See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/247166828

Development and validation of a Body Image Guilt and Shame Scale

Article  in  Personality and Individual Differences · January 2003

DOI: 10.1016/S0191-8869(02)00026-0

CITATIONS READS
51 3,681

3 authors, including:

Dale Dinnel
Western Washington University
28 PUBLICATIONS   849 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Academic achievement behaviours View project

All content following this page was uploaded by Dale Dinnel on 04 June 2018.

The user has requested enhancement of the downloaded file.

 Personality and Individual Differences 34 (2003) 59–75
www.elsevier.com/locate/paid

Development and validation of a Body Image Guilt

and Shame Scale
Ted Thompsona,*, Dale L. Dinnelb, Nicole J. Dilla
a
School of Psychology, University of Tasmania, GPO Box 252-30, Hobart, Tasmania, 7001, Australia
b
Western Washington University, USA

Received 19 March 2001; received in revised form 2 December 2001; accepted 21 January 2002

Abstract
The purpose of this study was to assess the psychometric properties of the Body Image Guilt and Shame
Scale (BIGSS). Undergraduate students (n=87 women, n=66 men) completed the BIGSS, the Test of Self-
conscious ffect (‘TOSCA’), scales assessing body image concern and importance, social physique anxiety,
and personal and social identity. A factor analysis indicated that a two-factor solution reflecting guilt and
shame was the most parsimonious explanation of the data from the BIGSS responses. An additional factor
analysis of the BIGSS and TOSCA responses indicated that the responses could be divided between gen-
eralised guilt, generalised shame, body image guilt, and body image shame dimensions. The BIGSS guilt
and shame dimensions also demonstrated strong internal consistency and item/item total correlations
especially when Item 10 was deleted. Consistent with predictions, BIGSS guilt and shame were positively
correlated with social physique anxiety, body image concern, and body image importance, demonstrating
construct validity. # 2002 Published by Elsevier Science Ltd.
Keywords: Guilt; Shame; Body image concern; Social physique anxiety; Body image importance; Personal and social
identity

1. Introduction

For many if not most women in Western societies, being a woman means experiencing some
apprehension over the extent to which one’s body shape matches the culturally defined ideal.
Where there is a perception that one’s body shape fails to approximate this ideal, dissatisfaction is
likely to result. As is now well-documented, this dissatisfaction with one’s body and vulnerability
to external sociocultural pressures has important implications for the development of a range
of clinical symptomatology and has been found to be the single strongest predictor of eating

* Corresponding author. Tel.: +61-3-6226-2887; fax: +61-3-6226-2883.

E-mail address: t.thompson@utas.edu.au (T. Thompson).

0191-8869/02/$ - see front matter # 2002 Published by Elsevier Science Ltd.

PII: S0191-8869(02)00026-0
60 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

disorders in young people (e.g. Phelps, Johnston, & Augustyniak, 1999). This link is particularly
strong in the case of women (e.g. Fallon & Rozin, 1985; Muth & Cash, 1997). In the view of
several researchers, current preoccupations with thinness and dieting in women may have con-
tributed to the large increase in eating disorders over the past several decades (e.g. Burney &
Irwin, 2000; Garner & Garfinkel, 1980; Thompson & Schwartz, 1982).

1.1. Body dissatisfaction as a sociocultural phenomenon

Biogenetic explanations have not been able to account for the current preoccupation with
thinness and dieting, nor the greater reporting of body dissatisfaction among women relative to men.
Several explanations based on sociocultural pressures and external socially mediated psychological
factors have been offered in an attempt to clarify the basis of what has been described as a ‘‘normative
discontent’’ among non-eating disordered women (Rodin, Silberstein, & Striegel-Moore, 1985).
Cash and Henry (1995) suggest that women are socialised from a very young age to be oriented
to others to receive validation of their self-worth, and that society tends to reward women with
thin bodies. Obesity is negatively sanctioned in women—far more so than in men—and is seen as
a product of lack of self-control. Bruch (1978) noted that there appears to be a ‘thinness=femi-
ninity=happiness’ life equation, with thinness being synonymous with happiness.
Brownmiller (1984) contends that the specific areas of physical attractiveness and weight may
be primary competitive domains where women are encouraged to contend with each other, citing
evidence that women report conducting fleeting, automatic comparisons with female occupants
upon entering a room in order to assess where they are positioned in terms of who is fatter, and
who is more attractive.

1.2. Gender differences in body image concerns

It is increasingly not only women who feel pressures to compare favourably with cultural ideals.
Many men are far from immune to maladaptive body image problems (Ricciardelli & Glynn,
1998), with body image discontent now recognised in men (e.g. Drewnowski & Yee, 1987; Faw-
kner & McMurray, 1999; Muth & Cash, 1997).
However, men and women may express their body image concerns in different ways. While
body esteem is a crucial factor in the development of dissatisfaction with one’s body (Phelps et
al., 1999) the effect of such dissatisfaction on one’s overall level of self-esteem is determined by
the relative importance of that domain to one’s self-definition. On this basis, failure to match an
ideal in an area of central importance to the individual (e.g. physique) will prove more damaging
to self-worth than inadequacy in a domain of lesser importance.
Contrary to findings for women, men’s body dissatisfaction is a two-tailed phenomenon invol-
ving both ends of a weight continuum. Empirical research has found that men who are either
above or below the acceptable range in their Body Mass Index scores, tend to be especially dis-
satisfied with their physical appearance (Drewnowski & Yee, 1987; Muth & Cash, 1997).

1.3. Social physique anxiety

Social physique anxiety—the anxiety experienced by individuals in response to others’ aesthetic

evaluations of their physique (Hart, Leary, & Rejeski, 1989)—has been found to be a stronger
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 61

predictor variable of body dissatisfaction than a number of sociocultural variables (e.g. appear-
ance orientation), physiological variables (e.g. body mass index) and body management variables
(e.g. dietary restraint) for both men and women (e.g. Risby, 1999).
While not discounting the anxiety that men experience over their body image, a number of
empirical studies have shown that body image is particularly important to women (e.g. Cash &
Henry, 1995; Frederick & Morrison, 1998; Garner, Garfinkel, Schwartz, & Thompson, 1980).
Women are more likely than men to describe themselves as overweight when by objective stan-
dards they are not (Gray, 1977). They also worry about being or becoming fat, and express
greater concern with dieting and body weight relative to men (Drewnowski & Yee, 1987; Fallon
& Rozin, 1985; Muth & Cash, 1997; Tiggemann, 1994).

