1

British Journal of Clinical Psychology (2015), 54, 115

2014 The British Psychological Society
www.wileyonlinelibrary.com

Social anxiety, submissiveness, and shame in men

and women: A moderated mediation analysis
Jacob Zimmerman, Amanda S. Morrison and Richard G. Heimberg*
Department of Psychology, Temple University, Philadelphia, Pennsylvania, USA
Objectives. Research suggests a positive relationship between social anxiety and
shame; however, few studies have examined this relationship or potential mechanisms.
Common behaviours of persons with social anxiety disorder (SAD), such as submissive
behaviours, may be more consistent with societal expectations of women than men and
therefore more likely to be associated with shame in socially anxious men than women.
We examined the hypothesis that submissive behaviours would mediate the relationship
between social anxiety and shame in men, but not in women, with SAD.
Design. Moderated mediation was examined in a cross-sectional dataset. Gender was
modeled to moderate the paths from social anxiety to submissive behaviours and from
submissive behaviours to shame. We also examined an alternative model of the
relationships among these variables and the potential contributory role of depression.
Methods. Men (n = 48) and women (n = 40) with SAD completed the Social
Interaction Anxiety Scale, Submissive Behaviour Scale, Internalized Shame Scale, and
Beck Depression Inventory.
Results. Analyses supported the hypothesized model. The relationship between
submissive behaviours and shame was greater in men than women with SAD; the
relationship between social anxiety and submissive behaviours was not. Controlling for
depression, moderation remained evident although diminished. Results for the comparison model did not support gender moderation.
Conclusions. Submissive behaviours mediated the relationship between social anxiety
and shame in men, but not women, with SAD. These findings provide preliminary
evidence for a model of shame in SAD and may help to further elucidate specific features
of SAD that differ between men and women.

Practitioner points
 Although researchers have argued that the display of submissive behaviours might allow the socially
anxious individual to limit or prevent attacks on the self, our results suggest that there are greater
costs, with regard to feelings of shame, associated with such behaviours for men.
 In men with SAD, the greater shame associated with submissive behaviours can be understood when
considering that socially anxious individuals appear to be particularly concerned with concealing
aspects of the self believed to violate perceived societal norms and that traditional masculine gender
roles revolve around the theme of dominance.
 Because the study was conducted in individuals with SAD, it is possible that the restricted range of
social anxiety severity may have precluded the observation of gender differences in the relationship
between social anxiety and submissive behaviour.
 Measures were administered in a cross-sectional design, which limits potential inferences of causality.
*Correspondence should be addressed to Richard G. Heimberg, Department of Psychology, Adult Anxiety Clinic, Temple
University, 1701 North 13th Street, Philadelphia, PA 19122-6085, USA (email: [email protected]).
DOI:10.1111/bjc.12057

Jacob Zimmerman et al.

Emerging evidence suggests that shame and submissive behaviours are both important
features of social anxiety (e.g., Weeks, Heimberg, & Heuer, 2011). Yet there is a lack of
research examining potential differences between socially anxious men and women on
these dimensions, despite theoretical suggestion for their consideration. Indeed, research
does highlight some gender differences in social anxiety disorder (SAD). Women with SAD
have reported greater social anxiety than their male counterparts (Turk et al., 1998), and
gender differences have also emerged when considering which specific social situations
provoke greater fear. For example, women reported greater fear and avoidance in
situations encountered in work settings (e.g., speaking up at a meeting), whereas men
with SAD endorsed greater fear urinating in public restrooms and returning goods to a
store (Turk et al., 1998). Epidemiologic studies have also revealed important gender
differences. Men with SAD reported more fears of dating (Xu et al., 2012) and more
frequently being single, unattached, and living alone than women with SAD (MacKenzie &
Fowler, 2013). Men were also more likely to have co-morbid externalizing disorders and
women more likely to have co-morbid internalizing disorders (Xu et al., 2012).
Considering that SAD is the fourth most common mental disorder, with a lifetime
prevalence of 12.1% (Kessler et al., 2005), and that most individuals with SAD report
moderate or greater impairment in basic areas of human functioning (Aderka et al., 2012),
it is important that potentially productive areas of inquiry are not left unexamined. Greater
awareness of potential differences between men and women with SAD has the potential
to stimulate research of a more contextual nature and ultimately may lead to more
individualized and effective treatments.
According to Gilberts (2001) ethological/psychobiological model of SAD, social
anxiety is triggered in environments in which individuals perceive themselves as
comparatively low in social status or at risk for losing status. From this perspective, a
primary aim of social anxiety is to facilitate non-violent interactions between individuals,
in part, via submissive behaviours, such as gaze avoidance or slumped posture. Submissive
behaviours inhibit, limit, or reduce subsequent attacks on the self by others (Gilbert,
2000). These behaviours may be related to feelings of shame or perceptions that the self
possesses unattractive physical attributes, personality characteristics, or has engaged in
unattractive behaviours (Gilbert, 2000).
Indeed, recent research has supported several of the ideas put forth by the
ethological/psychobiological model. Feelings of shame are positively associated with
social anxiety (Fergus, Valentiner, McGrath, & Jencius, 2010; Gilbert, 2000; Matos,
Pinto-Gouveia, & Gilbert, 2013; Weeks et al., 2011). Likewise, research has shown a
positive relationship between social anxiety and self-reported submissive behaviours
(Walters & Inderbitzen, 1998), as well as an inverse relationship between social anxiety
and actual displays of dominance (Walters & Hope, 1998). In one study, submissive
behaviours were examined in a sample of heterosexual male participants high and low in
social anxiety (Weeks et al., 2011). Participants engaged in a semi-structured role play
which involved competing with a male confederate for the attention of a female
confederate. Compared to their non-anxious counterparts, socially anxious male
participants displayed higher vocal pitch peaks, as well as greater body collapse both
indices of social submission. Interestingly, these two submissive behaviours, which were
moderately correlated with social anxiety, were also moderately correlated with a
self-report measure of shame.
Although shame has been considered, thus far, a likely precursor to submissive
displays, it is also plausible that submissive behavioural displays lead to increased feelings
of shame, and in particular, internalized shame. Whereas external shame refers to how

