Development and Validation of a Measure of Self ...

573300

research-article2015

ASMXXX10.1177/1073191115573300AssessmentSmart et al.

Review

Development and Validation of a Measure of Self-Critical Rumination

Assessment 1­–12 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1073191115573300 asm.sagepub.com

Laura M. Smart1, Jessica R. Peters1, and Ruth A. Baer1

Abstract Self-criticism is a form of negative self-evaluation that has strong associations with many forms of psychopathology. Rumination is a maladaptive form of repetitive thinking that is associated with many psychological disorders. Although measures of several different types of rumination (e.g., general rumination, depressive rumination, anger rumination) have been developed, none focuses specifically on self-critical rumination. An initial pool of items addressing self-critical rumination was developed by adapting items from existing rumination measures and through a writing task administered to both student and clinical samples. Following an evaluation of content validity, 24 items were administered to a large sample of undergraduates along with measures of related constructs. The final 10-item version of the Self-Critical Rumination Scale showed excellent internal consistency, a clear single-factor structure, convergent relationships with related constructs, and incremental validity over other measures of self-criticism and rumination in predicting both general distress and features of borderline personality disorder. Keywords self-criticism, rumination, shame, assessment Self-criticism is a form of negative self-evaluation in which judgmental, condemning, and attacking thoughts are directed to the self, especially in the context of perceived mistakes, failures, and inability to live up to one’s own or others’ standards (Gilbert, Clarke, Hempel, Miles, & Irons, 2004). Self-criticism has strong associations with many forms of psychopathology, including depression (Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982), social anxiety (Cox, Fleet, & Stein, 2004), posttraumatic stress disorder (Harman & Lee, 2009), and binge eating (Dunkley, Masheb, & Grilo, 2010). It has been shown to mediate relationships between childhood maltreatment and many problems in adulthood, including body dissatisfaction (Dunkley et al., 2010), nonsuicidal self-injury (Glassman, Weierich, Hooley, Deliberto, & Nock, 2007), impaired romantic relationships (Lassri & Shahar, 2012), and depression (Campos, Besser, & Blatt, 2010). These findings suggest that severe criticism in childhood may lead to internalized general tendencies to criticize oneself and to believe and submit to one’s own self-critical attacks (Gilbert et al., 2004). Selfcriticism is an element of shame, a painful negative emotion that includes feelings of inferiority, self-consciousness, and shrinking; a desire to hide or withdraw; and intensely selfcritical and self-devaluing thoughts (Tangney, Stuewig, & Mashek, 2007). Shame is also associated with a variety of psychological problems, including depression, anxiety, posttraumatic stress, suicidal behavior, and borderline personality disorder (Brown, Linehan, Comtois, Murray, &

Chapman, 2009; Fergus, Valentiner, McGrath, & Jencius, 2010; Hooley, Ho, Slater, & Lockshin, 2010; Kim, Thibodeau, & Jorgensen, 2011; Peters, Geiger, Smart, & Baer, 2013). Rumination is a maladaptive form of repetitive thinking that is common in many psychological disorders. Trapnell and Campbell (1999) described rumination as a neurotic form of self-focused attention characterized by recurrent thinking about perceived threats, losses, and injustices to the self and associated feelings of anxiety, depression, and anger. Response styles theory (Nolen-Hoeksema, 1991) focuses on depressive rumination and posits that people who dwell on their depressive symptoms by thinking passively and repetitively about their causes, consequences, and implications are subject to more severe and prolonged negative moods than those who do not ruminate. Depressive rumination has been shown to interfere with constructive thinking, problem solving, instrumental behavior, and interpersonal relationships. It is also associated with forms of psychopathology other than depression, including anxiety, binge eating, alcohol misuse, and self-harm (NolenHoeksema, Wisco, & Lyubomirsky, 2008). 1

University of Kentucky, Lexington, KY, USA

Corresponding Author: Laura M. Smart, Department of Psychology, University of Kentucky, 111J Kastle Hall, Lexington, KY 40506, USA. Email: [email protected]

2 While much of the literature has focused on depressive rumination, other types of rumination have shown significant relationships with specific forms of psychopathology. For example, anger rumination, or the tendency to dwell on experiences of anger, intensifies angry feelings (Rusting & Nolen-Hoeksema, 1998), increases aggressive behavior (Bushman, Bonacci, Pederson, Vasquesz, & Miller, 2005), and is uniquely related to symptoms of borderline personality disorder (BPD) after controlling for depressive rumination and anger (Baer & Sauer, 2011). Anxious rumination, or repetitive thinking about anxious states, predicts severity of current anxiety even after accounting for worry and depressive rumination (Rector, Antony, Laposa, Kocovski, & Swinton, 2008). Postevent processing, or rumination about social interactions, predicts anxiety about and avoidance of future performance and social situations (Brozovich & Heimberg, 2013). Overall, these findings suggest that both the prolonged, repetitive, unconstructive style of rumination and the content of the ruminative thoughts are important in accounting for the specific symptoms that characterize a wide range of problems and disorders.

Need for a Measure of Self-Critical Rumination Several factors suggest that a scale measuring self-critical rumination would be useful. First, existing measures of self-criticism, such as the Self-Criticism subscale of Depressive Experiences Questionnaire (DEQ; Blatt, D’Afflitti, & Quinlan, 1976) and the Levels of SelfCriticism Scale (LOSC; Thompson & Zuroff, 2004), are based on a developmental theory of depression that conceptualizes self-criticism as a broad personality construct encompassing perceptions of hostility and criticism from others, inability to cope with life, an avoidant way of dealing with problems, and repeated attempts to meet impossibly high goals. Thus, items on these scales go far beyond self-criticism itself (e.g., putting oneself down, calling oneself inadequate or a failure). Second, neither the DEQ nor the LOSC captures the prolonged and repetitive qualities that characterize ruminative thought. This is important because specific forms of rumination have been shown to predict variance in psychological symptoms after accounting for the presence of self-critical thoughts. For example, in a community sample of adults, O’Connor and Noyce (2008) showed that depressive rumination fully mediated the relationship between self-criticism at baseline and suicidal ideation 3 months later. In a long-term study of college students, Spasojevic and Alloy (2001) found that depressive rumination mediated the relationship between self-criticism at baseline and depression 2.5 years later. Powers, Koestner, and Zuroff (2007) found that a combination of rumination and procrastination mediated the relationship between self-criticism at baseline and goal progress over the course of a week.

