Research Methods in Psychology

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=vjrl20

The Journal of Psychology Interdisciplinary and Applied

ISSN: 0022-3980 (Print) 1940-1019 (Online) Journal homepage: http://www.tandfonline.com/loi/vjrl20

Faces of Shame: Implications for Self-Esteem, Emotion Regulation, Aggression, and Well-Being

Patrizia Velotti, Carlo Garofalo, Federica Bottazzi & Vincenzo Caretti

To cite this article: Patrizia Velotti, Carlo Garofalo, Federica Bottazzi & Vincenzo Caretti (2017) Faces of Shame: Implications for Self-Esteem, Emotion Regulation, Aggression, and Well-Being, The Journal of Psychology, 151:2, 171-184, DOI: 10.1080/00223980.2016.1248809

To link to this article: https://doi.org/10.1080/00223980.2016.1248809

© 2017 The Author(s). Published by Taylor & Francis© Patrizia Velotti, Carlo Garofalo, Federica Bottazzi, and Vincenzo Caretti.

Published online: 18 Nov 2016.

Submit your article to this journal

Article views: 5834

View Crossmark data

Citing articles: 6 View citing articles

Faces of Shame: Implications for Self-Esteem, Emotion Regulation, Aggression, and Well-Being

Patrizia Velottia, Carlo Garofalo b, Federica Bottazzia, and Vincenzo Carettic

aUniversity of Genoa; bTilburg University; cLUMSA University

ARTICLE HISTORY Received 28 June 2016 Accepted 9 October 2016

ABSTRACT There is an increasing interest in psychological research on shame experiences and their associations with other aspects of psychological functioning and well-being, as well as with possible maladaptive outcomes. In an attempt to confirm and extend previous knowledge on this topic, we investigated the nomological network of shame experiences in a large community sample (N D 380; 66.1% females), adopting a multidimensional conceptualization of shame. Females reported higher levels of shame (in particular, bodily and behavioral shame), guilt, psychological distress, emotional reappraisal, and hostility. Males had higher levels of self-esteem, emotional suppression, and physical aggression. Shame feelings were associated with low self- esteem, hostility, and psychological distress in a consistent way across gender. Associations between characterological shame and emotional suppression, as well as between bodily shame and anger occurred only among females. Moreover, characterological and bodily shame added to the prediction of low self-esteem, hostility, and psychological distress above and beyond the influence of trait shame. Finally, among females, emotional suppression mediated the influence of characterological shame on hostility and psychological distress. These findings extend current knowledge on the nomological net surrounding shame experiences in everyday life, supporting the added value of a multidimensional conceptualization of shame feelings.

KEYWORDS Aggression; emotion regulation; experiences of shame scale; psychological distress; shame; self-esteem; suppression

Shame is commonly defined as an intense negative emotion characterized by the perception of a global devaluation of the self (Tangney & Dearing, 2002). Shame feel- ings are often triggered by social events in which a drop of personal status or feelings of rejection are perceived. Of note, rather than representing a unidimensional construct, shame could actually refer to different aspects of the self, such as behaviors or body characteristics, as well to the broader identity (Andrews, Qian, & Valentine, 2002; Hej- denberg & Andrews, 2011). Specifically, a multidimensional conceptualization of shame has been proposed (Andrews et al., 2002) to identify: (a) experiences of characterologi- cal shame (i.e., regarding personal habits, manner with others, the kind of person one is, and personal skills); (b) experiences of behavioral shame (i.e., referred to doing

CONTACT Patrizia Velotti [email protected], Department of Educational Sciences, University of Genoa, Corso Andrea Podest�a, 2, 16126, Genoa, Italy. P. Velotti and C. Garofalo contributed equally to this work. © 2017 Patrizia Velotti, Carlo Garofalo, Federica Bottazzi, and Vincenzo Caretti. Published by Taylor & Francis. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http:// creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.

THE JOURNAL OF PSYCHOLOGY 2017, VOL. 151, NO. 2, 171–184 http://dx.doi.org/10.1080/00223980.2016.1248809

something wrong, saying something stupid, and failing in competitive contexts); and (c) bodily shame (i.e., referred to being ashamed of one’s physical appearance). There is substantial evidence linking shame with psychopathology in general, and internalizing symptoms in particular (Andrews et al., 2002; Velotti, Elison, & Garofalo, 2014). Yet, associations between specific experiences of shame and other psychological mechanisms have sparsely been investigated. In the current study, we sought to provide a fine- grained analysis of the nomological network surrounding shame feelings, adopting a multidimensional conceptualization of shame experiences. The concept of nomological network refers to a group of constructs that are theoretically or empirically expected to show consistent linkages, and as such nomological network analysis is used to measure construct validity (Cronbach & Meehl, 1955). Specifically, we aimed at confirming asso- ciations between shame feelings and psychopathological distress (Andrews et al., 2002). Furthermore, we sought to expand current knowledge broadening the scope of the nomological network of shame. To this end, we first reviewed prior literature to iden- tify possible correlates of shame experiences.

Shame and Self-Esteem

Frequent experiences of shame may eventually crystallize into trait-like shame prone- ness. Trait shame, in turn, involves a particularly painful, and often incapacitating, negative feeling involving a sense of inferiority, hopelessness, and helplessness, as well as a desire to hide personal flaws (Andrews et al., 2002). Accordingly, it has been pro- posed that experiences of shame are tightly linked with fluctuation in self-esteem, and it is plausible that frequent experiences of shame could be conceptually related to chronically low levels of self-esteem (Elison, Garofalo, & Velotti, 2014). Furthermore, low levels of self-esteem could increase the individual vulnerability to experience nega- tive emotional states, including shame. Accordingly, although the directionality of their association is not clear, several studies have reported a substantial relation between low self-esteem and negative emotions, such as guilt and shame (Garofalo, 2015; Marshall, Marshall, Serran, & O’Brien, 2009). Of note, both self-esteem and negative emotions have been linked to increased aggressive tendencies, yet few studies have empirically tested associations between shame and aggression.