1.4. Shame and guilt

When individuals do not measure up to their view of the ‘ideal’ physique, and when measuring
up to that ideal is of central concern, they are likely to experience feelings of shame (Silberstein,
Striegel-Moore, & Rodin, 1987). Rodin et al. (1985) assert that shame and social pressure lead to
women’s preoccupation with their appearance. While shame and guilt are both negative, self-
conscious emotions, they have been differentiated in recent empirical literature (e.g. Burney &
Irwin, 2000; Tangney, 1995).
With guilt, the focus is upon a specific action (or failure to act), is essentially a private matter,
involving one’s own conscience and personal standards. As such it is internally mediated. While
guilt may promote reparative action, shame is likely to prompt a desire to hide a defective self, or
as June Tangney puts it, ‘‘to sink into the floor and disappear’’ (Tangney, 1995).
However shame is essentially public in the sense that there is an implied, if not actual,
audience involved (e.g. Lutwak & Ferrari, 1996; Tangney, 1995). Relative to guilt, shame is
associated with greater psychopathology in the form of maladaptive and nonconstructive
responses to anger (Tangney, Price, Barlow, Wagner, Marscall, Borenstein, Sanftner, Mohr, &
Gramzow, 1996a, 1996b). It is also associated with escape avoidance and distancing as ways
of coping with stressors (Burggraf, 1995), fear of negative evaluation about oneself (Lutwak &
Ferrari, 1997), and pessimism and depression (Rayner, Singer, & O’Connor, 1995). On the
other hand guilt is positively related to moral behaviours, other-oriented empathy (Baumeis-
ter, Reiss, & Delespaul, 1995) and has been credited on the whole as more adaptive, facil-
itating empathic concern, directing the individual toward reparative action (e.g. Tangney,
1995).
Just as social physique anxiety is regarded as externally mediated, so too is shame. The level of
shame experienced about one’s body is viewed, in part, as a measure of the extent to which one
has internalised cultural standards (e.g. Cheek & Hogan, 1983; McKinley & Hyde, 1996; Noll &
Fredrickson, 1998; Phelps et al., 1999). Given the realities of the distributions of female body
shape relative to the culturally defined ideal, many women are bound to emerge from such com-
parisons with assessments of their body shape as flawed and on this basis, engage in self-debase-
ment (Rodin et al., 1985). Silberstein et al. (1987) suggest that such failures by comparison lie at
the core of women’s dissatisfaction with their bodies, while Burney and Irwin (2000) determine
that shame associated with eating behaviour is a primary predictor of the severity of eating-dis-
order symptomatology.
62 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

1.5. Shame, guilt and identity orientation

Associated with shame and guilt in a body-image context are two broad identity categories that
are assessed by the Aspects of Identity Questionnaire (AIQ-III; Cheek, Tropp, Chen, & Under-
wood, 1994). These are Personal Identity Orientation and Social Identity Orientation. Personal
Identity Orientation is a private conception of self involving subjective feelings of continuity and
uniqueness, and as such it is internally mediated. Social Identity Orientation, on the other hand,
refers to one’s public image presented through social roles and relationships, and is largely shaped
by the reactions of and comparisons with others (Cheek & Briggs, 1982). As such, like shame, it is
considered to be externally mediated.
For most individuals, self-definition doubtless involves both social and personal attributes.
However there may be marked individual differences in the relative importance of personal
compared with social aspects of identity. Lutwak, Ferrari, and Cheek (1998) found social identity
to be the best predictor of shame.
For both men and women, social identity has been found to significantly correlate with shame,
while Personal Identity orientation is correlated with guilt (Lutwak et al., 1998; Tobey & Tunnell,
1981). Sanftner, Barlow, Marschall, and Tangney (1995) also found that externally-mediated
feelings of shame are positively correlated with eating disorder symptoms, including body dis-
satisfaction.

1.6. The present study

While on the above bases it would appear important to be able to assess guilt and shame in
relation to body image concerns, we contend that measures used to assess shame in a body-rela-
ted context are inadequate, being either excessively global in nature (e.g. the Test of Self-Con-
scious Affect; Tangney, Wagner, & Gramzow, 1989) or specifically targeted at eating behaviours
(e.g. the Shame and Guilt Eating Scale; Frank, 1990). While the Objectified Body Consciousness
Scale of McKinley and Hyde (1996) contains subscales assessing Surveillance (viewing the body
as an outside observer), Appearance Control Beliefs, and Body Shame, there is no domain-spe-
cific measure that embraces both body image guilt and shame.
To date, shame and guilt in the body image and eating disorder literature has typically been
assessed using the Test of Self-Conscious Affect (TOSCA: Burney & Irwin, 2000; Lutwak &
Ferrari, 1996; Sanftner et al., 1995; Tangney, Wagner, Fletcher, & Gramzow, 1992). However,
the TOSCA may not be the most appropriate scale to investigate shame and guilt in body-related
contexts, as its focus is on interpersonal offences, and not general negative affects. Clearly, shame
and guilt may arise when a person reflects on his or her own body image, not because they have
insulted or offended someone. As such, the TOSCA and the BIGSS assess shame and guilt in
quite different contexts.
On this basis, the primary purpose of the present investigation was to develop a valid and reli-
able scale capable of assessing body-image shame and guilt that could be distinguished from
shame and guilt in relation to interpersonal transgressions. We expected a factor analysis of the
BIGSS would yield a scale with guilt and shame items aligning separately on the two factors that
would prove to be factorially different from TOSCA guilt and shame items. In addition, we
expected that higher BIGSS guilt and shame scores would be related to higher TOSCA guilt and
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 63

shame scores. Based on the findings of Crawford and Eklund (1994), we also expected that higher
BIGSS guilt and shame scores would be related to higher social physique anxiety.
Consistent with the findings of Lutwak et al. (1998) and Tobey and Tunnell (1981), we expected
high BIGSS guilt would be positively correlated with higher Personal Identity scores while BIGSS
shame scores would be positively correlated with Social identity scores. Mindful of the findings of
Ricciardelli and colleagues (e.g. Ricciardelli & Glynn, 1998; Ricciardelli & McCabe, 1999), we
expected that BIGSS guilt and shame scores would be positively related to scores on Body Image
Concern and Body Image Importance scores. Finally, we expected gender differences, with
women exhibiting higher BIGSS guilt and shame scores than men, reflecting the greater social
pressures for women to conform to a socially ideal body shape than men (Rodin et al., 1985).

