monitor the
affective ontcome.
In Session 5, the homework was reviewed and the cognitive therapy
concepts and their practical applications were further elaborated and
rehearsed. An audiotape discussed the notions of engaging in body-re-
lated behaviors that give one a sense ofmastery or pleasure (Beck, 1976)
and viewiag the body hOt only as an aesthetic object but also for its
capabilities (e.~, fimess, sensory experience, health). For homework,
clients completed the Body-Related Activities Inventory (modeled after
the Pleasant Events Sehedule; MacPhillamy & Lewinsohn, 1974), rat-
ing the frequency oftheir involvement in various physical activities and
their feelings of personal mastery and pleasure associated with each.
During the week, clients were assigned at ieast three such activities.
In Session 6, after the homework was reviewed, an audiotape reviewed
the skills taught in the program plus several strategies for posttreatment
maintenance of a more positive body image. These included the use of
stress-inoculation techniques (Meichenbaum, 1985) and relapse-pre-
vention strategies (Marlatt & Gordon, 1985) when confronted with an
Additional information and materials concerning treatment pro-
grams and assessment instruments are available by request from the
second author.
 BODY-IMAGE TREATMENT
apparent failure to cope with negative body-image thoughts and feel-
ings. FinaUy, the therapist and client reviewed her progress in the pro-
grain and discussed coping strategies tailored to her needs.
The high degree of structure in the program (i.e., liberal use of audio-
taped interventions, written handouts for clients, plus specific written
scripts for each session and each therapist) assured the delivery of a
consistent treatment package across clients and therapists. Each client
attended ail six sessions. Checks on the completion oftherapeutic home-
work assignments further verified clients' treatment compliance.
Waiting-list control (WLC) group. Control subjects were pre- and
posttested on the same schedule as the CBT group and began treatment
a week alter the second assessment. This treated waiting-list control
group (TWLC) received an abbreviated, 3-week treatment rather than
the full CBT treatment because of insuflicient time in the semester to
conduct two 6-week treatments plus assessments. FoUowing their sec-
ond assessment session, these individuals were asked to spend lime prior
to their first treatment session looking at themselves in the mirror, rating
their level of distress, and monitoring automatic thoughts as explained
in a handout. Each TWLC client received an extended 90-min flrst ses-
sion in order to prefacr the initial cognitive treatment session with the
audiotape of the program's rationale. Other than these differences and
a change in what was reviewed during the final session (because the
TWLC group had not received relaxation training or desensitization),
the t h r ~ TWLC sessions were idenlical to Sessions 4, 5, and 6 for the
CBT group.
Assessment Instruments a n d Measures
Ail subjects were assessed during Weeks 1, 8, and 13. Thus, the CBT
and TWLC groups' posttests were held 2 wer after the final treatment
session (to reduce demand eharacteristics); and follow-up assessment
was held 7 weeks afler the CBT group's final treatment session. AH as-
sessments afler mass testing were conducted on an individual basis by
the first author, who was n o t a therapist in the study. Honest responding
was stressed, and subjeet code numbers were used. Subjects were in-
formed that the therapists would not see these data.
Body--SelfRelations Questionnaire (BSRQ). The BSRQ (Cash et al.,
1986; Winstead & Cash, i 984) contains 140 items rated on 5-point Lik-
ert-type scales. Itis an attitudinal measure of body image that yields
data on three attitudinal dimensions (Evaluation, Attention/Impor-
tance, Activity) for three different somatic domains (Appearance, Fit-
ness, Health) plus a Sexuality Evaluation score. The four subscales per-
tinent to the study's hypothesized effects and their internal consisteneies
(Cronbach's alphas) in the present study were Appearanee Evaluation
(18 items), 0.82; Appearanoe Attention/Importance (20 items), 0.89;
Fitness Evaluation (19 items), 0.85; and Sexuality Evaluation (7 items),
0.70. Evaluation scales are scored in the positive direction; higher Ap-
pearance Attention/Importance scores reflect greater cognitive focus on
one's physieal appearance. Previous research has attested to the tempo-
rai stability and the convergent, discriminant, and construct validity of
these BSRQ subscales (e.g., Cash & Green, 1986; Cash et al., 1986;
Noies et al., 1985; Winstead & Cash, 1984).
Body Parts Satisfaction Scale (BPSS). The BPSS (Berscheid et al.,
1973; Bohrnstedt, 1977) is a 25-item measure ofbody satisfaction that
yields six factor seores relating to satisfaction with particular groups of
body parts as weU as a single overall appearance satisfaction rating. In
addition, an internally consistent (alpha = 0.86) total body parts satis-
faction score is obtained by computing the mean of ail 6-point ratings
(from extremely dissatisfied to extremely satisfied).
Mirror distress rating. This is a measure deveioped specifically for
this study to provide a simple, face-valid measure of body-image dis-
tress. With the examiner absent, the subject was instructed to gaze at
herself in a full-length mirror for 89min and then to rate the level of
anxiety experienced on a scale of subjective units of distress (SUDS;
