Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 29.
Published in final edited form as: J Psychopathol Behav Assess. 2019 Mar 22;41(2):257–270. doi: 10.1007/s10862-019-09734-1

Self-Evaluation and Depressive Symptoms: A Latent Variable Analysis of Self-esteem, Shame-proneness, and Self-criticism

Andrew C Porter 1,2,*, Rachel L Zelkowitz 1, Darcy C Gist 1, David A Cole 1
PMCID: PMC7523943  NIHMSID: NIHMS1525204  PMID: 32999528

Abstract

This paper examined the unique associations of latent self-esteem with symptoms of depression, over-and-above latent shame-proneness (study 1) and latent self-criticism (study 2), among two samples of undergraduate students. In study 1, confirmatory factor analysis (CFA) showed convergent and discriminant validity for most measures of shame-proneness and self-esteem. Shame-proneness and self-esteem (taken separately) were strongly related to depressive symptoms. Controlling for each other, self-esteem but not shame-proneness remained associated with depressive symptoms. In study 2, CFA showed convergent validity but not discriminant validity for measures of self-criticism and low self-esteem. Self-esteem and self-criticism (taken separately) were strongly related to depressive symptoms. Controlling for each other, however, neither construct was significantly associated with depressive symptoms. Findings suggest that (a) self-esteem and shame-proneness have good discriminant validity, (b) self-esteem is uniquely related to depressive symptoms above and beyond shame-proneness, and (c) self-report measures of self-criticism and self-esteem have poor discriminant validity.

Keywords: Depression, shame-proneness, self-esteem, self-criticism, validity


Major Depressive Disorder is a highly prevalent psychiatric disorder that affects roughly 4.7% of individuals worldwide (Ferrari et al., 2013), with many more individuals experiencing subthreshold, but clinically impairing depressive symptoms (Lewinsohn, Solomon, Seeley, & Zeiss, 2000). Depressive symptoms are associated with social and occupational impairment and depression is ranked one of the illnesses most contributing to disability worldwide (Mathers & Loncar, 2006). Identifying processes that maintain depressive symptoms will be critical for improving existing interventions for this debilitating disorder. One potential maintenance factor warranting closer attention is negative self-evaluation.

Negative self-evaluation is a complex collection of processes with strong theoretical ties to depression (Beck, 1963). Three facets of such evaluation are shame-proneness, self-criticism, and self-esteem. Examined separately, each has empirical support as a correlate and predictor of depressive symptoms (e.g., Auerbach, Ho, & Kim, 2014; Dunkley, Sanislow, Grilo, & McGlashan, 2009; Ehret, Joorman, & Berking, 2015; Kim, Thibodeau, & Jorgensen, 2011; Mills et al., 2015; Orth, Berking, & Burkhardt, 2006; Rice, Fallon, & Bambling, 2011; Sowislo & Orth, 2013). The degree to which these variables are uniquely associated with depressive symptomatology remains unclear, however. Examining these conceptually similar self-evaluation characteristics together will clarify the extent to which they overlap or diverge and how much each of these constructs is uniquely related to depression. Below, we outline areas of overlap and distinction among these constructs and detail the aims of the present study. We first define each construct and note areas of convergence across the literatures of low self-esteem, shame-proneness and self-criticism in terms of how each construct may relate to depressive symptoms. Next, we describe the conceptual and empirical distinctions between the constructs. Finally, we detail the aims of the current study.

Conceptual links between depressive symptoms and low self-esteem, shame-proneness, and self-criticism

Self-esteem is defined as the overall subjective appraisal of one’s worth (Orth, Robins, Meier, & Conger, 2016). Shame-proneness is defined as a propensity to experience shame, a self-conscious emotion typified by negative self-evaluation in response to perceived shortcomings and transgressions (typically social) across a variety of situations (Tangney & Dearing, 2002; Tracy, 2012). Self-criticism consists of negative self-evaluations often triggered by perceived discrepancies between the actual and ideal self (e.g., Beck, 1963; Blatt, 1974). Separate literatures exist for self-esteem (e.g., Sowislo & Orth, 2013), shame-proneness (e.g., Kim et al., 2011), and self-criticism (e.g., Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982) that support the utility of each characteristic for describing experiences of currently depressed individuals and predicting depressive symptoms.

First, low self-esteem is often described as a key vulnerability characteristic that influences the onset and maintenance of depression through a variety of mechanisms (Orth & Robins, 2013). For example, individuals with low self-esteem have a propensity for excessive reassurance seeking (Joiner, Katz, & Lew, 1999). They may seek positive feedback about their worth from others or may seek negative feedback consistent with their low self-esteem (Joiner, 1995; Joiner et al., 1999; Swann, Wenzlaff, & Tafarodi, 1992). Both forms of reassurance seeking can result in interpersonal rejection and further degrade self-esteem, thereby increasing risk for depression (Starr & Davila, 2008). Individuals with low self-esteem also often avoid and withdraw from social situations (Orth & Robins, 2013; Weinstock & Whisman, 2004). Thus, these individuals are less likely to reap the protective effects of social support and are more likely to feel isolated from others, elevating their risk for depression (Santini, Koyanagi, Tyrovolas, Mason, & Haro, 2015).

In addition, low self-esteem functions as a negative cognitive style that causes individuals to interpret negative experiences in a self-deprecatory manner (Kuster, Orth, & Meier, 2012). Both sociometer theory (Leary & Baumeister, 2000) and terror management theory (Pyszczynski, Greenberg, Solomon, Arndt, & Schimel, 2004) suggest that individuals low in self-esteem are more attuned to evidence that they are worthless, not accepted by others, and not measuring up to cultural or personal standards. Self-verification theory (Swann, 2012) further suggests that individuals seek or interpret feedback from others that is consistent with views of themselves, which further supports such a negative bias in interpersonal feedback. Finally, individuals who are low in self-esteem are especially likely to ruminate about their negative experiences and feelings about themselves, which has been found to elevate risk for increases in depressive symptoms (Kuster et al., 2012). For all of these reasons, low self-esteem constitutes a cognitive vulnerability for depression (Masselink, Roekel, & Oldehinkel, 2017; Orth & Robins, 2013; Sowislo, Orth, & Meier, 2014).

Second, shame-proneness (Andrews, Qian, & Valentine, 2002; Kim et al., 2011; Mills et al., 2015; Rice et al., 2011) and self-criticism (e.g., Abela, Webb, Wagner, Ho, & Adams, 2006; Auerbach et al., 2014; Blatt et al., 1982; Dunkley et al., 2009; Ehret et al., 2015) both have support as correlates and prospective predictors of depressive symptoms. Each may lead to depression through similar processes, as does low self-esteem. Shame-prone individuals, like individuals low in self-esteem, engage in problematic interpersonal behaviors, which could elevate their risk for depression. For instance, shame-prone individuals show deficits in effective interpersonal problem-solving (Covert, Tangney, Maddux, & Heeno, 2003; Zuroff & Duncan, 1999). Shame motivates individuals to respond to conflict in maladaptive ways, often involving withdrawal, conflict avoidance, and the criticism of self or others (Elison, Lennon, & Pulos, 2006). In addition, shame responses to interpersonal conflict predict poorer relationship satisfaction and depression (Black, Curran, & Dyer, 2013; Rice et al., 2011), even after controlling for rumination and affect regulation (Rice et al., 2011).

Self-critical individuals also experience difficulties in interpersonal relationships. For instance, Zuroff and Duncan (1999) found that individuals with highly self-critical cognitive styles experienced more extreme negative cognitive-affective reactions to dyadic conflict than did individuals who were lower in self-criticism. This, in turn, led to increased hostility in the individual’s partner and to deficits in conflict resolution. In addition, like individuals with low self-esteem, self-critical individuals often avoid intimacy out of fear of rejection (Zuroff, Moskowitz, & Côté, 1999), which also prevents them from developing strong social support networks. Self-criticism has also been found to mediate the relation between anxious and avoidant attachment styles and depressive symptoms (Cantazaro & Wei, 2010). The dysfunctional interpersonal behavior of self-critical individuals are suggested to “promote depressive vulnerability” (Vettese & Mongrain, 2000, p. 621).

