The Effect of Self-Compassion on Shame in Post-Event Processing

Abstract

Post-event processing (PEP) refers to negative and repetitive thinking following anxiety-inducing social situations and is a maintaining factor in social anxiety. Research has shown that state shame predicts PEP, beyond trait and state social anxiety. Therefore, it may be important for interventions to target shame during the post-event period. Self-compassion interventions have shown promise for PEP, and therefore, the primary purpose of the present study was to examine whether a self-compassion intervention reduces state shame during the post-event period. Participants (N = 73) completed a variety of baseline trait measures: social anxiety, depression, post-event processing, self-compassion, self-esteem, and shame. They then recalled a difficult social situation (e.g., presentation, first date, etc.) and rated their state PEP and shame in relation to that situation. They were then randomly assigned to a self-compassion writing intervention (n = 36) or control condition (n = 37). Following this, they rated their shame about the recalled situation and willingness to communicate in future social situations. Those in the self-compassion intervention reported significantly less shame post-intervention than those in the control condition. However, the two groups did not significantly differ on willingness to communicate in future social settings. Unexpectedly, the effect of the self-compassion intervention on shame or willingness to communicate was not moderated by trait social anxiety (or other baseline variables). The findings from the present study provide further evidence of the importance of shame in post-event processing. Given these findings, it may be beneficial to target shame in therapeutic settings, and self-compassion represents one effective avenue for doing so.

Keywords

self-compassion social anxiety post-event processing shame

Social anxiety is characterized by a fear of evaluation from others, resulting in distress surrounding one or more types of social situations (American Psychiatric Association, 2022). Cognitive models posit that social anxiety is maintained, in part, due to post-event processing (e.g., Clark & Wells, 1995; Brozovich & Heimberg, 2008). Post-event processing (PEP) refers to negative and reoccurring thoughts that follow anxiety-provoking situations (Brozovich & Heimberg, 2008). While self-reflection can be used as a tool to learn from one’s past mistakes, PEP is more than simple self-reflection; this post-event self-analysis involves repetitively reviewing the anxiety-provoking situation over in one’s mind and is negatively biased in nature. It is typically dominated by the self-perceived inadequacies one assumes were on full display for others to judge during the anxiety-provoking situation (Clark & Wells, 1995; Hofmann, 2007).

PEP typically occurs in response to anxiety-provoking social situations and has been implicated in the maintenance of social anxiety (e.g., Clark, 2001; Clark & Wells, 1995; Rapee & Heimberg, 1997), with research demonstrating that PEP predicts anxiety in subsequent social situations (Blackie & Kocovski, 2016; Brozovich & Heimberg, 2008; Campbell et al., 2016). PEP is also associated with a host of other negative processes, including negative performance evaluations (Brozovich & Heimberg, 2011; Dannahy & Stopa, 2007; Gavric et al., 2017), negative affect (Kashdan & Roberts, 2007), as well as avoidance behaviours and impression management (Kim et al., 2022).

Various interventions have been shown to be effective for reducing PEP (see Donohue et al., 2024 for a review). Some notable interventions include cognitive restructuring (Shikatani et al., 2014) and distraction (Blackie & Kocovski, 2016; Gavric et al., 2023), though some research contradicts the usefulness of distraction for PEP (Rowa et al., 2014). More recently, self-compassion has been implicated as a wellbeing factor that may be relevant to social anxiety.

Self-compassion refers to a warm and non-judgmental attitude toward oneself during difficult times (Neff, 2003). According to Neff (2003), self-compassion consists of three key bipolar elements: self-kindness versus self-judgment, common humanity versus isolation, and mindfulness versus over-identification. Self-kindness involves an attitude of gentleness and understanding toward oneself during difficult times, whereas self-judgment involves an attitude that is harsh and condemning. Common humanity refers to understanding that difficulties and failures are part of the larger human experience that all individuals experience, whereas isolation refers to the perception that others do not experience difficulties or hardships. Finally, mindfulness in the context of self-compassion involves a balanced awareness of difficult thoughts and feelings. Note that self-compassionate mindfulness is slightly different from the broader construct of mindfulness, which refers to an attentiveness to the present moment in a non-judgmental manner (Kabat-Zinn, 1994). Mindfulness is contrasted with overidentification, which refers to the act of becoming overly attached to and preoccupied with thoughts and feelings. Given that social anxiety is associated with less self-compassion (Blackie & Kocovski, 2018a; Werner et al., 2012) and post-event processing (Blackie & Kocovski, 2018a), researchers have been investigating the utility of self-compassion interventions for social anxiety.

Self-compassion has also shown promise as an intervention in social anxiety. For instance, in comparison to a wait-list control condition, self-compassion training led to significantly greater reductions in social anxiety (Gharraee et al., 2018; Teale Sapach & Carleton, 2023). Furthermore, Blackie and Kocovski (2018b) found that a self-compassion intervention led to reduced PEP in socially anxious individuals in comparison to a control. Self-compassion has also shown to lead to less distress prior to interacting with others (Szczyglowski & Kocovski, 2024). When comparing self-compassion to other active experimental conditions, self-compassion did not significantly differ from cognitive restructuring with respect to social anxiety (Stevenson et al., 2019) or distress following social situations (Burnham & Kocovski, 2024). Taken together, these findings support self-compassion as an effective intervention for social anxiety and negative thinking during the post-event period. However, there is currently no research that examines whether self-compassion decreases feelings of shame during the post-event period. Considering the positive elements of self-compassion and its usefulness in reducing post-event processing and feelings of distress during the post-event period (Burnham & Kocovski, 2024), it may extend to reducing feelings of shame during the post-event period as well.