Shame and submissive behaviour in sad

one thinks others see the self and the resulting negative feelings and behaviours related to
beliefs about potential rejection or attack by others, internalized shame refers to the
negative beliefs or feelings a person has about his or her own attributes or behaviour
(Allan, Gilbert, & Goss, 1994; Gilbert, 2000; Goss, Gilbert, & Allan, 1994). The chronic and
global nature of internalized shame can also be distinguished from state shame, which
refers to the experience of shame as a transient and acute reaction in specific situations
(Del Rosario & White, 2006). Perhaps unsurprisingly, chronic, global internalized shame
has been associated with a wide variety of negative outcomes including depression,
addiction, eating disorders, sexual abuse, and trauma, as well as feelings of abandonment
and rejection (see Del Rosario & White, 2006, for a review).
Individuals who believe that their core characteristics and behaviours are socially
unacceptable and who place a great deal of value on adhering to societal expectations and
norms may be at high risk for experiencing internalized shame. There are reasons to
suspect that socially anxious individuals are among this group. Moscovitch (2009)
suggests that socially anxious individuals are uniquely and primarily concerned about
characteristics of self that they perceive as being deficient or contrary to perceived
societal expectations or norms (p. 125, italics in original). Indeed, greater self-portrayal
concerns (i.e., concerns that perceived negative self-attributes may be exposed and
scrutinized by others in social situations) were uniquely associated with a diagnosis of SAD
compared to a diagnosis of another anxiety disorder or no diagnosis (Moscovitch et al.,
2013). Self-portrayal concerns likely depend on culturally sanctioned norms for social
behaviour, given differing self-portrayal concerns across individuals from different ethnic
backgrounds (Moscovitch & Huyder, 2011). In this sense, submissive behaviours, which
represent a key behavioural feature of SAD (Gilbert, 2001), may represent a mediator
between social anxiety and internalized shame in men more so than in women, given
differences in gender norms.
Traditional gender stereotypes of men cluster around the theme of dominance. Men
are generally expected to be successful and powerful, to remain stoic, and avoid acting in
ways that may be interpreted as feminine (see Moss-Racusin, Phelan, & Rudman, 2010, for
a review). Submissive behaviours appear to conflict sharply with this theme. Therefore,
socially anxious men, concerned with not violating societal norms (Moscovitch, 2009),
may detect an inconsistency between their submissive behavioural displays and their
gender role. This perceived social transgression could ultimately give rise to internalized
shame in socially anxious men, more so than socially anxious women.
In addition to suspecting a stronger relationship between submissive behaviours and
internalized shame in men than women, there are also reasons to consider whether the
relationship between social anxiety and submissive behaviours is greater in men than in
women. One evolutionary perspective considers the differential importance of competition across genders. According to the evolutionary model of parental investment
(Trivers, 1972), females tend to be more selective when choosing a mate because they
invest more in their offspring. Notably, . . .whichever sex invests more efforts and
material resources in offspring becomes, in effect, the limiting resource for the
reproduction of the less investing sex and hence an object of competition (Wilson &
Daly, 1992, p. 290). In this way, men are expected to devote significantly more time and
energy competing for mating opportunities than women. In fact, men engage in
intra-sexual competition more than women (Buunk & Massar, 2012; Saad & Gill, 2001).
Thus, socially anxious men may encounter dominance threats more frequently than
socially anxious women. Taken together with the finding that socially anxious individuals
view themselves as the subordinate members of social groups (e.g., Weeks et al., 2011), it

Shame and submissive behaviour in sad

15

Pula, K., Mcpherson, S., & Parks, C. D. (2012). Invariance of a two-factor model of social dominance
orientation across gender. Personality and Individual Differences, 52, 385389. doi:10.1016/j.
paid.2011.10.046
Rapee, R. M., & Heimberg, R. G. (1997). A cognitive-behavioral model of anxiety in social phobia.
Behaviour Research and Therapy, 35, 741756. doi:10.1016/S0005-7967(97)00022-3
Rodebaugh, T. L., Heimberg, R. G., Brown, P. J., Fernandez, K. C., Blanco, C., Schneier, F. R., &
Liebowitz, M. R. (2011). More reasons to be straightforward: Findings and norms for two scales
relevant to social anxiety. Journal of Anxiety Disorders, 25, 623630. doi:10.1016/j.janxdis.
2011.02.002
Rodebaugh, T. L., Woods, C. M., & Heimberg, R. G. (2007). The reverse of social anxiety is not always
the opposite: The reverse-scored items of the Social Interaction Anxiety Scale do not belong.
Behavior Therapy, 38, 192206. doi:10.1016/j.beth.2006.08.001
Saad, G., & Gill, T. (2001). Sex differences in the ultimatum game: An evolutionary psychology
perspective. Journal of Bioeconomics, 3, 171193. doi:10.1023/A:1020583425623
Sidanius, J., Sinclair, S., & Pratto, F. (2006). Social dominance orientation, gender, and increasing
educational exposure. Journal of Applied Social Psychology, 36, 16401653. doi:10.1111/j.
0021-9029.2006.00074.x
Stopa, L., Brown, M. A., Luke, M. A., & Hirsch, C. R. (2010). Constructing a self: The role of
self-structure and self-certainty in social anxiety. Behaviour Research and Therapy, 48, 955
965. doi:10.1016/j.brat.2010.05.028
Storch, E. A., Roberti, J. W., & Roth, D. A. (2004). Factor structure, concurrent validity, and internal
consistency of the Beck Depression Inventory-Second Edition in a sample of college students.
Depression and Anxiety, 19, 187189. doi:10.1002/da.20002
Trivers, R. L. (1972). Parental investment and sexual selection. In B. C. Campbell (Ed.), Sexual
selection and the descent of man, 18711971 (pp. 136179). Chicago, IL: Aldine.
Turk, C. L., Heimberg, R. G., Orsillo, S. M., Holt, C. S., Gitow, A., Street, L. L., . . . Liebowitz, M. R.
(1998). An investigation of gender differences in social phobia. Journal of Anxiety Disorders,
12, 209223. doi:10.1016/S0887-6185(98)00010-3
Walters, K. S., & Hope, D. A. (1998). Analysis of social behavior in individuals with social phobia and
nonanxious participants using a psychobiological model. Behavior Therapy, 29, 387407.
doi:10.1016/S0005-7894(98)80039-7
Walters, K. S., & Inderbitzen, H. M. (1998). Social anxiety and peer relations among adolescents:
Testing a psychobiological model. Journal of Anxiety Disorders, 12, 183198. doi:10.1016/
S0887-6185(98)00008-5
Weeks, J. W., Heimberg, R. G., & Heuer, R. (2011). Exploring the role of behavioral submissiveness
in social anxiety. Journal of Social and Clinical Psychology, 30, 217249. doi:10.1521/jscp.
2011.30.3.217
Weeks, J. W., Lee, C., Reilly, A. R., Howell, A. N., France, C., Kowalsky, J. M., & Bush, A. (2012). The
sound of fear: Assessing vocal fundamental frequency as a physiological indicator of social
anxiety disorder. Journal of Anxiety Disorders, 26, 811822. doi:10.1016/j.janxdis.2012.07.
005
Wilson, M., & Daly, M. (1992). The man who mistook his wife for a chattel. In J. H. Barkow, L.
Cosmides & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of
culture (pp. 289322). New York, NY: Oxford University Press.
Wilson, J. K., & Rapee, R. M. (2006). Self-concept certainty in social phobia. Behaviour Research
and Therapy, 44, 113136. doi:10.1016/j.brat.2005.01.006
Xu, Y., Schneier, F., Heimberg, R. G., Princisvalle, K., Liebowitz, M. R., Wang, S., & Blanco, C. (2012).
Gender differences in social anxiety disorder: Results from the national epidemiologic sample on
alcohol and related conditions. Journal of Anxiety Disorders, 26, 1219. doi:10.1016/j.janxdis.
2011.08.006
Received 5 October 2013; revised version received 8 April 2014