Other research suggests that shame-prone and self-critical individuals (like individuals with low self-esteem) are often excessively self-deprecatory and fearful of being devalued by others (Castilho, Pinto-Gouveia, & Duarte, 2017; Dorahy & Hanna, 2012). This, in turn, makes them prone to interpreting individual and interpersonal experiences in ways that generate beliefs of worthlessness and inadequacy (Zuroff & Mongrain, 1987). Such beliefs directly contribute to depression. In fact, two studies have shown that self-critical thinking accounts for a significant proportion of the relation between shame and internalizing symptoms (including depressive symptoms) in both clinical (Castilho et al., 2017) and undergraduate samples (Porter, Zelkowitz, & Cole, 2018). Taken together, there is substantial overlap between self-esteem, shame-proneness and self-criticism vís-a-vís depression. However, the reasons why each may be uniquely related to the disorder remains unclear.

Conceptual and empirical distinctions between low self-esteem, shame-proneness, and self-criticism

Although the three constructs appear to be similarly related to depressive symptoms, theoretical accounts of shame-proneness and self-criticism suggest that they may be distinct from low self-esteem in meaningful ways. For example, self-esteem represents one’s general evaluation of one’s worth in ways that are largely detached from specific situations, whereas shame involves negative affect and negative self-evaluation that occur in reaction to specific situations (Tangney & Dearing, 2002). By extension, shame-proneness describes individuals who have a propensity for such experiences (Tangney & Dearing, 2002). Like shame-proneness, self-criticism may constitute a more specific manifestation of negative self-evaluation than global low self-esteem. Namely, self-criticism describes individuals who have a propensity toward being overly harsh and self-punitive when they fail to meet the unreasonably high standards they set for themselves and who are hypervigilant and concerned about criticism from others (e.g., Blatt, 1995).

With respect to empirical differences between the constructs, zero-order correlations between self-esteem, shame-proneness, and self-criticism are typically strong but vary substantially between studies. Some studies report correlations between shame-proneness and self-esteem that are weaker than −.52 (Mosewich, Kowalski, Sabiston, Sedgwick, & Tracy, 2011; Orth, Robins, & Soto, 2010; Török, Szabó, & Boda-Ujlaky, 2014.) Although technically considered medium to large effects based on standard interpretations of Pearson’s r, such findings still indicate that the measures of the two constructs share less than 27% of their variance. In contrast, others have reported larger correlations between measures of these constructs (i.e., stronger than −.52; Birtel, Wood, & Kempa, 2017; Wells, Glickauf-Hughes, & Jones, 1999), and even as large as −.91 (Wood, Byrne, Burke, Enache, & Morrison, 2017), raising concerns about discriminant validity. Correlations between self-criticism and self-esteem have a similarly wide range across studies, from correlations hovering around −.40 (Grzegorek, Slaney, Franze, & Rice, 2004; Johnson, 2010; Kelly & Carter, 2013) to those stronger than −.58 (Dunkley, Masheb, & Grilo, 2010; Iancu, Bodner, & Ben-Zion, 2015; Stolow, Zuroff, Young, Karlin, & Abela, 2016), and even as high as −.83 (Iancu et al., 2015). Conceptually and empirically, the true strength of relation between self-esteem, shame-proneness, and self-criticism remains unclear.

The mixed findings from previous research may be due to two shared limitations. First, evidence for discriminant validity was based exclusively on Pearson correlations between manifest variables. Pearson correlations of manifest variables, however, are attenuated to the degree that measures involved are not perfectly reliable or valid, yielding spurious evidence of discriminant validity (Cole, 1987). Second, most studies examining self-esteem with shame-proneness or self-criticism (especially those that have examined each in relation to depression) have used only single measures of each construct. Comparing two variables using only single measures of each can lead to biased estimations of the true relations among the variables when the measures are not perfectly reliable or valid (Cole & Preacher, 2014).

Current Studies

The current paper seeks to address these limitations and extend prior research on self-evaluation and depression by presenting findings from two studies. In study 1, we examined the relations of shame-proneness and self-esteem to depressive symptoms. In study 2, we examined the relations of self-criticism and self-esteem to depressive symptoms. Both studies used multiple measures of self-esteem, self-criticism, and shame-proneness as indicators of latent variables (study 1 also includes a latent variable for depressive symptoms), thereby minimizing the effects of imperfect measure reliability and validity (Cole & Preacher, 2014). In each study, we used confirmatory factor analysis (CFA) to examine the overlap between the predictors.

CFA overcomes the limitation of relying on zero-order correlations between manifest variables in two key ways. First, CFA uses multiple measures of each construct to extract latent variables, thereby minimizing the effects of measurement error and providing less-biased estimates of the true relations between constructs (Cole, 1987). Second, CFA can be used to conduct actual statistical tests of discriminant validity (Cole, 1987; Kenny & Kashy, 1992). This is accomplished by creating a constrained model and examining its goodness of fit. The constrained model consists of the following. First, the relations between latent factors are fixed to unity (i.e., correlations are fixed at 1.0 or −1.0). Second, the relations of each factor to other factors such as depressive symptoms are constrained to equality (i.e., their correlations with other constructs are fixed to be equal in strength).1 A poor goodness of fit for the constrained model described above lends support for discriminant validity.

Following the CFA, we use path analysis in each study to examine the relations of self-esteem and either shame-proneness or self-criticism to depressive symptoms, controlling for the other construct. This analysis enables us to test the extent to which shame-proneness and self-criticism are uniquely related to depressive symptoms, over-and-above low self-esteem. The goals of both studies were to test the convergent validity of these measures, the discriminant validity of these constructs, and their incremental relations to depressive symptoms.

Study 1: Examining the unique associations of self-esteem and shame-proneness with symptoms of depression

Study 1 focused on shame-proneness, self-esteem, and symptoms of depression.

Method

Participants.

The sample consisted of 382 students recruited from the research subject pool at a mid-sized southern private university. Average age of the participants was 18.88 (SD = 1.43), and 76% of participants were female. The sample was 58% White, 19% Asian or American-Asian, 9% Black, and 3% other. Eleven percent of participants endorsed more than one ethnicity.

Measures.

We used three measures of self-esteem. The Rosenberg Self-Esteem Inventory (RSE; Rosenberg, 1965) is a self-report measure of global self-esteem, consisting of 10 items. Participants report the extent to which they agree with each statement using a 1 (Strongly Disagree) to 4 (Strongly Agree) Likert scale. Half of the statements are positively worded (e.g., “On the whole, I am satisfied with myself”), and half of the statements are negatively worded and reverse scored (e.g., “At times I think I am no good at all”). A total score is generated by summing all the items. The RSE has been validated for use across many different populations with coefficient alphas ranging from .79 to .86 (Supple, Su, Plunkett, Peterson, Bush, 2013). Internal consistency in the current sample was excellent (α = .90).

The Coopersmith Self-Esteem Inventory (CSEI; Coopersmith, 1959) is a self-report measure of global self-esteem with 25 statements of attitudes that individuals may have about themselves. Some of the statements are positively worded (e.g., “Things don’t usually bother me”); others are negatively worded (e.g., “I find it very hard to talk in front of a group”). Participants endorse statements as either “like me” or “unlike me.” Participants are assigned a score of “1” for each positively worded item that they endorse as “like me” and for each negatively worded item they endorse as “unlike me.” A total is the sum of these scores. The measure has shown strong internal consistency (Ahmed, Valliant, & Swindle, 1985). Internal consistency in the current sample was good (α = .84).

The Self-Liking Self-Competence Scale – Revised, Self-Liking Subscale (SLSCS-R_SL; Tafarodi & Swann, 2001) is a 16-item self-report measure assessing two types of self-esteem. Eight items assess “self-liking,” and the other eight items assess “self-competence.” Half of the items are positively worded, and half are negatively worded and reverse scored. An example self-liking item is “I am secure in my sense of self-worth.” An example self-competence item is “I am almost always able to accomplish what I try for.” Response options are 1 (Strongly Disagree) to 7 (Strongly Agree). To be consistent with the other self-esteem measures (which emphasize global evaluations of worth), only the self-liking subscale was used. The measure has been validated for use with university student samples (Tafarodi & Swann, 2001). In the current sample, internal consistency for the self-liking subscale was excellent (α = .92).

We also obtained three measures of shame-proneness. The Test of Self-Conscious Affect – 3 (TOSCA-3; Tangney, Dearing, Wagner, & Gramzow, 2000) consists of 16 items, each of which contains a brief scenario and several responses to the scenario. Each item in the TOSCA-3 lists four to six potential responses (shame, guilt, beta pride, alpha pride, and externalization) to a hypothetical scenario.2 Participants respond on 1 to 5 Likert scales. Only shame responses were used in the current study. An example of a scenario is “You break something at work and then hide it.” The shame response associated with this scenario is “You would think about quitting.” Total shame scores were the sum of the shame responses to the 16 scenarios. Alpha coefficients for the shame scale range from .77 to .88 (Rüsch et al., 2007). In the current study, coefficient alpha was good (α = .87).