Shame and Aggression

The link between shame and aggression was proposed by several scholars, and some went so far as to say that all forms of violence are anticipated by feelings of shame and humiliation (Gilligan, 1996). From this perspective, early experiences of rejection and abuse might lead to shame-proneness in adulthood. In turn, individuals with high levels of shame-proneness may believe that resorting to aggression and violence is the only possible way to get rid of their shame feelings. From an evolutionary perspective, the experience of shame early in the development may be later replaced by a condition of chronic anger, adopted as a means to keep others away so that shame feelings cannot be detected or triggered (Farmer & Andrews, 2009). Alternatively, the perception of self-devaluation – which is implicit in shame experiences – may represent the first step of a chain that connects shame and aggression (Elison et al., 2014). Specifically, the

172 P. VELOTTI ET AL.

sequence could begin with a devaluation of the self that causes shame; in turn, shame feelings can lead to the experience of substantial anger and to the expression of aggres- sive behavior towards the source of the initial devaluation (Elison et al., 2014; Velotti et al., 2014). In line with this assumption, in a recent study shame experiences were associated with trait anger, and this relation was accounted for by the role of angry reactions to criticisms (Hejdenberg & Andrews, 2011). Specifically, behavioral shame was linked to both proneness toward angry reactions and trait anger, whereas charac- terological and bodily shame were only related to angry reactions to criticisms (Hejden- berg & Andrews, 2011). In this chain, maladaptive emotion regulation may play a mediating role (Garofalo, Holden, Zeigler-Hill, & Velotti, 2016; Roberton, Daffern & Bucks, 2012). Yet, this possibility has not been empirically tested so far.

Shame and Emotion Regulation

Shame is considered among the emotions that are more difficult to regulate (Elison et al., 2014). Of note, the way people regulate emotions has important consequences for their well- being (Gross & John, 2003). Cognitive reappraisal and expressive suppression have been identified among the emotion regulation strategies that people use more often (Gross & John, 2003; Gross & Levenson, 1993). Cognitive reappraisal entails thinking about an upset- ting situation from a different angle in order to alter its meaning and modulate its emotional impact. Expressive suppression involves an attempt to inhibit or reduce the outward expres- sion of an ongoing emotional experience (i.e., not showing the emotion that one is feeling). In general, reappraisal and suppression are inversely related to a wide range of outcomes in the domains of subjective well-being, affectivity, and social relationships. Specifically, reap- praisal is typically associated with better, and suppression with poorer, outcomes (Gross & John, 2003). A recent experimental study has shown that trait shame was linked with emo- tional suppression (Lanteigne, Flynn, Eastabrook, & Hollenstein, 2014). This is important, because adopting maladaptive emotion regulation strategies for shame may ultimately lead to both internalizing (e.g., psychological distress) and externalizing (e.g., aggression) psycho- pathological symptoms (Elison, Pulos, & Lennon, 2006; Velotti et al., 2014). Therefore, it is possible that shame is associated with maladaptive emotion regulation, which in turn may explain the associations that shame has with psychopathological symptoms and aggression.

Overview of the Current Study

In the present study we sought to: (1) confirm and extend prior research on the nomological network of shame, investigating its associations with measures of psychological distress; self- esteem, aggression, and emotion regulation; and (2) examine the possible mediating role of maladaptive emotion regulation (i.e., emotional suppression) in the association between trait shame and external correlates (i.e., self-esteem, aggression and general psychopathology). Since gender differences in shame experiences (Andrews et al., 2002), emotion regulation (John & Gross, 2003), aggression (Fossati, Maffei, Acquarini, & Di Ceglie, 2003), and psychopathological distress (Prunas, Sarno, Preti, Madeddu, & Perugini, 2012) have consist- enly been reported, we also examined gender differences in mean levels and patterns of associations.

THE JOURNAL OF PSYCHOLOGY 173

Method

Participants and Procedures

The total sample comprised 380 adult participants (66.1% women, N D 251) recruited from the community. Participants were recruited by psychology graduate students with a snowball sam- pling technique: they started recruiting 5 participants from their acquaintances and asked them to recruit participants among their social networks. Men (Mage D 31.00, SD D 11.42, range 19–63) were slightly but significantly older than women (Mage D 28.50, SD D 9.88, range 18–58), t(225.85) D 2.11, p D .04. All participants were Italian. The majority of participants were university students (N D 225, 59.2%), whereas 6.3% (N D 24) were unemployed, 23.9% (N D 91) were employees, 9.5% (N D 36) were self-employed, and 1.1% (N D 4) were retired. Most participants (N D 313, 82.4%) reported to be (or to had been in the past) in a significant intimate relationship lasted at least 6 months. Finally, 64 participants (16.8%) did not have children, whereas 316 of them (83.2%) had at least one kid. All participants voluntarily and anonymously took part in the study and provided written informed consent. Participants were administered the questionnaires described below and returned them in a sealed envelope to ensure confidentiality. The local Institutional Review Board formally approved all procedures.

Measures

All measures were self-report Likert-type questionnaires. Participants were administered the Italian versions of the following measures. The translated items of all measures were obtained by the authors of the corresponding published Italian adaptations (see references below).