Shame and submissive behaviour in sad

Gender

SBS

SIAS-S

ISS

Figure 1. Conceptual diagram of the proposed model (adapted from Hayes, 2013). Both the first-stage
pathway (i.e., from SIAS-S to SBS) and the second-stage pathway (i.e., from SBS to ISS) are moderated by
gender. SIAS-S = Social Interaction Anxiety Scale Straightforward Total Score; SBS = Submissive
Behaviour Scale; ISS = Internalized Shame Scale.

identified as non-Hispanic and 10.2% identified as Hispanic, with 1.1% omitting this item.
In terms of marital status, 14.8% of participants reported being married or living with a
partner, whereas 83.0% reported being either single (never married) or divorced/
separated, and 2.3% declined to respond. The average number of years of education was
14.68 (SD = 2.14). Men and women did not differ in terms of age, t(86) = 0.66, p = .55;
years of education, t(86) = 0.13, p = .20; ethnicity, v2(3, N = 83) = 1.98, p = .81; or
marital status v2(1, N = 86) = 2.19, p = .14.

Measures
Social Interaction Anxiety Scale Straightforward Total Score (SIAS-S; Rodebaugh, Woods, &
Heimberg, 2007)
The SIAS (Mattick & Clarke, 1998) assesses trait fears of social interaction in dyads and
groups and consists of 20 items scored on a 5-point Likert-type scale ranging from 0 (not at
all characteristic or true of me) to 4 (extremely characteristic or true of me). Many
empirical studies have demonstrated the good to excellent reliability and good construct
validity of the SIAS (for a review see Heimberg & Turk, 2002). However, psychometric
performance of the SIAS has been shown to be weakened by the presence of three
reverse-scored items (Rodebaugh et al., 2007, 2011). Rodebaugh et al. (2007) recommend
the removal of these three items when computing a total score, as it appears to strengthen
both the construct and factorial validity of the measure. Therefore, only the sum of the
straightforwardly worded items (i.e., the SIAS-S) was included in the current analyses.

Submissive Behaviour Scale (SBS; Allan & Gilbert, 1997)

The SBS is a 16-item, self-report measure, designed to assess trait tendencies to engage in
submissive social behaviours, such as avoiding eye contact or repeatedly apologizing for
minor mistakes. Participants indicate the degree to which they engage in each behaviour

Jacob Zimmerman et al.

on a 5-point scale ranging from 0 (never) to 4 (always). A total score is calculated by

summing all items. The SBS has been shown to demonstrate good internal consistency and
good retest reliability (Allan & Gilbert, 1997; OConnor, Berry, Weiss, & Gilbert, 2002).

Internalized Shame Scale (ISS; Cook, 1987)

The ISS is a self-report questionnaire designed to assess chronic, negative, global
evaluations of the self, such as, Compared to other people, I feel like I somehow never
measure up or I feel as if I am somehow defective as a person, like there is something
basically wrong with me. Participants rate items on a 5-point Likert-type scale ranging
from 0 (never) to 4 (almost always). The ISS consists of two subscales: the primary
24-item scale which measures internalized shame and a secondary 6-item scale which
assesses self-esteem and serves to prevent response set bias (Del Rosario & White, 2006).
In the present study, a total shame score was calculated by summing ratings from the
24-item shame subscale. The self-esteem subscale was excluded from analyses. The ISS has
demonstrated good to excellent internal consistency and good retest reliability (Cook,
1993; Del Rosario & White, 2006).

Beck Depression Inventory-II (BDI-II; Beck, Steer, & Brown, 1996)

The BDI-II is a 21-item self-report instrument designed to assess the existence and severity
of depressive symptoms, such as sadness and loss of interest, over the past 2 weeks.
Participants rate the severity of each symptom on a 4-point scale ranging from 0 to 3, with
higher scores indicating greater severity. The BDI-II has been shown to demonstrate good
internal consistency in outpatient and undergraduate populations (e.g., Beck, Steer, Ball,
& Ranieri, 1996; Storch, Roberti, & Roth, 2004).