Assessment  Geiger, Peters, Sauer-Zavala, and Baer (2013) reported that anger rumination was a more consistent and robust predictor of BPD features than were the automatic thoughts and dysfunctional attitudes (many of which are self-critical) typical of BPD. These findings consistently suggest that self-criticism is distinct from the tendency to ruminate, and that rumination mediates the effects of self-criticism on psychological symptoms. Finally, most of the research on rumination as it relates to self-criticism has relied on measures of depressive or anger rumination. The former focus heavily on symptoms of depression and sadness, whereas the latter emphasize unfair situations and objectionable behavior by others; neither may fully capture the nature of repetitive thinking in people who are highly self-critical, some of whom may not have high levels of depression or anger. Yet self-critical rumination may be important in understanding a variety of psychopathologies (Gilbert & Procter, 2006). Because none of the existing rumination scales directly assess self-critical rumination, the purpose of the present study was to develop a self-report measure of this potentially important form of maladaptive repetitive thought. In developing this measure, we conceptualized self-criticism as a form of negative thinking that devalues the self, rather than the more global personality style described earlier and commonly measured by the DEQ and the LOSC. That is, we focused more narrowly on thoughts that criticize the self for perceived mistakes, failures, weaknesses, shortcomings, bad habits, or general inadequacy. We constructed the items to reflect ruminative qualities of thinking: frequent, prolonged, repetitive, and difficult to control. We expected our new measure to be strongly correlated with existing rumination scales for several reasons. First, the available scales also assess general qualities of ruminative thought (negative, frequent, repetitive, etc.). Second, the negative thoughts and feelings about which people might ruminate (e.g., sadness, anxiety, anger, social awkwardness) are all related to neuroticism and therefore highly correlated with each other and to self-critical tendencies (Costa & McCrae, 1992). Finally, as explained in more detail later, we adapted items from existing rumination measures to create items for our new scale. However, given previous findings showing that the specific content or focus of ruminative thought (sadness, anger, etc.) appears important in understanding a variety of psychological problems, we also expected that our new measure would show incremental validity over existing rumination scales in two ways. First, because self-criticism appears to be common to many disorders but is not well represented in the existing rumination questionnaires, we expected that our new measure would show incremental validity over the available rumination scales in predicting general distress. Second, because self-criticism is an element of shame, which is prevalent in BPD (Brown et al.,

3

Smart et al. 2009; Jacob et al., 2009), we predicted that our new measure would also show incremental validity over other rumination scales in predicting BPD features. We also tested several other hypotheses. We expected our new scale to be positively correlated with measures of shame. Although distinctions are often made between shame and guilt, with shame more focused on inadequacy of the self and guilt more focused on making amends for others’ distress, they tend to be significantly correlated. We therefore expected our new scale to show positive correlations with guilt, but smaller in magnitude than the correlations with shame. We also predicted a negative relationship between the new scale and self-compassion and a positive relationship with neuroticism. Other domains from the fivefactor model of personality (Costa & McCrae, 1992) were included to provide preliminary tests of discriminant validity. For extraversion, agreeableness, and conscientiousness (which tend to have small relationships with negative affect), we predicted negative correlations with our new scale but significantly smaller effects than for the relationship with neuroticism. For the openness domain, we predicted a nonsignificant correlation. Finally, because existing rumination scales have varying numbers of factors, we had no a priori expectations about the ultimate factor structure of our new scale.

Study 1: Item Development and Content Validity Item Development Three methods were used to develop an initial pool of 39 items for the Self-Critical Rumination Scale (SCRS). First, existing measures of other types of rumination were examined to identify items that could be adapted for use in the SCRS. For example, the Rumination on Interpersonal Offenses Scale (described in more detail below) includes items such as, “When I have been hurt or offended, I can’t stop thinking about how I was wronged by this person.” An item for the SCRS adapted from this scale is, “When I have made a mistake, I can’t stop thinking about how stupid I was.” Second, a sample of undergraduate students (n = 25) were asked to complete an open-ended writing task with the following instructions: We are interested in the types of thoughts that go through people’s minds when they are criticizing themselves. Please think of a time when you were feeling critical of yourself. You might have made a mistake or done something you thought was embarrassing, stupid, wrong, or foolish. Or you might have been feeling critical of an aspect of yourself, such as your appearance, intelligence, or likeability. Please write down the self-critical thoughts that went through your mind when this happened.

An example of an item developed using this method is, “I spend a lot of time wishing I were different.” Third, a sample of adults receiving outpatient treatment at a mental health clinic (n = 13) were provided the same open-ended writing task. Many clients at this treatment center have high levels of anxiety and depression, which are associated with self-criticism. Inclusion of this sample allowed for better representation of the high end of the distribution of self-criticism. An example of an item developed using this method is, “I often think about what a failure I am.” A total of 39 items were written using these methods.

Evaluation of Content Validity Content validity of the initial set of items was evaluated with a rating task. Raters were three advanced doctoral students in clinical psychology. They were given a written explanation of the nature of rumination in general and selfcritical rumination in particular. They were then given a list of items in random order that included the 39 new items, 10 items measuring general rumination, 15 items measuring specifically either depressive or anger rumination, and 15 items measuring shame and guilt but not rumination. They were asked to sort these into three categories: items assessing self-critical rumination, items assessing other forms of rumination, and items assessing shame or guilt but not rumination. Next, they were asked to rate each of the proposed new SCRS items for how well it reflects the construct of self-critical rumination, using a 4-point scale (1 = not at all, 4 = very well). Items that were misclassified by two or more raters were eliminated (n = 13), as well as items that received mean ratings less than 3.0 on the 4-point rating (n = 2) leaving 24 items in the item pool. The 24 remaining items averaged 3.77 on this scale indicating that on average raters believed that the items represented self-critical rumination clearly. Free-marginal kappa was calculated to assess interrater reliability on these remaining items and was .83, which indicates adequate reliability (Randolph, 2005).