Experience of Shame Scale (ESS; Andrews et al., 2002) The ESS is a 25-item questionnaire designed to capture the experience of shame across three components: characterological shame (sample item: ‘Have you felt ashamed of any of your personal habits?’; a D .85), behavioral shame (sample item: ‘Do you feel ashamed when you do something wrong?’; a D .85), and bodily shame (sample item: ‘Have you felt ashamed of your body or any part of it?’; a D .87). The sum of these three components provides an over- all index of shame feelings (a D .91). In the present study, the Italian version of the ESS was used (Caretti, Craparo, & Schimmenti, 2010), which substantially replicated the psychomet- ric properties of the original version.

Differential Emotions Scale-IV (DES-IV; Izard et al., 1993) The DES-IV consists of 36 items aimed at capturing the frequency of the experience of specific positive and negative emotions in the daily life. For the purpose of this study, only the six items assessing shame (e.g., ‘Feel embarrassed when anybody sees you make a mis- take’; a D .63) and guilt (e.g., ‘Feel regret, sorry about something you did’; a D .68) of the Italian adaptation of the DES-IV (Zavattini et al., 2015) were administered.

Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965) The 10-item RSES was used to assess the general level of self-esteem (e.g., ‘On the whole, I am satisfied with myself’). The Italian version of the RSES has shown sound psychometric properties (Prezza, Trombaccia, & Armento, 1997). Cronbach’s alpha was .88 in the present study.

174 P. VELOTTI ET AL.

Symptom Checklist-90-Revised (SCL-90-R; Derogatis, 1994) The SCL-90-R is a 90-item inventory designed to measure general psychopathological distress suffered in the past month. The SCL-90-R estimates a global index of psychopathol- ogy (Global Severity Index, GSI), by averaging all item scores (e.g., ‘To what extent do you feel/have you felt blue in the last month?’), rated on a Likert scale (a D .97). The Italian adaptation of the SCL-R-90 was used in this study (Prunas et al., 2012).

Emotion Regulation Questionnaire (ERQ; Gross & John, 2003) The 10-item ERQ was administered to assess individual differences in two emotion regula- tion strategies: cognitive reappraisal (e.g., ‘I control my emotions by changing the way I think about the situation I’m in’; a D .87) and expressive suppression (e.g., ‘I keep my emo- tions to myself’; a D .79). For the purpose of this study, we used the Italian version of the ERQ (Balzarotti, John, & Gross, 2010), which has shown adequate psychometric properties.

Aggression Questionnaire (AQ; Buss & Perry, 1992) The AQ is a 29-item instrument composed by four subscales: physical aggression (e.g., ‘Once in a while I can’t control the urge to strike another person’; a D .83); verbal aggression (e.g., ‘I can’t help getting into arguments when people disagree with me’; a D .76); anger (‘I some- times feel like a powder keg ready to explode’; a D 80); and hostility (e.g., ‘When people are especially nice, I wonder what they want’; a D .81). The AQ total score represent an index of trait aggression (a D 90). The Italian version of the AQ has shown good reliability and valid- ity (Fossati et al., 2003), and has been used in the present study.

Data Analytic Strategy

Descriptive statistics were computed for all study variables. Gender differences and associa- tions with socio-demographic variables of interest were evaluated with one-way between- groups univariate or multivariate analyses of covariance (i.e., ANCOVA or MANCOVA, respectively). Pillai’s Trace was used as the most robust test statistic, and Partial Eta squared (h2partial) was chosen as an estimate of the effect size of the univariate F tests. Pearson prod- uct-moment correlation coefficients among all measures were calculated. The homogeneity of correlation coefficients across gender was tested using the appropriate z statistic (Cohen, Cohen, West, & Aiken, 2003). To test whether the ESS scales explained additional variance in previously significant correlates, above and beyond the effect of DES-IV-assessed shame, hierarchical multiple regression analyses were conducted. The Variance Inflation Factor (VIF) was adopted to assess collinearity. Finally, to test the hypothesized indirect effect of shame on both aggression and psychopathological distress through the mediation of mal- adaptive emotion regulation, hierarchical regression and bootstrap analyses were conducted using the PROCESS Macro for SPSS (Hayes, 2013). All analyses were carried out holding constant the effect of age.

Results

Descriptive Analyses and Gender Differences

Descriptive statistics and gender differences are presented in Table 1.

THE JOURNAL OF PSYCHOLOGY 175

Females reported an overall greater level of ESS-assessed shame than males, and lower levels of self-esteem. Also, MANCOVA results revealed that there was a statistically signifi- cant difference between male and female participants on the combined ESS scale scores, F(3, 372) D 26.57, p < .001, Pillai’s Trace D .18, h2partial D .18. Specifically, females scored higher than males on behavioral and bodily shame. Females also reported significantly greater levels of DES-IV-assessed guilt and shame, multivariate F(2, 340) D 5.12, p < .01, Pillai’s Trace D .03, h2partial D .03. A subsequent ANCOVA revealed a significant gender dif- ference on the SCL-90-R GSI score. Furthermore, a significant multivariate effect of gender occurred on the combined ERQ scales, F(2, 375) D 9.52, p < .001, Pillai’s Trace D .05, h2par- tial D .05; in particular, females reported higher levels of cognitive reappraisal, and lower levels of expressive suppression than males. Finally, a significant difference across gender occurred when the AQ subscales were entered as combined dependent variables, F(4, 373) D 19.28, p < .001, Pillai’s Trace D .17, h2partial D .17, but not when males and females when compared on the AQ total score using an ANCOVA design. When the AQ subscales were considered separately, only two differences reached statistical significance, in opposite direc- tions: males reported higher levels of physical aggression, but lower levels of hostility, than females.

