Child Abuse & Neglect 31 (2007) 1143–1153

Shame, guilt, symptoms of depression, and reportedhistory of psychological maltreatment

Marcia Webb ∗, Dawn Heisler, Steve Call, Sarah A. Chickering,Trina A. Colburn

Seattle Pacific University, Department of Clinical Psychology, Seattle, WA, USA

Abstract

Objective: The purpose of the present study was to provide preliminary data extending earlier research on shame andguilt, examining their relationships both to symptoms of depression and to psychological maltreatment. Symptomsof depression were expected to correlate positively with shame, but not with guilt. Psychological maltreatment wasalso expected to correlate positively with shame. The relationship between psychological maltreatment and guiltwas examined on an exploratory basis.Method: Two hundred and eighty participants from a public community college and a private university completedscales assessing shame, guilt, depression, and history of childhood psychological maltreatment. Pearson correlationswere conducted with all data.Results: Results indicated that symptoms of depression were positively correlated with both shame and guilt.Partial correlations were then conducted in which the linear effects of shame were removed from guilt. In this latteranalysis, guilt was no longer positively correlated with symptoms of depression. Psychological maltreatment wasalso positively correlated with depression and with shame, but not with guilt.Conclusions: These results highlight the significance of psychological maltreatment in the relationship to theself-conscious emotions of guilt and shame. As in earlier studies, shame has been consistently correlated to poorpsychological functioning, while guilt appears to be relatively unrelated to pathological functioning.© 2007 Elsevier Ltd. All rights reserved.

Keywords: Shame; Guilt; Depression; Psychological maltreatment; Neglect

∗ Corresponding author address: Department of Clinical Psychology, Seattle Pacific University, Suite 107, 3307 Third AvenueWest, Seattle, WA 98119, USA.

0145-2134/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.chiabu.2007.09.003

1144 M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153

Introduction

Tangney and Dearing (2002), in their text Shame and Guilt, commented that the family was the “firstplace to look” (p. 146) for the origin of individual differences in the tendency to experience the moral,self-conscious emotions of shame and guilt. The family, given its powerful role in the development ofthe child, may be a prime vehicle for the transmission of both verbal and nonverbal messages aboutthe nature of good and bad, and about the origins of wrongdoing. Indeed, theorists have postulatedthat maladaptive patterns of interaction in the parent-child relationship may be the source of shame-based psychopathology in adult life (Kaufman, 1989; Loader, 1998; Nathanson, 1987). Despite theseassumptions, research examining the self-conscious emotions of shame and guilt in the context of thedysfunctional family has only recently begun.

Perhaps one reason for a delay in empirical investigation of these variables is the fact that the distinctionbetween the terms shame and guilt in the social science literature has only been clarified in the past fewdecades. In common, everyday parlance, these two terms are often used interchangeably. This may notbe surprising; some researchers have suggested that, in practical experience, the two emotions may be‘fused’ in one response, though they represent qualitatively different affective states (Tangney, Wagner,& Gramzow, 1992).

Thus, for both theoretical and research clarity, shame has increasingly been delineated as that negativeevaluation and affect directed toward the entirety of the self, often following a breach of social or moralconduct. In shame, individuals reject themselves, an appraisal which may be expressed in the statement,“I myself am bad” (Lewis, 1971; Tangney & Dearing, 2002). Research confirms that an acute sense ofsmallness, inadequacy, and worthlessness accompanies the negative self-appraisal of shame, as well asa desire to shrink away and remain unseen by others (Lewis, 1971; Loader, 1998; Tangney, 1990, 1995,1996; Tangney & Dearing, 2002). Tangney and Dearing (2002) further comment that it is this uniquephenomenological and affective component of shame which distinguishes it from the construct of lowself-esteem, which may be based upon self-descriptions or self-ratings alone.

In contrast, in the experience of guilt, the individual is able to make a cognitive distinction betweenthe self and behavior, and negative evaluation is restricted to the act of wrongdoing itself. Thus, whenexperiencing guilt, individuals reject not so much themselves, but instead their behaviors. This appraisalmay be expressed in the statement, “My behavior is bad” (Lewis, 1971; Tangney & Dearing, 2002).While the self remains relatively unscathed in guilt, the experience is still an unpleasant one. It is the verytensions of remorse associated with guilt that may lead the guilt-prone individual to act in constructiveways toward recompense and reconciliation (Tangney, 1998). Thus, in contrast to early, psychoanalyticassumptions about the deleterious impact of guilt, this emotion has been demonstrated in recent years tofacilitate prosocial, adaptive functioning (Baumeister, Stillwell, & Heatherton, 1994; Leith & Baumeister,1998), particularly when compared to shame.