Analytic plan
The predicted moderated mediation model corresponds to the first- and second-stage
moderation model outlined by Edwards and Lambert (2007). Moderation can occur in the
pathway from predictor to mediator (X to M) or from mediator to outcome (M to Y).
Moderated mediation is subsumed under the category of conditional indirect effects
(Preacher, Rucker, & Hayes, 2007), in which it is strongly advocated that bootstrapping be
used (see MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002; MacKinnon,
Lockwood, & Williams, 2004; see Preacher et al., 2007, for a review). For hypothesis
testing, if zero lies outside of the bootstrapped confidence interval, the null hypothesis of
no indirect effect is to be rejected (Preacher et al., 2007). Furthermore, the degree of an
indirect effect can vary at different levels of the moderator (Little, Card, Bovaird, Preacher,
& Crandall, 2007). In this way, a mediational model may hold for one group but not for
another, or there may be a differential effect in terms of the magnitude of an indirect effect
at different values of a moderator.
Analyses were performed using SPSS statistical software and macros provided by Hayes
and Matthes (2009) and Hayes (2013). Hayess (2013) SPSS macro for first- and
second-stage moderation models provides indices of the degree of moderation for the
first stage of the mediational model, but indices for the second stage of the mediational
model are conditional effects. In order to delineate which region (or regions) of the model
was responsible for an overall significant effect of moderated mediation, moderated
regression analyses were run for each pathway of the model. This moderation of effects

Shame and submissive behaviour in sad

(i.e., an interaction) occurs when the size or direction of a predictor variables effect on an
outcome variable depends on the value of the moderator variable (Hayes & Matthes,
2009).
Given that submissive behaviours and shame are also linked with depression and that
depression co-morbidity is common in SAD, we repeated our original analyses including
depression (i.e., the BDI-II) as a covariate. We also tested an alternative model in which the
roles of internalized shame and submissive behaviours were reversed, that is, internalized
shame was modelled to mediate the relationship between social anxiety and submissive
behaviours. Per the ethological/psychobiological model, it is feasible that feeling highly
ashamed of oneself would result in submissive behaviours to avoid competition with
dominant others. We had no hypotheses regarding whether this mediational path would
be moderated by gender; however, examination of this theoretically grounded alternative
model provided an opportunity to explore the utility of our originally hypothesized
pathways in men and women.

Results
Bivariate correlations
Men and women did not differ in their scores on the SIAS-S, t(86) = 0.87, p = .39; SBS,
t(86) = 0.31, p = .76; or ISS, t(86) = 0.57, p = .57. Correlations between the three
measures indicated strong positive bivariate relationships among social anxiety, submissive
behaviours, and internalized shame (SIAS-S/SBS, r = .49, p < .001; SBS/ISS, r = .61,
p < .001; SIAS-S/ISS, r = .59, p < .001). Correlations of social anxiety and submissive
behaviours remained significant and positive when examined separately for women
(r = .33, p < .05) and men (r = .63, p < .001). The relationship between submissive
behaviours and internalized shame also remained significant and positive for women
(r = .46, p < .01) and men (r = .74, p < .001). Social anxiety and internalized shame were
also significantly positively associated for both women (r = .46, p < .01) and men (r = .68,
p < .001). Two-tailed tests of the differences between independent correlations were
conducted using procedures outlined in Preacher (2002). The correlations for men and
women were not significantly different when examining the relationship between social
anxiety and submissive behaviours (SIAS-S/SBS, p = .07) or social anxiety and internalized
shame (SIAS-S/ISS, p = .13). However, there was a significant difference in the magnitude of
the correlations between submissive behaviours and internalized shame (SBS/ISS, p < .05).

Moderated mediation model

The test of moderated mediation with 5,000 bootstrapped samples revealed a conditional
indirect effect in the predicted direction (see Table 1). In both men and women,
the relationship between social anxiety and submissive behaviours was positive, as was
the relationship between submissive behaviours and internalized shame. For men, the
relationship between social anxiety and internalized shame was mediated by submissive
behaviours (indirect effect = 0.59; 95% bootstrap CI = 0.340.85). In contrast, this effect
was not significantly different from zero in women (indirect effect = 0.18; 95% bootstrap
CI = 0.000.57). In addition, analyses revealed a significantly greater indirect effect of
submissive behaviours in male than female participants (index = .32; 95% bootstrap
CI = 0.70 to 0.02).

Jacob Zimmerman et al.

Table 1. First- and second-stage moderated mediation model

Outcome
SBS

R
.50
B

R2
.25
SE

Constant
SIAS-S
Gender
SIAS-S 9 Gender

21.57
0.31
9.05
0.20

5.58
0.12
7.55
0.16

Outcome
ISS
Constant
SBS
SIAS-S
Gender
SBS 9 Gender

R2
.50
SE

R
.71
B
11.14
0.58
0.58
17.94
0.52

F
9.41
t

p
<.001
p

3.86
2.47
1.20
1.21

<.001
.02
.23
.23

F
20.95
t

p
<.001
p

1.23
2.68
3.86
1.61
1.71

.22
.01
<.001
.11
.09

9.03
0.22
0.15
11.16
0.31

Conditional indirect effect of SIAS-S on ISS

Mediator

Gender

Effect

Boot SE

Boot CI

SBS
SBS

Men
Women

0.56
0.18

0.13
0.14

.34.85
.00.57

Note. N = 88. Bootstrap sample size = 5,000. All path coefficients reported are unstandardized.
SIAS-S = Social Interaction Anxiety Scale Straightforward Total Score; SBS = Submissive Behaviour
Scale; ISS = Internalized Shame Scale; Boot CI = bootstrap confidence interval.

Moderated regression analyses

Results of additional moderated regression analyses at each pathway are presented in
Tables 2 and 3. The relationship between social anxiety and submissive behaviours was
not moderated by gender, B = 0.20, t(84) = 1.21, p > .05. However, the relationship
between submissive behaviours and shame was moderated by gender, B = 0.77,
t(84) = 2.40, p < .05; see Figure 2. Higher levels of submissive behaviours were
associated with higher internalized shame in both men and women, but this effect was
relatively stronger in men than women.