Study 2: Factor Structure, Internal Consistency, and Relationships With Related Constructs Method Participants and Procedure. Participants were 510 undergraduate students enrolled in introductory psychology courses at a large public university. They were recruited in two ways. First, the study was posted on a departmental website for research participation where it was open to the entire subject pool; most participants were recruited through this website without any prescreening. Second, we included the LOSC (Thompson & Zuroff, 2004) in a screening packet administered to the entire subject pool early in the semester;

4

Assessment 

this procedure allows students meeting specific criteria to be invited to participate in particular studies. We sent email messages to students scoring in the top 25% on the LOSC inviting them to sign up for this study, if they had not already done so, with the aim of increasing the likelihood that students in this part of the range would sign up for the study. This oversampling procedure increased the number of participants with above-average self-criticism scores (insuring that high levels would be reasonably well represented) but did not create an abnormal distribution on this variable. In the final sample, scores on the LOSC ranged from 1.14 to 6.5 (possible range: 1-7). All participants completed an online survey for course credit including demographic questions and all of the measures listed below (see Measures section). Following data screening procedures (detailed in the Results section), a final sample of 420 was used for all analyses. Participants ranged in age from 18 to 30 years (M = 18.99, SD = 1.44). The sample was predominately female (51.9%) and Caucasian (71.9%).

item when they are feeling anxious on a 4-point Likert-type scale from “never” to “always.” The Anxious Rumination Questionnaire demonstrated good internal consistency (α = .93) in the present sample. The Rumination on Interpersonal Offenses Scale (RIO; Wade, Vogel, Liao, & Goldman, 2008) is a 6-item measure of rumination about an interpersonal transgression (e.g., “I can’t stop thinking about how I was wronged by this person”). Participants respond using a 5-point Likert-type scale ranging from “strongly disagree” to “strongly agree.” The RIO demonstrated excellent internal consistency (α = .92) in the present sample. The Post-Event Processing Questionnaire–Revised (McEvoy & Kingsep, 2006) is an 8-item measure of the tendency to ruminate about past social situations (e.g., “After the event was over, did you find yourself thinking about it a lot?”). Participants rate themselves on a 100-point visual analogue scale with not at all (0) and totally agree (100) as the anchors. It demonstrated excellent internal consistency (α = .91) in the present sample.

Measures of Rumination. The rumination subscale of the Rumination and Reflection Questionnaire (RRQ; Trapnell & Campbell, 1999) has 12 items assessing a maladaptive form of repetitive thinking about the self (e.g., “Sometimes it is hard for me to shut off thoughts about myself”). Participants rate themselves on a 5-point Likert-type scale (1 = strongly disagree, 5 = strongly agree). The RRQ rumination subscale demonstrated excellent internal consistency (α = .93) in the present sample. The brooding subscale of the Ruminative Responses Scale (RRS; Nolen-Hoeksema & Morrow, 1991; Treynor, Gonzalez, & Nolen-Hoeksema, 2003) has 5 items measuring a maladaptive form of repetitive thought this is strongly associated with depression (e.g., “Think, what am I doing to deserve this?”). Participants rate themselves on a 4-point Likert-type scale ranging from “almost never” to “almost always.” The brooding subscale demonstrated good internal consistency (α = .87) in the present sample. The Anger Rumination Scale (Sukhodolsky, Golub, & Cromwell, 2001) is a 19-item measure of the tendency to ruminate on angry moods, past anger episodes, and the causes and consequences of anger episodes (e.g., “I keep thinking about events that angered me for a long time”). Participants respond on a 4-point Likert-type scale from “almost never” to “almost always.” The Anger Rumination Scale demonstrated excellent internal consistency (α = .95) in the present sample. The Anxious Rumination Questionnaire (Rector et al., 2008) is a 22-item measure of the tendency to ruminate on anxious moods, the perceived inability to cope with those moods, and how likely those moods are to interfere with activities (e.g., “These feelings will interfere with upcoming plans”). Participants indicate how often they think each

Measures of Self-Criticism and Shame.  The LOSC (Thompson & Zuroff, 2004) measures both comparative selfcriticism (e.g., “I fear that if people get to know me too well, they will not respect me”) and internalized self-criticism (e.g., “Failure is a very painful experience for me”). Participants rate each item on a 7-point Likert-type scale from “not at all” to “very well.” Each subscale of the LOSC demonstrated adequate to excellent internal consistency in the present sample (comparative self-criticism α = .80, internalized self-criticism α = .92). The Experiences of Shame Scale (Andrews, Qian, & Valentine, 2002) has 25 items measuring feelings of shame about a variety of areas including one’s body, personal habits, manner with others, personal ability, and whether one has said or done something wrong (e.g., “Have you felt ashamed of any of your personal habits?”). Participants respond using a 4-point Likert-type scale from “not at all” to “very much.” The Experiences of Shame Scale demonstrated excellent internal consistency (α = .96) in the present sample. The Test of Self-Conscious Affect (Tangney & Dearing, 2002) asks respondents to read 16 scenarios (11 negative, 5 positive). Using a 5-point Likert-type scale (1 = not very likely, 5 = very likely), they respond to four possible reactions to each scenario assessing shame-proneness, guiltproneness, detachment, and externalization. An example item is, “You break something at work and then hide it.” For this scenario, the possible reactions are “You would think: This is making me anxious. I need to either fix it or get someone else to” (indicating guilt-proneness); “You would think about quitting” (shame-proneness); “You would think: A lot of things aren’t made very well these days” (externalization); “You would think: It was only an accident”