2. Method

2.1. Participants

One-hundred and fifty-three undergraduate university students participated in the present study.
Each participant who completed the questionnaire package as part of a class requirement was reques-
ted to recruit an opposite sex peer to ensure approximately equal numbers of male and female partici-
pants. The average age of the participants was 21.26 years with ages ranging from 17 to 48. Despite our
efforts to obtain a gender-balanced sample, there were 66 men participants (M=21.30, SD=4.52) and
87 women participants (M=21.57, SD=5.07). Separate comparisons of men and women failed to
reveal differences between the scores of any scale in the questionnaire packet. On this basis, partici-
pants’ scores were combined across participant recruitment conditions for all subsequent analyses.

2.2. Materials

2.2.1. Body Image and Body Change Questionnaire (BIBCQ; Ricciardelli & McCabe, 1999)
The BIBCQ comprises 20 items comprising two subscales: Body Image Concern and Body
Image Importance. The 10 items on each of the two subscales make assessments of the same
aspects of physique (weight, body shape, muscle size, hips, thighs, chest, abdominal region/sto-
mach, shoulders legs and arms). However, the nature of the assessment differs for each subscale.
For Body Image Concern, respondents are asked to report, ‘‘How satisfied are you with your
[weight, body shape. . .etc]?’’, while for the Body Image Concern subscale respondents are asked
to report ‘‘How important to you is [what you weigh, the shape of your body . . . etc] compared with
other things in your life?’’
Respondents rate particular aspects of their physique on five-point Likert scales assessing Body
Image Concern with end-point designations extremely satisfied (5) and extremely dissatisfied (1).
For Body Image Importance, end-point designations were extremely important (1) and extremely
unimportant (5). For the purpose of the present study, responses for Body Image Importance
items were reverse-coded.
One of the strengths of this measure is that it is gender-neutral, having been developed in order
to address some of the issues regarding female bias in instruments purporting to measure body
satisfaction. McCabe and Ricciardelli (2000) reported an internal consistency of 0.90.
64 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

2.2.2. Social Physique Anxiety Scale (SPAS; Hart et al., 1989)

The SPAS comprises 15 items that assess the anxiety experienced by individuals in response to
the evaluations of others concerning their physique. As such, the SPAS is an affective rather than
a cognitive measure. Respondents register their responses on five-point Likert scales, with end-
point designations ranging from Not at all characteristic of me (1) to 5 Extremely characteristic of
me (5). A sample item is I worry about wearing clothes that might make me look thin or overweight.
Hart et al. (1989) report an inter-item reliability of r=0.90 and test–retest reliability of r=0.82.

2.2.3. Aspects of Identity Questionnaire (AIQ-III; Cheek & Hogan, 1983)

The AIQ-III comprsises three subscales, these being Personal Identity, Social Identity, and
Collective Identity. The 10-item Personal Identity subscale focuses on personal factors (e.g. My
feelings of being a unique person, being distinct from others). The seven-item Social Identity sub-
scale focuses on social issues (e.g. My popularity with other people) with the eight-item Collective
Identity subscale focusing on important referent group issues (e.g. Being a part of the many gen-
erations of my family). Items are formatted on five-point scales with end-point designations ran-
ging from Not important to my sense of who I am (1) to Extremely important to my sense of who I
am (5). Cheek et al. (1994) report internal consistencies from 0.68 to 0.84, and test–retest reli-
abilities ranging from 0.77 to 0.81. For the purpose of the present study, only the Personal and
Social Identity Orientation subscales were used.

2.2.4. Test of Self-Conscious Affect (TOSCA; Tangney et al., 1989)

The TOSCA is a scenario-based measure that yields indices of proneness to shame, proneness
to guilt, externalisation of blame, detachment-unconcern, alpha pride (pride in self), and beta
pride (pride in behaviour). Respondents are presented with a series of 15 scenarios that they may
encounter in daily life. A sample scenario from the TOSCA is At work, you wait until the last
minute to plan a project, and it turns out badly. The shame and guilt responses that follow are You
would feel incompetent (shame), and You would feel you deserve to be reprimanded (a guilt
response).
Respondents are required to rate each response on a five-point scale with end-point designa-
tions not likely (1) and very likely (5). For the purpose of the present study, only the shame and
guilt response items were scored. Tangney et al (1989) report internal consistencies of 0.76 and
0.66 for Shame and Guilt, respectively (Tangney et al., 1989).

2.2.5. BIGSS1
The BIGSS is a specific body-related, scenario-based measure that indicates proneness to shame
and proneness to guilt about one’s body and body-related behaviours. The structure of the
scale—the response alternatives and scenarios—were modelled on the TOSCA and trialed on
graduate students and colleagues who provided feedback on the wording of scenarios and
response options, and whether they classified as intended in terms of externalisation of blame,
detachment-unconcern, guilt and shame. Suggested revisions to response options and scenarios
were incorporated where these seemed soundly based. There was no disagreement in classifying
the response options as shame, guilt, externalisation, and detachment.

1
A copy of the BIGSS may be obtained from the principal author.
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 65

Following June Tangney’s lead with the TOSCA we operationalised guilt as a behavioural
response. While Tangney’s scale, the guilt response option involves reparative action in an inter-
personal sense, in the BIGGS the intended action is generally corrective of a past indulgence (e.g.
You would tell your friend you would stop eating junk food from now on). On the other hand in both
the BIGSS and TOSCA shame is operationalised as an emotional response, with no translation
into action. In attributional terms, the emotional reaction is global and frequently character-
ological (e.g. You would feel as though you want the ground to open up and swallow you).
Respondents are presented with a series of 15 items followed by four response options that they
rate on five-point scales with end-point designations not likely (1) and very likely (5). A sample
scenario from the BIGSS is You find that your clothes from last summer are very tight around your
waist, followed by four response options.
For this item, the response options are: You would feel undisciplined and overweight (shame
response), You would go out and buy a six-month membership to a gym (guilt response), You would
think: Well, it’s time to buy some new clothes anyway! (detachment), and You would think: I’ve
been very busy over the last year, with no time to exercise (externalisation/rationalisation).
The shame and guilt response options are the only two response options that are scored, with
externalisation/rationalisation and detachment serving as filler items. These response options are
randomised across the 15 items.