891
Woipe & I.azarus, 1966) from 0 to 100 (100 = extreme distress). She
was instructed to not fixate her gaze on any one body feature but to
look at ail ofher body features.
Photo self-ratings. Color photographs of individual subjects were
taken at the first assessment session using a Polaroid camera, with each
subjeet posed with a smile for a standard frontal, fuU-body photograph.
At eaeh assessment, each subject viewed her (pretest) photograph and
rated how attractive, on a 9-point scale (9 = extremelyphysically attrac-
tive), she expeeted peers would rate her photograph. This measure (Self-
PA) refleets subjeets' beliefs about how others view their physical attrac-
tiveness.
Observers" ratings of physical attractiveness. To determine how oth-
ers actually rated subjects' attractiveness, the four therapists, the experi-
menter, and the second author independently rated subjects' photo-
graphs using the same 9-point scale of physical attractiveness. This
truth-by-consensus method is the standard operational definition used
in physical attractiveness research (Cash, 198 l; Hatfield & Sprecher,
1986). Only the experimenter had met subjects at the lime ofhis ratings,
but he had not seen subjects' Self-PA ratings. The Cronbach's alpha for
judges' ralings was 0.88, indicating high interrater reliability.
Body Image Detection Device (BIDD). This overhead projector ap-
paratus provides an inexpensive and simple measure of an individual's
estimation ofthe width ofher body at different points (sec Ruff& Bar-
rios, 1986). When these estimations are compared to the actual widths,
a measure of body-size estimation accuracy is obtained. In addition, a
measure is obtained of how the subject judges her widths to compare
with those of most women her age and height (the Subjective Rating
Index, SRI). In the absence of available interjudge reliability data for
actuai body widths in this study, however, only the SRI is reported. For
this assessment, the subject rated her subjective appraisal of the width
of her face, chest, waist, hips, and thighs on scales of 0 (grossly br
the normal physical dimension given her height, age, and gender) to 100
(grossly above the norm), with 50 representing normal physical dimen-
sions. The SRI measure is caiculated for each body area as (subjective
ratingJ50) • 100 and is averaged across the rive areas (Ruff& Barrios,
1986). Thus, the SRI taps a cognitive aspect of body image--beliefs
about size normativeness.
PersonalAppearance Beliefs Test (PABT). This instrument contains
rive items developed for the present study to measure the extent to which
individuals subscribe to presumably dysfunctional beliefs about their
appearance. Items were constructed to conform to the specific cognitive
errors described by Beck, Rush, Shaw, and Emery (1979) or to other
cognitive processes known to diminish body image (Cash et al., 1983).
Subjects made 5-point Likert-type ratings of the extent of their agree-
ment with these statements: (a) "If there are people around who are
more attractive than I, then I don't feel attractive" (social comparison);
(b) " I f I feel unattractive, it usually means that I don't look attractive"
(emotional reasoning); (c) "IfI'm not beautiful, then I must be unattrac-
tive" (dichotomous thinking); (d) "If there's one flaw in my appextrance
then I don't feel attractive" (magnification); and (e) "How I feel about
myself is usually related to how 1 feel about my appearance" (selective
abstraction).
Texas Social Behavior Inventory (TSBI). This instrument consists
of 16 items rated on a 5-point Likert scale (Helmreich & Stapp, 1974).
It produces a single, internally consistent score for social self-esteem
(y = 0.91 ), with higher scores reflecting self-perceived confidence, social
dominance, and social compr
SCL-90-R. The SCL-90-R as a 90-item self-report inventory ofpsy-
chological symptoms (Derogatis, ! 977). Each item is rated on a 5-point
seale to indicate the distress occurring over a specified period. Although
this instrument can be seored for 10 symptom dimensions, a Global
Severity Index (GSI) can be computed as a mean rating across symp-
toms to provide a composite, internally consistent index of general psy-
chopathoiogy. At the first ~ m e n t , subjects rated their symptoms
892 JONATHAN W. BUTTERS AND THOMAS E CASH
over the past month, and T scores greater than or equai to 70 on the
Psychoticism and Paranoid Ideation scales were used to screen out se-
verely disturbed volunteers. Subsequently,the SCL-90-Rwas completed
in reference to a l-week period, and GSI scores were computed.
ClientSatisfactionQuestionnaire(CSQ). The 8-item CSQ is accepta-
bly reliable and valid as a general measure of client satisfaction with
health and human services (Larsen, Attkisson, Hargreave~ & Nguyen,
1979) and is scored with the composite of 4-point Likert-type scales
with higher scores indicating greater satisfaction. The item, "Ifyou were
to seek help again, would you corne back to out program?" was deleted
because the program was being offered only once. Cronbach's alphas
were 0.92 for the CBT group and 0.77 for the TWLC group.
Results
Outcome variables were organized into three conceptual
clusters: measures regarding affective evaluations of appear-
anee, measures pertaining to appearanee-related eognitions,
and measures related to other body-image domains or areas of
psychosocial functioning. The general strategy for data analysis
within each eluster involved 2 (CBT vs. WLC) • 2 (pretest vs.