Furthermore, we tested whether levels of shame differed between people with and without children, as well as between people who reported to be (or to had been in the past) in a long lasting romantic relationship and people who did not. Controlling for gender and age, par- ticipants with children reported significantly lower scores on the ESS total score, F(1, 373) D 3.95, p > .05, h2partial D .01. However, the multivariate main effect of the parental condition (i.e., children yes/no) on the combined ESS scale scores was only approaching significance, F(3, 371) D 2.22, p D .08, Pillai’s Trace D .02, h2partial D .02. Specifically, although partici- pants with children reported lower levels of shame on all three dimensions, these differences were significant for characterological shame, F(1, 373) D 3.98, p < .05, h2partial D .01,

Table 1. Means, Standard Deviations (SD), and Gender Comparisons (Controlling for Age) for All Study Variables (N D 380).

Male participants (N D 129)

Female participants (N D 251)

Whole sample (N D 380)

Mean (SD) Mean (SD) Mean (SD) F df h2 partial

ESS total score 46.89(11.79) 52.96(12.48) 50.89(12.57) 17.15��� 1, 375 .04 ESS characterological 21.63(6.09) 22.73(6.49) 22.36(6.37) 1.37 1, 375 .00 ESS behavioral 18.70(5.15) 20.65(5.41 19.99(5.40) 9.79�� 1, 375 .03 ESS bodily 6.57(2.53) 9.54(3.45) 8.53(3.46) 69.34��� 1, 375 .16 DES-IV shame 5.24(1.71) 6.03(2.11) 5.75(2.01) 9.85�� 1, 342 .03 DES-IV guilt 6.24(1.75) 6.79(1.97) 6.59(1.91) 4.45� 1, 342 .01 Self-esteem 32.00(4.75) 30.50(5.49) 31.01(5.29) 6.62� 1, 377 .02 SCL-90-R GSI .54(.38) .76(.53) .69(.49) 15.55��� 1, 377 .04 ERQ reappraisal 4.71(1.02) 4.97(1.05) 4.88(1.05) 4.93� 1, 377 .13 ERQ suppression 3.70(1.19) 3.18(1.24) 3.36(1.25) 12.98��� 1, 377 .03 AQ total score 63.57(16.51) 62.18(15.75) 62.65(16.00) 1.14 1, 377 .00 AQ physical aggression 17.42(6.50) 14.13(5.34) 15.25(5.96) 29.26��� 1, 377 .07 AQ verbal aggression 14.37(3.74) 13.67(4.24) 13.91(4.09) 2.59 1, 377 .01 AQ anger 15.47(5.24) 16.09(5.30) 15.88(5.28) 0.82 1, 377 .00 AQ hostility 16.33(5.02) 18.29(5.73) 17.63(5.57) 8.44�� 1, 377 .02

Note. ESS D Experience of Shame Scale; DES-IV D Differential Emotions Scale-IV; SCL-90-R D Symptom Checklist-90-Revised; ERQ D Emotion Regulation Questionnaire; AQ D Aggression Questionnaire. Bolded mean values are significantly greater than the corresponding value in the opposite gender.

�p < .05. ��p < .01. ���p < .001.

176 P. VELOTTI ET AL.

and behavioral shame, F(1, 373) D 4.79, p < .05, h2partial D .01. On the other hand, although there was not significant difference on the ESS total score, F(2, 372) D 1.49, p > .05, h2partial D .01, a significant multivariate effect between participants with and without a current or past intimate relationship occurred on the combined ESS scales, F(6, 742) D 2.57, p < .05, Pillai’s Trace D .04, h2partial D .02. Specifically, controlling for gender and age, individuals who had never had a long lasting intimate relationship reported significantly greater scores of bodily shame, F(2, 372) D 4.33, p < .05, h2partial D .02. The analyses in this paragraph were exploratory in nature, to test for associations between the ESS and demo- graphics. However, since an interesting pattern of results occurred, we opted for reporting and discussing them, to stimulate further research in this area.

ESS Nomological Network

Correlation analysis results are displayed in Table 2. Of note, the ESS scales were all strongly related to the ESS total score (controlling for age, partial rs were .89, .84. and .63 for the char- acterologial shame, behavioral shame, and bodily shame scales, respectively). Further, the ESS scales were significantly associated to each other (rs ranging between .33 and .60). All ESS scales were significantly and positively related to the DES-IV shame and guilt scales. Of note, homogeneity tests did not show significant differences in correlation coefficient values of any ESS scale with shame versus guilt scale from the DES-IV, min. z D 0.24, max z D 1.53, all ps > .05. Conversely, homogeneity tests across gender revealed that only 1 out 8 pairs of cor- relation coefficients (12.5%) significantly differed across gender, with the association between bodily shame and guilt being stronger among males (z D 2.27, p < .05). All the ESS scales were also significantly and negatively associated with levels of self-esteem. As for the associa- tions between the ESS and the ERQ, no significant correlations were found with cognitive reappraisal. Among females only, a significant positive correlations emerged between emo- tional suppression and both characterological shame and the ESS total score. Among males, no significant associations occurred between the ESS and the AQ total score; on the other hand, the AQ total score was significantly and positively related to the ESS bodily shame scale and the ESS total score among females. Physical aggression was negatively related with char- acterological shame among males. Among females, the direction of the association was inverse, but nonsignificant. Of note, correlation coefficients between physical aggression and ESS total and characterological shame scores significantly differed across gender (z D 2.12 and 2.41, respectively, both ps < .05). Verbal aggression was only significantly related to char- acterological shame (negatively) among females. Among females, anger showed a positive association with bodily shame. Moreover, the ESS total and subscale scores were significantly and positively related to hostility across gender. Finally, significant positive correlations were found between all ESS scales and the GSI scale of the SCL-90-R. Overall, only 3 out of 44 comparisons between correlation coefficients (across gender) turned out to be significant (6.82%), suggesting that the patterns of correlations between the ESS scales and external cor- relates are largely invariant across gender (min. z D 0.00, max z D 1.68, all ps > .05).