2.3. Procedure

Participants were informed that the purpose of the study was to explore issues involving shame,
identity factors, and levels of body dissatisfaction. Questionnaire packets were presented to par-
ticipants with the order of the five scales randomised to control for order effects. As comparisons
of scale scores on the basis of order of presentation yielded no significant differences on any of the
measures, scores were collapsed across presentation order for all analyses reported later.
The questionnaire packets were administered in a classroom setting with seating of participants
arranged to ensure privacy of responses. The instructions for each scale were provided at the
beginning of each scale. Participants were informed that there were no right or wrong answers
and urged to complete the scales in a way that was consistent with how they felt. Participants
took approximately 1 h to complete the questionnaire packages, and were allowed a coffee break
whenever they wished. Following completion of the study, participants were fully debriefed and
thanked for their participation.

3. Results

3.1. Reliability and item analysis

A measure of internal consistency was computed for the guilt and shame responses of the
BIGSS. The internal consistency was a=0.88 for the 15 items of the guilt responses and was
a=0.90 for the 15 items of the shame responses of the BIGSS. However, inspection of the cor-
rected item, item–total correlations revealed that for both the guilt and shame responses, the
correlations for Item 10 were below the criterion level of 0.30 (Table 1). Upon examination of this
66 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

Table 1
Corrected item–item total correlations and internal consistency (a) for the guilt and shame responses of the Body
Image Guilt and Shame Scale

Item Corrected item–item Alpha if item

total correlation deleted

Guilta
1 0.35 0.88
2 0.43 0.87
3 0.68 0.86
4 0.60 0.87
5 0.52 0.87
6 0.56 0.87
7 0.64 0.87
8 0.66 0.86
9 0.58 0.87
10 0.24 0.88
11 0.68 0.86
12 0.48 0.87
13 0.65 0.86
14 0.45 0.87
15 0.49 0.87
Shameb
1 0.36 0.91
2 0.48 0.90
3 0.57 0.90
4 0.66 0.90
5 0.64 0.90
6 0.65 0.90
7 0.73 0.89
8 0.60 0.90
9 0.64 0.91
10 0.21 0.91
11 0.65 0.90
12 0.71 0.89
13 0.68 0.90
14 0.58 0.90
15 0.67 0.90

N=153
a
a=0.88 for all 15 guilt items
b
a=0.90 for all 15 shame items

item, we determined that the low correlations were likely due to the fact that this item referred to
another person’s body image and not the respondent’s body image as for the other 14 items.
Given these considerations, we decided to eliminate Item 10 from all subsequent analyses. Coef-
ficient alphas for the guilt and shame responses when Item 10 was removed were a=0.88 and
a=0.91, respectively.
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 67

3.2. Factor analysis

To determine whether the 14 items from the revised BIGSS represented differentiated guilt and
shame dimensions, we subjected the data to a factor analysis using a principal axis extraction with
an oblique rotation. An oblique rotation was used because guilt and shame have been shown to
be consistently correlated irrespective of the guilt and shame dimensions that are being measured.
We concluded from the scree test that the data were best described by two factors. The first factor
accounted for 37.51% of the variability in the item responses and seemed most related to a guilt
factor. Eleven of the 14 guilt responses loaded on this factor while four of the 14 shame items
loaded on this factor.
The second factor accounted for 4.69% of the variability in the item responses and seemed most
related to a shame factor. Ten of the 14 shame responses loaded on this factor while one of the 14
guilt item responses loaded on this factor. Two of the 14 guilt responses failed to load on either
factor (Table 2). On these bases, the BIGSS seems to differentiate between body image guilt and
body image shame. It should be noted that the guilt and shame scores of the BIGSS were sig-
nificantly correlated (r=0.59) and the guilt and shame factors from the factor analysis were cor-
related at r=0.52, reinforcing the decision to use an oblique rotation.
To determine if the BIGSS guilt and shame responses could be distinguished from the TOSCA
guilt and shame responses, we submitted the responses to a principal axis factor analysis with an
oblique rotation. Once again, an oblique rotation was used because the guilt and shame factors of
scales have been shown to be correlated. From the scree test, we concluded that the data were
best described by four factors. The first factor accounted for 22.35% of the variance in item
responses and seemed most related to body image guilt. Of the 19 items that loaded on this factor,
17 items loaded on only this factor. Thirteen of the items that loaded on this factor were BIGSS
guilt responses (12 items loaded uniquely on this factor and one item also loaded on the fourth
factor). Four of the items that loaded on this factor were BIGSS shame responses (three items
loaded uniquely on this factor and one item also loaded on the fourth factor), and one item from
the TOSCA guilt responses. One item from the TOSCA shame responses also loaded uniquely on
this factor.
The second factor accounted for 5.39% of the variance in the item responses and seemed most
related to self-conscious shame. Fifteen items loaded on the second factor with 12 of these items
loading solely on the second factor. Eleven of the items that loaded on the self-conscious shame
factor were TOSCA shame responses. (Ten items loaded uniquely on this factor and one item also
loaded on the third factor). Three of the items that loaded on this factor were TOSCA guilt
responses. (Two items loaded uniquely on this factor and one item also loaded on the fourth
factor). One BIGSS guilt item loaded on this factor and the first factor.
The third factor accounted for 3.92% of the variability in item responses and seemed most
related to self-conscious guilt. Of the nine items that loaded on this factor, eight items load on it
uniquely. Eight of the TOSCA guilt item responses loaded on this factor (seven items loaded on
this factor uniquely and one item also loaded on the second factor) and one TOSCA shame item
loaded uniquely on this factor.
The fourth factor accounted for 2.85% of the variability in item responses and seemed most
related to body image shame. Of the 14 items that loaded on this factor, 10 loaded uniquely on it.
Ten of the BIGSS shame responses loaded on this factor. (Nine items loaded uniquely on this
68 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