The Guilt-and-Shame-Proneness Scale (GASP; Cohen, Wolf, Panter, & Insko, 2011) is a 16-item scenario-based measure that asks respondents to report the likelihood with which they would respond to a series of hypothetical scenarios. Response options reflect either shame or guilt. The measure has four subscales, two assessing guilt-proneness and two assessing shame-proneness (shame-withdrawal and shame-negative self-evaluation). Both shame subscales were used. Internal consistency for the withdraw subscale was adequate (α = .68); internal consistency for the negative self-evaluation subscale was also adequate (α = .71).

The Internalized Shame Scale (ISS; Cook, 1994, 2001) is a 30-item measure of trait shame (24 items) and self-esteem (6 items). We only used the shame subscale. Response options were on 0 (never) to 4 (almost always) Likert scales, reflecting the frequency with which participants experienced thoughts and emotions such as “I feel insecure about others’ opinion of me.” Previous studies report coefficient alphas of .90 to .97 (Del Rosario & White, 2006). In the current sample, internal consistency was excellent (α = .96).

Finally, we obtained three measures of depressive symptoms. The Self-Rating Depression Scale (SDS; Zung, 1965) is a 20-item measure of depressive symptoms. Items ask how often respondents generally have certain depressive experiences: 1 (a little of the time) to 4 (most of the time) scale. An example of an item is “I feel down-heartened and blue.” Previous studies report good internal consistency (e.g., alphas of .88 to .93; Gabrys & Peters, 1985) and validity (e.g., Cole, 1987) for the SDS. Internal consistency for this sample was good (α = .85).

The Center for Epidemiological Studies – Depression Scale (CES-D; Radloff, 1997) is a 20-item measure asking how often people have experienced depression-related symptoms (such as negative thinking, sadness, and loneliness) in the past week using a 0 (rarely or none of the time) to 3 (all of the time) scale. Previous studies report coefficient alpha of .82 (Lewinsohn, Seeley, Roberts, & Allen, 1997). Internal consistency in the current sample was excellent (α = .91).

The Beck Depression Inventory-II (BDI-II, Beck, Steer, & Brown, 1996) is a commonly-used, well-validated measure designed to assess severity of depressive symptoms. It asks respondents to rate their experience of 21 depressive symptoms over the past two weeks on 0 to 3 scales, with higher scores indicating increased levels of symptoms. Due to concerns raised by the Institutional Review Board, we removed the item assessing suicidality. The measure has been validated in a university population where coefficient alpha was .91 (Dozois, Dobson, & Ahnberg, 1998). In the current sample, internal consistency was excellent (α = .90).

Procedure

Participants independently completed measures of self-esteem, shame-proneness, and depressive symptoms via the Qualtrics online survey system. Graduate research assistants contacted participants who reported elevated depressive symptoms and provided them information about online resources and ways to contact the university counseling center. Participants received course credit in exchange for their participation.

Results

Descriptive statistics.

We conducted all data analyses using Mplus Version 8.0 (Muthén & Muthén, 2017). Table 1 reports the descriptive statistics for all study variables. Approximately 11%, 18%, and 25% of participants scored above clinical cutoffs for moderate depressive symptom severity on the SDS, BDI-II, and CES-D, respectively (Beck et al., 1996; Dugan et al., 1998; Radloff, 1997). The correlations between measures of self-esteem and shame-proneness ranged from −.12 to −.80, consistent with the mixed findings from previous research.

Table 1.

Correlations and Descriptive Statistics for Study 1 Measures

Measure 1 2 3 4 5 6 7 8 9 10
1. RSE -
2. CSEI .76 -
3. SLSC-R_SL .82 .70 -
4. TOSCA-3 −.51 −.45 −.51 -
5. GASP_W −.26 −.27 −.23 .45 -
6. GASP_NSE −.24 −.12 −.28 .43 .26 -
7. ISS −.80 −.74 −.79 .61 .34 .32 -
8. SDS −.67 −.68 −.61 .42 .30 .15 .72 -
9. CES-D −.64 −.63 −.62 .42 .27 .19 .79 .72 -
10. BDI-II −.65 −.63 −.58 .40 .23 .17 .76 .81 .72 -
M 30.34 15.75 25.67 49.02 2.92 5.80 32.69 37.95 16.84 9.23
SD 6.19 5.29 7.45 11.31 1.13 1.07 19.91 8.96 10.27 8.53
Skewness −.40 −.45 −.04 −.04 .67 −1.40 .62 .62 .62 1.41
Kurtosis −.53 −.55 −.93 −.55 −.03 2.54 −.20 −.20 −.19 2.17
Minimum 12 1 9 23 1 1 0 20 0 0
Maximum 40 25 40 78 7 6.5 96 67 48 48

Note. All correlations significant at p < .05; RSE = Rosenberg Self-Esteem Inventory; CSEI = Coopersmith Self-Esteem Inventory; SLSCS-R_SL = Self-Liking Self-Competence Scale – Revised – Self-Liking Subscale; TOSCA-3 = Test of Self-Conscious Affect – 3; GASP_W = Guilt and Shame-Proneness Scale – Withdraw Subscale; GASP_NSE - = Guilt and Shame-Proneness Scale – Negative Self-Evaluation Subscale; ISS = Internalized Shame Scale; SDS = Self-Rating Depression Scale; CES-D = Center for Epidemiological Studies – Depression Scale; BDI-II = Beck Depression Inventory-II.

Goal 1: To test the discriminant and convergent validity of self-esteem and shame-proneness measures.

First, we tested discriminant and convergent validity for all measures of self-esteem and shame-proneness. With the ISS included, however, the measurement model fit poorly (χ2(13, N=389) = 108.33 (p < .001), CFI = .94, SRMR = .063, RMSEA = .137, 95% CI [.114, .162]), and examinations of modification indices revealed strong evidence that the ISS cross-loaded more strongly onto the self-esteem latent factor than it did onto the shame-proneness latent factor. Because the ISS showed poor convergence with the other shame-proneness measures and showed poor discriminant validity with self-esteem, we excluded the ISS from subsequent analyses. The model with the ISS removed is depicted in Figure 1 and reflects an oblique, two-factor confirmatory factor analysis (CFA) between self-reported self-esteem and shame-proneness. Without cross-loadings or correlated residuals, the model fit the data well (χ2(8, N = 389) = 22.075 (p = .005), CFI = .99, SRMR = .024, RMSEA = .067, 95% CI [.035, .102]). All measures had significant and sizeable loadings onto their respective factors, demonstrating convergent validity (Cole, 1987). Notably, the TOSCA-3 (λ = .93, 95% CI [.86, 1.0]) had a significantly larger loading than the GASP_W (λ = .48, 95% CI [.39, .57]) and the GASP_NSE (λ = .46, 95% CI [.37, .55]) onto latent shame-proneness. In addition, the CSEI (λ = .81, 95% CI [.77, .85]) had a significantly smaller factor loading than the RSE (λ = .93, 95% CI [.91, .96]) and the SLCS-R_SL (λ = .88, 95% CI [.85, .91]), onto latent self-esteem. The correlation between self-esteem and shame-proneness was −.60, 95% CI [−.69, −.51]. This correlation was significantly different from unity and suggests only 36% shared variance between the two latent variables. Collectively, these findings provide evidence of convergent validity for the measures and discriminant validity between the latent factors (Cole, 1987).

Figure 1.

Figure 1.

Confirmatory factor analysis of shame-proneness and self-esteem. All estimates are standardized and significant, ps < .001. RSE = Rosenberg Self-Esteem Inventory; CSEI = Coopersmith Self-Esteem Inventory; SLSCS-R_SL = Self-Liking Self-Competence Scale – Revised – Self-Liking Subscale; TOSCA-3 = Test of Self-Conscious Affect – 3, Shame-proneness subscale; GASP_W = Guilt and Shame-Proneness Scale – Withdraw Subscale; GASP_NSE - = Guilt and Shame-Proneness Scale – Negative Self-Evaluation Subscale.

Goal 2: To examine the associations of self-esteem and shame-proneness with symptoms of depression.

