Self-discrepancy, Depression, Anxiety, and Psychological Well-Being: The Role of Affective Style and Self-efficacy

Maladaptive emotion regulation strategies have been identified as transdiagnostic key variable across several mental disorders (Aldao et al., 2010; Cludius et al., 2020; Gross et al., 2019). The above-mentioned meta-analysis called for an investigation of different aspects of emotion regulation as important variables implicated in the relationship between self-discrepancies and psychopathology (Mason et al., 2019). Indeed, some recent studies found selective moderation effects for actual:ought discrepancies and emotion regulation on anxiety symptoms (Gürcan-Yıldırım & Gençöz, 2022) and selective mediation effects of rumination on the relationship between actual:ideal discrepancies and depressive and anxious symptoms (Dickson et al., 2019). Altogether, research demonstrates that self-discrepancies may lead to unfolding negative emotional cascades that may require emotional regulation to maintain psychosocial functioning (Mason et al., 2019). In this regard, emotion regulation may play a mediating and/or moderating role in the relationship between self-discrepancies and psychopathology. Prior research has identified several styles of affect regulation, including concealing or avoiding emotions, readjusting or balancing emotions, and tolerating or accepting emotions (Hofmann et al., 2012).

Concealing describes a response-focused suppressive regulative strategy to conceal and avoid negative emotions (Hofmann & Kashdan, 2010; Totzeck et al., 2018). In line with regulatory focus theory (Higgins, 1997), concealing aligns with prevention-focused self-regulation as opposed to promotion-focused self-regulation. Recall that the former is defined as a self-regulation avoidant strategy that is sensitive to negative outcomes and associated with anxious agitation, whereas the latter is defined as a self-regulation strategy that is reactive to the absence of positive outcomes and therefore associated with dejection-related emotions, such as depression or dissatisfaction. Consequently, actual:ought discrepancies may be especially prone to initiate concealing when individuals aim to avoid acknowledging failures of meeting duties, responsibilities, or obligations. This would align with the initial theoretical framework in which actual:ought discrepancies were conceptualized to evoke anxious emotions (Higgins, 1987). Empirical evidence indicates that individuals with anxiety disorders engage in more expressive suppression than individuals with affective disorders (D’Avanzato et al., 2013). However, in a clinical sample there were no differences between the concealing affective style when comparing individuals with anxiety and mood disorders (Totzeck et al., 2018).

Adjusting on the other hand describes the capability to modulate unfolding negative emotions and adjust according to the situational demands to maintain psychosocial functioning (Hofmann & Kashdan, 2010; Totzeck et al., 2018). Previous research found evidence that patients suffering from depressive disorder use reappraisal strategies as an adjusting strategy less frequently compared to patients suffering from anxiety disorders (D’Avanzato et al., 2013). Likewise, patients with affective disorders reported less affective adjusting than patients with anxiety disorders (Totzeck et al., 2018). Conflicting cognitive representations of desired versions of the self and the actual self should lead to emotional distress. This, in turn, should lead to a greater self-focus and greater difficulties adjusting to the unfolding emotional distress, which may contribute to the development or maintenance of depression or anxiety. Also, when individuals believe that they are too discrepant from their desired self, this may be especially problematic for individuals with a lower habitual tendency to respond with adjusting when faced with arising negative emotions (Hofmann et al., 2012). This may be because adjusting constitutes an antecedent-focused strategy that is applied in the early phase of the emotion regulation process (Hofmann et al., 2012). Thus, a combination of perceived high discrepancy and low levels of adjusting may be associated with high levels of depression or anxiety.