Table 2
Factor analysis for the items on the Body Image Guilt and Shame Scale (BIGSS) with Item 10 removeda
Item I II
1. BIGSS—Guilt – –
2. BIGSS—Guilt – –
3. BIGSS—Guilt 0.52 –
4. BIGSS—Guilt 0.50 –
5. BIGSS—Guilt 0.56 –
6. BIGSS—Guilt – 0.51
7. BIGSS—Guilt 0.69 –
8. BIGSS—Guilt 0.59 –
9. BIGSS—Guilt 0.67 –
11. BIGSS—Guilt 0.75 –
12. BIGSS—Guilt 0.33 –
13. BIGSS—Guilt 0.51 –
14. BIGSS—Guilt 0.45 –
15. BIGSS—Guilt 0.78 –
1. BIGSS—Shame 0.43 –
2. BIGSS—Shame – 0.56
3. BIGSS—Shame – 0.68
4. BIGSS—Shame – 0.68
5. BIGSS—Shame – 0.69
6. BIGSS—Shame – 0.73
7. BIGSS—Shame – 0.74
8. BIGSS—Shame 0.62 –
9. BIGSS—Shame – 0.61
11. BIGSS—Shame – 0.55
12. BIGSS—Shame – 0.73
13. BIGSS—Shame – 0.61
14. BIGSS—Shame 0.52 –
15. BIGSS—Shame 0.55 –

Eigenvalue 11.05 1.87

Percent of variance 37.51 4.69
a
A criterion of 0.30 was used for factor loadings. N=153

factor while one item also loaded on the first factor). Two BIGSS guilt items loaded on this factor
as well as on the first factor, and two TOSCA shame items loaded on the fourth factor. (One item
loaded uniquely on this factor whereas one item also loaded on the second factor). The results of
this factor analysis are summarised in Table 3. On these bases, it appears that both the BIGSS
and the TOSCA result in responses that can be differentiated on the basis of guilt and shame.
However, it is evident that the BIGSS and the TOSCA are also measuring qualities of guilt and
shame that are unique to the particular scale being used.

3.3. Construct validity

Evidence for the construct validity of the BIGSS was established by significant zero order cor-
relations consistent with theoretical positions. In particular, the guilt score of the BIGSS was
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 69

Table 3
Factor analysis for the items on the Test of Self-Conscious Affect (TOSCA) and the Body Image Guilt and Shame
Scale (BIGSS) with Item 10 on the BIGSS removeda

Item I II III IV

1. TOSCA—Guilt – – 0.32 –
2. TOSCA—Guilt – – – –
3. TOSCA—Guilt – – – –
4. TOSCA—Guilt – – – –
5. TOSCA—Guilt – – 0.69 –
6. TOSCA—Guilt – – 0.32 –
7. TOSCA—Guilt 0.48 – – –
8. TOSCA—Guilt – – 0.41 –
9. TOSCA—Guilt – – 0.38 –
10. TOSCA—Guilt – 0.33 0.31 –
11. TOSCA—Guilt – 0.43 – –
12. TOSCA—Guilt – – 0.53 –
13. TOSCA—Guilt – – 0.62 –
14. TOSCA—Guilt – 0.35 – –
15. TOSCA—Guilt – – – –
1. TOSCA—Shame – 0.40 – –
2. TOSCA—Shame – 0.32 – –
3. TOSCA—Shame – – – 0.34
4. TOSCA—Shame – 0.47 – –
5. TOSCA—Shame – 0.48 – –
6. TOSCA—Shame – 0.53 – –
7. TOSCA—Shame 0.50 – – –
8. TOSCA—Shame – 0.49 – –
9. TOSCA—Shame – 0.42 – –
10. TOSCA—Shame – 0.44 – –
11. TOSCA—Shame – 0.52 – –
12. TOSCA—Shame – – – –
13. TOSCA—Shame – – 0.51 –
14. TOSCA—Shame – 0.33 – 0.35
15. TOSCA—Shame – 0.31 – –
1. BIGSS—Guilt – – – –
2. BIGSS—Guilt 0.34 – – –
3. BIGSS—Guilt 0.55 – – 0.31
4. BIGSS—Guilt 0.45 0.33 – –
5. BIGSS—Guilt 0.49 – – –
6. BIGSS—Guilt 0.42 – – 0.37
7. BIGSS—Guilt 0.65 – – –
8. BIGSS—Guilt 0.57 – – –
9. BIGSS—Guilt 0.69 – – –
11. BIGSS—Guilt 0.40 – – –
12. BIGSS—Guilt 0.56 – – –
14. BIGSS—Guilt 0.48 – – –
15. BIGSS—Guilt 0.70 – – –

(continued on next page)

70 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

Table 3 (continued)

Item I II III IV

1. BIGSS—Shame 0.40 – – –
2. BIGSS—Shame – – – 0.47
3. BIGSS—Shame – – – 0.67
4. BIGSS—Shame – – – 0.49
5. BIGSS—Shame – – – 0.58
6. BIGSS—Shame – – – 0.53
7. BIGSS—Shame – – – 0.56
8. BIGSS—Shame – – – –
9. BIGSS—Shame – – – 0.50
11. BIGSS—Shame 0.38 – – 0.42
12. BIGSS—Shame – – – 0.55
13. BIGSS—Shame – – – 0.48
14. BIGSS—Shame 0.59 – – –
15. BIGSS—Shame 0.62 – – –
Eigenvalue 13.53 3.82 2.92 2.27
Percent of variance 22.35 5.39 3.91 2.85
a
A criterion of 0.30 was used for factor loadings. N=153