posttest) multivariate analyses of variance (MANOVAS using
Wilks's lambda criterion) that were followed, where appropri-
ate, by univariate analyses of variance (ANOVAS) to examine
significant simple effects. These analyses permitted evaluation
of any pretest differences between groups and of pre- to posttest
changes for both treatment and control groups. With pretest
scores as covariates, analyses of covariance (ANCOW,s) were
used to test between-groups differences at posttest. One-way re-
peated-measures ANOVASfollowed by Newman-Keuls analyses
evaluated the follow-up maintenance of any significant treat-
ment changes. 2
C B T Effects on Body-Image Affect
Four measures of body-image affect were analyzed to deter-
mine the CBT's effeetiveness in altering clients' body-image dis-
satisfaction: the BSRQ Appearance Evaluation subseale, the
BPSS rating of appearanee satisfaction, the BPSS eomposite
score for body parts satisfaction, and the mirror SUDS rating.
Comparing the CBT and WLC groups at pre- and posttest, the
2 • 2 MANOW, yielded a significant (p < .003) Group • Time
interaction, which was also significant for each dependent vari-
able. Simple effects analyses revealed no between-groups
differences at pretest and no pre- to posttest changes for the
WLC group.
As shown in Table 1, the CBT produced significant pre- to
posttest increases on BSRQ Appearance Evaluation (p < .002),
BPSS appearance satisfaction (p < .001), and BPSS body parts
satisfaction (p < .001) as well as significant reductions in dis-
tress on the mirror SUDS ratings (p < .001). Between-groups
ANCOVAcomparisons at posttest indicated that thesr improve-
ments produced significant (p < .005) differences in relation to
the WLC group.
Similar multivariate and univariate analyses were performed
to ascertain effects for the six BPSS factor scores: Face, Extremi-
ries, Lower Torso, Midtorso, Height, and Breasts. The results
indicated significant increases in satisfaction for the CBT group
on ail factors (p < .005) except Height. Relative to the WLC
group at posttest, CBT4reated clients reported greater satisfac-
tion with Face (p < .001), Extremities (p < .001 ), Breasts (p <
.001), Midtorso (p < .001), and Lower Torso (p < .006). Thus,
with the exception of the Height factor (whieh was rarely a
source of body-image dissatisfaction), the CBT program en-
hanced clients' reported satisfaction with ail bodily regions.
Analyses ofthe 7-week foUow-up data for the CBT group (sec
Table 1) were conducted to evaluate the maintenance of treat-
ment gains on body-image affect. As predicted, ail one-way, re-
peated-measures ANOVAS were significant (ps < .001), and
Newman-Keuls eomparisons revealed differences between pre-
test and follow-up means but not between posttest and follow-
up means on each ofthe four body-image affeet measures. Sim-
ilarly, treatment effeets were maintained for the BPSS factor
stores.
C B T Effects on Body-Image Cognitions
Several ofthe outcome variables were essentially cognitive in
nature, including (a) beliefs about how physically attractive one
is to one's peers (Self-PA); (b) appraisals of the normativeness
of one's body size (BIDD SRI); (c) cognitive focus (attending
to, thinking about, and placing importance) on physical appear-
ance (BSRQ Appearance Attention/Importance); and (d) cog-
nitive errors concerning personal appearance (the rive PABT
items). Means, standard deviations, and effects are presented in
Table 2.
For the MANOVA and for each of the eight measures, the
Group • Time effect was significant (p < .03 t o p < .001). Sim-
ple effects analyses indicated that, although there were no pre-
to posttest changes for the WLC group, the CBT group changed
significantly on ail measures. Following treatment, CBT clients
believed that others would find them more attractive (p < .002),
appraised their body size as smaller and doser to the norm
(p < .03), reported less cognitive focus on appearance (p < .03),
and espoused each cognitive error to a weaker extent (i.e., social
comparison, p < .001; emotional reasoning, p < .009; dichoto-
mous thinking, p < .02; magnification, p < .003; selective ab-
straction, p < .002). Comparisons of CBT and WLC groups at
posttest were significant for all these cognitive variables, except
for a marginal effeet (p < .07) on dichotomous thinking.
Additional analysis was carried out to examine the veridical-
ity of clients" judgments of their physical attractiveness in rela-
tion to actual observer ratings. Groups did not differ on observ-
er-rated attractiveness, with CBT subjects averaging 5.2 (SD =
1.0) and WLC subjects averaging 5.0 (SD = 1.4). Discrepancy
scores (PA discrepancy) were calculated by subtracting the ob-
server rating from the Self-PA rating and conducting a 2 • 2
ANOVA, which produced a significant Group • Time interac-
tion (p < .001). The two groups did not differ significantly at
pretest, when they underestimated their attractiveness (Ms =
- 0 . 2 0 and - 0 . 4 0 for CBT and WLC groups, respectively), but
differed significantly at posttest (p < .02), with no change for
the control subjects (M = -0.52) and a significant shift to over-
estimation (M = 1.33) for treated subjects (p < .002).
2 Initial data analyses included the therapist factor as part ofthe de-
sign. However,becausr the multivariate main and interactive effects of
therapists were consistently nonsignificant, the reported analyses were
collapsed across this factor.
 BODY-IMAGE TREATMENT 893