Incremental Variance of ESS Dimensions on Relevant Outcomes

We then tested whether the ESS scales significantly explained a portion of additional vari- ance in self-esteem, hostility, and psychopathological distress (as these were the variables

THE JOURNAL OF PSYCHOLOGY 177

Ta b le 2.

Pa rt ia lC

or re la ti on s (C on tr ol lin g fo r A g e)

of th e ES S To ta la nd

Sc al e Sc or es

w it h th e D ES -I V G ui lt an d Sh am

e Sc al es ,S el f- Es te em

Le ve l, Em

ot io n Re g ul at io n

St ra te g ie s, an d A g g re ss io n D im en si on s, in Bo th

M al e (N

D 12 9) an d Fe m al e (N

D 25 1) Pa rt ic ip an ts (T ot al N D

38 0) .

D ES -I V

sh am

e D ES -I V

g ui lt

Se lf-

es te em

ER Q

re ap p ra is al

ER Q

su p p re ss io n

A Q

to ta l

A Q p hy si ca l

ag g re ss io n

A Q ve rb al

ag gr es si on

A Q

an g er

A Q

ho st ili ty

SC L- 90 -R

G SI

ES S to ta l

M al es

.4 5�

�� .5 6�

�� – .4 2�

�� .0 1

.0 3

.0 4

– .1 6a

– .0 4

.0 2

.3 7�

�� .4 3�

��

Fe m al es

.5 1�

�� .5 4�

�� – .5 1�

�� – .0 6

.1 9�

� .1 6�

.0 7a

– .1 2

.0 7

.4 1�

�� .5 2�

��

ES S c ha ra ct er ol og ic al

M al es

.4 0�

�� .5 4�

�� – .4 6�

�� .0 2

.1 2

.0 1

– .2 1�

a – .0 3

– .0 1

.3 4�

�� .4 5�

��

Fe m al es

.5 0�

�� .5 9�

�� – .5 4�

�� – .0 5

.2 2�

� .1 2

.0 5a

– .1 5�

.0 5

.3 4�

�� .5 2�

��

ES S be ha vi or al

M al es

.3 9�

�� .4 0�

�� – .2 5�

�� .0 2

– .0 4

.0 6

– .0 9

– .0 1

.0 5

.3 1�

�� .2 6�

��

Fe m al es

.3 8�

�� .3 6�

�� – .3 0�

�� – .0 1

.1 2

.0 9

.0 3

– .1 2

– .0 2

.3 3�

�� .3 3�

��

ES S bo di ly

M al es

.3 3�

� .4 9�

�� a

– .3 5�

�� – .0 4

– .0 6

.0 4

– .0 6

– .1 0

.0 1

.2 9�

� .3 6�

��

Fe m al es

.3 0�

�� .2 8�

�� a

– .3 7�

�� – .1 1

.0 9

.2 2�

�� .1 2

.0 3

.1 7�

� .3 3�

�� .3 9�

��

N ot e. ES S D

Ex pe ri en ce

of Sh am

e Sc al e; D ES -I V D

D iff er en ti al Em

ot io ns

Sc al e- IV ;E RQ

D Em

ot io n Re g ul at io n Q ue st io nn

ai re ;A Q D

A g gr es si on

Q ue st io nn

ai re ;S C L- 90 -R

D Sy m p to m C he ck lis t- 90 -

Re vi se d ;G

SI D

G lo b al Se ve ri ty In d ex

sc al e of th e SC L- 90 -R .

a S ig ni fi ca nt

d iff er en ce

in co rr el at io n co ef fi ci en ts b et w ee n m al e an d fe m al e p ar ti ci p an ts .

� p <

.0 5. �� p <

.0 1. �� � p

< .0 01 .

178 P. VELOTTI ET AL.

most strongly associated at the bivariate level with the ESS scales), above and beyond the influence of DES-IV-assessed shame (see Table 3). None of the VIF values suggested that collinearity among predictors could have biased regression results. Hierarchical multiple regression analyses revealed that, controlling for age and gender, the ESS scales significantly explained a portion of incremental variance in all of the outcomes considered. Specifically, after removing the shared variance among ESS scales, characteriological and bodily shame scales significantly and independently predicted self-esteem level (negatively) and global psy- chopathological distress (positively). Further, bodily shame uniquely and positively pre- dicted hostility, over and above the influence of the shame scale of the DES-IV.

Does Emotional Suppression Account for an Indirect Relationship between Sshame and Maladaptive Outcomes?

Based on partial correlation results (see Table 2), the significance of statistical indirect effects was assessed only among female participants (i.e., because among males the association between shame and emotional suppression was not significant, ruling out the possibility of any indirect effect of the former through the latter). A summary of the indirect effect analy- ses conducted is presented in Table 4. A total of seven indirect effects were tested, four of which yielded significant results with small effect size. Specifically, emotion suppression mediated the effect of ESS total score and characterological shame on both hostility and psy- chopathological distress. The other indirect effects (involving self-esteem and overall trait aggression) did not reach statistical significance.

Discussion

The aim of the current study was to confirm and extend previous knowledge on the nomo- logical net surrounding the multidimensional construct of shame, also testing for invariance across gender. Further, the possible mechanisms underlying the association between shame and maladaptive outcomes were investigated, examining the mediating role of maladaptive emotion regulation (i.e., emotional suppression).

Table 3. Hierarchical Multiple Regression Analyses Predicting Self-Esteem, Hostility, and Psychopathologi- cal Distress (N D 380).