moderately, significantly correlated with general self-conscious guilt (TOSCA-G; r=0.48), gen-
eral self-conscious shame (TOSCA-S; r=0.41), social physique anxiety (SPA; r=0.60), body
image concern (BIBCQ-C; r=0.43), body image importance (BIBCQ-I; r=0.39) and social
identity (AIQ-SI; r=0.32). The BIGSS guilt score were only somewhat significantly correlated with
personal identity (AIQ-PI; r=0.19). Furthermore, shame scores of the BIGSS were moderately, sig-
nificantly correlated with general self-conscious guilt (TOSCA-G; r=0.41), general self-conscious
shame (TOSCA-S; r=0.51), social physique anxiety (SPA; r=0.73), body image concern (BIBCQ-
C; r=0.56), body image importance (BIBCQ-I; r=0.40) and social identity (AIQ-SI; r=0.44).
It should be noted that the correlation between body image shame and body image guilt was
relatively large (r=0.59). Following the procedure established by Tangney, Burggraf, and
Wagner (1995), we conducted part correlations to isolate the unique variance of guilt and shame.
These correlations with the other variables are presented in Table 4 and support the validity of
the BIGSS.
We also noted the fairly strong correlations between body image guilt (BIGSS guilt) and guilt
associated with moral affects (TOSCA guilt) as well as between body image shame (BIGSS
shame) and shame associated with moral affects (TOSCA shame). Thus, we also conducted part
correlations to isolate the unique variance of body image guilt and guilt associated with moral
affects, and the unique variance of body image shame and shame associated with moral affects.
These correlations are also presented in Table 4 and support the validity of the BIGSS.

3.4. Gender differences

Since many of the scales that measure attitudes and behaviours associated with body image and
disordered eating have demonstrated gender differences, we compared the scores of university
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 71

Table 4
Relationship of body image guilt and shame to other measures of guilt, shame, body concerns, and identity (N=153)

BIGSS-Ga BIGSS-Sa BIGSS-Gb BIGSS-Sc BIGSS-Gd BIGSS-Se

1. BIGSS-G – 0.59** – – – 0.44**

2. BIGSS-S 0.59** – – – 0.45** –
3. TOSCA-G 0.48** 0.41** 0.30** 0.16* – 0.13*
4. TOSCA-S 0.41** 0.56** 0.14* 0.39** 0.17** –
5. SPA 0.60** 0.73** 0.21** 0.46** 0.48** 0.62**
6. BIBCQ-C 0.43** 0.56** 0.12* 0.38** 0.36** 0.53**
7. BIBCQ-I 0.39** 0.40** 0.19* 0.21** 0.27** 0.33**
8. AIQ-PI 0.19* 0.07 0.18* 0.05 0.04 0.04
9. AIQ-SI 0.32** 0.44** 0.07 0.31** 0.12* 0.31**

Coefficient alphas for data from the present study are included in parentheses after each scale below: BIGSS-G
(0.88)=Body Image Guilt and Shame Scale Guilt Score; BIGSS-S (0.91)=Body Image Guilt and Shame Sca-
le Shame Score; TOSCA-G (0.72)=Test of Self-Conscious Affect Guilt Score; TOSCA-S (0.79)=Test of Self-Con-
scious Affect Shame Score; SPA (0.93)=Social Physique Anxiety; BIBCQ-C (0.87)=Body Image/Body Concern
Questionnaire-Concern Score; BIDCQ-I (0.84)=Body Image/Body Concern Questionnaire-Importance Score; AIQ-PI
(0.77)=Aspects of Identity Questionnaire Personal Identity Score; AIQ-SI (0.86)=Aspects of Identity Ques-
tionnaire Social Identity Score
a
Zero order correlations
b
Part correlations in which shame was factored out from guilt
c
Part correlations in which guilt was factored out from shame
d
Part correlations in which TOSCA guilt was factored out from BIGSS guilt
e
Part correlations in which TOSCA shame was factored out from BIGSS shame
* P< 0.05
** P< 0.01

men and women on the BIGSS guilt and shame responses. The Levene test for equality of var-
iances indicated that the assumption of homogeneity of variance was violated for both the guilt
and the shame responses. As such, we adjusted the degrees of freedom and the standard error of
mean differences of each independent samples t-test to account for this violation. There was a
gender difference in the BIGSS guilt response scores: t (150.782)=4.13, P<0.001, SEM=1.64,
Z2=0.09, with women (M=42.89, SD=11.74) scoring higher on body image guilt than men
(M=36.11, SD=8.56). There was also a gender difference in the BIGSS shame response scores: t
(145.693)=8.02, P<0.001, SEM=1.56, Z2=0.27, with women (M=42.33, SD=11.85) scoring
higher on body image shame than men (M=29.82, SD=7.34). These results seem to reflect gen-
der differences in socialisation patterns of men and women relative to body image issues as well as
in relation to guilt and shame. Significant gender differences on the TOSCA guilt and shame
responses, the SPA Scale, the Personal and Social Identity subscales of the AIQ, and the BIBCQ
concern subscale provide support for this (Table 5).

4. Discussion

The results of the present study indicate that the BIGSS is a reliable and valid instrument for
measuring the level of guilt and shame associated with body image concerns. The factor analysis
72 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

Table 5
Tests for gender differences in age and the scores on the scales of the questionnaire packet

Variable t df P SEM Z2 CI Men (N=66) Women (N=87)

M(SD) M(SD)
a
BIGSS-G 4.13 150.78 < 0.001 1.64 0.09 ( 10.16, 3.40) 36.11(8.56) 42.89(11.74)
BIGSS-Sa 8.02 145.69 < 0.001 1.56 0.27 ( 15.60, 9.43) 29.82(7.34) 42.33(11.85)
TOSCA-G 4.48 151 < 0.001 1.06 0.12 ( 6.82, 2.64) 56.20(7.17) 60.93(5.90)
TOSCA-S 3.77 151 < 0.001 1.36 0.09 ( 7.85, 2.45) 39.05(8.51) 44.20(8.25)
SPAa 5.60 150.66 < 0.001 1.50 0.16 ( 11.32, 5.41) 28.73(7.75) 37.09(10.74)
AIQ-PI 2.04 151 0.043 0.83 0.03 ( 3.35, 0.54) 38.14(5.14) 39.84(5.09)
AIQ-SI 2.38 151 0.019 0.80 0.04 ( 3.50, 0.33) 21.92(4.72) 23.84(5.08)
BIBCQ-Ca 2.94 151.99 0.004 1.09 0.05 ( 5.83, 1.06) 24.98(5.64) 28.21(7.41)
BIBCQ-I 0.74 151 0.484 1.17 <0.01 ( 2.93, 1.33) 29.92(6.87) 29.10(7.37)
Age 0.34 151 0.730 0.79 <0.01 ( 1.83, 1.29) 21.30(4.52) 21.57(5.07)