Table 1
Affective Body-Image Means, Standard Deviations, and Treatment Effects
Assessment time"

1 2 3
Intergroup Intragroup
Measure and condition M SD M SD M SD comparisonsb comparisonsc

BSRQ Appr Evaluation 19.97"**

CBT 2.52 0.37 3.33 0.62 3.24 0.63 16.30***
WLC/TWLC 2.43 0.53 2.50 0.59 3.15 0.54 53.80***
BPSS Appearance Satisfaction 17.24***
CBT 2.83 0.96 4.27 0.80 3.93 0.88 22.25***
WLC]TW'LC 3.00 1.21 3.06 1.18 4.06 1.06 40.00***
BPSS Body Parts Satisfaction 25.81 ***
CBT 3.50 0.57 4.39 0.50 4.28 0.54 24.25***
WLC/TWLC 3.64 0.63 3.66 0.44 4.23 0.66 45.37***
Mirror SUDS rating 12.75"*
CBT 54.1 25.8 19.2 13.6 23.8 19.8 14.59"**
WLCfrWLC 54.2 23.2 43.6 18.9 26.8 20.3 7.64*

Note. BPSS = Body Parts Satisfaction Scale; BSRQ = Body-Self Relations Questionnaire; CBT = cognitive-behavioral treatment; SUDS = subjec-
tive units ofdistress; TWLC = treated waiting-list eontrol; WLC = waiting-fist control.
a For the CBT group, Assessments 1, 2, and 3 were pretest, posttest, and follow-up. For the WLC group, Assessments 1 and 2 were pre- and posttest
and Assessment 3 was posttest for the treated control subjects, b Comparisons between CBT and WLC groups at posttest were adjusted for pretest
stores (df= 1, 28). c For the CBT group, pre- and posttest and follow-up were compared (df= 2, 28). For the TWLC group, pre- and posttest were
compared ( d f = 1, i 5).
*p<.05. **p<.Ol. ***p<.O01.

Repeated-measures analyses that included follow-up data in-
dicated that ail significant outcomes were maintained, without
further gains or subsequent losses, on the cognitive variables,
except for a decline (p < .01) in Self-PA rating (and in the auto-
correlated PA discrepancy index). Although hOt significantly
different from baseline, Self-PA ratings were somewhat more
accurate at follow-up ( M = 0.20) than at the posttest level of
overestimation.
Generalization of CBT Effects
Four measures concerned the generalization of treatment
effects beyond appearance-related parameters to self-evalua-
tions of physical fitness, sexuality, social self-esteem, and gen-
eral psychosocial functioning. Respectively, these effects were
examined on the BSRQ Fitness Evaluation and Sexuality Eval-
uation subscales, the TSBI, and the SCL-90-R (see Table 3). Be-
cause of differences in the administration of the SCL-90-R at
pretest screening and other assessments, comparisons between
these particular administrations could not be made.
A multivariate Group x Time effect was significant (p < .03),
and 2 • 2 ANOVASyielded significant interaction effects on Fit-
ness Evaluation (p < .002), Sexuality Evaluation (p < .001),
and TSBI (p < .005). Analyses of simple effects revealed sig-
nificant pre- to posttest differences for the CBT group but not
for the WLC group on Fitness Evaluation (p < .001), Sexuality
Evaluation (p < .001), and TSBI (p < .001). Between-groups
posttest comparisons were also significant for Fitness Evalua-
tion (p < .006), for Sexuality Evaluation (p < .003), and for
TSBI (p < .02). Groups failed to differ on GSI posttest scores
o f the SCL-90-R.
Maintenance of the three observed effects for the CBT group
was supported at foUow-up, when stores differed from pretest
levels (ps < .01) but did not differ from posttest. No change
from posttest was evident for the GSI measure.
Client Satisfaction With the CBT Program
On the 7-item CSQ, the CBT group rated the program with
a mean of 3.5 (SD = 0.5) on a 4-point scale. Specifically, 93%
rated the services good to excellent, 100% definitely or generally
received the services desired, 80% indicated their needs were
met, 87% would recommend the program to friends, 87% were
satisfied with the help received, 93% felt they were dealing more
effectively with their body-image problems, and 100% reported
overall satisfaction with the service they obtained. Client satis-
faction was maintained at follow-up.
Effects of Treatment of the Control Group
The abbreviated treatment of the control group (i.e., the
TWLC group) was evaluated for each set of measures by one-
way MANOVASand ANOVAS comparing clients at pretest (sec-
ond assessment) versus posttest (third assessment) and compar-
ing posttests ofthe two treated groups. Regarding the latter, no
posttest differenees between CBT and TWLC groups were
found. However, the 3-week treatment did produee significant
multivariate changes in each set of measures.
For body-image affect (see Table 1), ail four measures re-
flected improvement: BSRQ Appearance Evaluation (p <
.001), BPSS appearance satisfaction (p < .001), BPSS body
parts satisfaction (p < .001), and mirror SUDS (p < .02). The
brief treatment increased satisfaction on rive of the six BPSS
factor scores (ps < .01) and showed a nonsignifieant trend for
the Height factor (p < .09). On the cognitive measures (see Ta-
ble 2), the TWLC group showed several signifieant changes in
894 JONATHAN W. BUTI'ERS AND THOMAS F. CASH

Table 2
Cognitive Body-Image Means, Standard Deviations, and Treatment Effects
Assessmenttime"

1 2 3
Intergroup Intragroup
Measure and condition M SD M SD M SD comparisonsb comparisonsc