Self-esteem AQ Hostility GSI B b b

Step 1 Age .21��� –.20��� –.10 Gender .13� –.15�� –.20���

Adjusted R2change .06��� .06��� .05���

Step 2 DES-IV shame –.38��� .60��� .54���

Adjusted R2change .13��� .33��� .27���

Step 3 ESS characterological shame –.39��� –.02 .30���

ESS behavioral shame –.09 .08 –.08 ESS bodily shame –.17��� .15�� .18���

Adjusted R2change .13��� .02�� .10���

Note. AQ D Aggression Questionnaire; GSI D Global Severity Index of the Symtpom Checklist-90-Revised; DES-IV D Differential Emotions Scale-IV; ESS D Experience of Shame Scale. Gender was dummy-coded with 0 D females.

THE JOURNAL OF PSYCHOLOGY 179

Ta b le 4.

Su m m ar y of

Bo ot st ra p p in g A na ly se s Ex am

in in g th e In d ir ec t Ef fe ct of

Sh am

e on

Se lf- Es te em

,A g g re ss io n, an d Ps yc ho p at ho lo g ic al D is tr es s Th ro ug

h th e Ro le

of Em

ot io na lS up

p re ss io n, C on tr ol lin g fo r A g e (A m on g Fe m al es

O nl y; N D

25 1; 5, 00 0 Bo ot st ra p Sa m p le s) .

In d ir ec t ef fe ct (b ia s

co rr ec te d in te rv al s)

Pr ed ic to r

V ar ia bl e( PV )

M ed ia ti ng

va ri ab le (M

) C ri te ri on

V ar ia bl e (C V )

Ef fe ct of

PV on

M (a )

Ef fe ct of

M on

C V ,

co nt ro lli ng

fo r th e PV

(b )

To ta le ff ec t

(c )

D ir ec t ef fe ct

(c ’)

(a £

b) 95 % C I

ab cs [9 5%

b ia s

co rr ec te d C I]

ES S to ta l

Em ot io na ls up

p re ss io n

Se lf- es te em

.0 2�

� – .3 9

– .2 3�

�� – .2 2�

�� – .0 07

[– .0 21 ,. 00 0]

ES S to ta l

Em ot io na ls up

p re ss io n

A Q to ta l

.0 2�

� .6 8

.2 1�

� .2 0�

.0 12

[– .0 15 ,. 05 3]

ES S ch ar ac te ro lo g ic al

sh am

e Em

ot io na ls up

p re ss io n

A Q to ta l

.0 4�

� .7 5

.3 1�

.2 8

.0 28

[– .0 28 ,. 11 1]

ES S to ta ls co re

Em ot io na ls up

p re ss io n

A Q ho st ili ty

.0 2�

� .8 7�

� .2 0�

�� .1 8�

�� .0 16

[.0 05 ,. 03 5]

.0 34

[.0 09 ,. 07 5]

ES S ch ar ac te ro lo g ic al

sh am

e Em

ot io na ls up

p re ss io n

A Q ho st ili ty

.0 4�

� .9 0�

� .3 2�

�� .2 8�

�� .0 34

[.0 12 ,. 07 5]

.0 39

[.0 13 ,. 08 4]

ES S to ta ls co re

Em ot io na ls up

p re ss io n

G SI

.0 2�

� .1 0�

�� .0 2�

�� .0 2�

�� .0 02

[.0 01 ,. 00 3]

.0 41

[.0 12 ,. 08 1]

ES S ch ar ac te ro lo g ic al

sh am

e Em

ot io na ls up

p re ss io n

G SI

.0 4�

� .0 9�

�� .0 4�

�� .0 4�

�� .0 03

[.0 01 ,. 00 7]

.0 42

[.0 15 ,. 08 5]

N ot e. ES S D

Ex pe ri en ce

of Sh am

e Sc al e; A Q D

A g g re ss io n Q ue st io nn ai re ;G

SI D

G lo b al Se ve ri ty In d ex

sc al e of th e Sy m p to m C he ck lis t- 90 -R ev is ed ;C ID

C on fi d en ce

In te rv al s. U ns ta nd

ar d iz ed

co ef -

fi en ts ar e re p or te d .a b c

s D

co m p le te ly st an d ar d iz ed

in d ir ec t ef fe ct ,m

ea su re of th e ef fe ct si ze

fo r si g ni fi ca nt

in d ir ec t ef fe ct s.

� p <

.0 5.

�� p <

.0 1.

�� � p

< .0 01 .

180 P. VELOTTI ET AL.

In line with the expectations, women reported greater levels of shame (specifically: behav- ioral and bodily shame), and lower levels of self-esteem, than men. Women also reported greater levels of psychopathological distress, cognitive reappraisal, and hostility, whereas men had greater levels of emotional suppression and physical aggression. Taken together, these findings are in line with prior findings (Andrews et al., 2002; Fossati et al., 2003; John & Gross, 2003) on gender differences in emotional experience, emotion regulation styles, as well as in the expression of aggressive tendencies. Notably, levels of characterological and behavioral shame were higher among people who did not have kids, and levels of bodily shame were higher among people who never had a significant romantic relationship. The correlational nature of the study did not allow to speculate about causal effects, but the inter- esting link between intimate attachment relationships (with children and/or romantic part- ners) and shame feelings in different domains warrants future investigation.