Additional analyses
Results of a moderated mediation model with 5,000 bootstrapped samples which
included depression as a covariate also revealed the hypothesized conditional indirect
effect. For men, the relationship between social anxiety and internalized shame was
mediated by submissive behaviours (indirect effect = 0.29; 95% bootstrap CI = 0.11
0.53). In contrast, this effect was not significantly different from zero in women (indirect
effect = 0.09; 95% bootstrap CI = 0.02 to 0.35). However, the magnitude of the indirect
effect did not differ between men and women (index = .21; 95% bootstrap CI = 0.47 to
0.03).

Shame and submissive behaviour in sad

Table 2. Regression results for simple mediation of first-stage pathway

Predictor
SIAS-S

Outcome
SBS

Constant
SIAS-S
Gender
SIAS-S 9 Gender

R
.50
B

R2
.25
SE

21.57
0.31
9.05
0.20

5.58
0.12
7.55
0.16

F
9.41
t
3.86
2.47
1.20
1.21

p
<.001
p
<.001
.02
.23
.23

Conditional effect of SIAS-S on SBS

Gender

SE

Men
Women

0.50
0.31

0.11
0.12

4.69
2.47

<.001
.02

Note. All regression coefficients reported are unstandardized. SIAS-S = Social Interaction Anxiety Scale
Straightforward Total Score; SBS = Submissive Behaviour Scale.

The test of the alternative moderated mediation model, in which internalized shame
mediated the relationship between social anxiety and submissive behaviours, with 5,000
bootstrap samples, revealed that neither path was moderated by gender.1

Discussion
Although previous research has established a positive association between social anxiety
and internalized shame, little is known about mechanisms involved in this relationship.
Consistent with study hypotheses, submissive behaviours mediated the positive relationship between social anxiety and internalized shame in men, but not in women, with SAD.
Follow-up tests to examine each stage of the moderated mediation revealed the second
pathway, that is, from submissive behaviours to internalized shame, to be moderated by
gender. In contrast, and inconsistent with study hypotheses, the relationship between
social anxiety and submissive behaviours was not moderated by gender.
Although the exhibition of submissive behaviours might allow the socially anxious
individual to limit or prevent attacks on the self (Gilbert, 2001), our results suggest that
there are gender differences in the costs associated with such behaviours. The stronger
positive relationship between submissive behaviours and internalized shame among men
than women with SAD provides initial support for the conceptualization of shame in social
anxiety as being associated with perceiving that one is not acting in a manner that is
socially acceptable (Gilbert, 2001). Traditional gender roles are differentiated for men and
women, with masculine gender roles revolving around the theme of dominance
(Moss-Racusin et al., 2010); thus, the behavioural submissiveness associated with social
anxiety would conflict more sharply with the expected behaviour for men than for
women. Given that socially anxious individuals are thought to be uniquely concerned
with concealing characteristics of the self that they feel are deficient or that violate
perceived societal norms (Moscovitch, 2009), the transgression of engaging in a

Results of the analyses of the alternative model are available from the corresponding author on request.

10

Jacob Zimmerman et al.

Table 3. Regression results for simple mediation of second-stage pathway

Predictor
SBS

Outcome
ISS

R2
.41
SE

R
.64
B

Constant
SBS
Gender
SBS 9 Gender

29.28
0.79
25.64
0.77

8.32
0.23
11.85
0.32

F
19.73
t

p
<.001
p

3.52
3.46
2.16
2.39

<.001
<.001
.03
.02

Conditional effect of SBS on ISS

Gender

SE

Men
Women

1.56
0.79

0.23
0.23

6.83
3.46

<.001
<.001

Note. All path coefficients reported are unstandardized. SBS = Submissive Behaviour Scale; ISS = Internalized Shame Scale.

100

Men
Women

ISS Total

80

60

40

20
10

20

30

40

50

60

SBS Total

Figure 2. Significant interaction of submissive behaviours as measured by the Submissive Behaviour

Scale (SBS) and gender, in predicting internalized shame as measured by the Internalized Shame Scale (ISS)
in the regression results for the simple mediation of the second-stage pathway.

submissive behaviour might elicit greater negative beliefs or feelings about the self (i.e.,
internalized shame) in men relative to women.
A complementary explanation also arises in considering how externalized shame (i.e.,
negative feelings resulting from how one thinks others see the self) might contribute to
internalized shame. Cognitive behavioural models of SAD highlight the important role of
negative self-perceptions in the development and maintenance of social anxiety (Clark &