Researchers have also begun to consider those family environments that promote either the moreconstructive emotion of guilt, or the more destructive emotion of shame in children. Pulakos (1996)provided empirical support for the relationship between dysfunctional family environments and shameamong family members. In her research, shame-proneness was negatively correlated with family cohesionand positively correlated with conflict; guilt, however, was not correlated in any way with these variables.

Preliminary studies have further indicated that dysfunctional parenting styles in the family may beinvolved in the assimilation of a shame-based sense of self. Shame-proneness in adults has been correlatedpositively with recall of criticism from parental figures in childhood (Gilbert, Allan, & Goss, 1996) and

M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153 1145

with report of incidents of “emotional abuse” by a parent (Hoglund & Nicholas, 1995). Shame-pronenesshas also been inversely correlated with positive reports of maternal and paternal caregiving in childhood(Lutwak & Ferrari, 1997).

In more recent studies, childhood guilt has been correlated with parental disciplinary techniques whichfocus upon inappropriate behavior as the target for intervention, rather than on the assumption of aninherent failure in the child’s self. In contrast, shame-proneness in children has been correlated withparental disgust, withdrawal of love, and disciplinary messages focused on the rejection of the child’sself (Tangney & Dearing, 2002).

What are the specific mechanisms by which dysfunctional family interactions, particularly thoseinvolved in child maltreatment, promote this shame-based sense of self? Loader (1998) has written:

The abused child, irrespective of the form of the parental mistreatment, is left to conclude that thepain [he] feels at the hands of the parent is his fault, deserved because of the way he is. This conclusionmay well be supported, even insisted upon, by the parent overtly stating that their treatment of thechild is for his own good. The child’s overriding need to believe that his parents care about him driveshim to conclude that there is something fundamentally wrong with him – and thus to experience apervasive, and often life-long, sense of shame. Child abuse is all about shame. . .. The family curseof child abuse is the curse of shame (pp. 52–53, italics his).

Perhaps it is no surprise, then, that childhood psychological maltreatment has been correlated withadult onset depression (Bifulco, Moran, Baines, Bunn, & Stanford, 2002; Braver, Bumberry, Green, &Rawson, 1992; Gross & Keller, 1992; Stone, 1993). Some studies have indicated that psychological mal-treatment may in fact be a stronger predictor of depression than physical abuse alone (Zelikovsky & Lynn,1994, 2002). Shame—perhaps itself a consequence of the experience of psychological maltreatment—hasalso been positively correlated with depression, unlike the emotion of guilt (Tangney & Dearing,2002).

The present study: purpose and hypotheses

Research examining the relationship between the self-conscious emotions of shame and guilt andthe character of the family environment has just recently begun. The purpose of the present study,then, was to extend this research regarding shame and guilt, considering its potential associationsboth to depression and to psychological maltreatment. Based upon previous research, we expected thatdepressive symptomology would be correlated positively with shame (Hypothesis A), but not with guilt(Hypothesis B).

In addition, while most previous research has focused upon the broader constructs of family envi-ronment and parenting style in the development of self-conscious emotion, our research specificallytargeted psychological maltreatment of the child. With the use of the Psychological Maltreatment Inven-tory, we delineated three forms of maltreatment, those of rejection, neglect, and isolation. Based uponstudies examining family environment, parenting style, and shame, we expected that all forms of psy-chological maltreatment would correlate positively with shame (Hypothesis C). At present, the researchliterature provides less information about the potential relationships (or the lack thereof) between fam-ily environment, parenting style, and guilt. Due to the relative paucity of research in this area, weinvestigated the relationship between forms of psychological maltreatment and guilt on an exploratorybasis.

1146 M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153

Method

Participants

The study included a convenience sample of 280 participants. The research team recruited 212 par-ticipants from a private Christian university in the Pacific Northwest. While affiliated with the Wesleyantradition, at the time of the study the university practiced open enrollment in its admission policy, accept-ing students from a variety of religious and nonreligious backgrounds. The research team also recruited68 participants from a public community college in the area, resulting in a total sample of 280.