Figure 2 depicts a model in which a latent depressive symptoms variable was regressed onto latent self-esteem and latent shame-proneness. The chi-square test for fit was significant χ2(24, 389) = 71.569, (p < .001). Other fit indices suggested good fit to the data: CFI = .98, SRMR = .031, RMSEA = .071, 95% CI [.053, .091]. Factor loadings were .46 to .92 for shame-proneness measures, .82 to .93 for self-esteem measures, and .88 to .91 for depression measures. Latent variable correlations were −.60 (p < .001) between shame-proneness and self-esteem, .50 (p < .001) between depression and shame-proneness, and −.79 (p < .001) between depression and self-esteem. The standardized path coefficient between shame-proneness and depression was small and nonsignificant (.04, 95% CI [−.07, .14]). The standardized path coefficient between self-esteem and depressive symptoms was large and significant (−.77, 95% CI [−.86, −.70]). Self-esteem accounted for almost all (92%) of the correlation between shame-proneness and depressive symptoms. Unstandardized effects are depicted in Table 3.

Figure 2.

Figure 2.

Path model depicting shame-proneness and self-esteem as predictors of depression. All estimates are standardized. RSE = Rosenberg Self-Esteem Inventory; CSEI = Coopersmith Self-Esteem Inventory; SLSCS-R_SL = Self-Liking Self-Competence Scale – Revised – Self-Liking Subscale; TOSCA-3 = Test of Self-Conscious Affect – 3, Shame-proneness subscale; GASP_W = Guilt and Shame-Proneness Scale – Withdraw Subscale; GASP_NSE - = Guilt and Shame-Proneness Scale – Negative Self-Evaluation Subscale; SDS = Self-Rating Depression Scale; CES-D = Center for Epidemiological Studies – Depression Scale; BDI-II = Beck Depression Inventory-II. Estimates with asterisks are significant, ps < .001. All other coefficients are not significant at α = .05. Readers should note that, because the structural model is just-identified, the fit indices only provide information of the fit of the measurement model (Tomarken & Waller, 2003).

Table 3.

Path Estimates for Structural Models

Predictor of depressive
symptoms
Unstandardized
path
S.E. 95% CI Standardized
path
Study 1
1. Self-esteem −1.02 0.08 [−1.16,−0.87] −.78*
2. Shame-proneness 0.58 0.79 [−0.97,2.13] .04
Study 2
1. Self-esteem −0.77 0.82 [−1.00,1.24] −.56
2. Self-criticism 0.12 0.57 [−2.38,0.84] .13

Note. For Study 1, the dependent variable consists of a latent depressive symptoms factor represented by three indicators: the Self-Rating Depression Scale, the Center for Epidemiological Studies – Depression Scale, and the Beck Depression Inventory-II. For study 2, the dependent variable is represented solely by the Beck Depression Inventory-II.

*

p < .001.

To examine the degree to which the self-evaluation items in the depression measures may have driven this finding, we ran a second analysis in which all items from the BDI-II, CES-D, and SDS that reflected self-evaluation were removed (see supplemental material for removed items and results table for this analysis). The unique relation between self-esteem and depressive symptoms was somewhat reduced (from −.78 to −.72) but remained large and significant (p < .001). The unique relation between shame-proneness and depressive symptoms remained nonsignificant (.04, p > .47). In the second analysis, self-esteem accounted for 80% of relation between shame-proneness and depressive symptoms.

Summary of Study 1 Findings

In study 1, two key findings emerged. First, CFA revealed convergent validity and discriminant validity for most measures of shame-proneness and self-esteem.3 The correlation between shame-proneness and self-esteem was large but not so large (at −.60) as to call into question the discriminant validity of these two constructs. Again, we note that this finding implies shared variance of only 36%. Second, depressive symptoms were strongly associated with both self-esteem and shame-proneness. Self-esteem remained significantly related to depressive symptoms when shame-proneness was statistically controlled; however, shame-proneness was unrelated to depressive symptoms when self-esteem was controlled.

Study 2: Examining the unique associations of self-esteem and self-criticism with symptoms of depression

Study 2 focused on self-criticism, self-esteem, and symptoms of depression.

Method

Participants.

The sample consisted of 193 students recruited from the research subject pool at the same university as in study 1. Average age of the participants was 19.21 (SD = 1.89), and 79% of participants were female. The sample was 54% White, 19% Asian or American-Asian, 12% Black, 2% Hispanic, and 3% other. Ten percent of participants endorsed more than one ethnicity.

Measures

We used the same three measures of self-esteem that were used in study 1 (see pages 9 and 10 for descriptions of these measures). Internal consistency in study 2 was excellent (α = .92) for the RSE, good for the CSEI (α = .86), and excellent for the SLSC-R_SL (α = .92).

We used three measures of self-criticism. The Self-Rating Scale (SRS; Hooley, Ho, Slater, & Lockshin, 2002) is an eight-item measure of self-criticism that asks people to rate their agreement with statements such as “others are justified in criticizing me.” Internal consistency has previously been found to be acceptable (alpha > .70; Glassman, Weierich, Hooley, Deliberto, & Nock, 2007). Internal consistency in the current study was excellent (α = .90).

The Levels of Self-Criticism Scale (LOSCS; Thompson & Zuroff, 2004) is a 22-item measure that assesses self-criticism due to negative comparisons to others and to internalized ideals. Participants indicate how well particular statements describe themselves using 1 (not at all) to 7 (very well) Likert scales. Representative items include “I often worry that other people will find out what I’m really like and be upset with me” (comparative self-criticism; LOSCS_CSC) and “I get very upset when I fail” (internalized self-criticism; LOSCS_ISC). Internal consistency in the current study was excellent for internalized self-criticism (α = .90) and good for comparative self-criticism (α = .80).

The Depressive Experiences Questionnaire (DEQ; Zuroff, Quinlan, & Blatt, 1990) is a 66-item measure developed by Blatt et al. (1976) to measure feelings and cognitions associated with depression. It uses 1 (strongly disagree) to 7 (strongly agree) Likert scales. The measure has been validated among undergraduates and shown acceptable internal consistency in this population (alphas range from .73 to .81). Only the self-criticism subscale was administered (19 items). Representative items include “I have a difficult time accepting weaknesses in myself” and “I tend to be very critical of myself.” Internal consistency in the current study was good (α = .85).

We used the BDI-II, as in study 1. Due to concerns raised by the Institutional Review Board, we again removed the item assessing suicidality. In the study 2 sample, internal consistency was excellent (α = .90).

Procedure

Participants independently completed measures of self-esteem, self-criticism, and depression via the Qualtrics online survey system. Graduate research assistants contacted participants who reported elevated depressive symptoms and provided them information about online resources and ways to contact the university counseling center. Participants received course credit in exchange for their participation.

Results

Descriptive statistics.

We conducted all data analyses using Mplus Version 8.0 (Muthén & Muthén, 2017). Table 2 reports the means, standard deviations, and correlations for all study variables. On the BDI-II, 13.3% percent of participants scored at or above the clinical cutoff of 20 for moderate depressive symptoms (Beck et al., 1996). The correlations between measures of self-esteem and self-criticism were large, ranging from −.51 to −.75, with the majority of correlations being stronger than −.65. Thus, unlike previous research that found moderate to strong correlations, the findings of this study were more in agreement with studies finding strong effects.

Table 2.

Correlations and Descriptive Statistics for Study 2 Measures

Measure 1 2 3 4 5 6 7 8
l. SRS -
2. DEQ 0.61 -
3. LOSCS-CSC 0.64 0.70 -
4. LOSCS-ISC 0.53 0.69 0.56 -
5. RSE −0.73 −0.71 −0.75 −0.55 -
6. SLSC-R_SL −0.69 −0.72 −0.74 −0.56 0.82 -
7. CSEI −0.68 −0.65 −0.72 −0.51 0.77 0.76 -
8. BDI-II 0.55 0.53 0.54 0.37 −0.63 −0.63 −0.59 -
M 23.54 78.04 38.02 45.55 30.51 26.45 15.83 9.67
SD 11.06 16.60 10.77 11.79 6.49 7.50 5.60 8.28
Skewness .66 .14 .40 −.21 −.29 −.14 −.42 1.19
Kurtosis −.23 −.34 −.70 −.42 −.90 −.84 −.61 1.46
Minimum 8 32 17 16 15 10 2 0
Maximum 56 126 73 70 40 40 25 41

Note. All correlations significant at p < .001; SRS = Self Rating Scale; DEQ = Depressive Experiences Questionnaire; LOSCS_ISC = Levels of Self-Criticism Scale – Internalized Self-Criticism; LOSCS_CSC = Levels of Self-Criticism Scale – Comparative Self-Criticism; RSE = Rosenberg Self-Esteem Scale; SLCS-R_SL – Self-Liking/Self-Competence Scale – Self-Liking Subscale; CSEI = Coopersmith Self-Esteem Inventory; BDI-II = Beck Depression Inventory Scale – II.