Self-PArating 24.61 ***

CBT 5.00 1.31 6.53 1.12 5.40 1.92 8.56**
WLC/TWLC 4.56 1.71 4.44 1.90 5.00 1.75 1.06
BIDD SRI 6.34*
CBT 127.4 15.6 114.9 13.2 119.9 18.3 6.11"*
WLC/TWLC 124.8 25.6 124.1 21.5 116.7 18.3 10.42"*
BSRQ Attention/Importance 8.94**
CBT 4.20 0.36 3.93 0.42 4.01 0.41 4.02*
WLCfrWLC 4.32 0.39 4.34 0.39 4.08 0.36 10.48"*
PABT SocialComparison 15.31"**
CBT 4.20 0.78 2.53 0.92 3.13 1.06 18.76***
WLCfrWLC 4.31 0.87 3.81 0.83 3.00 0.97 18.78"**
PABT Emotional Reasoning 13.43**
CBT 4.00 1.07 2.93 1.03 3.07 1.03 6.49"*
WLCfrWLC 3.88 0.88 4.00 0.63 3.00 1.03 20.00***
PABT Diehotomous Thinking 2.24
CBT 3.00 1.20 2.00 0.93 2.27 1.34 4.34*
WLC/TWLC 2. i 9 1.11 2.12 0.88 1.94 0.85 1.90
PABT Magnification 8.84**
CBT 3.40 1.18 2.20 0.68 2.33 0.90 8.38***
WLC/TWLC 3.19 1.11 3.12 0.96 2.62 1.02 3.75
PABT SeleetiveAbstraction 9.92**
CBT 4.33 0.62 3.27 0.96 3.60 0.74 9.57***
WLCfrWLC 4.19 0.54 4.25 0.78 3.75 0.93 2.50

Note. BIDD SRI = Body Image Detection Deviee SubjectiveRating Index; BSRQ = Body-Self Relations Questionnaire; CBT = cognitive-behav-
ioral treatment; PABT = Personal Appearanee BeliefsTest; PA = physical attractiveness; TWLC = treated waiting-listcontrol; WLC = waiting-list
control.
"For the CBT group, Assessments 1, 2, and 3 were pretest, posttest, and follow-up. For the WLC group, Assessments I and 2 were pre- and posttest
and Assessment3 was posttest for the treated control subjects, bComparisons between CBT and WLC groups at posttest were adjusted for pretest
stores (df= 1,28). c For the CBT group, pre- and posttest and follow-upwere compared (df= 2, 28). For the TWLC group, pre- and posttest were
compared (df= 1, 15).
*p<.05. **p<.01. ***p<.001.

the predicted direction: BIDD SRI (p < .007), BSRQ Appear-
ance Attention/Importance (p < .007), and PABT social com-
parison and emotional reasoning items (ps < .00 I), with a trend
shown for the magnification item (p < .07). No significant
changes were found for Self-PA ratings or the remaining two
PABT items. Among the four generalization variables (see Table
3), the TWLC group improved significantly on each: Fitness
Evaluation (p < .001), Sexuality Evaluation (p < .01), TSBI
(p < .00 I), and the GSI ofthe SCL-9O-R (p < .008).
The CSQ scores reflected a high level of client satisfaction
with the 3-week program (M = 3.5, SD = 0.4) that was very
similar to that for the longer CBT program. Ninety-four percent
of the TWLC participants indicated the program met most or
ail of their needs.
Discussion
The results confirm the therapeutic efficacy of cognitive-be-
havioral procedures for altering the dysfunctional affective and
cognitive aspects ofa negative body image. Relative to their own
pretreatment levels as wcll as to the untreated control subjects,
the CBT clients made significant gains on the multiple measures
of their presenting target complaint--feelings of unattractive-
ness and dissatisfaction with their physical appearance. As a
result of the treatment program, participants reported more
positive evaluations of their appearance, feelings of greater sat-
isfaction with various aspects oftheir physical appearance, and
less discomfort when privately viewing themselves in a mirror.
These outcomes were maintained approximately 2 months af-
ter the final therapy session.
Consistent with a cognitive interpretation ofbody-image dis-
satisfaction (e.g., Cash & Butters, in press; Fallon & Rozin,
1985; Gainer & Bemis, 1982; Noles et al., 1985), the treatment
program had a significant impact on several cognitive parame-
ters of body image. Following treatment and at follow-up, cli-
ents judged their body size to be smallcr and closer to the norm,
whereas before therapy they believed they wcre over 25% wider
than peers their age, gender, and height. 3 The CBT program also
significantlyreduced clients' extremely high level of preoccupa-
tion with appearance, as reflected by reductions in BSRQ Ap-
3Although women typically appraise their body size as somewhat
wider or heavier than the norm of their actual size (Cash & Brown, in
 BODY-IMAGE TREATMENT 895

Table 3
Generalization Variable Means, Standard Deviations, and Treatment Effects

Assessmenttime"

1 2
Intergroup Intragroup
Measure and condition M SI) M SI) M SD comparisons b comparisonsr
TSBI Self-Esteem 7.31"
CBT 31.3 9.2 40.3 8.2 38.7 9.4 16.43"**
WLC/TWLC 34.9 5.1 36.4 3.8 42.2 5.5 22.88***
SCL-90-R GSI
CBT B
0.55 0.43 0.48 0.34
WLC/TWLC m
0.64 0.41 0.44 0.23 9.93**
BSRQ Fitness Evaluation 9.68**
CBT 2.72 0.65 3.34 0.60 3.33 0.60 20.75***
WLC/TWLC 2.89 0.46 3.01 0.45 3.35 0.37 22.41 ***
BSRQ SexualityEvaluation 11.57"*
CBT 3.14 0.56 3.81 0.70 3.70 0.67 13.17"**
WLC/TWLC 3.39 0.57 3.48 0.50 3.70 0.52 9.14"*