Sixty-five percent of all respondents were women. Respondents’ ages ranged from 18 to 44, with amean of 20.9 (SD = 4.6). The sample was predominantly Caucasian American (76.1%). Ethnicity forthe rest of the sample was as follows: .4% African American, 3.2% Asian American, 1.8% HispanicAmerican, .4% Native American, and 17.7% other. Five respondents (1.8%) had missing data for theethnicity item.

Procedure

Researchers obtained IRB approval for the research from Seattle Pacific University. The researchteam then attended a variety of undergraduate classes and provided to students a brief (5 min) statementregarding the study, including its focus on past stressful experiences, psychological health, and relationalissues. The research team clarified the anonymous and voluntary nature of participation. Students weretold they could earn extra credit for their research participation or for their completion of an equivalentproject. The researchers distributed consent forms and questionnaire packets. Participants were permittedto complete the questionnaires at home, and were given a date during the following week when theresearchers would return to their classes to pick up the questionnaires.

Approximately 1 week later, the researchers returned to classes to collect informed consent forms andquestionnaires. Further information about the study was distributed, including the name and number ofthe principal investigator, as well as the name and number of a local counseling clinic, given the sensitivenature of some questionnaire items.

Measures

The test of self-conscious affect (TOSCA). Participant’s levels of shame and guilt were assessed withthe TOSCA (Tangney, Wagner, & Gramzow, 1989). As a scenario-based measure, the TOSCA providedparticipants with brief narratives of 15 imagined events. For example, one scenario described a situationin which participants imagine that they have procrastinated on a project, which then turns out poorly.Each scenario was followed by four or five potential responses. On a 5-point Likert scale, participantsindicated how likely it was that they would react as stated in the response items. Reactions includedcognitive, affective, and behavioral responses falling within five theoretical categories of self-consciousemotion. The current study employed only two of these response categories, shame and guilt.

In previous studies, Cronbach’s alpha for the shame subscale has ranged from .73 to .80; for the guiltsubscale, Cronbach’s alpha has ranged from .62 to .83 (Tangney et al., 1989). In the present analysis,estimates of internal consistency fell within these same ranges. For the shame subscale, internal consis-tency was .79, and for the guilt subscale, it was .68. Nunnelly (1978) has suggested that reliabilities of .70

M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153 1147

or greater are sufficient for research purposes; it should be noted that the reliability of the guilt subscaledoes not exceed, but instead only approaches, this figure.

In an effort to boost ecological validity, the original construction of the 15 events described in theTOSCA were based upon the written accounts of personal experiences of shame, guilt, and pride fromseveral hundred adults, including both college and non-college participants (Tangney, 1990; Tangney,Burgraff, Hamme, & Domingos, 1988).

Unlike other scales designed to assess shame and guilt, the development of the TOSCA was groundedtheoretically in Lewis’ (1971) distinctions between shame as self-focused and guilt as behavior-focused.Ferguson and Crowley’s research (1997) demonstrated that the construct of shame was adequately mea-sured by the shame scale of the TOSCA. They found that the TOSCA shame scale was positively andsignificantly correlated with the shame scale of the Personality Factor Questionnaire-2.

Center for epidemiological studies—depression scale (CESD). Radloff (1977) originally designed theCESD, a 20-item self-report measure of symptoms of depression, with a 4-point Likert scale to assesssymptom frequency. During its construction, its internal consistency ranged from .85 to .90, when testedacross a variety of participant samples. In the current analysis, internal consistency was determined at.88.

Radloff (1977) also provided evidence for the construct validity of the CESD, with significant, pos-itive correlations with the Hamilton Clinician’s Rating Scale (Hamilton, 1960). The CESD was able todifferentiate between clinical samples and samples from the general population. The CESD has also beenused in adolescent and college populations. Radloff (1991) examined data from approximately 4,000participants, comparing CESD scores of junior high, high school, college, general adult, and clinicallydepressed persons, and concluded that the CESD is a reliable instrument for college students.