Goal 1: To test the discriminant and convergent validity of self-esteem and self-criticism measures.

We tested discriminant and convergent validity between measures of self-esteem and self-criticism using the oblique, two-factor CFA depicted in Figure 3.4 Without cross-loadings or correlated residuals, the model fit the data well, χ2(8, 193) = 14.91 (p > .06), CFI = .99, SRMR = .016, RMSEA = .067, 95% CI [.000, .120]. All measures had significant and sizeable loadings onto their respective factors, demonstrating convergent validity. The loadings of the DEQ (λ = .84, 95% CI [.79, .90]), SRS (λ = .78, 95% CI [.72, .85]) and CSC (λ = .84, 95% CI [.79, .89]) onto the self-criticism factor were similar in strength. The CSEI (λ = .84, 95% CI [.79, .89]), however, had a significantly smaller factor loading than both the RSE (λ = .93, 95% CI [.91, .96]) and the SL (λ = .92, 95% CI [.90, .95]) onto latent self-esteem. This finding is consistent with the finding from study 1. The correlation between self-esteem and self-criticism was very large (−.98, p <.01) with a 95% CI that included −1.0 [−1.0, −.94].

Figure 3.

Figure 3.

Confirmatory factor analysis of self-criticism and self-esteem. All estimates are standardized and significant, ps < .001. SRS = Self Rating Scale; DEQ = Depressive Experiences Questionnaire; LOSCS_CSC = Levels of Self-Criticism Scale – Comparative Self-Criticism; RSE = Rosenberg Self-Esteem Scale; CSEI = Coopersmith Self-Esteem Inventory; SLCS-R_SL – Self-Liking/Self-Competence Scale – Self-Liking Subscale. All estimates are significant, ps < .001.

Collectively, these findings provide evidence of convergent validity for all measures. However, the correlation between the two latent variables was quite strong and raises concerns about the discriminant validity of self-criticism and low self-esteem.

Goal 2: To examine the unique associations of self-esteem and self-criticism with symptoms of depression.

Figure 4 depicts a model in which depressive symptoms are regressed onto latent self-esteem and latent self-criticism. The fit indices suggested a good fit for the model: χ2(12, 193) = 15.963, p > .19; CFI = 1.0; SRMR = .015; RMSEA = .042, 95% CI [.000, .090]. Standardized factor loadings were .79 to .84 for self-criticism measures and .84 to .94 for self-esteem measures. Correlations between the BDI-II and latent self-esteem (r = −.68) and latent self-criticism (r = .67) were strong and significant (ps < .001). The standardized path coefficients between self-esteem and the BDI-II (γ = −.56, p > .35) and self-criticism and the BDI-II (γ = .13, p > .84), each controlling for the other factor, were not significant. Unstandardized effects are depicted in Table 3.

Figure 4.

Figure 4.

Path model depicting self-esteem and self-criticism as predictors of depressive symptoms. All estimates are standardized. SRS = Self Rating Scale; DEQ = Depressive Experiences Questionnaire; LOSCS-CSC = Levels of Self-Criticism Scale – Comparative Self-Criticism; RSE = Rosenberg Self-Esteem Scale; SLCS-R_SL – Self-Liking/Self-Competence Scale – Self-Liking Subscale; CS = Coopersmith Self-Esteem Inventory; BDI-II = Beck Depression Inventory Scale – II. Estimates with asterisks are significant, ps < .001. All other coefficients are not significant at α = .05. Readers should note that, because the structural model is just-identified, the fit indices only provide information of the fit of the measurement model (Tomarken & Waller, 2003).

As an additional test of discriminant validity, we fit a model in which (a) the latent correlation between self-criticism and self-esteem was fixed at −1.0 and (b) the correlation between self-criticism and depressive symptoms was fixed at the absolute value of the correlation between self-esteem and depressive symptoms. This model did not show a worse fit to the data than when these parameters were freely estimated (Δχ2(2) = 2.59, P > .100), providing additional evidence of poor discriminant validity between self-criticism and self-esteem. To examine the degree to which the self-evaluative items in the BDI-II were driving this finding, we removed all self-evaluative items from the BDI-II and re-ran the previous model. Results did not meaningfully change. The unique relation of self-esteem and the BDI-II (γ = −.52, p > .36) and the unique relation between self-criticism and the BDI-II (γ = .08, p > .89) remained nonsignificant.

Summary of Study 2 Findings

Two key findings emerged from Study 2. First, CFA demonstrated convergent validity but not discriminant validity for measures of self-criticism and self-esteem. The correlation between self-criticism and self-esteem was nearly perfect (at −.98). Second, depressive symptoms were strongly associated with both self-esteem and self-criticism when each factor was considered uniquely. Neither self-esteem nor self-criticism showed incremental utility over the other factor in relation to depressive symptoms.

General Discussion

This is the first study to use structural equation modeling and confirmatory factor analysis to examine the unique relations of self-esteem with depressive symptoms, over-and-above two conceptually similar constructs: shame-proneness (study 1) and self-criticism (study 2). With respect to the shame-proneness-self-esteem relation, confirmatory factor analysis (CFA) suggested that the Internalized Shame Scale, unlike the other measures of shame-proneness, had weak discriminant validity with self-esteem. This may be due to the method used in the ISS. Like the measures of self-esteem, the ISS asks respondents to indicate the degree to which self-evaluative statements are true for them (e.g., “I feel like I am never quite good enough”). This format may overly emphasize broad self-evaluative beliefs that typify low self-esteem. The other indices of shame-proneness did show discriminant validity with self-esteem and suggests that the two constructs are distinct but significantly related, which is consistent with previous theoretical work and some empirical findings based on zero-order correlations between single measures of each construct (e.g., Tangney & Dearing, 2002). A CFA in study 2, however, raised concerns about discriminant validity between self-criticism and low self-esteem.

Despite the overlap between self-esteem and shame-proneness, the identified correlation still supports discriminant validity of these constructs. That is, there was substantially more unique variance (64%) in each latent factor than there was shared variance (36%) between them. Previous research on shame-proneness and self-esteem offers insight into the difference between these two constructs. Shame is a discrete experience that occurs in response to specific situations and involves acute negative affect, excessive concern of evaluation from others, and motivation to withdraw from and avoid situations similar to the shame experience (Cohen et al., 2011; Tracy, 2012). Shame-prone individuals, by extension, frequently respond in these ways to a variety of situations. Because such responses are maladaptive, each shame experience could incrementally degrade a person’s interpersonal functioning (e.g., via poor conflict management) or psychological wellbeing (e.g., via increasing feelings of worthlessness). Many people with low self-esteem, however, do not frequently experience shame. Rather, self-esteem represents a more general and enduring self-belief that is largely independent of specific situations. Because of these differences, shame-proneness and self-esteem may predict different outcomes or relate to the same outcome via different mechanisms.

The second finding of study 1 showed that, after controlling for self-esteem, the relation between shame-proneness and depressive symptoms was negligible; in fact, the association between shame-proneness and depressive symptoms was almost completely accounted for by self-esteem. In other words, even though shame-proneness was found to be distinct from self-esteem, the unique characteristics of shame-proneness were not associated with depressive symptoms. The relation between self-esteem and depressive symptoms, however, was minimally reduced after controlling for shame-proneness.

The findings from study 2 suggested that self-report measures of self-criticism and low self-esteem have weak discriminant validity. As a result of the sizable collinearity between self-criticism and low self-esteem, neither predictor was uniquely related to depressive symptoms, controlling for the effects of the other.

Collectively, these studies suggest that the shared characteristics (i.e., negative self-evaluation) between low self-esteem, shame-proneness, and self-criticism may be the principal drivers of the correlation between shame-proneness and depressive symptoms and self-criticism and depressive symptoms. One explanation for this finding is that individuals high in shame-proneness may be at risk for depression because of the negative self-evaluation triggered by their shame experiences. Similarly, a self-critical response style may place individuals at risk for depression because of the wear that such responses may have on a person’s self-esteem. Alternately, depression itself may diminish individuals’ self-esteem. Future longitudinal research is needed to clarify these potentially prospective relations.