Note. BSRQ = Body-Sr Relations Questionnaire; CBT = cognitive-behavioral treatment; GSI = Global Severity Index; SCL-90-R = Symptom
Checklist-90-Revised; TSBI = TexasSocial Behavior Inventory; TWLC = treatr waitin_g-listcontroi; WLC = waiting-listcontrol.
a For the CBT group, Assessments 1, 2, and 3 were pretest, posttest, and follow-up. For the WLC group, Assessments 1 and 2 were pre- and posttest
and Assessment3 was posttest for the treated control subjects, b Comparisons betwer CBT and WLC groups at posttest were adjusted for pretest
scores (dr= 1, 28). r For the CBT group, pre- and posttest and foUow-upwere compared (dr= 2, 28). For the TWLC group, pre- and posttest were
compared (df= 1, 15).
*p<.05. **p<.Ol. ***p<.O01.

pearance Attention/Importance subscale scores from the 70th
percentile at pretest to the 50th percentile at follow-up. More-
over, after treatment and at follow-up, CBT clients were less
likely to endorse a variety of dysfunctional or erroneous beliefs
about physical appearance that paraUel more general cognitive
errors formulated by Beck et al. (1979). The only cognitive out-
corne not maintained at follow-up was clients' judgments of
how attractive peers would rate their photo. Participants under-
estimated their actual physical attractiveness before treatment
and greafly overestimated it immediately after CBT. At follow-
up, clients' estimations were more accurate albeit not signifi-
cantly different from pretreatment levels. The accuracy of self-
estimates of attractiveness, although associated with a variety
of affective states, may be a more fragile, less stable measure
than other body-image indices (Cash, 1985; Cash et al., 1983;
Noies et al., 1985). 4
The treatment program exerted generalized effects beyond
the focal outcomes on appearance-related aspects of body im-
age. Treatment enhanced clients' social self-esteem, improved
their sexual interest and feelings, and promoted more positive
evaluations of physical fitness. Examination of these changes
in relation to local norms for college women (Butters, 1985;
Winstead & Cash, 1984) indicated that clients became quite
similar to their average peer on each oftbese measures.
With several minor variations, the abbreviated cognitive-be-
havioral treatment of control subjects generally replicated the
outcomes produced by the 6-week program, supporting the
effectiveness of cognitive-behavioral therapy in the ameliora-
tion of a negative body image. As consumers, participants re-
ported a high level of satisfaction and goal-attainment with each
program. Any specific comparisons of outcomes for the two
treatments must be ruade cautiously due to differences in the
components, duration, and time ofyear oftreatments; the num-
ber of pretreatment assessments; and the experience of thera-
pists in conducting each program. However, the combination of
the 6-week CBT program desensitization procedures with the
largely cognitive interventions of the 3-week program make the
conclusions of a recent meta-analytic review noteworthy (Ber-
man, Miller, & Massman, 1985): that cognitive and desensitiza-
tion treatments are roughly equal and not additive in efi9
Of course, differential effectiveness may depend on the particu-
lar problems treated, and none ofthe 25 meta-analyzed studies
concerned treatment ofbody-image dissatisfaction. Until com-
parative, dismantling research is conducted (Kazdin, 1983) that
equates groups on ail variables except the components ofstudy,
the tentative conclusion from the present study is that both vari-
ations of the treatment are effective. Future researchers should
evaluate the CBT's therapeutic efl9 in relation to other in-
terventions that serve as better controls than a waiting-list group
for nonspecific and demand-characteristic effects (Kazdin,
1983; Kendall & Norton-Ford, 1982). Self-report assessments
should be supplemented by the development of unobtrusive be-
havioral and Ix'er-report indices of body image. Longer term