Psychological maltreatment inventory (PMI). Childhood history of psychological maltreatment wasassessed with the Psychological Maltreatment Inventory (Engels & Moisan, 1994). The original scale wasadministered to an adult clinical outpatient population and reduced through principal component analysiswith varimax rotation. Initially, 11 factors were isolated, accounting for approximately 72% of the totalvariance. Of these factors, the researchers retained only the first three, with 25 items from an original list of47. The factor accounting for the most variance (31.8%) was labeled emotional neglect. It included 12 itemswhich indicated that parental figures or childhood caregivers “[were] detached/uninvolved,” “showed littleaffection,” or “failed to provide support” (p. 604). A second factor, hostile rejection accounted for 7.3% ofthe variance. It consisted of seven items in which childhood caregivers had, for example, “defined you asa failure,” “called you names (dummy, stupid, monster)” or “screamed/raged at you” (p. 604). Lastly, thethird factor, isolation, accounted for 6.2% of the variance. It contained six items about caregiver’s actionswhich, for example, “encouraged you to withdraw from opportunities for social contact” or “didn’t allowyou to go outside” (p. 604). The entire scale of 25 items was presented with a 6-point Likert scale.

During the initial construction of the scale, estimates of test-retest reliability for each of the factorsranged from .75 to .78 (Engels & Moisan, 1994). Cronbach’s alpha for emotional neglect was .91, forhostile rejection was .90, and for isolation was .85. Reliability estimates examining internal consistencyin the current study were comparable to these earlier figures. In the present analysis, Cronbach’s alphafor emotional neglect was .90, for hostile rejection, .82, and for isolation, .85.

Engels and Moisan (1994) provided evidence for convergent reliability for these subscales throughsignificant correlations with factors of Rohner’s (1991) Adult Parental Acceptance and Rejection Ques-

1148 M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153

Table 1Descriptive statistics for all variables

Minimum score Maximum score M SD

Shame 16 67 43.91 9.19Guilt 32 95 59.49 7.20CESD 0 55 17.69 10.30Hostile rejection 0 35 6.59 7.00Isolation 0 35 4.41 6.34Emotional neglect 0 59 8.16 10.82

tionnaire (PARQ). The PMI factor, emotional neglect, was positively correlated with the PARQ factor,neglect/indifference. The PMI’s factor hostile rejection was also positively correlated with the PARQfactor, aggression/hostility. Finally, the PMI’s factor, isolation, was positively correlated with both thePARQ aggression/hostility factor and the PARQ neglect/indifference factor. Predictive validity was fur-ther supported for the PMI, with significant correlations between the factor hostile rejection, a historyof psychological treatment, and a DSM-III-R Axis II diagnosis (i.e., a personality disorder). The PMIfactor, emotional neglect, was significantly correlated with a history of psychological treatment. The PMIfactor, isolation, was not significantly correlated with either of these variables.

Data analysis

Initial review of the data indicated that missing data ranged from .4% to 4.3% of participant responsesto items. Sample means for items were then determined, and used to replace missing data, in order toproduce a full sample of 280 participants for all analyses. Descriptive statistics were also calculated foreach of the variables, including shame, guilt, CESD, and PMI factors. These scores are presented inTable 1.

To examine relationships between all variables considered, including shame, guilt, CESD, and PMIfactors, Pearson correlations were conducted using SSPS 12.0 for Windows. Given the resulting correlationbetween shame and guilt, partial correlations were also conducted. For example, in order to determine thepotential association of guilt with both depression and psychological maltreatment, partial correlationswere conducted to remove the linear effects of shame from guilt in some analyses. Partial correlationsremoving the linear effects of guilt from shame were also conducted.

Results

Hypothesis A was confirmed. As can be seen in Table 2, shame was positively correlated with CESDscores. However, contrary to Hypothesis B, guilt was also positively associated with CESD scores. Furtherreview of the analyses suggested the possible impact of the shared variance between the variables of shameand guilt (22%). This potentially confounding overlap between shame and guilt is not new to the researchliterature; Tangney and Fisher (1995) described the importance of the use of partial correlations to assessthe contributions of each of these variables apart from the other. Thus, in order to assess further thevariables in Hypothesis B, partial correlations were conducted, as modeled in Tangney et al. (1992), in

M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153 1149

Table 2Intercorrelations among all variables

1 2 3 4 5 6

1. Shame –2. Guilt .47** –3. CESD .46** .21** –4. Hostile rejection .22** .00 .30** –5. Isolation .24** .02 .22** .64** –6. Emotional neglect .18** .03 .29** .76** .60** –

** p < .001.

Table 3Bivariate and partial correlations of shame and guilt

CESD Hostile rejection Isolation Emotional neglect

Shame (bivariate) .46** .22** .24** .18**

Shame (partial) .42** .24** .26** .19**

Guilt (bivariate) .21** .00 .02 .03Guilt (partial) −.01 −.12 −.11 −.07

** p < .001.

which the linear effects of shame were removed from guilt. In this latter analysis, guilt was no longerpositively correlated with symptoms of depression, as seen in Table 3.