5

Smart et al. (detachment). Only the shame and guilt reactions to the negative scenarios were used in the present study. They demonstrated adequate to good internal consistencies in the present sample (shame α = .79; guilt α = .83). The Personal Feelings Questionnaire–2 (Harder & Zalma, 1990) has two subscales: shame-proneness (e.g., “feelings of blushing”) and guilt-proneness (e.g., “worry about hurting or injuring someone”). Participants rate statements on a 4-point Likert-type scale from “never experience this feeling” to “experience the feeling continuously or almost continuously.” The two subscales demonstrated good internal consistency in the present sample (αs = .84 and .80, respectively). Measures of Other Constructs.  The Self-Compassion Scale (Neff, 2003) has 26 items measuring compassion toward oneself (e.g., “I’m tolerant of my own flaws and inadequacies”). Participants rate statements on a 5-point Likert-type scale from “almost never” to “almost always.” Internal consistency of the Self-Compassion Scale was good (α = .87) in the present sample. The Personality Assessment Inventory–Borderline Personality Disorder subscale (PAI-BOR; Morey, 1991) has 24 items rated on a 4-point Likert-type scale from “false, not at all true” to “very true.” It has four subscales tapping aspects of BPD pathology including self-harm (“When I’m upset, I typically do something to hurt myself”), relationship difficulties (“My relationships have been stormy”), identity problems (“My attitude about myself changes a lot”), and affective instability (“My mood can shift quite suddenly”). It demonstrated good internal consistency (α = .89) in the present sample. The NEO Five Factor Inventory (Costa & McCrae, 1992) is a 60-item, well-validated measure of the domains of personality according to the Five Factor Model. Items are scored on a 5-point Likert-type scale from “strongly disagree” to “strongly agree.” It provides domain scores for each of the five factors including neuroticism (“I often feel tense and jittery”), extraversion (“I enjoy parties with lots of people”), openness (“I often try new and foreign foods”), agreeableness (“I generally try to be thoughtful and considerate”), and conscientiousness (“I have a lot of self-discipline”). The five domain scores showed adequate to good internal consistency in the present sample (αs = .85, .79, .66, .70, and .82, respectively). The Depression Anxiety Stress Scales (DASS; Lovibond & Lovibond, 1993) is a 42-item measure of stress (“I found it difficult to relax”), anxiety (“I felt I was close to panic”), and depression (“I felt downhearted and blue”). Participants respond using a 4-point Likert-type scale from “did not apply to me at all” to “applied to me very much, or most of the time.” In the present sample, each subscale demonstrated excellent internal consistency (αs = .94, .92, and .94 for depression, anxiety, and stress, respectively).

Results and Discussion Data Screening.  The data were analyzed using SPSS 19.0. Each measure included an added validity item such as “Please choose” ‘strongly disagree’ for this question” to ensure that participants were attending to both item content and the response scale. A final validity question was included at the end of the survey which asked participants if they answered the entire survey honestly (“I have tried to answer all of these questions honestly and accurately”). Participants who responded to the final validity question with “somewhat disagree” or “strongly disagree” (n = 73) or who responded incorrectly to more than 10 of the 19 validity questions embedded in the survey (n = 17) were excluded from analyses. These procedures resulted in the exclusion of 90 participants; therefore, data from 420 participants were used for analyses. Item-Level Analyses and Internal Consistency. Prior to conducting factor analysis, item-total and interitem correlations were examined. No items had corrected item-total correlations below .50 or interitem correlations less than .40. Several items (n = 14) had corrected item-total correlations greater than .80. These appeared to include redundant content and were eliminated. The coefficient alpha for the remaining 10 items was .92. Of these 10, 6 were skewed (i.e., had a skew value greater than three times its standard error) but were retained because they were necessary for content coverage; these items are likely to be less skewed in a clinical sample. Three of the 10 items were based on the writing task responses from the clinical sample, 1 on the writing task responses from the student sample, and 6 were based on adaptations of items from existing measures of rumination. Instructions for the scale were: “Please rate how well each item describes you.” The Likert-type scale ranged from 1 to 4 (1 = not at all, 2 = a little, 3 = moderately, 4 = very much). Exploratory Factor Analysis.  Responses to the 10 remaining items were subjected to an exploratory factor analysis using principal axis factoring with an oblique rotation to allow any possible factors to correlate. A single factor emerged with an eigenvalue greater than 1 and examination of the scree plot revealed a clear single-factor solution to the data. This factor accounted for 58.41% of the variance. All factor loadings were well above .40. Factor loadings are presented in Table 1. Correlations With Other Measures.  Correlations between all study variables are shown in Table 2. Rumination. As predicted, the SCRS was significantly correlated with each of the other measures of rumination (r =.53-.81), indicating that individuals who are likely to

6

Assessment 

Table 1.  Item Content and Factor Structure of the Self-Critical Rumination Scale in an Undergraduate Sample (n = 420). Item 1. M  y attention is often focused on aspects of myself that I’m ashamed of. 2. I always seem to be rehashing in my mind stupid things that I’ve said or done. 3. S ometimes it is hard for me to shut off critical thoughts about myself. 4. I can’t stop thinking about how I should have acted differently in certain situations. 5. I spend a lot of time thinking about how ashamed I am of some of my personal habits. 6. I criticize myself a lot for how I act around other people. 7. I wish I spent less time criticizing myself. 8. I often worry about all of the mistakes I have made. 9. I spend a lot of time wishing I were different. 10. I often berate myself for not being as productive as I should be. Percentage of variance accounted for

ruminate self-critically are also likely to ruminate about other topics, including depression, anger, anxiety, interpersonal offenses, and uncomfortable social interactions. Self-criticism, shame, and guilt.  Scores on the SCRS were significantly and positively associated with all measures of self-criticism and shame as predicted. Contrary to prediction, the SCRS was also about equally strongly correlated with the shame and guilt subscales of the PFQ (t = 0.83, ns). However, the expected difference was found for the Test of Self-Conscious Affect, with the SCRS demonstrating a much smaller correlation with the guilt subscale than the shame subscale (t = 10.69, p < .001). These dissimilar relationships between self-critical rumination and guilt may be because of differences in how guilt is conceptualized in the PFQ versus the TOSCA. The TOSCA provides examples of responses believed to represent guilt but not shame (e.g., trying to make amends), whereas the PFQ focuses directly on feelings of guilt, regret, and remorse. The distinction between shame and guilt may not be widely understood among undergraduates; therefore, it is possible that our participants were conflating shame and guilt, which drove up the correlation between the SCRS and the guilt subscale of the PFQ. Other related constructs. As expected, scores on the SCRS were negatively related to scores on the SCS indicating that individuals who engage in self-critical rumination tend to be less compassionate toward themselves. Scores on the SCRS were also positively associated with scores on the PAI-BOR indicating that individuals who ruminate self-critically tend to endorse more borderline personality features. The SCRS was also positively associated with the DASS total score as expected, indicating that individuals who engage in more self-critical thinking tend to experience more general distress. Correlations with the DASS