Given the nature of PEP, feelings of shame during and after social situations may be particularly important to its development and maintenance. Shame is a self-conscious emotion that arises during negative self-evaluation and is characterized by a sense of exposure and feelings of worthlessness (Tangney et al., 2014). Feelings of pain typically involve a distressing focus on the self in general, rather than on a specific behavior (Tracy & Robins, 2004). Past research has shown that shame is positively associated with social anxiety (for a review, see Swee et al., 2021). Recently, research has also shown that shame is a strong predictor of post-event processing, even when controlling for trait and state social anxiety (e.g., Cândea & Szentágotai-Tătar, 2018). Because shame involves a sense of self-defectiveness and diminished self-worth, it is possible that it may amplify post-event processing in social anxiety. Therefore, when considering the importance of shame in social anxiety and PEP, it is important to examine potential interventions for reducing shame during the post-event period.

The Present Study: Self-Compassion as an Intervention for Shame During PEP

As an initial step in the present research, we examined baseline correlations, aiming to replicate the relationship between post-event processing and shame (Cândea & Szentágotai-Tătar, 2018). Past research has shown the importance of state shame in predicting post-event processing (Cândea & Szentágotai-Tătar, 2018), and additional research has shown that individuals with social anxiety disorder tend to report greater feelings of shame (Hedman et al., 2013) and less self-compassion (Werner et al., 2012) than healthy controls. Thus, the primary purpose of the present study was to examine whether self-compassion as an intervention reduces state shame after recalling a difficult social task (e.g., interview, first date) during the post-event period. Additionally, given that past research has shown that a self-compassion intervention led to increased willingness to communicate in future social situations (Blackie & Kocovski, 2018b), we hypothesized that our brief self-compassion intervention would increase willingness to communicate in future hypothetical social situations in the present study. A final aim was to examine whether dispositional variables moderate the effect of the intervention on shame and willingness to communicate. Given the significant relationships among social anxiety, PEP, and self-compassion in past research (Blackie & Kocovski, 2018a), we hypothesized that trait measures of these constructs would serve as important moderators of the self-compassion intervention. To investigate these aims, several hypotheses were developed:

(1) Baseline levels of state shame (prior to the intervention) would predict state PEP (prior to the intervention), even when controlling for trait social anxiety, trait PEP, as well as state anxiety experienced during the recalled situation.

(2) An ultra-brief, five-minute self-compassion intervention would lead to reduced state shame during the post-event period, and a greater willingness to engage in future social situations in comparison to a control condition.

(3) Trait social anxiety, trait PEP, and trait self-compassion would (separately) interact with the intervention (self-compassion versus control) to predict levels of state shame during the post-event period, as well as willingness to communicate in future social situations. More specifically, it was hypothesized that those with higher trait social anxiety, trait PEP, and lower trait self-compassion (independently) would experience significantly less shame and more willingness to communicate when in the self-compassion condition than when in the control condition. However, those with lower levels of trait social anxiety, lower trait PEP, and heightened trait self-compassion would not significantly differ on shame or willingness to communicate when in the self-compassion versus the control condition. In other words, those with the highest levels of trait social anxiety, highest trait PEP, and lowest levels of trait self-compassion were expected to benefit from the self-compassion intervention, whereas those with low social anxiety, low PEP, and high self-compassion would not.

Method

Participants

A student sample of 79 participants took part in the present study. Participants in the present study were not pre-screened for elevated social anxiety or a social anxiety diagnosis.

Measures

Trait Social Anxiety

Trait social anxiety was assessed using the Social Phobia Inventory (SPIN; Connor et al., 2000). This 17-item measure uses a 5-point Likert-type scale to assess the extent to which participants agree or disagree with each statement (0 = Not at All; 4 = Extremely). Higher scores on the scale represent higher social anxiety. The scale has shown excellent reliability and validity in past research (Connor et al., 2000), and excellent internal consistency in the present study (α = .92).

Depression

The Beck Depression Inventory – II (BDI-II; Beck et al., 1996) was used to assess depression in the present study. This is a commonly used measure of depression, which consists of 21 items, each rated on a scale from 0-3, with higher numbers reflecting higher symptom severity. Past research has demonstrated excellent psychometric properties for this scale (e.g., Beck et al., 1996). The internal consistency of the scale was excellent in the present study (α = .93).

Post-Event Processing

The Post-Event Processing Inventory (PEPI; Blackie & Kocovski, 2017) was employed to assess both trait and state PEP. Each form of the scale contains 12 items, represented by 3 factors (frequency, intensity, and self-judgment), which, in turn, are represented by the higher-order, global PEP factor. Thus, both subscale scores and total scale scores can be used on each form of the scale. Items on each scale are rated from 1-5 (1 = Strongly Disagree, 5 = Strongly Agree), with higher scores representing higher trait or state PEP. Both forms of the scale have excellent psychometric properties (Blackie & Kocovski, 2017, 2019). The internal consistencies for each form of the PEPI were excellent in the present study (PEPI-T α = .93; PEPI-S α = .93).

Self-Compassion

The short-form of the Self-Compassion Scale (SCS-SF; Raes et al., 2011) is a 12-item measure that assesses three bipolar aspects of self-compassion: self-kindness versus self-judgment, common humanity versus isolation, and mindfulness versus over-identified. Items on this scale are rated from 1-5 (1 = Almost Never, 5 = Almost Always), with higher scores representing higher self-compassion. The internal consistency reliability for the total scale score was good in past research (e.g., Raes et al., 2011), as well as in the present study (α = .79).

Self-Esteem

Trait self-esteem was assessed via The Single-Item Self-Esteem Scale (SISE; Robins et al., 2001), which is a one-item scale rated from 1-7 (1 = Not Very True of Me, 7 = Very True of Me) that measures global self-esteem. The psychometric properties of this scale are very good. For instance, the SISE correlates highly with the Rosenberg Self-Esteem Scale (Rosenberg, 1965), and these two measures of self-esteem correlate at similar strengths with other related variables (Robins et al., 2001). The SISE has also demonstrated very good test-retest reliability (Robins et al., 2001).