prr Cash et al., 1986), participants in the present study judged their 4 With respect to participants' observer-rated physical attractiveness,
size as much wider at prctest than the average appraisals for samples of it is noteworthy that they were by no means uniformly unattractive.
normal control subjectsas wellas widcr than bulimics described by Ruff Among these women who sought therapy to feel more attractive, 55%
and Barrios (1986) or binge eatcrs describr by Koeton et al. (1987). were actually rated as slightlyto very attractive.
896 JONATHAN W. BUTTERS AND THOMAS E CASH
follow-up will also be important in future replications and re-
finements.
The quasidinical nature ofthis investigation also necessitates
caution in the extrapolation offindings to various patient popu-
lations. Although clients were volunteers whose scores on the
BSRQ and otber baseline body-image measures reflected sub-
stantial unhappiness with appearance, 5 in actual dinical set-
tings, body-image problems are most often presented as a facet
of the phenomenology or symptomatology of other psychologi-
cal disorders (Cash, 1985). Accordingly, one imperative direc-
tion for future clinical research is to consider the utility of the
cognitive-behavioral body-image interventions as a component
of a broader therapeutic regimen for affective, anxiety, sexual,
and eating disorders. For example, becausr residual body-image
disturbances are predictive of relapse in the treatment o f an-
orexia and bulimia nervosa (Cash & Brown, in press; Garner &
Bemis, 1982; Garner & Garfinkel, 1981), body-image t r o t -
ment may contribute significantly to clinical outcomes with
these patients. Although several clinical researchers have incor-
porated procedures for altering a dysfunctional body image in
their eating-disorder treatment programs (e.g., Garner, 1986;
Weiss, Katzman, & Wolchik, 1985; Wooley & Kearney-Cooke,
1986), the specific incremental efficacy of these procedures
awaits controlled investigation.
Tbe scientific study ofhow best to help the diverse individuals
who surfer from distressing thoughts and feelings of physical
unattractiveness is one of considerable opportunity. In record
numbers today, persons consult plastic surgeons for anomalies
of physical appearance that range from major congenital or
traumatic disfigurements to minor deviations from aesthetic
norms to imagined defects in appearance (Cash & Horton,
1983; Goin & Goin, 198 l). Cancer treatments can bave appear-
ance-altering effects that adversely affect body image and t r o t -
ment compliance (Mulready & Lamb, 1985). For many people,
the pursuit of strategies for weight loss is perhaps less for the
purpose ofshedding pounds than for shedding social stigma and
a negative body image (Cash, 1985). Our hope is that the present
investigation and its findings will encourage much-needed re-
search on a problem that is far-reaching.
In fact, subjects"pretest scooeson the BSRQ Appearance Evaluation
subscale and on BPSS measures were approximately 1 standard dr
tion Iowerthan means obtained locallyfor moderately depressed college
students (Noies et al., 1985) and for binge eaters (Keeton et al., 1987).
Subjects' BSRQ prr162 mean stores placed them at the 7th percentile
for collr women. Self-r was at the ! 6th percentile on the TSBI
(Butters, 1985), and global psychological symptomatology on the SCL-
90-R was about t standard deviation higher than the norm for nonpa-
tient women (Derogatis, 1977).
References
Archer, R. P., & Cash, T. E (1985). Physical attractiveness and malad-
justment among psychiatrieinpatients. JournalofSocialand Clinical
Psycholog~, 3, 170-180.
Bock, A. T. (1976). Cognitive therapy and the emotional disorders. New
York: International Universities Press.
Beck, A. T., Rush, A. J., Shaw, B. E, & Emery, G. (1979). Cognitive
therapy ofdepression. New York: Guilford Prr
Berman, J. S., Miller, R. C., & Massman, P. J. (1985). Cognitive therapy
versus systematic dr Is one treatment superior? Psycho-
logical Bulletin, 97. 451--461.
Br E., Walster, E., & Bohrnstr G. (1973). The happy Ameri-
can body: A survey report. Psychology Today, 7, 119-131.
Bohrnstedt, G. W. (1977). On measuring body satisfaction. Unpub-
lished manuscript, Institute of Social Research, Indiana University,
Bloomington, IN.
Butters, J. W. (1985). The controlled evaluation of a cognitive-behav-
ioral treatment program for body-image dissatisfaction in college
women, Unpublished doctoral dissertation, Virginia Consortium for
Professional Psychology, Norfolk, VA.
Cash, T. E (1981). Physical attractiveness: An annotated bibliography
of theory and research in the behavioral sciences. JSAS Catalog of
Selected Documents in Psycholog~, / /(Ms. No. 2370).
Cash, T. E (1985). Physical appearance and mental health. In J. A. Gra-
haro & A. Kligman (Eds.), The psychology of cosmetic treatments (pp.
196-2 ! 6). New York: Praeger.
Cash, T. E, & Brown, T. A. (in press). Body image in anorexia nervosa
and bulimia nervosa: A review of the literature. Behavior Modifica-
tion.
Cash, T. E, & Butters, J. W. (in press). Poor body image: Helping the
patient change. Medical Aspects of Human Sexuality.
Cash, T. E, Cash, D. W., & Butters, J. W. (1983). "Mirror, mirror on
thr w a l l . . . ? : " Contrast effects and sr162 of physical at-
t~ctiveness. Personality and Social Psychology Bulletin, 9, 351-358.
Cash, T. E, & Green, G. K. (1986). Body weight and body image among
college women: Perception, cognition, and affect. Journal of Personal-
ity Assessment, 50, 290-301.
Cash, T. E, & Horton, C. E. (I 983). Aesthetic surgery: Effects of rhin~
plasty on the social perception of patients by others. Plastic and Re-
constructive Surger~, 72, 543-548.
Cash, T E, Winstead, B. A., & Janda, L. H. (1986). The grr American