The nomological network surrounding shame experiences was largely consistent across gender. Behavioral, characterological, and bodily shame feelings were all related with trait shame and trait guilt. Furthermore, shame feelings were associated with decreased self- esteem, and with higher levels of hostility and psychological distress. This is consistent with previous literature (Garofalo, 2015; Elison, 2005; Gilligan, 1996; Marshall et al., 2009) and suggests that the experience of shame does not come in isolation. Rather, feelings of shame are likely associated with a more general lack of confidence in the self as well as in the out- side world. As such, it is not surprising that feelings of shame and lack of confidence in the self and the others can be accompanied by overall psychological distress. Some gender differ- ences in the nomological net emerged, suggesting that the experience of shame might have more pronounced consequences in women. Indeed, characterological shame was related with emotional suppression, indicating that women who are more prone to experience feel- ing of shame about their own personality may tend to adopt maladaptive emotion regulation strategies (Nystr€om & Mikkelsen, 2013). This might be due to an attempt to protect them- selves from the effects of such unbearable feelings (Elison et al., 2014), which would probably increase the experience of shame itself. Therefore, it appears that women who experience characterological shame are likely to suppress, rather than show, their own emotions. In line with this, higher levels of characterological shame were also associated with lower levels of verbal aggression. Further, bodily shame was associated with angry feelings, suggesting that the perception of flaws in their physical aspect could be a trigger for angry and aggressive outbursts especially in females (Hejdenberg & Andrews, 2011). Finally, physical aggression was negatively related to characterological shame in men, indicating a possible inhibiting effect of shame toward aggressive tendencies. This is consistent with the idea that shame feel- ings are not always bad, and that the experience of shame might have healthy and adaptive consequences (Farmer & Andrews, 2009). For example, feeling ashamed about a previous aggressive behavior can have the positive consequence of inhibiting the same behavior in the future. As a whole, although all others coefficients were nonsignificant, the trends seem to indicate that shame is negatively associated with aggression among men, but positively related (or unrelated) to aggression among women. Future examinations of the possible dif- ferential effects of shame on aggression across gender are required to obtain a deeper insight into the shame-aggression link.

Our findings also supported the importance of adopting a multidimensional conceptuali- zation of shame (Andrews et al., 2002). Indeed, over and above the influence of trait shame, characterological and bodily shame were independently associated with low self-esteem and

THE JOURNAL OF PSYCHOLOGY 181

psychological distress. Further, bodily shame was independently related to hostility. This suggests that, even after controlling for individual differences in the experience of shame in daily life, specific feelings of shame referred to one’s identity or body might add to the expla- nation of internalizing and externalizing symptoms. As a last step, we aimed at exploring possible mechanisms linking shame feelings with maladaptive outcomes. Mediation analyses revealed that—among women—the associations of shame feelings (and, in particular, char- acterological shame) with hostility and psychological distress was accounted for by emo- tional suppression. This might indicate that both the externalization (i.e., hostility) and internalization (i.e., psychological distress) of shame feelings could be explained by poor emotion regulation, rather than being an effect of shame per se. In other words, shame feel- ings are likely to increase the individual difficulty in regulating emotions, and this in turn could lead to an increase in hostile attitudes and psychopathological symptoms (Garofalo et al., 2016; Velotti et al., 2014). On the other hand, the same pattern did not explain the associ- ation between shame and low self-esteem, suggesting that their shared variance was not accounted for by emotional suppression.

This study presented several limitations. First, the reliance on self-report questionnaires may have inflated correlations due to the spurious effect of common method variance. Sec- ond, the generalization of these results is limited by the recruitment of a convenience sample of community individuals, the majority of whom where university students. One possible problem in relying on convenience samples comprising a strong component of university students is a restriction in the variance of several demographic (e.g., Socio Economic Status) and personality (e.g., impulsive or antisocial traits) characteristics that may be linked to some of the variables examined in the present study (e.g., aggression). Third, the correla- tional design prevents us from drawing conclusions about the causal relations among study variables. Nevertheless, the present findings may help design longitudinal studies to test pro- spective associations among shame experiences and related constructs over time. Finally, the use of single measures for each construct of interest raises the possibility that results would not generalize to other operationalizations of the same constructs.

Nonetheless, we believe these findings extend current knowledge on the nomological net- work of shame feelings, providing novel insight on the role of specific shame experiences. Specifically, the present findings suggest that conceptualizing shame as a multidimensional construct may be helpful in delineating the associations between shame and maladaptive outcomes. Notably, our study also provides additional support for the use of the ESS as a brief multidimensional measure of shame experiences. From a clinical point of view, this is important as a focus on specific ‘faces’ of shame can be appropriate to target psychopatho- logical symptoms and aggression. Finally, the current study advanced prior knowledge indi- cating that fostering the use of adaptive emotion regulation strategies could be an important target to prevent or reduce psychopathological symptoms and aggression.

Author Notes

Patrizia Velotti, PhD, PsyD, is Associate Professor at the Department of Educational Sciences of the University of Genoa, Italy. She studies the role of emotion regulation and other intersecting mechanisms linked with mental health and psychopathology. She is the head of the Emotion Regu- lation Interpersonal and Intergroup Relations Lab where she conducts research projects address- ing these issues in laboratory, field, and clinical settings.

182 P. VELOTTI ET AL.

Carlo Garofalo, PhD, is Assistant Professor at the Department of Developmental Psychology of Tilburg University, The Netherlands. His main research interests concern the role of emotion reg- ulation in the development and manifestation of personality pathology and antisociality, with a specific focus on psychopathic traits.

Federica Bottazzi, MSc, recently received her Master’s degree in Psychology from the University of Genoa, Italy. She plans to pursue a PsyD in clinical psychology, with research interests in aggres- sion and emotion.

Vincenzo Caretti, PhD, is Full Professor at the Department of Human Sciences of LUMSA Univer- sity in Rome, Italy. His research interests currently focus on psychopathy and addictive behaviors. He has been scientific advisor to several academic and inter-institutional programs aimed at understanding, preventing, and treating violent behaviors.

ORCID

Carlo Garofalo http://orcid.org/0000-0003-2306-6961

References

Andrews, B., Qian, M., & Valentine, J. D. (2002). Predicting depressive symptoms with a new measure of shame: The experience of shame scale. British Journal of Clinical Psychology, 41, 29–42.