N=153
a
Degrees of freedom and standard error of mean differences were adjusted to account for violations of the
assumption of homogeneity of variance.

of the responses on the BIGSS yielded a two-factor solution that seemed most related to body image
guilt and body image shame. An additional factor analysis of the guilt and shame responses on the
BIGSS and the TOSCA yielded a four-factor solution that provided the most parsimonious expla-
nation of the data. The four-factor solution seemed most related to body image guilt, general self-
conscious shame, general self-conscious guilt, and body image shame. However, the BIGSS guilt and
shame scores were positively related to the TOSCA guilt and shame scores. As a consequence, the
BIGSS and the TOSCA appear to be measuring different, yet related, dimensions of guilt and shame.
The reliability of the guilt and shame responses of the BIGSS was demonstrated at a high level
on the basis of internal consistency as determined by coefficient alpha. However, corrected item,
item–total correlations indicated that Item 10 should be removed for both the guilt and shame
responses based on low correlations.
The BIGSS guilt and shame dimensions also appear to have acceptable levels of construct
validity. Consistent with Crawford and Eklund’s (1994) conclusions, we found that both body
image guilt and body image shame were moderately correlated with social physique anxiety. Since
social physique anxiety is defined as the anxiety experienced by individuals in response to other
people’s aesthetic evaluations of their physiques (Hart et al., 1989), it seems that shame and guilt
about our body images may develop in response to the evaluations of others. Alternatively, as
individuals feel increased shame and guilt about their body images, they may be more sensitive to
other peoples’ evaluations of their physiques.
Consistent with Lutwak et al.’s (1998) and Tobey and Tunnell’s (1981) findings, construct
validity was also demonstrated by a low correlation between personal identity and BIGSS shame
scores, and a positive, moderate correlation between social identity and BIGSS shame scores.
Consistent with our predictions, BIGSS guilt and shame scores were moderately, positively
correlated with body image concern and body image importance. These results are also consistent
with the findings of Ricciardelli and colleagues (e.g. McCabe & Ricciardelli, 2000; Ricciardelli &
Glynn, 1998; Ricciardelli & McCabe, 1999). As levels of concern and the perceived importance of
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 73

the individual’s body parts increases, individuals may be increasingly concerned about the
inadequacy of those parts. As they focus concern on their various body parts, they may experi-
ence body image shame deriving from perceived discrepancies between their body shape relative
to the cultural ideal as well as relative to their personal standard. Depending on the mediation of
personality variables related to guilt and shame such as self-esteem (e.g. Tangney et al., 1995),
these discrepancies may register as either internal, global consonant with characterological self-
blame, manifest as shame, or as an internal, unstable attributions consonant with behavioural
self-blame, manifest as guilt.
Despite our attempts to construct a scale that was gender-neutral, gender differences were
found with respect to the BIGSS guilt and shame dimensions. However, this may be an indication
of true gender differences as opposed to a gender-bias implicit in the scale since gender differences
were also found on all other measures used in the present study with the exception of body image
importance. Despite recent studies showing that men are being more impacted than in the past
(Garner & Kearney-Cooke, 1997; Grogan, 1999), this difference may be an indication that the
sociocultural factors implicit in body image issues influence women to a greater degree than men
(e.g. Cash & Henry, 1995; Drewnowski & Yee, 1987; Fallon & Rozin, 1985; Frederick & Morri-
son, 1998; Garner et al., 1980; McCabe & Ricciardelli, 2000; Muth & Cash, 1997).

References

Baumeister, R. F., Reiss, H. T., & Delespaul, P. (1995). Subjective and experiential correlates of guilt in daily life.
Personality and Social Psychology Bulletin 21, 1256–1268.
Brownmiller, S. (1984). Femininity. New York: Simon and Schuster.
Bruch, H. (1978). The golden cage: the enigma of anorexia nervosa. Cambridge, MA: Harvard University Press.
Burggraf, S. A. (1995, April). Coping styles related to shame- and guilt-proneness. Poster presented at the annual meet-
ing of the Eastern Psychological Association, Boston, MA.
Burney, J., & Irwin, H. (2000). Shame and guilt in women with eating disorder symptomatology. Journal of Clinical
Psychology, 56, 51–61.
Cash, T., & Henry, P. (1995). Women’s body images: the results of a national survey in the USA. Sex Roles: A Journal
of Research, 33, 19–26.
Cheek, J. M., & Briggs, S. R. (1982). Self-consciousness and aspects of identity. Journal of Research in Personality, 16,
401–408.
Cheek, J. M., & Hogan, R. (1983). Self-concepts, self-presentations, and moral judgments. In J. Suls, &
A. G. Greenwald (Eds.), Psychological perspectives on the self (Vol. 2). Hillsdale, NJ: Lawrence Erlbaum Associates.
Cheek, J.M., Tropp, L.R., Chen, L.C., & Underwood, M.K. (1994, August). Identity orientations: personal, social, and
collective aspects of identity. Paper presented at the Annual Meeting of the American Psychological Association, Los
Angeles.
Crawford, S., & Eklund, R. (1994). Social physique anxiety, reasons for exercise, and attitudes towards exercise set-
tings. Journal of Sport and Exercise Psychology, 16, 431–448.
Drewnowski, A., & Yee, D. K. (1987). Men and body image: are males satisfied with their body weight? Psychosomatic
Medicine, 49, 626–634.
Fallon, A., & Rozin, P. (1985). Sex differences in perceptions of desirable body shape. Journal of Abnormal Psychology,
94, 102–105.
Fawkner, H.J., & McMurray, N. (1999). Body image: What men think and feel. Paper presented at the 34th Annual
Conference of the Australian Psychological Society, Hobart, Tasmania, Australia.
Frank, E. S. (1990). Shame and guilt in eating disorders. (Doctoral dissertation, Harvard Graduate School of Educa-
tion, Cambridge, MA, 1989). Dissertation Abstracts International, 50, 3896A.
74 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75