Shame and submissive behaviour in sad

11

Wells, 1995; Heimberg, Brozovich, & Rapee, 2010; Rapee & Heimberg, 1997). However,
the self-perceptions of individuals with SAD might actually be unduly influenced by their
perception of how they are evaluated by others. Social anxiety has been associated with
reduced clarity or certainty about the self (Stopa, Brown, Luke, & Hirsch, 2010; Wilson &
Rapee, 2006), which may contribute to a greater reliance on others evaluation when
thinking about ones global self-concept (Wilson & Rapee, 2006). At the same time, men
have been found to encounter prejudice from others when they exhibit behaviour
inconsistent with their gender stereotype (Moss-Racusin et al., 2010). Thus, the current
finding of a stronger positive relationship between submissive behaviours and internalized shame in men than women with SAD may be due in part to the interaction of the
socially anxious individuals reliance on others evaluation in delineating ones own
self-concept coupled with the likely negative feedback a socially anxious man would
receive when exhibiting submissive behaviours. Although the current study cannot
address the complexity of these hypotheses, future research should consider the potential
role of externalized shame in the aforementioned relationships.
In an effort to better understand the validity of our preliminary interpretation of the
stronger positive relationship between submissive behaviours and shame in men than in
women with SAD, an alternative model was tested which reversed these two constructs in
the mediational chain. The results of this analysis revealed that neither path in the overall
model was moderated by gender. These results could perhaps be viewed as a slight
bolstering of confidence in our preliminary model and interpretations of the hypothesized
significant finding, but of course these interpretations should be considered preliminary
given the cross-sectional nature of our data.
To further elucidate the limits of our originally hypothesized model, self-reported
depression was included in the original model as a covariate. It is important to note that
depression is frequently co-morbid with SAD (e.g., Kessler, Stang, Wittchen, Stein, &
Walters, 1999) and that both submissive behaviours and shame are considered core
features of depression (Gilbert, 2000). The results of this modified analysis provided partial
support for the moderated mediation model. The indirect effect of submissive behaviours
on internalized shame remained significant only in men; however, the magnitude of the
indirect effect was no longer significantly different between men and women. This
suggests that the stronger relationship between submissive behaviours and internalized
shame in men than in women with SAD may be partially accounted for by depression.
In contrast with study hypotheses, the first-stage pathway of the mediation, that is, the
relationship between social anxiety and submissiveness, was not moderated by gender. In
both men and women with SAD, social anxiety was significantly positively associated with
submissive behaviours. There are several potential explanations for this, some methodological and others theoretical. Although one strength of the current study is the use of a
treatment-seeking sample of individuals with SAD, this aspect of design likely restricted
the range of scores on the social anxiety measure. It may be that gender differences in the
relationship between social anxiety and submissive behaviours would be observed if the
full range of social anxiety were examined. In other words, perhaps the relationship
between social anxiety and submissive behaviours is strong for men across the social
anxiety spectrum, whereas for women the relationship grows as social anxiety increases.
Alternatively, we may not have adequately addressed the relationship between social
anxiety and submissive behaviours. We suggested that this relationship may be stronger in
socially anxious men than women because men may encounter dominance threats more
frequently and therefore display submissiveness more frequently. Although it may be the
case that men encounter dominance threats more frequently, thereby eliciting more

12

Jacob Zimmerman et al.

frequent submissive behaviours, this argument may not speak to the relationship
between social anxiety and submissive behaviours. In other words, it does not explain
why the frequency of submissive behaviours increases as social anxiety increases, even
among men. In addition, this theory does not provide sufficient explanation of the
relationship between social anxiety and submissive behaviours in women. Most prior
work on the relationship between social anxiety and submissiveness has been conducted
either in combined samples of men and women (e.g., Walters & Hope, 1998; Walters &
Inderbitzen, 1998) or in samples of men only (Weeks et al., 2011).
A final consideration on this issue is that gender alone may not have adequately
captured our hypothesis regarding intra-sexual competition for mating opportunities. It
may be that a consideration of marital status in addition to gender would have been a
better targeted test of this hypothesis. However, analyses including marital status as a
second moderator were precluded by the low frequency of married men and women in
our sample (n = 4 and n = 9, respectively). In sum, further research is warranted to
investigate the causes and consequences of submissive behaviours in women, in addition
to delineating how the relationship between social anxiety and submissiveness may differ
between men and women with SAD.
The present study has several limitations that should also be noted. As mentioned
above, the restricted range of social anxiety severity may have precluded the observation
of gender differences in the relationship between social anxiety and submissiveness. The
study also relied on self-reports, rather than physiological or behavioural measures. Future
studies would benefit from the use of more objective measures of submissive behaviours
(e.g., Weeks et al., 2011, 2012). Also, measures were administered in a cross-sectional
design, which limits potential inferences of causality. Future studies would benefit from
adopting a longitudinal design to assess the temporal relationships of these constructs.
Results of the present study suggest several additional areas for future research.
Submissive behaviours were a stronger predictor of internalized shame in men than women
with SAD, which may suggest that internalized shame is a result of social norm violations.
However, it may also suggest that men who engaged in a high frequency of submissive
behaviours were more discouraged by their seemingly subordinate position than women,
perhaps due to a greater orientation towards social dominance (Pula, Mcpherson, & Parks,
2012; Sidanius, Sinclair, & Pratto, 2006). Future research is needed to explore these
potential explanations. Examining the effects of gender-atypical behaviour on internalized
shame in women with SAD would likely help to clarify this issue. Further, our results may
provide indirect support for the notion that socially anxious individuals may be particularly
vulnerable to the opinions of others in the development of a self-concept (Wilson & Rapee,
2006). Men who deviate from their dominance-oriented gender stereotype encounter
prejudice from others (Moss-Racusin et al., 2010), which may explain the stronger
relationship between submissive behaviours and the global negative self-evaluations of
internalized shame in men with SAD observed in the current analyses.
The present study provides preliminary evidence for the mediating role of submissive
behaviours in the relationship between social anxiety and internalized shame in men, but
not women, with SAD. Real or perceived social norm violations may instigate or
perpetuate feelings of shame in socially anxious individuals. Although causal interpretations of our findings are premature, the results of the present study build upon previous
research of gender differences in SAD (MacKenzie & Fowler, 2013; Turk et al., 1998; Xu
et al., 2012) and may help to extend emerging research on ethologically oriented
perspectives on SAD (Weeks et al., 2012) and notions of the self as it functions in SAD
(Moscovitch et al., 2013; Stopa et al., 2010).