Shame

The External and Internal Shame Scale (EISS; Ferreira et al., 2022) measures both internal and external aspects of trait shame. External shame involves feeling a sense of being judged by others, whereas Internal shame involves a sense of judging oneself. Although this measure consists of two subscales, a total, global sense of shame can also be calculated. All items on the scale are rated from 0-4 (0 = Never, 4 = Always), and greater scores reflect greater shame. Past research has shown that the scale demonstrates good psychometric properties (e.g., Ferreira et al., 2022). The internal consistency of the EISS (global score) was very good in the present study (α = .82).

Minor revisions were made to the EISS (Ferreira et al., 2022) to also assess shame from a state perspective, specifically in relation to a speech. For example, the original EISS (Ferreira et al., 2022) includes an item that assesses, “Other people are judgmental and critical of me,” whereas the revised state version assesses, “Other people are judgmental and critical of my speech.” In the present study, the internal consistency for the EISS-S was very good at time 1 (α = .85), as well as time 2 (α = .86).

Willingness to Engage in Future Social Situations

The Willingness to Communicate Scale (WTC; McCroskey, 1992) measures willingness to initiate communication in various hypothetical situations (e.g., public speaking, speaking with acquaintances, speaking with friends, etc.). Although the scale consists of 20 items, 8 items serve as fillers, which serve to distract attention away from the 12 scored items. The subscales consist of willingness to communicate in four contexts (interpersonal, groups, meetings, and public speaking), with three receiver types (friends, acquaintances, and strangers). For each item, participants rate the percentage of time they would initiate communication (0% = Never, 100% = Always). Subscale scores or total scale scores may be used. A total score was used in the present study, which showed excellent internal consistency (α = .90).

State Anxiety

The Subjective Units of Distress Scale (SUDS; Wolpe, 1969) was used to rate the level of state anxiety participants experienced during the recalled situation. Participants rated this single item using a scale ranging from 0 (no anxiety) to 100 (highest possible anxiety).

Manipulation Check

Following Blackie and Kocovski (2018b), the manipulation-check for the self-compassion intervention consisted of three items: “I was supportive and nice to myself in relation to the situation I experienced,” “I reminded myself that many people have a hard time in social situations,” and “I considered both the positive and negative aspects of the situation I experienced.” These three items were intended to capture the three key elements of self-compassion (self-kindness, common humanity, and mindfulness, respectively). Each item was rated on a scale from 1-5 (1 = Strongly Disagree, 5 = Strongly Agree), with higher scores on each item representing a higher state of self-kindness, common humanity, and mindfulness.

Procedure

After obtaining informed consent, a questionnaire booklet was administered to participants in several psychology courses at the beginning of class in exchange for one bonus point toward the course. Participants completed a demographics questionnaire, as well as trait measures of social anxiety (SPIN), depression (BDI-II), PEP (PEPI-T), self-compassion (SCS-SF), self-esteem (SISE), and shame (EISS). Following this, participants recalled a difficult, anxiety-provoking social situation that occurred within the previous two weeks (e.g., presentation, first date, etc.). They briefly described the situation, when it occurred, and level of state anxiety experienced during the situation (SUDS). All information pertaining to the recalled situation was written by participants in their survey booklet, in a private manner. Participants then rated their state levels of shame (EISS-S) and state levels of post-event processing (PEPI-S) since encountering the situation. Next, participants were randomized to either the self-compassion writing condition (n = 36) or the control condition (n = 37). Those in the self-compassion condition completed a five-minute writing exercise developed by Blackie and Kocovski (2018b), which was designed to elicit the three main components of self-compassion, whereas those in the control condition were asked to sit and wait for 5 minutes. Because participants were run in a mass setting, they were instructed to keep track of their allotted 5 minutes on their assigned task by using the clock displayed on the projected computer screen at the front of the room. Once the 5 minutes had passed, they completed a manipulation check. They then reported their state levels of shame surrounding the recalled situation (EISS-S) for the second time, as well as a measure of willingness to communicate in future social situations (WTC).

Results

Missing Data

Overall, there was 0.41% missing values within the dataset. In the present sample, six participants had significant missing data and were thus excluded, leaving a final sample of 73 participants. The remaining missing values were estimated using mean imputation for the given questionnaire in which the values were missing.

Sociodemographic Information

The average age in the present study was 22.19 (SD = 4.98). In terms of gender, 35.62% (n = 26) reported being male, 61.64% (n = 45) reported being female, and 2.74% (n = 2) reported being non-binary. The breakdown of ethnicity was as follows: White (50.7%), Asian (8.2%), Black/African Canadian (16.4%), Indian/West Asian (20.5%), Indigenous/First Nations (2.7%), and missing/prefer not to say (1.4%).

Comparing Conditions on Baseline Measures

Independent samples t-tests were used to compare conditions on baseline variables prior to the manipulation. As shown in Table 1, there were no significant differences between conditions on any baseline variables.

Table 1.