Hypothesis C was confirmed. As seen in Table 2, all three factors of the PMI, including hostilerejection, isolation, and emotional neglect, were positively correlated with shame. However, no significantrelationships were found between factors of the PMI and guilt. Due to the shared variance between shameand guilt, partial correlations were again conducted. These partial correlations appear in Table 3. After thelinear effects of shame were removed from guilt, relationships between guilt and PMI factors continuedto be nonsignificant, but shifted toward the negative direction.

Discussion

The present study demonstrated that psychological maltreatment was positively correlated with symp-toms of depression and with shame, but not with guilt. Positive correlations were also found betweensymptoms of depression and both shame and guilt. However, partial correlations demonstrated thatshame-free guilt was not positively correlated with symptoms of depression.

The present study supports previous research reporting a relationship between symptoms of depres-sion, psychological maltreatment, and shame, but not between symptoms of depression, psychologicalmaltreatment, and guilt (Tangney & Dearing, 2002; Tangney et al., 1992). While traditional psycho-analytic formulations of personality have pointed to the maladaptive role of guilt in the developmentof psychopathology, the present study does not provide evidence for this phenomenon. Yet use of theTOSCA provided a means by which to assess guilt apart from the deleterious self-focus of shame. Whenusing partial correlations, in which the linear effects of shame are removed from the variable of guilt, guiltis not correlated with symptoms of depression. In contrast, regardless of whether or not the linear effects

1150 M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153

of guilt are removed from the variable of shame, shame remains positively correlated with symptoms ofdepression.

Shame and guilt as conceptualized by Lewis (1971) and assessed through the TOSCA may represent,at one level, forms of cognitive appraisal that fall on opposing ends of a continuum. While, at one endof the continuum, shame represents appraisals focused on the self’s enduring role in moral choice, at theother end, guilt represents appraisals focused on specific, time-limited, behavior. If individuals experienceshame and guilt fused as one, as has been suggested (Tangney et al., 1992), this may explain the positivecorrelation often noted between these two affective states, and demonstrated in the current research. Evenso, their relationship to one another does not negate that the two are distinct emotions. As an analogy, arainbow is a continuum of color, and yet each hue may remain a distinct perceptual experience for theindividual.

This study further extends previous knowledge about the distinctions between guilt and shame by itsanalysis of these variables in relationship to participants’ reports of childhood psychological maltreatment.Statistical analysis indicated that report of psychological maltreatment, whether by direct, open hostility,by neglect, or by enforcement of social isolation on the child, was not associated with guilt-proneness inrespondents. Instead, endorsement of shame-proneness increased with report of each of these forms ofchildhood psychological maltreatment. Thus, this study supports and extends previous research indicatingthat dysfunctional family interactions are associated with shame-proneness in adulthood (Gilbert et al.,1996; Hoglund & Nicholas, 1995; Lutwak & Ferrari, 1997; Pulakos, 1996). The study also focused uponthe specific dynamics of childhood psychological maltreatment in the home.

Additionally, this study supports Loader’s (1998) contention that child abuse, considered here in theform of psychological maltreatment, may, indeed, be “all about shame” (p. 53). Perhaps, in accordancewith more recent conceptions of shame, childhood psychological maltreatment incites in the developingchild a sense of ‘badness’ not simply about certain behavioral choices or courses of action, but about thetotality of the self. If children have the perception that psychological maltreatment occurs in response notsimply to behavior, but also to themselves as persons, this perception itself may contribute to problematicdevelopment, including the possibility of depression.

For example, this sense of ‘badness’ might eventually permeate the self-schema of the adult. As seen inthe current study, adults who reported via the PMI that they were “defined . . . as a failure,” or “called . . .

names (dummy, stupid, monster)” in childhood (and who thus indicated that parental interactions includednegative and rejecting appraisals of the totality of the child’s self) also demonstrated a tendency towardshaming, negative self-appraisal in their understanding of everyday foibles, as seen by their responses tothe TOSCA. In the PMI items described above, it is important to note that it is the ‘self’ of the child, andnot the child’s behavior, which is the focus of maltreatment.