Shame and submissive behaviour in sad

13

References
Aderka, I. M., Hofmann, S. G., Nickerson, A., Hermesh, H., Gilboa-Schechtman, E., & Marom, S.
(2012). Functional impairment in social anxiety disorder. Journal of Anxiety Disorders, 26,
393400. doi:10.1016/j.janxdis.2012.01.003
Allan, S., & Gilbert, P. (1997). Submissive behaviour and psychopathology. British Journal of
Clinical Psychology, 36, 467488. doi:10.1111/j.2044-8260.1997.tb01255.x
Allan, S., Gilbert, P., & Goss, K. (1994). An exploration of shame measures-II: Psychopathology.
Personality and Individual Differences, 17, 719722. doi:10.1016/0191-8869(94)90150-3
Beck, A. T., Steer, R. A., Ball, R., & Ranieri, W. F. (1996). Comparison of Beck Depression
Inventories-IA and -II in psychiatric outpatients. Journal of Personality Assessment, 67, 588
597. doi:10.1207/s15327752jpa6703_13
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for Beck Depression Inventory-II. San
Antonio, TX: Psychological Corporation.
Buunk, A. P., & Massar, K. (2012). Intrasexual competition among males: Competitive towards men,
prosocial towards women. Personality and Individual Differences, 52, 818821. doi:10.1016/
j.paid.2012.01.010
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G. Heimberg, M. R.
Liebowitz, D. A. Hope & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and
treatment (pp. 6993). New York, NY: Guilford Press.
Cook, D. (1987). Measuring shame: The Internalized Shame Scale. Alcoholism Treatment
Quarterly, 4, 197215. doi:10.1300/J020v04n02_12
Cook, D. (1993). The Internalized Shame Scale manual. Menomonie, WI: Channel Press.
Del Rosario, P. M., & White, R. M. (2006). The Internalized Shame Scale: Temporal stability, internal
consistency, and principal components analysis. Personality and Individual Differences, 41,
95103. doi:10.1016/j.paid.2005.10.026
Di Nardo, P. A., Brown, T. A., & Barlow, D. H. (1994). Anxiety Disorders Interview Schedule for
DSM-IV: Lifetime version (ADIS-IV-L). New York, NY: Oxford University Press.
Edwards, J. R., & Lambert, L. (2007). Methods for integrating moderation and mediation: A general
analytical framework using moderated path analysis. Psychological Methods, 12, 122. doi:10.
1037/1082-989X.12.1.1
Fergus, T. A., Valentiner, D. P., McGrath, P. B., & Jencius, S. (2010). Shame- and guilt- proneness:
Relationships with anxiety disorder symptoms in a clinical sample. Journal of Anxiety
Disorders, 24, 811815. doi:10.1016/j.janxdis.2010.06.002
Gilbert, P. (2000). The relationship of shame, social anxiety and depression: The role of the
evaluation of social rank. Clinical Psychology and Psychotherapy, 7, 174189. doi:10.1002/
1099-0879(200007)7:3<174::AID-CPP236>3.0.CO;2-U
Gilbert, P. (2001). Evolution and social anxiety: The role of attraction, social competition, and social
hierarchies. Psychiatric Clinics of North America, 24, 723751. doi:10.1016/S0193-953X(05)
70260-4
Goss, K., Gilbert, P., & Allan, S. (1994). An exploration of shame measures-I: The other as shamer
scale. Personality and Individual Differences, 17, 713717. doi:10.1016/0191-8869(94)
90149-X
Hayes, A. F. (2013). An introduction to mediation, moderation, and conditional process analysis:
A regression-based approach. New York, NY: Guilford Press.
Hayes, A. F., & Matthes, J. (2009). Computational procedures for probing interactions in OLS and
logistic regression: SPSS and SAS implementations. Behavior Research Methods, 41, 924936.
doi:10.3758/BRM.41.3.924
Heimberg, R. G., Brozovich, F. A., & Rapee, R. M. (2010). A cognitive-behavioral model of social
anxiety disorder: Update and extension. In S. G. Hofmann & P. M. DiBartolo (Eds.), Social
anxiety: Clinical, developmental, and social perspectives (2nd ed., pp. 395422). New York,
NY: Academic Press. doi:10.1016/B978-0-12-375096-9.00015-8

14

Jacob Zimmerman et al.

Heimberg, R. G., & Turk, C. L. (2002). Assessment of social phobia. In R. G. Heimberg & R. E. Becker
(Eds.), Cognitive-behavioral group therapy for social phobia: Basic mechanisms and clinical
strategies (pp. 107126). New York, NY: Guilford Press.
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime
prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey
Replication. Archives of General Psychiatry, 62, 593602. doi:10.1001/archpsyc.62.6.593
Kessler, R. C., Stang, P., Wittchen, H. U., Stein, M., & Walters, E. E. (1999). Lifetime co-morbidities
between social phobia and mood disorders in the US National Comorbidity Survey.
Psychological Medicine, 29, 555567. doi:10.1017/S0033291799008375
Little, T. D., Card, N. A., Bovaird, J. A., Preacher, K. J., & Crandall, C. S. (2007). Structural equation
modeling of mediation and moderation with contextual factors. In T. D. Little, J. A. Bovaird & N.
A. Card (Eds.), Modeling contextual effects in longitudinal studies (pp. 207230). Mahwah, NJ:
Lawrence Erlbaum.
MacKenzie, M. B., & Fowler, K. F. (2013). Social anxiety disorder in the Canadian population:
Exploring gender differences in sociodemographic profile. Journal of Anxiety Disorders, 27,
427434. doi:10.1016/j.janxdis.2013.05.006
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of
methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83
104. doi:10.1037/1082-989X.7.1.83
MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect:
Distribution of the product and resampling methods. Multivariate Behavioral Research, 39,
99128. doi:10.1207/s15327906mbr3901_4
Maner, J. K., Miller, S. L., Schmidt, N. B., & Eckel, L. A. (2008). Submitting to defeat: Social anxiety,
dominance threat, and decrements in testosterone. Psychological Science, 19, 764768. doi:10.
1111/j.1467-9280.2008.02154.x
Matos, M., Pinto-Gouveia, J., & Gilbert, P. (2013). The effect of shame and shame memories on
paranoid ideation and social anxiety. Clinical Psychology and Psychotherapy, 20, 334349.
doi:10.1002/cpp.1766
Mattick, R. P., & Clarke, J. (1998). Development and validation of measures of social phobia scrutiny
fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455470. doi:10.
1016/S0005-7967(97)10031-6
Mehta, P. H., & Josephs, R. A. (2006). Testosterone change after losing predicts the decision to
compete again. Hormones and Behavior, 50, 684692. doi:10.1016/j.yhbeh.2006.07.001
Moscovitch, D. A. (2009). What is the core fear in social phobia? A new model to facilitate
individualized case conceptualization and treatment. Cognitive and Behavioral Practice, 16,
123134. doi:10.1016/j.cbpra.2008.04.002
Moscovitch, D. A., & Huyder, V. (2011). The Negative Self-Portrayal Scale: Development, validation,
and application to social anxiety. Behavior Therapy, 42, 183196. doi:10.1016/j.beth.2010.04.
007
Moscovitch, D. A., Rowa, K., Paulitzki, J. R., Ierullo, M. D., Chiang, B., Antony, M. M., & McCabe, R. E.
(2013). Self-portrayal concerns and their relation to safety behaviors and negative affect in social
anxiety disorder. Behaviour Research and Therapy, 51, 476486. doi:10.1016/j.brat.2013.05.
002
Moss-Racusin, C. A., Phelan, J. E., & Rudman, L. A. (2010). When men break the gender rules: Status
incongruity and backlash against modest men. Psychology of Men & Masculinity, 11, 140151.
doi:10.1037/a0018093
OConnor, L. E., Berry, J. W., Weiss, J., & Gilbert, P. (2002). Guilt, fear, submission, and empathy in
depression. Journal of Affective Disorders, 71, 1927. doi:10.1016/S0165-0327(01)00408-6
Preacher, K. J. (2002). Calculation for the test of the difference between two independent
correlation coefficients [Computer software]. Retrieved from http://www.quantpsy.org
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses:
Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185227. doi:10.
1080/00273170701341316