Comparing Conditions on Variables Assessed Prior to Manipulation

Measure	Self-compassion		Control		t	df	p
Measure	M	SD	M	SD	t	df	p
Trait Measures
Social anxiety (SPIN)	24.90	14.49	24.31	14.42	0.18	71	.86
Depression (BDI-II)	18.70	12.74	16.38	11.64	0.81	71	.42
Post-event processing (PEPI-T)	35.34	10.53	33.49	9.73	0.78	71	.44
Self-compassion (SCS-SF)	34.11	8.11	34.29	8.32	−0.09	71	.93
Self-esteem (SISE)	4.05	1.82	4.24	1.61	−0.47	71	.64
Shame (EISS-T)	14.15	7.28	14.11	6.25	0.02	71	.98
State Measures (before manipulation)
State post-event processing (PEPI-S)	36.11	12.35	35.61	11.25	0.18	71	.86
State shame (EISS-S)	14.28	9.17	13.59	5.97	0.38	71	.71
Measures Regarding Recalled Situation
Days since situation occurred	9.03	8.28	9.70	5.59	−0.38	71	.71
Anxiety during situation	64.47	24.43	64.19	26.71	0.05	71	.96
Shame regarding situation	14.23	9.26	13.59	5.97	0.46	71	.65
PEP in relation to situation	36.11	12.35	35.62	11.25	0.18	71	.86

Note. SPIN = Social Phobia Inventory. BDI-II = Beck Depression Inventory II. PEPI = Post-Event Processing Inventory. SCS-SC = Self-Compassion Scale – Short Form. SISE = Single Item Self-Esteem Scale. EISS = External and Internal Shame Scale. T = Trait. S = State.

Baseline Correlations

Two-tailed Pearson correlations were computed among baseline variables assessed prior to the manipulation. As shown in Table 2, both trait social anxiety (SPIN) and trait PEP (PEPI-T) were negatively correlated with trait self-compassion (SCS-SF), and positively correlated with trait shame (EISS), state shame (EISS state), and state PEP (PEPI-S). Additionally, state shame (EISS state) and state PEP (PEPI-S) were negatively correlated with trait self-compassion and positively correlated with trait shame.

Table 2.

Correlations Among Baseline Variables Measured Prior to the Manipulation

Measures	SPIN	BDI-II	PEPI-T	SCS-SF	SISE	EISS	PEPI-S
Trait social anxiety (SPIN)
Trait depression (BDI-II)	.59***
Trait post-event processing (PEPI-T)	.54***	.38***
Trait self-compassion (SCS-SF)	−.49***	−.54***	−.30**
Trait self-esteem (SISE)	−.59***	−.57***	−.38***	.49***
Trait shame (EISS)	.59***	.67***	.60***	−.58***	−.57***
State post-event processing (PEPI-S)	.42***	.38**	.71***	−.26***	−.32**	.48***
State shame (EISS-S)	.42***	.49***	.48***	−.35*	−.36***	.66***	.63***

Note. *p < .05. **p ≤ .01. ***p ≤ .001. SPIN = Social Phobia Inventory. BDI-II = Beck Depression Inventory II. PEPI = Post-Event Processing Inventory. SCS-SC = Self-Compassion Scale – Short Form. SISE = Single Item Self-Esteem Scale. EISS = External and Internal Shame Scale. T = Trait. S = State.

Recalled Situation

Independent samples t-tests were used to compare conditions on several variables related to the recalled situation. As shown in Table 1, those in the self-compassion and control conditions did not significantly differ on levels of state anxiety experienced during the situation, shame felt during the situation, or post-event processing since the situation occurred.

Predictors of State PEP Prior to Manipulation

A hierarchical regression analysis was computed to examine the unique contribution of state shame in predicting levels of state PEP, beyond levels of trait social anxiety, trait PEP, trait depression, and situational state anxiety. Trait social anxiety, trait PEP, state anxiety, and trait depression were entered into step 1 of the regression (F (4, 68) = 22.81, p < .001, R² = .57), and state shame was entered into step 2 (F (5, 67) = 24.89, p < .001, R² = .65). As expected, state shame predicted unique variance in state PEP when controlling for these other variables (ΔR² = .08, p < .001). All variables, except for trait social anxiety and depression, were significant predictors of PEP in step 2 of the regression (see Table 3).

Table 3.

Assessing Shame as a Unique Predictor of State PEP

Measure	B	t	p
Regression Step 1
Social anxiety (SPIN)	−.05	−0.45	.65
Trait post-event processing (PEPI-T)	.57	5.69	<.001
State anxiety (SUDS)	.26	2.81	.01
Depression (BDI-II)	.10	0.95	.35
Regression Step 2
Social anxiety (SPIN)	−.06	−0.65	.52
Trait post-event processing (PEPI-T)	.47	4.97	<.001
State anxiety (SUDS)	.23	2.65	.01
Depression (BDI-II)	−.01	−0.11	.91
State shame (EISS-S)	.34	3.84	<.001

State PEP assessed via the Post-Event Processing Inventory – State form. SPIN = Social Phobia Inventory. PEPI-T = Post-Event Processing Inventory – Trait. SUDS = Subjective Units of Distress Scale. BDI-II = Beck Depression Inventory II. EISS = External and Internal Shame Scale – State.

Manipulation Check for the Intervention

Independent samples t-tests were used to compare the self-compassion and control conditions on the three manipulation check items. Levene’s test for equality of variances was violated for each of these items. Therefore, we used the Welch’s adjustment to estimate the corresponding t- and p-values. As shown in Table 4, those in the self-compassion condition scored significantly higher on all three manipulation check items (Self-Kindness, Common Humanity, and Mindfulness) than those in the control condition.

Table 4.

Items to Assess Self-Compassion Manipulation

Manipulation check items	Self-compassion		Control		t	p	Cohen’s d
Manipulation check items	M	SD	M	SD	t	p	Cohen’s d
Self-kindness	2.94	0.79	2.35	1.11	2.62	.01	0.62
Common humanity	3.11	0.95	2.03	1.28	4.12	<.001	0.97
Mindfulness	3.14	0.76	2.17	1.23	4.03	<.001	0.58

Intervention: The Effect of Self-Compassion on Shame and Willingness to Communicate

A one-way multivariate analysis of covariance (MANCOVA) was used to compare differences in the self-compassion and control conditions on post-intervention State Shame and Willingness to Communicate. Baseline state shame (pre-intervention) and depression were added to the model as covariates. We used the Bonferroni adjustment for multiple comparisons.