This negative focus upon the self is reminiscent of current knowledge regarding explanatory style,as it is described in attribution theory. Attribution theory details the significance of certain explanatorystatements individuals create about life events. For example, individuals who blame negative life eventson themselves, who believe these events have enduring consequences, and who believe these eventsaffect multiple areas of their lives are more likely to struggle with depression. Tangney et al. (1992) hasalso described the contribution of attributional components in the development of shame. Yapko (2000)has further written about the transmission of attibutional styles in family systems. The current studylends support to the possibility of the transmission of an explanatory style within the context of childhoodpsychological maltreatment. For example, the item in which participants were queried about whether theywere “defined . . . as . . . failure[s]” is itself an example of an attribution which is internal (self-focused),

M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153 1151

stable (enduring), and global (involving multiple areas of life); over time, this explanatory style firstlearned in childhood may become the dominant attributional framework for the adult who emerges fromthe abusive environment. Thus, it may be that one of the destructive outcomes of childhood psychologicalmaltreatment is the adoption of explanatory styles in childhood which are predictive of both shame anddepression.

Clinical implications of the current study include the importance of the distinction between shameand guilt in the therapeutic setting. When listening to clients’ recall of negative experiences, cliniciansmight carefully appraise clients’ tendencies to experience shame, guilt, or the fusion of both emotions.Due to the nature of the shame experience, and the inherent tendency for individuals to wish to hidewhen shamed, some clients may struggle to express this particular emotion, or any events associatedwith it. Clinicians may, therefore, need to consider multiple verbal and nonverbal cues that suggest aclient’s potential shame, including perhaps an apparent difficulty discussing certain issues, a failure tomaintain eye contact, or a tendency to speak at times at barely audible voice levels (Tangney & Fisher,1995). Clinicians who work with individuals who report histories of childhood maltreatment might alsobe particularly alerted to the potential for these clients to experience shame.

A therapeutic environment which is characterized by acceptance and empathy may be, in itself, anopportunity for the healing of shame. A clinician’s nonjudgmental and understanding response to theserevelations may begin the process of the client’s reassessment of these events and of themselves (Tangney& Fisher, 1995). With the establishment of greater therapeutic rapport and trust over time, therapeuticstrategies for shame-prone individuals might also include assisting clients to distinguish between thesetwo emotions, providing them with clues to recognize these emotions within their own attributionalstrategies, and facilitating the re-conceptualization of negative evaluations of the self as demonstrated inshame.

This study’s findings must be regarded with caution, however, due to study limitations. These limitationsinclude the availability of information about the sample. More detailed information about the participantpool (including family structure, prior history of psychopathology, or substance use history) would furtherassist in the interpretation of research findings. The study also employed self-report data. The use of self-report data always carries with it certain complications, such as the possibility of response bias. In thecurrent sample, there were also strong intercorrelations among the PMI factor scores, indicating greaterrelationships between these factors, and a potential underlying construct, than demonstrated in earlierresearch with this measure. Finally, the internal consistency for the TOSCA guilt scores was assessedat .68; further testing is needed to confirm the results of analyses with this measure given this marginalreliability score.

Conclusion

This preliminary study highlights the significance of various parental-child interactions in thedevelopment of the self-conscious emotions of guilt and shame. Shame, in this study and in earlierresearch, has been consistently correlated with poor psychological functioning, while shame-free guiltappears to be relatively unrelated to pathological functioning both inter- and intrapersonally. Futureresearch might examine the possibility that shame impacts an individual’s overall well-being, qual-ity of life, and physiological health, in accordance with a comprehensive biopsychosocial model ofpersons.

1152 M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153

References

Baumeister, R. F., Stillwell, A. M., & Heatherton, T. F. (1994). Guilt: An interpersonal approach. Psychological Bulletin, 115(2),243–267.

Bifulco, A., Moran, P. M., Baines, R., Bunn, A., & Stanford, K. (2002). Exploring psychological abuse in childhood: II.Association with other abuse and adult clinical depression. Bulletin of the Menninger Clinic, 66(3), 241–258.

Braver, M., Bumberry, J., Green, K., & Rawson, R. (1992). Childhood abuse and current psychological functioning in a universitycounseling center population. Journal of Counseling Psychology, 39(2), 252–257.

Engels, M., & Moisan, D. (1994). The Psychological Maltreatment Inventory: Development of a measure of psychologicalmaltreatment in childhood for use in adult clinical settings. Psychological Reports, 74, 595–604.