Frederick, C., & Morrison, C. (1998). A mediational model of social physique anxiety and eating disordered beha-
viours. Perceptual and Motor Skills, 86, 139–145.
Garner, D. M., & Garfinkel, P. (1980). Socio-cultural factors in the development of anorexia nervosa. Psychological
Medicine, 10, 647–656.
Garner, D. M., Garfinkel, P., Schwartz, D., & Thompson, M. (1980). Cultural expectations of thinness in women.
Psychological Reports, 47, 483–491.
Garner, D., & Kearney-Cooke, A. (1997). The 1997 body image survey results. Psychology Today, 30, 30–76.
Gray, S. (1977). Social aspects of body image: perception of normalcy of weight and affect of college undergraduates.
Perceptual and Motor Skills, 45, 1035–1040.
Grogan, S. (1999). Body image: understanding body dissatisfaction in men, women, and children. London: Routledge.
Hart, E. A., Leary, M. R., & Rejeski, W. (1989). The measurement of social physique anxiety. Journal of Sport and
Exercise Psychology, 11, 94–104.
Lutwak, N., & Ferrari, J. (1996). Moral affect and cognitive processes: differentiating shame from guilt among men
and women. Personality and Individual Differences, 21, 891–896.
Lutwak, N., & Ferrari, J. (1997). Understanding shame in adults: recalling perceptions of parental-bonding during
childhood. Journal of Nervous and Mental Diseases, 185, 595–598.
Lutwak, N., Ferrari, J., & Cheek, J. (1998). Shame, guilt, and identity in men and women: the role of identity orien-
tation and processing style in moral affects. Personality and Individual Differences, 25, 1027–1036.
McCabe, M., & Ricciardelli, L. (2000, October). Increasing muscle size among adolescents: differences between males
and females. Paper presented at the Australian Psychological Society annual conference, Canberra, Australia.
McKinley, N. M., & Hyde, J. S. (1996). The objectified body consciousness scale: development and validation. Psy-
chology of Women Quarterly, 20, 181–215.
Muth, J. L., & Cash, T. (1997). Body image attitudes: what difference does gender make? Journal of Applied Social
Psychology, 27, 1438–1452.
Noll, S., & Fredrickson, B. (1998). A mediational model linking self-objectification, body shame, and disordered eat-
ing. Psychology of Women Quarterly, 22, 623–636.
Phelps, L., Johnston, L., & Augustyniak, K. (1999). Prevention of eating disorders: identification of predictor vari-
ables. Eating Disorders: The Journal of Treatment and Prevention, 7, 99–108.
Rayner, T. Z., Singer, J. A., & O’Connor, L. E. (1995, April). Attributional style, proneness to guilt, and depression in
adolescents in recovery from substance abuse. Paper presented at the annual meeting of the Eastern Psychological
Association, Boston, MA.
Ricciardelli, L., & Glynn, H. (1998). Body dissatisfaction and disordered eating in men. Proceedings of the Body Image
Research Forum, 3, 25–35.
Ricciardelli, L., & McCabe, M. (1999). The Body Image Change Questionnaire. Obtainable from the authors, School of
Psychology, Deakin University, Burwood, Victoria, Australia.
Risby, J. (1999). Prediction of body dissatisfaction in males and females using physiological, body management and
sociocultural variables. Unpublished honours thesis, University of Tasmania.
Rodin, J., Silberstein, L., & Striegel-Moore, R. (1985). Women and weight: a normative discontent. In T. B. Sonderegger
(Ed.), Psychology and gender: Nebraska symposium on motivation (1984). Lincoln, NE: University of Nebraska Press.
Sanftner, J. L., Barlow, D., Marschall, D., & Tangney, J. P. (1995). The relation of shame and guilt to eating disorder
symptomatology. Journal of Social and Clinical Psychology, 14, 315–324.
Silberstein, L., Striegel-Moore, R., & Rodin, J. (1987). Feeling fat: a woman’s shame. In H. B. Lewis (Ed.), The role of
shame in symptom formation (pp. 89–108). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Tangney, J. P. (1995). Shame and guilt in interpersonal relationships. In J. P. Tangney, & K. W. Fischer (Eds.), Self-
conscious emotions: shame, guilt, embarrassment, and pride (pp. 114–139). New York: Guilford Press.
Tangney, J. P., Burggraf, S. A., & Wagner, P. E. (1995). Shame-proneness, guilt-proneness, and psychological symp-
toms. In J. P. Tangney, & K. W. Fischer (Eds.), Self-conscious emotions: shame, guilt, embarrassment, and pride
(pp. 343–367). New York: Guilford Press.
Tangney, J. P., Price, J., Barlow, D., Wagner, P., Marscall, D., Borenstein, J-K., Sanftner, J., Mohr, T., & Gramzow,
R. (1996a). Assessing individual differences in constructive versus destructive responses to anger across the lifespan.
Journal of Personality and Social Psychology, 70, 780–796.
 T. Thompson et al. / Personality and Individual Differences 34 (2003) 59–75 75

Tangney, J. P., Price, J., Barlow, D., Wagner, P., Marscall, D., Borenstein, J-K., Sanftner, J., Mohr, T., & Gramzow,
R. (1996b). Relation of shame and guilt to constructive versus destructive responses to anger across the lifespan.
Journal of Personality and Social Psychology, 70, 797–809.
Tangney, J. P., Wagner, P., Fletcher, C., & Gramzow, R. (1992). Shamed into anger? The relation of shame and guilt
to anger and self-reported aggression. Journal of Personality and Social Psychology, 62, 669–675.
Tangney, J., Wagner, P., & Gramzow, R. (1989). The test of self-conscious affect (TOSCA). Fairfax, VA: George
Mason University.
Thompson, M. G., & Schwartz, M. (1982). Life adjustment in women with anorexia and anorexia-like behaviour.
International Journal of Eating Disorders, 1, 47–60.
Tiggemann, M. (1994). Gender differences in the interrelationship between weight dissatisfaction, restraint, and self-
esteem. Sex Roles: A Journal of Research, 30, 319–327.
Tobey, E. L., & Tunnell, G. (1981). Predicting our impressions on others: effects of public self-consciousness and act-
ing, a self-monitoring subscale. Personality and Social Psychology Bulletin, 7, 661–669.

View publication stats