shape-up: Body image survey report. Psychology Today, 20(4), 30-
37.
Derogatis, L. R. (1977). SCL-90: Administration, scoring and proce-
dures manual for the revised version. Baltimore, MD: Johns Hopkins
University Sehool of Medidne.
Ellis, A. (1962). Reason and emotion in psyrhotherap~. Secaucus, NJ:
Citadel Press.
Fallon, A. E., & Rozin, P. (1985). Sex differences in perceptions ofdesir-
able body shape. JournalofAbnormal Psycholog~, 94, 102-105.
Farina, A., Burns, G. L., Austad, C., Bugglin, C., & Fischer, E. H.
(1986). The foie of physical attractiveness in the rr of dis-
charged psychiatrie patients. Journal of Abnormat Psychologs 95,
139-143.
Farina, A., Fischer, E., Sherman, S., Smith, W., Groh, T., & Mermin, P.
(1977). Physical attractiveness and mental illnr Journal of Abnor-
mal Psycholog~, 86, 510-517.
Gartinkel, la. E., & Garner, D. M. (1982). Anorexia nervosa: A muhidi-
mensionalperspective. New York: Bruner-Mazel.
Garner, D. M. (1986). Cognitive therapy for anorexia nervosa. In K. D.
Brownell & J. P. Foreyt (Eds.), Handbook of eating disorders (pp.
301-327). New York: Basie Books.
Garner, D. M., & Br K. (1982). A cognitive-br approach
to anorexia nervosa. Cognitive TherapyandResearch, 6, 1-27.
Garner, D. M., & Garfinkel, P. E. (1981). Body image in anorexia ner-
vosa: Measurement, theory, and clinical implications. International
Journal of Psychiatry in Medicine, 1t, 263-284.
Goin, J. M., & Goin, M. K. (1981). Changing the body: Psychological
effects of plastic surgers Baltimore, MD: Williams & Williams.
Goldfried, M. R., & Davison, G. C. (1976). Clinical behavior therap)~.
New York: Holt, Rinehart & Winston.
Hatfield, E., & Sprecher, S. (1986). Mirror, mirror...: The importance
oflooks in •veryday lire. Albany: State University of New York Press.
 BODY-IMAGE TREATMENT
Heimreich, R., & Stapp, J. (1974). Short forms ofthe Texas Social Be-
havior Inventory (TSBI): An objective measure of self-esteem. Bulle-
tin of the Psychonomic Society, 4, 473--475.
Kazdin, A. E. (1983). Treatment research: The investigation and evalu-
ation of psychotherapy. In M. Hersen, A. E. Kazdin, & A. S. Bellack
(Eds.), The clinicalpsychology handbook. New York: Pergamon Press.
Kazdin, A. E., & Wilson, G. T. (1978). Evaluation ofbehavior therapy:
Issues, evidence and research strategies. Lincoln: University of Ne-
braska Press.
Keeton, E, Cash, T. E, & Lewis, R. J. (1987, March). The multidimen-
sional assessment of body image among college students. Paper pre-
sented at the Southeastr Psyehological Convention, Aflanta, GA.
Kendall, E C., & Hollon, S. D. (Eds.). (1979). Cognitive-behavioral in-
terventions: Theors research, and procedures. New York: Academie
Press.
Kendall, E C., & Norton-Ford, J. D. (1982). Therapy outcome researeh
methods. In P. C. Kendall & J. N. Butcher (Eds.), Handbook ofre-
search methods in clinicalpsychology (pp. 429--460). New York: Wi-
ley.
Larsen, D. L., Attkisson, C. C., Hargreaves, W. A., & Nguyen, T. D.
(1979). Assessment of client/patient satisfaction: Development of a
general scale. Evaluation and Program Planning, 2, 197-207.
Lerner, R. M., Orlos, J. B., & Knapp, J. R. (1976). Physical attractive-
ness, physical effectiveness and self-concept in iate adolescents. Ad-
olescence, 1I, 313-326.
MacPhiUamy, D. J., & Lewinsohn, E M. (1974). Depression as a func-
tion of desired and obtained pleasure. Journal ofAbnormal Psychol-
ogy. 83, 651-657.
Marlatt, G. A., & Gordon, J. R. (Eds.). (1985). Relapse prevention:
Maintenance strategies in the treatment of addictive behaviors. New
York: Guilford Press.
McCrea, C. W., Summerfield, A. B., & Rosen, B. (1982). Body image:
A selective review of existing measurement techniques. British Jour-
nal of Medical Psychologs 55, 225-233.
Meichenbaum, D. (1985). Stress inoculation training. New York: Perga-
mon Press.
897
Metropolitan Life Insurance Company. (1983). Metropolitan life insur-
ance table of desirable weight. New York: Author.
Miller, R. C., & Berman, J. S. (1983). The elticacy of cognitive behavior
therapies: A quantitative review ofthe research evidence. Psychologi-
cal Bulletin, 94, 39-53.
Mulready, P. M., & Lamb, J. M. 0985). Cosmetics therapy for female
chemotherapy patients. In M. R. Solomon (Ed.), The psychology of
fashion (pp. 255-263). Lexington, MA: Lexington Books.
Noies, S. W., Cash, T. E, & Winstead, B. A. (1985). Body image, physi-
cal attractiveness, and depression. Journal of Consulting and Clinical
Psychologs 53, 88-94.
O'Grady, K. E. (1982). Sex, physical attractiveness, and perceived risk
for mental illness. Journal of Personality and Socml Psychologs 43,
1064-1071.
Rosen, G. M., & Ross, A. O. (1968). Relationship ofbody image to self-
concept. Journal of Consulting and Clinical Psychology, 32, l O0.
Ruff, G. A., & Barrios, B. A. (1986). Realistic assessment ofbody image.
Behavioral Assessment, 8, 237-251.
Secord, E E, & Jourard, S. M. (1953). The appraisal of body-cathexis:
Body-cathexis and the self. Journal of Consulting Psychologs 17,
343-347.
Weiss, L., Katzman, M., & Wolehik, S. (1985). Treating bulimia: A psy-
choeducational approach. New York: Pergamon Press.
Winstead, B. A., & Cash, T. E (1984, March). Reliability and validity
of the Body Self-Relations Questionnaire: A new measure of body im-
age. Paper presented at the Southeastern Psychological Association
Convention, New Orleans, LA.
Wolpe, J. (1973). Thepracticeofbehaviortherapy(2nd ed.). New York:
Pergamon Press.
Wolpe, J., & Lazarus, A. A. (1966). Behavior theral~ techniques. New
York: Pergamon Press.
Wooley, S. C., & Kearney-Cooke, A. (1986). Intensive treatment ofbu-
limia and body-image disturbance. In K. D. Brownell & J. P. Foreyt
(Eds.), Handbook ofeating disorders (pp. 476-502). New York: Basic
Books.
Received September 21, 1986
Revision received March 16, 1987
Accepted March 30, 1987 9