The overall MANCOVA showed significant multivariate effects for pre-intervention shame (Pillai’s trace = .52, F (2, 68) = 37.11, p < .001, partial η² = .52) and condition (Pillai’s trace = .22, F (2, 68) = 9.50, p < .001, partial η² = .22), but not for depression (Pillai’s trace = .04, F (2, 68) = 1.45, p = .24, partial η² = .04). At the univariate level, there was a significant effect for post-intervention shame (F (1, 69) = 17.05, p < .001, partial η² = .20), such that those in the self-compassion condition (M = 9.32, SD = 7.06) reported significantly less shame than those in the control condition (M = 13.60, SD = 7.03; p < .001). Unexpectedly, willingness to communicate did not significantly differ between conditions at the univariate level (F (1, 69) = 0.05, p = .82, partial η² = .001).

Moderators of Intervention

Hayes PROCESS extension bundle, version 5.0 for SPSS, model 1 (Hayes, 2022), was used to examine whether each trait measure (social anxiety, post-event processing, and self-compassion) independently interacted with condition to predict post-intervention State Shame and Willingness to Communicate. A bias-corrected percentile bootstrap with 5000 samples was selected. For each of the three moderating variables (Trait Social Anxiety, Trait Post-Event Processing, and Trait Self-Compassion), two separate analyses were conducted, one with State Shame as the primary outcome and one with Willingness to Communicate as the primary outcome. In each analysis, the condition variable was binary-coded (0 = control, 1 = self-compassion), and we selected the option to mean-center components of products. We also controlled for depression in each analysis. To account for multiple moderation analyses, a Bonferroni correction was applied. The conventional alpha level (p = .05) was divided by the number of analyses conducted (k = 6), resulting in an adjusted significance threshold of p = .008.

Trait Social Anxiety

Shame

The overall model with shame as the primary outcome was significant (R = .58, R² = .33, F (4, 68) = 8.47, p < .001). There were no significant main effects, nor was the interaction between condition and trait social anxiety significant (Bonferroni correction at p < .008). All parameter estimates are shown in Table 5.

Table 5.

Moderation Coefficients for Post-intervention State Shame

Predictor variable	Coeff.	Std. Error	t	p	95% Confidence interval
Predictor variable	Coeff.	Std. Error	t	p	Lower	Upper
Constant	8.37	1.47	5.68	<.001	5.45	11.31
Condition	4.15	1.43	2.90	.01	1.30	7.01
Social anxiety (SPIN)	0.13	0.06	2.15	.04	0.01	0.26
SPIN x condition	−0.12	0.10	−1.26	.21	−0.032	0.07
Depression (BDI-II)	0.18	0.07	2.40	.02	0.03	0.32

Note. The Bonferroni correction was applied at p = .008. Condition was coded such that 1 = Self-Compassion and 0 = Control. Coeff. = coefficient. Std. Error = Standard Error. SPIN = Social Phobia Inventory. BDI-II = Beck Depression Inventory – II.

Willingness to Communicate

The model with willingness to communicate as the primary outcome was significant (R = .58, R² = .33, F (4, 68) = 8.53, p < .001). There was a significant main effect of trait social anxiety, such that higher social anxiety predicted less willingness to communicate, but all other main effects and the interaction term were non-significant (Bonferroni correction at p < .008). All parameter estimates are shown in Table 6.

Table 6.

Moderation Coefficients for Post-intervention Willingness to Communicate

Predictor variable	Coeff.	Std. Error	t	p	95% Confidence interval
Predictor variable	Coeff.	Std. Error	t	p	Lower	Upper
Constant	53.05	4.54	11.68	<.001	43.99	62.11
Condition	2.17	4.41	0.49	.62	−6.63	10.98
Social anxiety (SPIN)	−0.95	0.19	−4.99	<.001	−1.34	−0.57
SPIN x condition	−0.59	0.31	−1.91	.06	−1.20	0.03
Depression (BDI-II)	0.25	0.23	1.09	.28	−0.20	0.70

Other Moderating Variables

In separate analyses, following the same procedures described above, trait PEP, as well as trait Self-Compassion, were independently assessed as moderating variables for the effect of condition (self-compassion vs control conditions) on State Shame and on Willingness to Communicate. Unexpectedly, the interaction term was non-significant in each of these analyses (all ps > .12)

Discussion

The findings from the present study add to the growing body of research, providing evidence that self-compassion serves a well-being factor. In this study, we found that a very brief self-compassion intervention was effective in reducing feelings of shame during the post-event period in comparison to a control. Unexpectedly, however, the self-compassion intervention did not impact willingness to communicate in hypothetical, future-oriented scenarios. Furthermore, trait levels of social anxiety, PEP, nor self-compassion interacted with condition in predicting each of these outcomes (shame and willingness to communicate in future settings). These findings are interpreted and discussed in detail below.

As an initial step in our study, we provided evidence that shame correlates with social anxiety and PEP, and that shame is a strong predictor of PEP. As expected, trait and state shame were positively correlated with trait social anxiety, as well as trait and state PEP. Moreover, we showed that state levels of shame experienced during an anxiety-inducing social situation predicted post-event processing, even when controlling for trait social anxiety, trait PEP, trait depression, and situational state anxiety. Nonetheless, these variables combined predicted nearly two-thirds of the variance in PEP. This finding shows the importance of feelings of shame during the post-event period, and is similar to findings reported by Cândea and Szentágotai-Tătar (2018).