Shame and submissive behaviour in sad

15

Pula, K., Mcpherson, S., & Parks, C. D. (2012). Invariance of a two-factor model of social dominance
orientation across gender. Personality and Individual Differences, 52, 385389. doi:10.1016/j.
paid.2011.10.046
Rapee, R. M., & Heimberg, R. G. (1997). A cognitive-behavioral model of anxiety in social phobia.
Behaviour Research and Therapy, 35, 741756. doi:10.1016/S0005-7967(97)00022-3
Rodebaugh, T. L., Heimberg, R. G., Brown, P. J., Fernandez, K. C., Blanco, C., Schneier, F. R., &
Liebowitz, M. R. (2011). More reasons to be straightforward: Findings and norms for two scales
relevant to social anxiety. Journal of Anxiety Disorders, 25, 623630. doi:10.1016/j.janxdis.
2011.02.002
Rodebaugh, T. L., Woods, C. M., & Heimberg, R. G. (2007). The reverse of social anxiety is not always
the opposite: The reverse-scored items of the Social Interaction Anxiety Scale do not belong.
Behavior Therapy, 38, 192206. doi:10.1016/j.beth.2006.08.001
Saad, G., & Gill, T. (2001). Sex differences in the ultimatum game: An evolutionary psychology
perspective. Journal of Bioeconomics, 3, 171193. doi:10.1023/A:1020583425623
Sidanius, J., Sinclair, S., & Pratto, F. (2006). Social dominance orientation, gender, and increasing
educational exposure. Journal of Applied Social Psychology, 36, 16401653. doi:10.1111/j.
0021-9029.2006.00074.x
Stopa, L., Brown, M. A., Luke, M. A., & Hirsch, C. R. (2010). Constructing a self: The role of
self-structure and self-certainty in social anxiety. Behaviour Research and Therapy, 48, 955
965. doi:10.1016/j.brat.2010.05.028
Storch, E. A., Roberti, J. W., & Roth, D. A. (2004). Factor structure, concurrent validity, and internal
consistency of the Beck Depression Inventory-Second Edition in a sample of college students.
Depression and Anxiety, 19, 187189. doi:10.1002/da.20002
Trivers, R. L. (1972). Parental investment and sexual selection. In B. C. Campbell (Ed.), Sexual
selection and the descent of man, 18711971 (pp. 136179). Chicago, IL: Aldine.
Turk, C. L., Heimberg, R. G., Orsillo, S. M., Holt, C. S., Gitow, A., Street, L. L., . . . Liebowitz, M. R.
(1998). An investigation of gender differences in social phobia. Journal of Anxiety Disorders,
12, 209223. doi:10.1016/S0887-6185(98)00010-3
Walters, K. S., & Hope, D. A. (1998). Analysis of social behavior in individuals with social phobia and
nonanxious participants using a psychobiological model. Behavior Therapy, 29, 387407.
doi:10.1016/S0005-7894(98)80039-7
Walters, K. S., & Inderbitzen, H. M. (1998). Social anxiety and peer relations among adolescents:
Testing a psychobiological model. Journal of Anxiety Disorders, 12, 183198. doi:10.1016/
S0887-6185(98)00008-5
Weeks, J. W., Heimberg, R. G., & Heuer, R. (2011). Exploring the role of behavioral submissiveness
in social anxiety. Journal of Social and Clinical Psychology, 30, 217249. doi:10.1521/jscp.
2011.30.3.217
Weeks, J. W., Lee, C., Reilly, A. R., Howell, A. N., France, C., Kowalsky, J. M., & Bush, A. (2012). The
sound of fear: Assessing vocal fundamental frequency as a physiological indicator of social
anxiety disorder. Journal of Anxiety Disorders, 26, 811822. doi:10.1016/j.janxdis.2012.07.
005
Wilson, M., & Daly, M. (1992). The man who mistook his wife for a chattel. In J. H. Barkow, L.
Cosmides & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of
culture (pp. 289322). New York, NY: Oxford University Press.
Wilson, J. K., & Rapee, R. M. (2006). Self-concept certainty in social phobia. Behaviour Research
and Therapy, 44, 113136. doi:10.1016/j.brat.2005.01.006
Xu, Y., Schneier, F., Heimberg, R. G., Princisvalle, K., Liebowitz, M. R., Wang, S., & Blanco, C. (2012).
Gender differences in social anxiety disorder: Results from the national epidemiologic sample on
alcohol and related conditions. Journal of Anxiety Disorders, 26, 1219. doi:10.1016/j.janxdis.
2011.08.006
Received 5 October 2013; revised version received 8 April 2014