Exploratory factor loadings

Confirmatory factor loadings

.73 .74 .77 .77

.80 .81 .75 .42

.76

.75

.65 .72 .77 .74 .68 58.41

.73 .75 .83 .70 .68  

subscale scores (not shown in the table to conserve space) were similar (depression: r = .54; stress: r = .57; anxiety: r = .46; all ps < .001). Relationships between the SCRS and each of the five domains of personality were also analyzed. As expected, higher scores on the SCRS were associated with higher levels of neuroticism (r = .66, p < .001). As hypothesized, the SCRS was negatively associated with extraversion, agreeableness, and conscientiousness (r = −.33, −.24, and −.27, respectively; all p < .01). Each of these three correlations was significantly smaller than the correlation with neuroticism (t = 7.69, 8.44, and 8.45, respectively, all p < .001), providing evidence of discriminant validity. In addition, the correlation between the SCRS and the Openness domain of personality, though statistically significant (r = .11, p < .05) was much smaller than the relationship with neuroticism, providing additional evidence of discriminant validity. Incremental Validity of the SCRS Over Other Measures. Incremental validity was examined with a series of hierarchical regression analyses. In each case, variance inflation factors were well below the suggested maximum value of 10 (Myers, 1990), suggesting that multicollinearity was not problematic. Preliminary analyses showed that gender was not a significant predictor of any of the dependent variables; therefore, gender was not included in the analyses described here. As noted in Table 2, the SCRS was strongly correlated with the RRQ-rumination scale (r = .81, p < .001). The SCRS was designed to address self-critical rumination in particular, whereas the RRQ-rumination scale measures general rumination that is self-focused but not necessarily self-critical. Nevertheless, the high correlation raised concern about whether the two scales measures distinct constructs. For this reason, we first examined the incremental validity of the SCRS over the RRQ-rumination scale in

7

2

3

4

5

6

7

8

9

10

11

12

13

18

19

1.15 0.50

2.35 0.66

2.83 0.52

2.08 0.56

1.64 0.58

17

2.49 0.49

16

.58** .66** −.33** .11* .50** .61** −.28** .03 .59** .63** −.24** .10* .61** .56** −.31** .04 .55** .54** −.11* .11* .42** .48** −.15** .04 .57** .58** −.22** .01 .62** .71** −.35** .06 .58** .67** −.20** .07 .41** .53** −.09 .10 −.03 .11* .32** .19** .54** .64** −.26** .06 .57** .54** −.29** .11* −.52** −.71** .32** −.07 .68** .69** −.32** .12* .67** −.35** .04 −.35** .10 .11*

15

−.62** .65** −.62** .63** −.58** .67**. −.53** .71** −.45** .56** −.47** .63** −.51** .61** −.69** .66* −.63** .62** −.47** .44** −.07 −.04 −.51** .60** −.46** .59** −.65**

14

2.93 0.54

−.24** −.17** −.14** −.21** −.08 −.08 −.14** −.25** −.12* −.02 .41** −.17** −.17** .23** −.31** −.24** −.15** .50** .06

20

−.27** −.28** −.18** −.38** −.10* −.22** −.22** −.37** −.17** −.02 .47** −.14** −.14** .21** −.44** −.34** −.26** .44** .15** .59**   2.45 0.46

21

Note. SCRS = Self-Critical Rumination Scale; RRQ-r = rumination subscale of the Ruminative Responses Questionnaire; RRS-b = brooding subscale of the Ruminative Response Styles Questionnaire; ARS = Anger Rumination Scale; ARQ = Anxious Rumination Questionnaire; RIO = Rumination on Interpersonal Offenses scale; PEPQ = Post-Event Processing Questionnaire; ESS = Experiences of Shame Scale; TOSCA-s = shame subscale of the Test of Self-Conscious Affect scale; TOSCA-g = guilt subscale of the Test of Self-Conscious Affect scale; PFQ-s = shame subscale of the Personal Feelings Questionnaire; PFQ-g = guilt subscale of the Personal Feelings Questionnaire; SCS = Self-Compassion Scale; BPD = borderline features scale of Personality Assessment Inventory; DASS = Depression Anxiety Stress Scales total score; Neur = neuroticism; Extra = extraversion; Open = openness; Consc = conscientiousness; Agree = agreeableness. *p < .05. **p < .01.

1. SCRS .81** .68** .67** .59** .53** .65** .81** .73** .55** .12* .66** .64** 2. RRQ-r .63** .69** .51** .61** .67** .78** .64** .49** .12* .60** .57** 3. RRS-b .72** .64** .62** .62** .65** .66** .49** .15** .59** .54** 4. ARS .58** .71** .70** .64** .62** .43** .01 .56** .59** 5. ARQ .51** .63** .57** .64** .50** .22** .54** .51** 6. RIO .57** .54** .52** .41** .13* .49** .45** 7. PEPQ .64** .68** .50** .09** .57** .54** 8. LOSC .72** .55** .04 .61** .59** 9. ESS .59** .20** .69** .61** 10. TOSCA-s .50** .54** .48** 11. TOSCA-g .16** .11* 12. PFQ-s .77** 13. PFQ-g 14. SCS 15. BPD 16. DASS 17. Neur 18. Extra 19. Open 20. Consc 21. Agree Mean 2.17 3.07 2.22 1.85 2.14 2.85 2.71 3.69 2.21 2.75 3.59 2.21 2.07 SD 0.73 0.90 0.82 0.63 0.61 1.06 0.98 1.05 0.71 0.62 0.63 0.54 0.58

1

Table 2.  Relationships Between the Self-Critical Rumination Scale and Related Constructs (N = 420).

8

Assessment 

Table 3.  Hierarchical Regression Analyses Showing Prediction of General Distress and Borderline Personality Features by General Rumination and Self-Critical Rumination. Dependent variable DASS total

PAIBOR total

Step 1   2     1   2    

Predictor(s) RRQ rumin Model total RRQ rumin SCRS Model total RRQ rumin Model total RRQ rumin SCRS Model total

R2

ΔR2

.249

.249**

.334

.084**

.389

.389**

.452

.063**

β .499*   .086 .505*   .623**   .275** .430**  

Note. DASS = Depression Anxiety Stress Scale; RRQ rumin = rumination scale from the Rumination Reflection Questionnaire; SCRS = Self-Critical Rumination Scale. *p < .05. **p < .01.