Given the importance of shame in predicting PEP, we examined whether a self-compassion intervention would be effective in reducing shame during the post-event period. As expected, those assigned to an ultra-brief, five-minute self-compassion intervention reported significantly lower levels of shame surrounding an anxiety-inducing social situation during the post-event period in comparison to control. Given the role of shame in PEP, it is important to understand whether interventions aimed at decreasing PEP can effectively target factors that may contribute to the development and maintenance of PEP. This finding builds upon Blackie and Kocovski (2018b), who found that a self-compassion intervention reduced levels of PEP. An interesting next step may be to examine whether reductions in shame serve as a mechanism through which self-compassion exerts its effects on PEP. With respect to willingness to communicate, the self-compassion intervention did not show evidence of increasing this variable. This contradicts Blackie and Kocovski (2018b), who found that a ten-minute self-compassion intervention led to increased willingness to communicate in a sample of individuals with elevated social anxiety. However, it is possible that the five-minute intervention employed in our non-selected, healthy student sample was insufficient to produce an effect. Blackie and Kocovski (2018b) employed a longer intervention and pre-selected participants based on elevated anxiety scores. The discrepant finding between studies may be understandable given these differences.

An additional aim of the present research was to examine whether trait levels of social anxiety, trait PEP, or trait self-compassion would (independently) moderate the effect of self-compassion on shame and willingness to communicate in future social situations. It was hypothesized that those with higher levels of social anxiety and trait PEP, and lower levels of trait self-compassion, would benefit most from the self-compassion intervention. On the other hand, those with lower social anxiety and PEP, and higher self-compassion would not benefit from the intervention, as they would have little room for improvement in their shame and willingness to communicate. Unexpectedly, none of these variables significantly interacted with condition (self-compassion versus control) in predicting either shame or willingness to communicate. These findings are consistent with Burnham and Kocovski (2024), who also found that trait social anxiety did not influence the effect of self-compassion on feelings of distress during the post-event period. This may suggest that the self-compassion intervention was equally effective, regardless of levels of social anxiety, baseline PEP, or self-compassion. It is possible that because all participants recalled a difficult social situation, baseline social anxiety, PEP, and self-compassion do not play a role. In other words, the situation was difficult for everyone, not just those with heightened social anxiety and PEP, and lowered self-compassion. In this regard, it would be important for future research to expose all participants to a consistent situation, rather than having all participants, with varying levels of social anxiety, recall their most difficult situation. Finally, given that the intervention was not successful in increasing willingness to communicate, it seems reasonable that none of these baseline variables altered the non-significant effects.

Although the findings from the present study are encouraging, it is important to note several limitations. One limitation was the inclusion of a student sample who was not pre-selected based on social anxiety scores or a clinical diagnosis. Although those with the highest levels of social anxiety would likely benefit most from self-compassion interventions, it is possible that self-compassion is difficult to induce when anxiety becomes too severe. Therefore, our intervention may not be as effective amongst those who would benefit from it most. Although we assessed trait social anxiety as a moderator, those with higher social anxiety (+1SD) may not have been as severe as those at diagnostic levels. Thus, it would be important to replicate this study using a sample of individuals diagnosed with social anxiety disorder to determine the effectiveness of the intervention for those with more severe social anxiety. An additional limitation in the present study was that participants recalled an anxiety-inducing social situation, rather than having all participants exposed to the same situation. Despite this, however, prior to the manipulation, participants in both conditions reported non-significantly different levels of state anxiety during the situation, as well as feelings of shame and post-event processing. Another limitation in the present study relates to the control condition. Given that participants in the control condition were asked to sit and wait, it is possible that some may have been distracted with other tasks. Past research has shown that distraction leads to less-post-event processing (e.g., Blackie & Kocovski, 2016). Despite this, those in the self-compassion condition reported less shame than those in the control, even if they were distracted, rather than sitting quietly. An additional limitation is that 61% of the sample identified as female, and an equal balance of gender identity would make generalization of findings more accurate. However, it is notable that our sample was fairly diverse in terms of ethnicity, which represents a strength in the present study. A final limitation is the relatively small sample size, which may have limited the power to detect effects.

Given the findings from the present study, future research should continue to examine shame in post-event processing and social anxiety. Shame may be an important factor that contributes to the development and maintenance of PEP and social anxiety. This should be experimentally explored in further studies to determine whether a causal link is present. Research should also continue to explore interventions for shame in post-event processing to help limit this negative and repetitive thinking. Currently, researchers have been comparing self-compassion to other active interventions. For instance, Burnham and Kocovski (2024) and Stevenson et al. (2019) compared self-compassion to cognitive restructuring for social anxiety and distress following social situations and found that the interventions did not significantly differ from one another. Future research should continue to examine self-compassion to other active interventions and examine under what conditions and for whom certain interventions may be most effective. Researchers may also examine whether shame serves as a mechanism of change to understand how interventions exert their effects.

Taken together, the findings from the present study highlight the importance of shame in post-event processing. Therefore, it may be important to target this variable in therapeutic settings through the use of self-compassionate exercises. Our findings also demonstrate that the effect of self-compassion on shame did not vary based on social anxiety severity (among other baseline variables). Therefore, it may be important for self-compassion interventions to target shame amongst all individuals struggling with social anxiety and post-event processing.

Footnotes

Ethical Considerations

This study was reviewed and approved by the Cape Breton University Research Ethics Board (protocol number: 2022083) on November 17, 2022.

Consent to Participate

Respondents gave written consent for review and signature before participation for partial course credit (one bonus point).

Consent for Publication

Informed consent for publication was provided.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a Start-Up Research Grant from Cape Breton University, which was awarded to the first author.

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.

Data Availability Statement

The data set from the present study is available at: https://doi.org/10.6084/m9.figshare.29717873.v1 (Blackie & Pelley, 2025).