These findings have implications for the conceptualization of shame-proneness, self-criticism, and low self-esteem. Study 1 re-affirmed that self-esteem and shame-proneness are distinct characteristics. Although we did not find that shame-proneness was related to depressive symptoms after controlling for self-esteem, the unique aspects of shame may be related to other important clinical outcomes. For example, the withdraw/avoidance component of shame-proneness may influence or maintain anxiety, substance abuse, disordered eating, and nonsuicidal self-injury (Aldao, Nolen-Hoeksema, & Schweizer, 2010; Anderson & Crowther, 2012).

Study 2, however, found substantial overlap between self-criticism and low self-esteem, enough to call into question the discriminant validity of the two constructs. If there are important theoretical distinctions between the two constructs, they may not be captured by self-report measures of self-criticism. If these findings are replicated, researchers should revisit the conceptualization of self-criticism as a construct distinct from low self-esteem.

Certain limitations of this study suggest avenues for future research. First, the design of this study was cross-sectional. Therefore, the directionality of the observed relations remains unclear. Findings from prospective research could (a) identify whether depressive symptoms are a common underlying cause of shame-proneness, low self-esteem, and self-criticism, thereby explaining the overlap between these constructs and (b) clarify the degree to which shame, self-criticism, and self-esteem are affected by each other and possibly mediate or moderate their respective associations with depressive symptoms.

Second, although we used latent variable statistics to minimize the effects of measurement error, the latent variables were represented solely by self-report measures. Thus, the latent variables may still misrepresent true shame-proneness, self-esteem, and depression. For example, they may contain variance due to both the construct of interest and self-report method. The fact that the depressive symptoms variable in study 2 was represented by a single indicator makes this limitation even more salient for the findings of study 2. That is, using a single indicator for depressive symptoms means that the variance in our outcome for study 2 may be due, in part, to characteristics that are specific to the Beck Depression Inventory – II (in addition to general self-report method and the construct of interest). Moreover, the scenario-based format of the self-report shame-proneness measures is qualitatively different than that of the self-esteem and depression measures. The difference between scenario-based measures and those used in the self-esteem and depression measures may have influenced the relation between latent shame-proneness and self-esteem and depression. Future research should seek to extend the current findings by constructing latent variables of shame-proneness, self-criticism, self-esteem, and depressive symptoms using more diverse methods (e.g., interviews, implicit measures), as suggested by Campbell and Fiske (1959) in their treatise on convergent and discriminant validity.

Third, both studies used college-based samples. Although a modest proportion of participants in each sample endorsed moderate to severe depressive symptoms, and depression is a current public health concern for university students (Eisenberg, Gollust, Golberstein, & Hefner, 2007), future research would benefit from replicating the findings of this study in a clinical sample. The role of shame and self-criticism in depressive symptoms may be more pronounced in individuals with major depressive disorder or more severe manifestations of depressive symptoms.

Conclusion

In conclusion, study 1 found (a) convergent and discriminant validity between measures of shame-proneness and self-esteem, (b) that shame-proneness and self-esteem were highly correlated but distinct, and (c) that shame-proneness and self-esteem were significantly correlated with depressive symptoms, but only self-esteem remained significant when each was tested while controlling for the other. Study 2 found (a) convergent validity but not discriminant validity between measures of self-criticism and self-esteem, and (b) that neither self-criticism nor self-esteem was significantly associated with depressive symptoms, controlling for the effects of each other.

Supplementary Material

10862_2019_9734_MOESM1_ESM

Funding:

There is no funding to report for study 1. Funding for study 2 was provided by National Institute of Mental Health (NIMH) research service award F31[award number redacted for masked review]. NIMH had no role in the study design, collection, analysis or interpretation of the data, writing of the manuscript, or the decision to submit the paper for publication.

Footnotes

Ethical Approval: Ethical approval for both studies was granted by Vanderbilt University’s institutional review board. All procedures performed in study 1 and study 2 were in accordance with the ethical standards of Vanderbilt University’s institutional review board and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent: Informed consent was obtained from all individual participants included in the study prior to their participation.

Conflict of Interest: None of the authors have any conflicts of interest to report.

1

The rationale for the second step is that, if two constructs fail to demonstrate discriminant validity (i.e., are the same construct), their relations to other constructs should be identical. When two constructs overlap substantially, this analysis makes for a stricter test of discriminant validity.

2

Examples of items from other subscales of the TOSCA-3 are as follows:

Scenario: “You and a group of coworkers worked very hard on a project. Your boss singles you out for a bonus because the project was a success.”

Shame Response: “You would feel alone and apart from your colleagues.”

Guilt Response: “You would feel you should not accept it.”

Alpha Pride Response: “You would feel competent and proud of yourself.”

Beta Pride Response: “You would feel your hard work had paid off.”

Externalization Response: “You would feel the boss is rather short-sighted.”

3

The Internalized Shame Scale did not show discriminant validity with self-esteem.

4

We repeated these analyses, replacing the DEQ with the ISC. Results did not differ between the two models.