Table 4.  Hierarchical Regression Analyses Showing Prediction of General Distress and Borderline Personality Features by Trait-Level Self-Criticism and Self-Critical Rumination. Dependent variable DASS total

PAIBOR total

Step 1     2       1     2      

Predictor(s) LOSC CSC LOSC ISC Model total LOSC CSC LSOC ISC SCRS Model total LOSC CSC LOSC ISC Model total LOSC CSC LOSC ISC SCRS Model total

R2

ΔR2

.380

.380**

.396

.017**

.432

.432**

.474

.042**

β .295** .408**   .197** .295** .226**   .326** .426**   .171** .248** .355**  

Note. DASS = Depression Anxiety Stress Scale; LOSC CSC = Levels of Self-Criticism Scale Comparative Self-Criticism; LOSC ISC = Levels of SelfCriticism Scale Internal Self-Criticism; SCRS = Self-Critical Rumination Scale; PAIBOR = Personality Assessment Inventory Borderline Features Scale. *p < .05. **p < .01.

predicting both general distress and borderline personality features (shown in Table 3). In the first analysis, the RRQrumination scale entered at Step 1 and accounted for significant variance in general distress (R2 = .249, p < .001). The SCRS entered at Step 2 and resulted in a significant increase in R2 to .334 (p = .001). In the final model, only the SCRS was a significant predictor of general distress. This analysis was repeated with borderline personality features as the dependent variable. At Step 1, RRQ-rumination accounted for significant variance in BPD features (R2 = .389, p < .001). The SCRS entered at Step 2, resulting in a significant increase in R2 to .452 (p = .001). In the final model, both RRQ-rumination and the SCRS were significant predictors of borderline personality features. These analyses suggest

that, despite the high correlation between the two scales, the SCRS has significant incremental validity over the RRQ in predicting general distress and borderline personality features. Next, we examined whether the SCRS is a unique predictor of overall distress and borderline personality features after accounting for trait-level self-criticism. These analyses are shown in Table 4. In the first analysis, the two subscales of the LOSC were entered simultaneously at Step 1 and accounted for significant variance in overall distress (R2 = .380, p < .001). The SCRS was entered at Step 2 and resulted in a significant increase in R2 to .396 (p = .001). In the final model, each of the subscales of the LOSC and the SCRS were significant predictors of overall distress.

9

Smart et al. Table 5.  Hierarchical Regression Analyses Showing Prediction of General Distress and Borderline Personality Features by Specific Forms of Rumination. Dependent variable DASS total

PAIBOR total

Step

Predictor(s)

1           2             1           2            

RRS brood ARS ARQ RIO PEPQ Model total RRS brood ARS ARQ RIO PEPQ SCRS Model total RRS brood ARS ARQ RIO PEPQ Model total RRS brood ARS ARQ RIO PEPQ SCRS Model total

R2

ΔR2

.448

.448**

.459

.011**

.581

.581**

.600

.019**

β .189** .320** .180** −.088 .157**   .146* .284** .160** −.092 .123* .157**   .235** .302** .084 .151** .115   .179** .253** .058 .147** .070 .208**  

Note. DASS = Depression Anxiety Stress Scales; RRS brood = brooding subscale of the Ruminative Response Styles Questionnaire; ARS = Anger Rumination Scale; ARQ = Anxious Rumination Questionnaire; RIO = Rumination on Interpersonal Offenses scale; PEPQ = Post-Event Processing Questionnaire; SCRS = Self-Critical Rumination Scale; PAIBOR = Personality Assessment Inventory Borderline Features Scale. *p < .05. **p < .01.

This analysis was repeated with borderline personality features as the dependent variable. At Step 1, the two subscales of the LOSC were entered simultaneously and accounted for significant variance in BPD features (R2 = .432, p < .001). The SCRS entered at Step 2, resulting in a significant increase in R2 to .474 (p < .001). In the final model, each of the subscales of the LOSC and the SCRS were significant predictors of BPD features. These analyses suggest that the SCRS has significant incremental validity over a measure of trait-level self-criticism in predicting general distress and borderline personality features. Finally, we examined whether the SCRS is a unique predictor of general distress and borderline personality features after accounting for other specific forms of rumination. In the first, measures of specific types of rumination (depressive brooding, anger rumination, etc.) were entered simultaneously at Step 1 and accounted for significant variance in general distress (R2 = .448, p < .001). The SCRS entered at Step 2, resulting in a significant increase in R2 to .459 (p = .006). In the final model, each of the different types of

rumination was a significant predictor of general distress except the RIO (see Table 5). This analysis was repeated with borderline personality features as the dependent variable. At Step 1, specific forms of rumination accounted for significant variance in BPD features (R2 = .581, p < .001). The SCRS entered at Step 2 and resulted in a significant increase in R2 to .600 (p < .001). In the final model, each of the different types of rumination was a significant predictor of general distress except anxious rumination and postevent processing. Overall, these findings suggest that the SCRS has significant incremental validity over measures of other specific forms of rumination in predicting general distress and borderline personality features.

Study 3: Confirmatory Factor Analysis Method A new sample was collected to confirm the factor structure of the SCRS. Participants in this sample were 143 undergraduate students enrolled in introductory psychology

10 classes who completed the SCRS in person via a paper and pencil packet. The sample was predominantly female (69.9%), Caucasian (72.2%), and ranged in age from 18 to 30 years old (M = 19.00, SD = 1.46).

Results and Discussion Confirmatory factor analysis provides fit indices, which indicate the degree to which the proposed model fits the data. We used three fit indices: the model chi square divided by degrees of freedom (χ2/df), the comparative fit index (CFI; Bentler, 1990), and the root mean square error of approximation (RMSEA; Marsh, Balla, & Hau, 1996). The model chi square is frequently reported; however, this fit index is calculated using the sample size such that larger sample sizes increase the likelihood that this fit index will be poor. To correct for this, the model chi square can be divided by the degrees of freedom in the model. Generally, χ2/df values below 2 are considered good, while values below 3 are acceptable. Values above .95 for the CFI indicate excellent fit while values above .90 indicate acceptable fit. For the RMSEA, values below .05 indicate good fit, values below .08 indicate acceptable fit, and values below .10 indicate marginal fit (Browne & Cudeck, 1993). The hypothesized single-factor model fit the data well: χ2/df = 1.69, CFI = .97, and RMSEA = .07.