Balzarotti, S., John, O. P., & Gross, J. J. (2010). An Italian adaptation of the Emotion Regulation Ques- tionnaire. European Journal of Psychological Assessment, 26, 61–67.

Buss, A. H., & Perry, M. (1992). The aggression questionnaire. Journal of Personality and Social Psy- chology, 63, 452–459. doi: 10.1037/0022-3514.63.3.452

Caretti, V., Craparo, G., & Schimmenti, A. (2010). Il ruolo della disregolazione affettiva, della dissocia- zione e della vergogna nei disturbi del comportamento alimentare. [The role of affective dysregula- tion, dissociation and shame in eating disorders]. In V. Caretti, & D. La Barbera (eds.), Addiction. Aspetti Biologici e di Ricerca [Addiction. Biological and Research Aspects] (pp. 135–165). Milano, IT: Raffaello Cortina Editore.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/ correlation analysis for the behavioral sciences (3rd ed.). Hillsdale, NJ: Erlbaum.

Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52(4), 281–302.

Derogatis, L. R. (1994). SCL-90-R, administration, scoring and procedures manual for the revised ver- sion (3rd ed.). Minneapolis, MN: National Computer Systems.

Elison, J. (2005). Shame and guilt: A hundred years of apples and oranges. New Ideas in Psychology, 23, 5–32.

Elison, J., Garofalo, C., & Velotti, P. (2014). Shame and aggression: Theoretical considerations. Aggres- sion and Violent Behavior, 19, 447–453.

Elison, J., Pulos, S., & Lennon, R. (2006). Shame-focused coping: An empirical study of the compass of shame. Social Behavior and Personality, 34, 161–168.

Farmer, E., & Andrews, B. (2009). Shameless yet angry: shame and its relationship to anger in male young offenders and undergraduate controls. Journal of Forensic Psychiatry & Psychology, 20, 48– 65.

Fossati, A., Maffei, C., Acquarini, E., & Di Ceglie, A. (2003). Multigroup confirmatory component and factor analyses of the Italian version of the aggression questionnaire. European Journal of Psycho- logical Assessment, 19, 54–65.

Garofalo, C. (2015). Emozionalit�a negativa ed autostima in un campione di offender detenuti: Uno studio preliminare [Negative emotionality and self-esteem in an incarcerated offender sample: Pre- liminary investigation]. Giornale Italiano di Psicologia [Italian Journal of Psychology], 42, 363–371.

THE JOURNAL OF PSYCHOLOGY 183

Garofalo, C., Holden, C. J., Zeigler-Hill, V., & Velotti, P. (2016). Understanding the connection between self-esteem and aggression: The mediating role of emotion dysregulation. Aggressive Behavior, 42, 3–15.

Gilligan, J. (1996). Violence. Reflections on a National Epidemic. New York, NY: Vintage Books. Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implica-

tions for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85, 348– 362.

Gross, J. J., & Levenson, R. W. (1993). Emotional suppression: Physiology, self-report, and expressive behavior. Journal of Personality and Social Psychology, 64, 970–986.

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regres- sion-based approach. New York, NY: Guilford Press.

Hejdenberg, J., & Andrews, B. (2011). The relationship between shame and different types of anger: A theory-based investigation. Personality and Individual Differences, 50, 1278–1282.

Izard, C. E., Libero, D. Z., Putnam, P., Haynes, O. M. (1993). Stability of emotion experiences and their relations to traits of personality. Journal of Personality and Social Psychology, 64(5), 847–860.

Lanteigne, D. M., Flynn, J. J., Eastabrook, J. M., & Hollenstein, T. (2014). Discordant patterns among emotional experience, Arousal, and expression in adolescence: Relations with emotion regulation and internalizing problems. Canadian Journal of Behavioural Science / Revue canadienne des scien- ces du comportement, 46, 29–39.

Marshall, W. L., Marshall, L. E, Serran, G. A., & O’Brien, M. D. (2009). Self-esteem, shame, cognitive distorsions and empathy in sexual offenders: Their integration and treatment implications. Psychol- ogy, Crime & Law, 15, 217–234.

Nystr€om, M. B. T., & Mikkelsen, F. (2013). Psychopathy-related personality traits and shame manage- ment strategies in adolescents. Journal of Interpersonal Violence, 28, 519–537.

Prezza, M., Trombaccia, F. R., & Armento, L. (1997). La scala dell’autostima di Rosenberg: Traduzione e validazione Italiana. [The Rosenberg self-esteem Scale: Italian translation and validation.]. Bollettino di Psicologia Applicata [Applied Psychology Bulletin], 223, 35–44.

Prunas, A., Sarno, I., Preti, E., Madeddu, F., & Perugini, M. (2012). Psychometric properties of the Italian version of the SCL-90-R: A study on a large community sample. European Psychiatry, 27, 591–597.

Roberton, T., Daffern, M., & Bucks, R. S. (2012). Emotion regulation and aggression. Aggression and Violent Behavior, 17, 72–82.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York, NY: Guilford Press. Velotti, P., Elison, J., & Garofalo, C. (2014). Shame and aggression: Different trajectories and implica-

tions. Aggression and Violent Behavior, 19, 454–461. Zavattini, G. C., Garofalo, C., Velotti, P., Tommasi, M., Romanelli, R., Espirito Santo, H., & Saggino,

A. (2015). Dissociative experiences and psychopathology among inmates in Italian and Portuguese prisons. International Journal of Offender Therapy and Comparative Criminology. Advance online publication. doi.org/10.1177/0306624X15617256.

184 P. VELOTTI ET AL.