References

  1. Abela JRZ, Webb CA, Wagner C, Ho MR, & Adams P (2006). The role of self-criticism, dependency, and hassles in the course of depressive illness: A multiwave longitudinal study. Personality and Social Psychology Bulletin, 32(3), 328–338. doi: 10.1177/0146167205280911 [DOI] [PubMed] [Google Scholar]
  2. Ahmed SMS, Valliant PM, & Swindle D (1985). Psychometric Properties of Coopersmith Self-Esteem Inventory. Perceptual and Motor Skills, 61(3f), 1235–1241. doi: 10.2466/pms.1985.61.3f.1235 [DOI] [Google Scholar]
  3. Aldao A, Nolen-Hoeksema S, & Schweizer S (2010). Emotion-regulation strategies across psychopathology: A meta-analytic review. Clinical Psychology Review, 30(2), 217–237. doi: 10.1016/j.cpr.2009.11.004 [DOI] [PubMed] [Google Scholar]
  4. Anderson NL, & Crowther JH (2012). Using the experiential avoidance model of non-suicidal self-injury: Understanding who stops and who continues. Archives of Suicide Research, 16(2), 124–134. doi: 10.1080/13811118.2012.667329 [DOI] [PubMed] [Google Scholar]
  5. Andrews B, Qian M, & Valentine JD (2002). Predicting depressive symptoms with a new measure of shame: The Experience of Shame Scale. British Journal of Clinical Psychology, 41(1), 29–42. doi: 10.1348/014466502163778 [DOI] [PubMed] [Google Scholar]
  6. Auerbach RP, Ho MR, & Kim JC (2014). Identifying cognitive and interpersonal predictors of adolescent depression. Journal of Abnormal Child Psychology, 42(6), 913–924. doi: 10.1007/s10802-013-9845-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Beck AT (1963). Thinking and depression: Idiosyncratic content and cognitive distortions. Archives of General Psychiatry, 9, 324–333.doi: 10.1001/archpsyc.1963.01720160014002 [DOI] [PubMed] [Google Scholar]
  8. Beck AT, Steer RA, & Brown GK (1996). Beck Depression Inventory—Second Edition manual. San Antonio: The Psychological Corporation. [Google Scholar]
  9. Black RSA, Curran D, & Dyer KFW (2013). The impact of shame on the therapeutic alliance and intimate relationships. Journal of Clinical Psychology, 69(6), 646–654. doi: 10.1002/jclp.21959 [DOI] [PubMed] [Google Scholar]
  10. Blatt SJ (1995). The destructiveness of perfectionism: Implications for the treatment of depression. American Psychologist, 50(12), 1003–1020. doi: 10.1037/0003-066X.50.12.1003 [DOI] [PubMed] [Google Scholar]
  11. Blatt SJ (1974). Levels of object representation in anaclitic and introjective depression. Psychoanalytic Study of the Child, 29, 107–157 [PubMed] [Google Scholar]
  12. Blatt SJ, D’Afflitti JP, & Quinlan DM (1976). Experiences of depression in normal young adults. Journal of Abnormal Psychology, 85, 383–389. doi: 10.1037/0021843X.85.4.383 [DOI] [PubMed] [Google Scholar]
  13. Blatt SJ, Quinlan DM, Chevron ES, McDonald C, & Zuroff D (1982). Dependency and self-criticism: Psychological dimensions of depression. Journal of Consulting and Clinical Psychology, 50(1), 113–124. doi: 10.1037/0022-006X.50.1.113 [DOI] [PubMed] [Google Scholar]
  14. Campbell DT & Fiske DW (1959). Convergent and discriminant validation by the multitrait- multimethod matrix. Psychological Bulletin, 56(2), 81–105. [PubMed] [Google Scholar]
  15. Castilho P, Pinto-Gouveia J, & Duarte J (2017). Two forms of self-criticism mediate differently the shame–psychopathological symptoms link. Psychology and Psychotherapy: Theory, Research and Practice, 90(1), 44–54. doi: 10.1111/papt.12094 [DOI] [PubMed] [Google Scholar]
  16. Cohen TR, Wolf ST, Panter AT, & Insko CA (2011). Introducing the GASP scale: A new measure of guilt and shame proneness. Journal of Personality and Social Psychology, 100(5), 947–966. doi: 10.1037/a0022641 [DOI] [PubMed] [Google Scholar]
  17. Cole DA (1987). Utility of confirmatory factor analysis in test validation research. Journal of Consulting and Clinical Psychology, 55(4), 584–594. doi: 10.1037/0022-006X.55.4.584 [DOI] [PubMed] [Google Scholar]
  18. Cole DA, & Preacher KJ (2014). Manifest variable path analysis: Potentially serious and misleading consequences due to uncorrected measurement error. Psychological Methods, 19(2), 300–315. doi: 10.1037/a0033805 [DOI] [PubMed] [Google Scholar]
  19. Cook DR (1994, 2001). Internalized shame scale: technical manual. Multi-Health Systems, Inc., North Tonawanda, NY. [Google Scholar]
  20. Coopersmith S (1959). A method for determining types of self-esteem. The Journal of Abnormal and Social Psychology, 59(1), 87–94. doi: 10.1037/h0048001 [DOI] [PubMed] [Google Scholar]
  21. Covert MV, Tangney JP, Maddux JE, & Heleno NM (2003). Shame-proneness, guilt-proneness, and interpersonal problem solving: A social cognitive analysis. Journal of Social and Clinical Psychology, 22(1), 1–12. doi: 10.1521/jscp.22.1.1.22765 [DOI] [Google Scholar]
  22. Del Rosario PM, & White RM (2006). The Internalized Shame Scale: Temporal stability, internal consistency, and principal components analysis. Personality and Individual Differences, 41(1), 95–103. doi: 10.1016/j.paid.2005.10.026 [DOI] [Google Scholar]
  23. Dorahy MJ, & Hanna D (2012). Shame, intimacy and self-definition: An assessment of the emotional foundation and intimate relationship consequences of an introjective personality orientation. Journal of Nervous and Mental Disease, 200(8), 699–704. doi: 10.1097/NMD.0b013e318261427b [DOI] [PubMed] [Google Scholar]
  24. Dozois DJA, Dobson KS, & Ahnberg JL (1998). A psychometric evaluation of the Beck Depression Inventory–II. Psychological Assessment, 10(2), 83–89. doi: 10.1037/1040-3590.10.2.83 [DOI] [Google Scholar]
  25. Dugan W, McDonald MV, Passik SD, Rosenfeld BD, Theobald D, & Edgerton S (1998). Use of the Zung Self-Rating Depression Scale in cancer patients: Feasibility as a screening tool. Psycho-Oncology, 7, 483–493. doi: [DOI] [PubMed] [Google Scholar]
  26. Dunkley DM, Masheb RM, & Grilo CM (2010). Childhood maltreatment, depressive symptoms, and body dissatisfaction in patients with binge eating disorder: The mediating role of self-criticism. International Journal of Eating Disorders, 43(3), 274–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Dunkley DM, Sanislow CA, Grilo CM, & McGlashan TH (2009). Self-criticism versus neuroticism in predicting depression and psychosocial impairment for 4 years in a clinical sample. Comprehensive Psychiatry, 50(4), 335–346. doi: 10.1016/j.comppsych.2008.09.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ehret AM, Joormann J, & Berking M (2015). Examining risk and resilience factors for depression: The role of self-criticism and self-compassion. Cognition and Emotion, 29(8), 1496–1504. doi: 10.1080/02699931.2014.992394 [DOI] [PubMed] [Google Scholar]
  29. Eisenberg D, Gollust SE, Golberstein E, & Hefner JL (2007). Prevalence and correlates of depression, anxiety, and suicidality among university students. American Journal of Orthopsychiatry, 77(4), 534–542. doi: 10.1037/0002-9432.77.4.534 [DOI] [PubMed] [Google Scholar]
  30. Elison J, Lennon R, & Pulos S (2006). Investigating the compass of shame: The development of the compass of shame scale. Social Behavior and Personality, 34(3), 221–238. doi: 10.2224/sbp.2006.34.3.221 [DOI] [Google Scholar]
  31. Ferrari AJ, Somerville AJ, Baxter AJ, Norman R, Patten SB, Vos T, & Whiteford HA (2013). Global variation in the prevalence and incidence of major depressive disorder: A systematic review of the epidemiological literature. Psychological Medicine, 43(3), 471–481. doi: 10.1017/S0033291712001511 [DOI] [PubMed] [Google Scholar]
  32. Gabrys JB, & Peters K (1985). Reliability, discriminant and predictive validity of the Zung Self-Rating Depression Scale. Psychological Reports, 57(3), 1091–1096. doi: 10.2466/pr0.1985.57.3f.1091 [DOI] [PubMed] [Google Scholar]
  33. Glassman LH, Weierich MR, Hooley JM, Deliberto TL, & Nock MK (2007). Child maltreatment, non-suicidal self-injury, and the mediating role of self-criticism, Behaviour Research and Therapy, 45(10), 2483–2490. doi: 10.1016/j.brat.2007.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Grzegorek JL, Slaney RB, Franze S, & Rice KG (2004). Self-criticism, dependency, self-esteem, and grade point average satisfaction among clusters of perfectionists and nonperfectionists. Journal of Counseling Psychology, 51(2), 192–200. doi: 10.1037/0022-0167.51.2.192 [DOI] [Google Scholar]
  35. Hooley JM, Ho DT, Slater JA, & Lockshin A (2002). Pain insensitivity and self-harming behavior. Paper presented at the annual meeting of the Society for Research in Psychopathology. San Francisco, CA. [Google Scholar]
  36. Iancu I, Bodner E, & Ben-Zion I (2015). Self esteem, dependency, self-efficacy and self-criticism in social anxiety disorder. Comprehensive Psychiatry, 58, 165–171. doi: 10.1016/j.comppsych.2014.11.018 [DOI] [PubMed] [Google Scholar]
  37. Johnson M (2010). Depressive styles, self-esteem structure, and health: A dynamic approach to differential vulnerability in self-criticism and dependency. Individual Differences Research, 8(1), 45–66. [Google Scholar]
  38. Joiner TE (1995). The price of soliciting and receiving negative feedback: Self-verification theory as a vulnerability to depression theory. Journal of Abnormal Psychology, 104(2), 364–372. doi: 10.1037/0021-843X.104.2.364 [DOI] [PubMed] [Google Scholar]