Study 4: Test–Retest Reliability Method A final sample was collected to assess test-retest reliability of the SCRS. Participants in this sample were 85 undergraduate students enrolled in introductory psychology classes who completed an online survey for course credit. Of the initial 85 participants, 15 did not respond to the invitation to complete Time 2, leaving a sample of 70 who participated in both time points. The sample was predominantly female (89.9%) and Caucasian (91.3%). Participants were asked to complete a demographic questionnaire and the SCRS at Time 1. Two weeks later they were emailed invitations to recomplete the SCRS. Participants completed Time 2 between 13 and 37 days after completing Time 1 (M = 16.80 days, SD = 4.02).

Results and Discussion Correlations between scores on Time 1 and Time 2 were calculated. The test-retest correlation for the SCRS was .86, indicating good test-retest reliability. A paired sample t test indicated that on average, participants’ scores on Time 1 (M = 1.90, SE = .08) were not significantly different from their scores on Time 2 (M = 1.83, SE = .08), t(68) = 1.81, p = .07.

Assessment 

General Discussion The purpose of this project was to develop a measure of self-critical rumination. Several methods were used to generate items, including a writing task administered to student and clinical samples as well as adapting items from other validated measures of rumination. Using these methods, a total of 39 items were written. Content validity ratings led to the elimination of several items. Other items were eliminated if they demonstrated excessively high item-total correlations. The remaining 10 items were then subjected to factor analyses, which indicated that a single-factor solution best fit the data. The final scale demonstrated excellent internal consistency and significant correlations in the expected directions with measures of related constructs. Discriminant validity of the SCRS was supported by a very small relationship with the openness domain of personality and correlations with extraversion, agreeableness, and conscientiousness that were significantly smaller, as predicted, than the correlation with neuroticism. A series of regression analyses showed incremental validity of the SCRS in predicting both general psychological distress and borderline personality features above and beyond other measures of rumination or self-criticism. Although the incremental validity coefficients for the SCRS were small, they were statistically significant even after accounting for a wide range of other measures of selfcriticism or rumination. This pattern of findings suggests that self-critical rumination may have value as an important predictor of psychological symptoms and distress even after accounting for other types of rumination and nonruminative self-criticism. Future work should seek to clarify further the role of self-critical rumination in psychological distress and BPD as well as in other psychological disorders. Intervention approaches that focus on the reduction of rumination, including cognitive-behavioral and mindfulness-based approaches, may be enhanced by helping clients identify the content of their typical ruminative thought patterns (depressive, angry, self-critical, etc.) as a way of recognizing when they are ruminating, understanding the consequences of this thought pattern, and taking constructive steps to reduce rumination. Several limitations of this study must be noted. While both student and clinical samples were used during the creation of the initial item pool, only a student sample was used to validate these items. Psychometric properties should be examined in clinical and community samples. Future research also should include additional measures that may be strongly correlated with the SCRS. These include the Automatic Thoughts Questionnaire (Hollon & Kendall, 1980), which includes many self-critical items (“I’m no good,” “I’m a loser”) but does not explicitly ask about the prolonged, repetitive style of thinking that characterizes rumination; and the Depressive Experiences Questionnaire,

Smart et al. which also measures self-criticism but not rumination. Another potential limitation is that the SCRS was strongly correlated with the general rumination scale of the RRQ, raising concern about whether it represents a distinct construct. We suggest that this high correlation is not surprising, as people who ruminate in a general way are also likely to ruminate on a variety of specific topics. In addition, several SCRS items were adapted from RRQ items and retained very similar structure, so measurement factors are likely included in the overlapping variance. The utility of assessing rumination on specific topics, and self-critical rumination specifically, was demonstrated with significant incremental validity of the SCRS over the RRQ-rumination scale in accounting for variance in both general distress and borderline personality features. Finally, discriminant validity of the SCRS was tested only with domains of the fivefactor model of personality. Broader tests of discriminant validity may be informative. Despite these concerns, the ability of the SCRS to predict both general distress and borderline personality features above and beyond these other measures is indicative of the SCRS’s incremental validity and suggests that selfcritical rumination merits additional research. Taken together, these results indicate that SCRS is a valuable addition to the literature on the different forms of rumination and can contribute to the understanding of self-criticism, which is common in many forms of psychopathology but is relatively understudied. Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.

References Andrews, B., Qian, M., & Valentine, J. D. (2002). Predicting depressive symptoms with a new measure of shame: The Experiences of Shame Scale. British Journal of Clinical Psychology, 41, 29-42. Baer, R. A., & Sauer, S. E. (2011). Relationships between depressive rumination, anger rumination, and borderline personality features. Personality Disorders: Theory, Research and Treatment, 2, 142-150. Bentler, P. M. (1990). Comparative fit indices in structural models. Psychological Bulletin, 107, 238-246. Blatt, S., Quinlan, D., Chevron, E., McDonald, C., & Zuroff, D. (1982). Dependency and self-criticism: Psychological dimensions of depression. Journal of Consulting and Clinical Psychology, 50, 113-124. Blatt, S. J., D’Afflitti, J. P., & Quinlan, D. M. (1976). Experiences of depression in normal young adults. Journal of Abnormal Psychology, 85, 383-389.