Ferguson, T. J., & Crowley, S. L. (1997). Measure for measure: A multitrait-multimethod analysis of guilt and shame. Journalof Personality Assessment, 69(2), 425–441.

Gilbert, P., Allan, S., & Goss, K. (1996). Parental representations, shame, interpersonal problems, and vulnerability to psy-chopathology. Clinical psychology and Psychotherapy, 3, 23–34.

Gross, A. B., & Keller, H. R. (1992). Long-term consequences of childhood physical and psychological maltreatment. AggressiveBehavior, 18, 171–185.

Hamilton, M. (1960). A rating scale for depression. Journal of Neurologic Neurosurgical Psychiatry, 23, 56–62.Hoglund, C. L., & Nicholas, K. B. (1995). Shame, guilt, and anger in college students exposed to abusive family environments.

Journal of Family Violence, 10, 141–157.Kaufman, G. (1989). The psychology of shame. New York: Springer.Leith, K. P., & Baumeister, R. F. (1998). Empathy, shame, guilt, and narratives of interpersonal conflicts: Guilt-prone people are

better at perspective taking. Journal of Personality, 66(1), 1–37.Lewis, H. B. (1971). Shame and guilt in neurosis. New York: International Universities Press.Loader, P. (1998). Such a shame—a consideration of shame and shaming mechanisms in families. Child Abuse Review, 7, 44–

57.Lutwak, N., & Ferrari, J. R. (1997). Understanding shame in adults: Retrospective perceptions of parental-bonding during

childhood. The Journal of Nervous Disease, 185(10), 595–598.Nathanson, D. L. (Ed.). (1987). The many faces of shame. New York: Guilford.Nunnelly, J. C. (1978). Psychometric theory. San Francisco: McGraw-Hill Publishing.Pulakos, J. (1996). Family environment and shame: Is there a relationship? Journal of Clinical Psychology, 52, 617–623.Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psycho-

logical Measurement, 1(3), 385–401.Radloff, L. S. (1991). The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults. Journal

of Youth and Adolescence, 20(2), 149–166.Rohner, R. P. (1991). Handbook for the study of parental acceptance and rejection. Storrs, CT: University of Connecticut.Stone, N. (1993). Parental abuse as a precursor to childhood onset depression and suicidality. Child Psychiatry and Human

Development, 24(1), 13–24.Tangney, J. P. (1990). Assessing individual differences in proneness to shame and guilt: Development of the self-conscious affect

and attribution inventory. Journal of Personality and Social Psychology, 59(1), 102–111.Tangney, J. P. (1995). Recent advances in the empirical study of shame and guilt. American Behavioral Scientist, 38(8),

1132–1145.Tangney, J. P. (1996). Conceptual and methodological issues in the assessment of shame and guilt. Behaviour Research and

Therapy, 34(9), 741–754.Tangney, J. P. (1998). How does guilt differ from shame? In J. Bybee (Ed.), Guilt and children (2nd ed., pp. 1–14). San Diego:

Academic Press.Tangney, J. P., Burgraff, S. A., Hamme, H., & Domingos, B. (1988). Assessing individual differences in proneness to shame

and guilt: The Self Conscious Affect and Attribution Inventory. Poster presented at the meeting of the Eastern PsychologicalAssociation, Buffulo, NY.

Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York: Guilford.Tangney, J. P., & Fisher, K. W. (1995). Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride.

New York: Guilford.

M. Webb et al. / Child Abuse & Neglect 31 (2007) 1143–1153 1153

Tangney, J. P., Wagner, P. E., & Gramzow, R. (1989). The Test of Self-Conscious Affect (TOSCA). Fairfax, VA: George MasonUniversity.

Tangney, J. P., Wagner, P. E., & Gramzow, R. (1992). Proneness to shame, proneness to guilt, and psychopathology. Journal ofAbnormal Psychology, 101(3), 469–478.

Yapko, M. D. (2000). Hand-me-down blues: How to stop depression from spreading in families. New York: St. Martin’s.Zelikovsky, N., & Lynn, S. J. (1994). The aftereffects and assessment of physical and psychological abuse. In S. J. Lynn & J. W.

Rhue (Eds.), Dissociation: Clinical and theoretical perspectives (pp. 190–214). New York: Guilford Press.Zelikovsky, N., & Lynn, S. J. (2002). Childhood psychological and physical abuse: Psychopathology, dissociation, and Axis I

diagnosis. Journal of Trauma and Dissociation, 3(3), 27–58.