  39. Joiner TE., Katz J, & Lew A (1999). Harbingers of depressotypic reassurance seeking: Negative life events, increased anxiety, and decreased self-esteem. Personality and Social Psychology Bulletin, 25(5), 632–639. doi: 10.1177/0146167299025005008 [DOI] [Google Scholar]
  40. Kelly AC, & Carter JC (2013). Why self-critical patients present with more severe eating disorder pathology: The mediating role of shame. British Journal of Clinical Psychology, 52(2), 148–161. doi: 10.1111/bjc.12006 [DOI] [PubMed] [Google Scholar]
  41. Kim S, Thibodeau R, & Jorgensen RS (2011). Shame, guilt, and depressive symptoms: A meta-analytic review. Psychological Bulletin, 137(1), 68–96. doi: 10.1037/a0021466. [DOI] [PubMed] [Google Scholar]
  42. Kuster F, Orth U, & Meier LL (2012). Rumination mediates the prospective effect of low self-esteem on depression: A five-wave longitudinal study. Personality and Social Psychology Bulletin, 38(6), 747–759. doi: 10.1177/0146167212437250 [DOI] [PubMed] [Google Scholar]
  43. Leary MR, & Baumeister RF (2000). The nature and function of self-esteem: Sociometer theory. Advances in Experimental Social Psychology, 32, 1–62. doi: 10.1016/S0065-2601(00)80003-9 [DOI] [Google Scholar]
  44. Lewinsohn PM, Seeley JR, Roberts RE, & Allen NB (1997). Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults. Psychology and Aging, 12(2), 277–287. doi: 10.1037/0882-7974.12.2.277 [DOI] [PubMed] [Google Scholar]
  45. Lewinsohn PM, Solomon A, Seeley JR, & Zeiss A (2000). Clinical implications of “subthreshold” depressive symptoms. Journal of Abnormal Psychology, 109(2), 345–351. doi: 10.1037/0021-843X.109.2.345 [DOI] [PubMed] [Google Scholar]
  46. Masselink M, Roekel E, & Oldehinkel AJ (2017). Self-esteem in early adolescence as predictor of depressive symptoms in late adolescence and early adulthood: The mediating role of motivational and social factors. Journal of Youth and Adolescence. doi: 10.1007/s10964-017-0727-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Mathers CD, & Loncar D (2006). Projections of global mortality and burden of disease from 2002 to 2030. PLoS Medicine, 3(11), 2011–2030. doi: 10.1371/journal.pmed.0030442 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Mills SD, Fox RS, Merz EL, Clements PJ, Kafaja S, Malcarne VL, Furst DE, & Khanna D (2015). Evaluation of the Satisfaction with Appearance Scale and its short form in systemic sclerosis: Analysis from the UCLA Scleroderma Quality of Life Study. Journal of Rheumatology, 42(9), 1624–1630. doi: 10.3899/jrheum.141482 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Moroz M, & Dunkley D (2015). Self-critical perfectionism and depressive symptoms: Low self-esteem and experiential avoidance as mediators. Personality and Individual Differences, 87, 174–179. doi: 10.1016/j.paid.2015.07.044 [DOI] [Google Scholar]
  50. Muthén LK, & Muthén BO (1998–2017). Mplus User’s Guide. Eighth Edition. Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  51. Orth U, & Robins RW (2013). Understanding the link between low self-esteem and depression. Current Directions in Psychological Science, 22(6), 455–460. doi: 10.1177/0963721413492763 [DOI] [Google Scholar]
  52. Orth U, Berking M, & Burkhardt S (2006). Self-conscious emotions and depression: Rumination explains why shame but not guilt is maladaptive. Personality and Social Psychology Bulletin, 32(12), 1608–1619. [DOI] [PubMed] [Google Scholar]
  53. Orth U, Robins RW, Meier LL, & Conger RD (2016). Refining the vulnerability model of low self-esteem and depression: Disentangling the effects of genuine self-esteem and narcissism. Journal of Personality and Social Psychology, 110(1), 133–149. doi: 10.1037/pspp0000038 [DOI] [PubMed] [Google Scholar]
  54. Porter A, Zelkowitz R, & Cole D (2018). The unique associations of self-criticism and shame-proneness to symptoms of disordered eating and depression. Eating Behaviors, 29, 64–67. doi: 10.1016/j.eatbeh.2018.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Pyszczynski T, Greenberg J, Solomon S, Arndt J, & Schimel J (2004). Why do people need self-esteem? A theoretical and empirical review. Psychological Bulletin, 130(3), 435–468. [DOI] [PubMed] [Google Scholar]
  56. Radloff LS (1997). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. [Google Scholar]
  57. Rice S, Fallon B, & Bambling M (2011). Men and depression: The impact of masculine role norms throughout the lifespan. Australian Educational and Developmental Psychologist, 28(02), 133–144. doi: 10.1375/aedp.28.2.133 [DOI] [Google Scholar]
  58. Rosenberg M (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. [Google Scholar]
  59. Rüsch N, Corrigan PW, Bohus M, Jacob GA, Brueck R, & Lieb K (2007). Measuring shame and guilt by self-report questionnaires: A validation study. Psychiatry Research, 150(3), 313–325. doi: 10.1016/j.psychres.2006.04.018 [DOI] [PubMed] [Google Scholar]
  60. Santini ZI, Koyanagi A, Tyrovolas S, Mason C, & Haro JM (2015). The association between social relationships and depression: A systematic review. Journal of Affective Disorders, 175, 53–65. doi: 10.1016/j.jad.2014.12.049 [DOI] [PubMed] [Google Scholar]
  61. Sowislo JF, & Orth U (2013). Does low self-esteem predict depression and anxiety? A meta-analysis of longitudinal studies. Psychological Bulletin, 139, 213–240. doi: 10.1037/a0028931. [DOI] [PubMed] [Google Scholar]
  62. Sowislo JF, Orth U, & Meier LL (2014). What constitutes vulnerable self-esteem? Comparing the prospective effects of low, unstable, and contingent self-esteem on depressive symptoms. Journal of Abnormal Psychology, 123(4), 737–753. doi: 10.1037/a0037770 [DOI] [PubMed] [Google Scholar]
  63. Starr L, & Davila J (2008). Excessive reassurance seeking, depression, and interpersonal rejection: A meta-analytic review. Journal of Abnormal Psychology, 117(4), 762–775. doi: 10.1037/a0013866 [DOI] [PubMed] [Google Scholar]
  64. Stolow D, Zuroff DC, Young JF, Karlin RA, & Abela JRZ (2016). A prospective examination of self-compassion as a predictor of depressive symptoms in children and adolescents. Journal of Social and Clinical Psychology, 35(1), 1–20. doi: 10.1521/jscp.2016.35.1.1 [DOI] [Google Scholar]
  65. Supple AJ, Su J, Plunkett SW, Peterson GW, & Bush KR (2013). Factor structure of theRosenberg Self-Esteem Scale. Journal of Cross-Cultural Psychology, 44(5), 748–764. doi: 10.1177/0022022112468942 [DOI] [Google Scholar]
  66. Swann WB Jr. (2012). Self-verification theory In Van Lange PAM, Kruglanski AW, & Higgins ET (Eds.), Handbook of theories of social psychology (pp. 23–42). Thousand Oaks, CA: Sage Publications Ltd. [Google Scholar]
  67. Swann WB Jr., Wenzlaff RM, & Tafarodi RW (1992). Depression and the search for negative evaluations: More evidence of the role of self-verification strivings. Journal of Abnormal Psychology, 101(2), 314–317. [DOI] [PubMed] [Google Scholar]
  68. Tafarodi RW, & Swann WB Jr. (2001). Two-dimensional self-esteem: Theory and measurement. Personality and Individual Differences, 31, 653–673. doi: 10.1016/S0191-8869(00)00169-0 [DOI] [Google Scholar]
  69. Tangney JP, & Dearing RL (2002). Shame and Guilt. New York, NY: Guilford Press. [Google Scholar]
  70. Tangney JP, Dearing RL, Wagner PE, & Gramzow R (2000). The Test of Self-Conscious Affect–3 (TOSCA-3). Fairfax, VA: George Mason University. [Google Scholar]
  71. Thompson R, & Zuroff DC (2004). The Levels of Self-Criticism Scale: Comparative self-criticism and internalized self-criticism. Personality and Individual Differences, 36(2), 419–430. doi: 10.1016/S0191-8869(03)00106-5 [DOI] [Google Scholar]
  72. Tracy JL (2012). Self-Conscious Emotions In Leary MR & Tangney JP (Eds.), Handbook of Self and Identity (2nd ed.) (446–478). New York, NY: Guilford Press. [Google Scholar]
  73. Vettese LC, & Mongrain M (2000). Communication about the self and partner in the relationships of dependents and self-critics. Cognitive Therapy and Research, 24(6), 609–626. doi: 10.1023/A:1005587009681 [DOI] [Google Scholar]
  74. Weinstock LM, & Whisman MA (2004). The self-verification model of depression and interpersonal rejection in heterosexual dating relationships. Journal of Social and Clinical Psychology, 23(2), 240–259. doi: 10.1521/jscp.23.2.240.31020 [DOI] [Google Scholar]
  75. Zung WK (1965). A Self-Rating Depression Scale, Archives of General Psychiatry, 12(1), 63–70. doi: 10.1001/archpsyc.1965.01720310065008. [DOI] [PubMed] [Google Scholar]
  76. Zuroff DC, & Duncan N (1999). Self-criticism and conflict resolution in romantic couples. Canadian Journal of Behavioural Science/Revue Canadienne Des Sciences Du Comportement, 31(3), 137–149. doi: 10.1037/h0087082 [DOI] [Google Scholar]
  77. Zuroff DC, & Mongrain M (1987). Dependency and self-criticism: Vulnerability factors for depressive affective states. Journal of Abnormal Psychology, 96(1), 14–22. doi: 10.1037/0021-843X.96.1.14 [DOI] [PubMed] [Google Scholar]
  78. Zuroff DC, Quinlan DM, & Blatt SJ (1990). Psychometric properties of the Depressive Experiences Questionnaire in a college population. Journal of Personality Assessment, 55(1–2), 65–72. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

10862_2019_9734_MOESM1_ESM

RESOURCES