11 Brown, M. Z., Linehan, M. M., Comtois, K., Murray, A., & Chapman, A. L. (2009). Shame as a prospective predictor of self-inflicted injury in borderline personality disorder: A multimodal analysis. Behaviour Research & Therapy, 47, 815-822. Browne, M. W., & Cudeck, E. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136-162). Newbury Park, CA: Sage. Brozovich, F., & Heimberg, R. (2013). Mental imagery and postevent processing in anticipation of a speech performance among socially anxious individuals. Behavior Therapy, 44, 701-716. Bushman, B. J., Bonacci, A. M., Pedersen, W. C., Vasquez, E. A., & Miller, N. (2005). Chewing on it can chew you up: Effects of rumination on triggered displaced aggression. Journal of Personality and Social Psychology, 88, 969-983. Campos, R., Besser, A., & Blatt, S. (2010). The mediating role of self-criticism and dependency in the association between perceptions of maternal caring and depressive symptoms. Depression and Anxiety, 27, 1149-1157. Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Lutz, FL: Psychological Assessment Resources. Cox, B., Fleet, C., & Stein, M. (2004). Self-criticism and social phobia in the US national comorbidity survey. Journal of Affective Disorders, 82, 227-234. Dunkley, D., Masheb, R., & Grilo, C. (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, 274-281. Fergus, T. A., Valentiner, D. P., McGrath, P. B., & Jencius, S. (2010). Shame- and guilt-proneness: Relationships with anxiety disorder symptoms in a clinical sample. Journal of Anxiety Disorder, 24, 811-815. Geiger, P. J., Peters, J. R., Sauer-Zavala, S. E., & Baer, R. A. (2013). Relationships between maladaptive cognitive content, dysfunctional cognitive processes, and borderline personality features. Journal of Personality Disorders, 27, 457-464. Gilbert, P., Clarke, M., Hempel, S., Miles, J., & Irons C. (2004). Criticizing and reassuring oneself: An exploration of forms, styles, and reasons in female students. British Journal of Clinical Psychology, 43, 31-50. Gilbert, P., & Proctor, S. (2006). Compassionate mind training for people with high shame and self-criticism: Overview and pilot study of a group therapy approach. Clinical Psychology and Psychotherapy, 13, 353-379. Glassman, L., Weierich, M., Hooley, J., Deliberto, T., & Nock, M. (2007). Child maltreatment, non-suicidal self-injury, and the mediating role of self-criticism. Behaviour Research and Therapy, 45, 2483-2490. Harder, D. H., & Zalma, A. (1990). Two promising shame and guilt scales: A construct validity comparison. Journal of Personality Assessment, 55, 729-745. Harman, R., & Lee, D. (2010). The role of shame and self-critical thinking in the development and maintenance of current threat in post-traumatic stress disorder. Clinical Psychology and Psychotherapy, 17(1), 13-24.

12 Hollon, S., & Kendall, P. (1980). Cognitive self-statements in depression: Development of an automatic thoughts questionnaire. Cognitive Therapy and Research, 44, 383-395. Hooley, J. M., Ho, D. T., Slater, J., & Lockshin, A. (2010). Pain perception and non-suicidal self-injury: A laboratory investigation. Personality Disorders: Theory, Research and Treatment, 1, 170-179. Jacob, G. A., Hellstern, K., Ower, N., Pillmann, M., Scheel, C., Rüsch, N., & Lieb, K. (2009). Emotional reactions to standardized stimuli in women with borderline personality disorder. Journal of Nervous and Mental Disease, 197, 808-815. Kim, S., Thibodeau, R., & Jorgensen, R. (2011). Shame, guilt, and depressive symptoms: A meta-analytic review. Psychological Bulletin, 137(1), 68-96. Lassri, D., & Shahar, G. (2012). Self-criticism mediates the link between childhood emotional maltreatment and young adults’ romantic relationships. Journal of Social and Clinical Psychology, 31, 289-311. Lovibond, S. H., & Lovibond, P. F. (1993). Manual for the Depression Anxiety Stress Scales (DASS) (Psychology Foundation Monograph). Sydney, New South Wales, Australia: University of New South Wales. Marsh, H., Balla, K., & Hau, K. (1996). An evaluation of incremental fit indices: A clarification of mathematical and empirical properties. In G. A. Marcoulides & R. E. Schumacker (Eds.), Advanced structural equation modeling: Issues and techniques (pp. 315-353). Mahwah, MJ: Lawrence Erlbaum. McEvoy, P. M., & Kingsep, P. (2006). The post-event processing questionnaire in a clinical sample with social phobia. Behaviour Research and Therapy, 44, 1689-1697. Morey, L. C. (1991). Personality Assessment Inventory Professional Manual. Odessa, FL: Psychological Assessment Resources. Myers, R. (1990). Classical and modern regression with applications (2nd ed.). Boston, MA: Duxbury. Neff, K. D. (2003). Development and validation of a scale to measure self-compassion. Self and Identity, 2, 223-250. Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100, 569-582. Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61, 115-121. Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3, 400-424. O’Connor, R. C., & Noyce, R. (2008). Personality and cognitive processes: Self-criticism and different types of rumination

Assessment  as predictors of suicidal ideation. Behaviour Research and Therapy, 46, 392-401. Peters, J. R., Geiger, P. J., Smart, L. M., & Baer, R. A. (2013). Shame and borderline personality features: The potential mediating role of anger and anger rumination. Personality Disorders: Theory, Research, & Treatment, 5, 1-9. doi:10.1037/per0000022 Powers, T., Koestner, R., & Zuroff, D. (2007). Self-criticism, goal motivation, and goal progress. Journal of Social and Clinical Psychology, 26, 826-840. Randolph, J. J. (2005, October). Free-marginal multirater kappa: An alternative to Fleiss’ fixed-marginal multirater kappa. Paper presented at the Joensuu University Learning and Instruction Symposium 2005, Joensuu, Finland. Rector, N., Antony, M., Laposa, J., Kocovski, N., & Swinson, R. (2008). Assessing content domains of repetitive thought in the anxiety spectrum: Rumination and worry in nonclinical and clinically anxious samples. International Journal of Cognitive Therapy, 1, 352-377. Rusting, C. L., & Nolen-Hoeksema, S. (1998). Regulating responses to anger: Effects of rumination and distraction on angry mood. Journal of Personality and Social Psychology, 74, 790-803. Spasojevic, J., & Alloy, L. (2001). Rumination as a common mechanism relating depressive risk factors to depression. Emotion, 1, 25-37. Sukhodolsky, D. G., Golub, A., & Cromwell, E. N. (2001). Development and validation of the anger rumination scale. Personality and Individual Differences, 31, 689-700. Tangney, J. P., & Dearing, R. (2002). Shame and guilt. New York, NY: Guilford. Tangney, J. P., Stuewig, J., & Mashek, D. (2007). Moral emotions and moral behavior. Annual Review of Psychology, 58, 345-372. Thompson, R., & Zuroff, D. C. (2004). The Levels of SelfCriticism Scale: Comparative self-criticism and internalized self-criticism. Personality and Individual Differences, 36, 419-430. Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness and the five-factor model of personality: Distinguishing rumination from reflection. Journal of Personality and Social Psychology, 76, 284-304. Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A psychometric analysis. Cognitive Therapy and Research, 27, 247-259. Wade, N. G., Vogel, D. L., Liao, K., & Goldman, D. B. (2008). Measuring state-specific rumination: Development of the Rumination About an Interpersonal Offense Scale. Journal of Counseling Psychology, 55, 419-426.