Author Biographies

Rebecca A. Blackie is an Assistant Professor at Cape Breton University, Nova Scotia, Canada. Her research focuses on social anxiety, post-event processing, and self-compassion. She has secondary research interests in psychometry, meditation, general anxiety, and wellbeing.

Ainsley B. Pelley graduated with an Honours degree in Psychology from Cape Breton University. Her research interests focus on meditation, wellbeing, and anxiety. She is also a candidate for the Bachelor of Social Work program at Cape Breton University.

References

American Psychiatric Association . (2022). Diagnostic and statistical manual of mental disorders. (5th ed., text rev.). https://doi.org/10.1176/appi.books.9780890425787

Beck

A. T.

Steer

R. A.

Brown

G. K.

(1996). Manual for the beck depression inventory–II. Psychological Corporation.

Blackie

R. A.

Kocovski

N. L.

(2016). Letting go of yesterday: Effect of distraction on post-event processing and anticipatory anxiety in a socially anxious sample. Cognitive Behaviour Therapy, 45(1), 60–72. https://doi.org/10.1080/16506073.2015.1104383

Blackie

R. A.

Kocovski

N. L.

(2017). Development and validation of the trait and state versions of the post-event processing inventory. Anxiety, Stress & Coping, 30(2), 202–218. https://doi.org/10.1080/10615806.2016.1230668

Blackie

R. A.

Kocovski

N. L.

(2018a). Examining the relationships among self-compassion, social anxiety, and post-event processing. Psychological Reports, 121(4), 669–689. https://doi.org/10.1177/0033294117740138

Blackie

R. A.

Kocovski

N. L.

(2018b). Forgive and let go: Effect of self-compassion on post-event processing in social anxiety. Mindfulness, 9(2), 654–663. https://doi.org/10.1007/s12671-017-0808-9

Blackie

R. A.

Kocovski

N. L.

(2019). Confirmatory factor analysis of the Post-Event Processing Inventory in a community sample seeking self-help for social anxiety. Anxiety, Stress & Coping, 32(2), 196–201. https://doi.org/10.1080/10615806.2019.1571194

Blackie

R. A.

Pelley

A. B.

(2025). Self-compassion and shame in PEP FINAL.sav [Dataset]. https://doi.org/10.6084/m9.figshare.29717873.v1

Brozovich

Heimberg

R. G.

(2008). An analysis of post-event processing in social anxiety disorder. Clinical Psychology Review, 28(6), 891–903. https://doi.org/10.1016/j.cpr.2008.01.002

10.

Brozovich

Heimberg

R. G.

(2011). The relationship of post-event processing to self-evaluation of performance in social anxiety. Behavior Therapy, 42(2), 224–235. https://doi.org/10.1016/j.beth.2010.08.005

11.

Burnham

C. M.

Kocovski

N. L.

(2024). Coping with past social stress: Comparing brief self-compassion, cognitive reappraisal and control writing exercises. Mindfulness, 15(1), 203–216. https://doi.org/10.1007/s12671-023-02268-4

12.

Campbell

C. G.

Bierman

K. L.

Molenaar

C. M. P.

(2016). Individual day-to-day process of social anxiety in vulnerable college students. Applied Developmental Science, 20(1), 1–15. https://doi.org/10.1080/10888691.2015.1026594

13.

Cândea

D. M.

Szentágotai-Tătar

(2018). Shame as a predictor of post-event rumination in social anxiety. Cognition & Emotion, 31(8), 1684–1691. https://doi.org/10.1080/02699931.2016.1243518

14.

Clark

D. M.

(2001). A cognitive perspective on social phobia. In Crozier

W. R.

Alden

L. E.

(Eds.), International handbook of social anxiety: Concepts, research and interventions relating to the self and shyness (pp. 405–430): John Wiley.

15.

Clark

D. M.

Wells

(1995). A cognitive model of social phobia. In Heimberg

Liebowitz

M. R.

Hope

D. A.

Schneier

F. R.

(Eds.), Social phobia: Diagnosis, assessment and treatment. Guilford Press.

16.

Connor

K. M.

Davidson

J. R.

Churchill

L. E.

Sherwood

Foa

Weisler

R. H.

(2000). Psychometric properties of the Social Phobia Inventory (SPIN): New self-rating scale. The British Journal of Psychiatry, 176(4), 379–386. https://doi.org/10.1192/bjp.176.4.379

17.

Dannahy

Stopa

(2007). Post-event processing in social anxiety. Behaviour Research and Therapy, 45(6), 1207–1219. https://doi.org/10.1016/j.brat.2006.08.017

18.

Donohue

H. E.

Modini

Abbott

M. J.

(2024). Psychological interventions for pre-event and post-event rumination in social anxiety: A systematic review and meta-analysis. Journal of Anxiety Disorders, 102, 102823. https://doi.org/10.1016/j.janxdis.2023.102823

19.

Ferreira

Moura-Ramos

Matos

Galhardo

(2022). A new measure to assess external and internal shame: Development, factor structure, and psychometric properties of the External and Internal Shame Scale. Current Psychology: A Journal for Diverse Perspectives on Diverse Psychological Issues, 41(4), 1892–1901. https://doi.org/10.1007/s12144-020-00709-0

20.

Gavric

Cameron

Waechter

Moscovitch

D. A.

McCabe

R. E.

Rowa

(2023). Just do something: An experimental investigation of brief interventions for reducing the negative impact of post-event processing in social anxiety disorder. Jornal of Anxiety Disorders, 98, 102744. https://doi.org/10.1016/j.janxdis.2023.102744

21.

Gavric

Moscovitch

D. A.

Rowa

McCabe

R. E.

(2017). Post-event processing in social anxiety disorder: Examining the mediating roles of positive meta-cognitive beliefs and perceptions of performance. Behaviour Research and Therapy, 91, 1–12. https://doi.org/10.1016/j.brat.2017.01.002

22.

Gharraee

Zahedi Tajrishi

Ramezani Farani

Bolhari

Farahani

(2018). A randomized controlled trial of compassion focused therapy for social anxiety disorder. Iranian Journal of Psychiatry and Behavioral Sciences, 12(4), e80945. https://doi.org/10.5812/ijpbs.80945

23.

Hayes

A. F.

(2022). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (3rd ed.). The Guilford Press. 0.5.7, 0.5.14.

24.

Hedman

Ström

Stünkel

Mörtberg

(2013). Shame and guilt in social anxiety disorder: Effects of cognitive behavior therapy and association with social anxiety and depressive symptoms. PLoS One, 8(4), e61713. https://doi.org/10.1371/journal.pone.0061713

25.

Hofmann

S. G.

(2007). Cognitive factors that maintain Social Anxiety Disorder: A comprehensive model and its treatment implications. Cognitive Behaviour Therapy, 36(4), 193–209. https://doi.org/10.1080/16506070701421313

26.

Kabat-Zinn

(1994). Wherever you go, there you are: Mindfulness meditation in everyday life: Hyperion.

27.

Kashdan

T. B.

Roberts

J. E.

(2007). Social anxiety, depressive symptoms, and post-event rumination: Affective consequences and social contextual influences. Journal of Anxiety Disorders, 21(3), 284–301. https://doi.org/10.1016/j.janxdis.2006.05.009

28.

Kim

Yang

H. Y.

Ahn

J. K.

(2022). The relationship between three sub-types of safety behaviours and social anxiety: Serial mediating effects of state and trait post-event processing. Frontiers in Psychology, 13, 987426. https://doi.org/10.3389/fpsyg.2022.987426

29.

McCroskey

J. C.

(1992). Reliability and validity of the Willingness to Communicate Scale. Communication Quarterly, 40(1), 16–25. https://doi.org/10.1080/01463379209369817

30.

Neff

K. D.

(2003). Self-compassion: An alternative conceptualization of a healthy attitude toward oneself. Self and Identity, 2(2), 85–101. https://doi.org/10.1080/15298860309032

31.

Raes

Pommier

Neff

K. D.

Van Gucht

(2011). Construction and factorial validation of a short form of the self-compassion scale. Clinical Psychology & Psychotherapy, 18(3), 250–255. https://doi.org/10.1002/cpp.702

32.

Rapee

R. M.

Heimberg

R. G.

(1997). A cognitive-behavioral model of anxiety in social phobia. Behaviour Research and Therapy, 35(8), 741–756. https://doi.org/10.1016/S0005-7967(97)00022-3

33.

Robins

R. W.

Hendin

Trzesniewski

K. H.

(2001). Measuring global self-esteem: Construct validation of a single-item measure and the Rosenberg Self-Esteem Scale. Personality and Social Psychology Bulletin, 27(2), 151–161. https://doi.org/10.1177/0146167201272002

34.

Rosenberg

(1965). Society and the adolescent self-image. Princeton University Press.

35.

Rowa

Antony

M. M.

Swinson

R. P.

McCabe

R. E.

(2014). A preliminary attempt to experimentally induce post-event processing in social anxiety disorder. Behaviour and Cognitive Psychotherapy, 42(2), 238–242. https://doi.org/10.1017/S1352465813000143

36.

Shikatani

Antony

M. M.

Kuo

J. R.

Cassin

S. E.

(2014). The impact of cognitive restructuring and mindfulness strategies on post-event processing and affect in social anxiety disorder. Journal of Anxiety Disorders, 28(6), 570–579. https://doi.org/10.1016/j.janxdis.2014.05.012

37.

Stevenson

Mattiske

J. K.

Nixon

R. D. V.

(2019). The effect of a brief online self-compassion versus cognitive restructuring intervention on trait social anxiety. Behaviour Research and Therapy, 123(4), 103492. https://doi.org/10.1016/j.brat.2019.103492

38.

Swee

M. B.

Hudson

C. C.

Heimberg

R. G.

(2021). Examining the relationship between shame and social anxiety disorder: A systematic review. Clinical Psychology Review, 90(4), 102088. https://doi.org/10.1016/j.cpr.2021.102088

39.

Szczyglowski

K. A.

Kocovski

N. L.

(2024). To avoid or not to avoid: Impact of self-compassion on safety behaviors in social anxiety. Anxiety, Stress & Coping: International Journal, 37(5), 587–601. https://doi.org/10.1080/10615806.2024.2307466

40.

Tangey

J. P.

Stuewig

Martinez

A. G.

(2014). Two faces of shame: The roles of shame and guilt in predicting recidivism. Psychological Science, 25(3), 799–805. https://doi.org/10.1177/0956797613508790

41.

Teale Sapach

M. J. N.

Carleton

R. N.

(2023). Self-compassion training for individuals with social anxiety disorder: A preliminary randomized controlled trial. Cognitive Behaviour Therapy, 52(1), 18–37. https://doi.org/10.1080/16506073.2022.2130820

42.

Tracy

J. L.

Robins

R. W.

(2004). Putting the self into self-conscious emotions: A theoretical model. Psychological Inquiry, 15(2), 103–125. https://doi.org/10.1207/s15327965pli1502_0

43.

Werner

K. H.

Jazaieri

Goldin

P. R.

Ziv

Heimberg

R. G.

Gross

J. J.

(2012). Self-compassion and social anxiety disorder. Anxiety Stress Coping, 25(5), 543–558. https://doi.org/10.1080/10615806.2011.608842

44.

Wolpe

(1969). The practice of behavior therapy: Pergamon Press.