Tolerating is an acceptance-based affective style (Hofmann et al., 2012). Given that the ideal self and the ought self are desirable states, it may be difficult for individuals with high levels of self-discrepancy to maintain a tolerating or accepting attitude towards negative emotions resulting from these discrepancies. Both types of discrepancy may lead to a greater self-focus, which may instigate difficulties to maintain a tolerating and accepting attitude towards one’s emotions (Dickson et al., 2019). Therefore, high levels of discrepancy may be associated with lower levels of tolerating, which, in turn, may increase emotional vulnerability. In a previous study, mindfulness mediation helped reduce the negative effects of actual:ideal discrepancies (Ivtzan et al., 2011). In this vein, tolerating may also be a protective factor, which could be reflected in a moderation effect. This aligns with findings that depressed and anxious people encounter more emotion regulation difficulties (Hofmann & Kashdan, 2010). Accordingly, a combination of greater perceived discrepancy and low levels of tolerating arousing negative emotions may be associated with high levels of depression and anxiety.

Self-efficacy, defined as individuals’ belief in their own capacity to exercise control over their functioning and over events that affect their life (Bandura, 1977), may also play a significant role in the relationship between self-discrepancy and psychopathology. Low self-efficacy is associated with both depressive (Sawatzky et al., 2012) and anxiety symptoms (Goldin et al., 2009), and mediates the effect of daily stress on mental health outcomes (Schönfeld et al., 2019). The nature of self-efficacy is future-orientated and goal directed (Bandura, 1977). Accordingly, self-efficacy may be severely undermined by self-discrepancies when individuals feel that they cannot operate in the world as they wish or feel obligated to. In this vein, discrepancies may be associated with a lower sense of agency, which in turn, can be debilitating for subsequent mental health. Specifically, self-efficacy determines initial decision making on whether to perform a specific behavior (Bandura, 1977). When individuals, however, believe that they cannot operate effectively enough as abstractly evinced in self-discrepancies, this may lead to passivity or avoidance resulting from a lower sense of self-efficacy. This, in turn, could contribute to depression and anxiety symptoms that both have been found to be associated with self-efficacy (Goldin et al., 2009; Sawatzky et al., 2012). Indeed, a previous study reported that a lower sense of agency explained the relationship between both actual:ideal and actual:ought discrepancies and purpose in life (Stanley & Burrow, 2015).

In the present study, we examined the relationship of actual:ideal as well as actual:ought discrepancies with depression, anxiety, and psychological well-being. To elucidate more granulated aspects of these relationships, we further investigated whether affective styles (concealing, adjusting, and tolerating) and self-efficacy are implicated in these relationships as mediator or moderator variables. To define self-discrepancies in the present study, we focus on perceived discrepancies in the Big Five (extraversion, agreeableness, conscientiousness, emotional stability, and openness to experience) taxonomy of personality (Rammstedt & John, 2007). This framework was chosen because it operationalizes the core parts of individual’s personality, and discrepancies perceived in these relatively stable traits may be especially linked to mental health problems (Hayes & Joseph, 2003). Based on prior literature (Mason et al., 2019), we expected that both types of discrepancies would be associated with depression, anxiety, and well-being. We tested the affective styles (concealing, adjusting, and tolerating) competitively in a pathway model to examine whether they uniquely account for differential effects of actual:ideal or actual:ought discrepancies and our outcome variables. In addition, we tested moderation models to gauge potential effects of these variables beyond their main effects. Concerning our pathway models and the moderation models, we had no specific a priori hypotheses.

The link to our study was disseminated via online panel provider Prolific Researcher (Palan & Schitter, 2018). The survey was open to all panel members fluent in English above 17 years of age. In total, N = 596 healthy individuals participated in the current study. Mean age was 24.83 (SD = 7.15) years and the majority of the participants were female (74%, n = 441). Of the participants, n = 81 indicated to have a graduate degree, n = 158 had a bachelor’s degree, n = 18 had an associate degree (above high school education but below a bachelor’s degree), n = 126 had some college education but no degree, n = 204 had a high school (or equivalent) degree, and n = 9 had no high school degree. Most participants (n = 466) were single or never married followed by being married (n = 126). The Ethics Committee of the University of Münster approved the study protocol. The survey material, the anonymized data, and R analysis code can be found in the OSF supplement at https://​osf.​io/​c8d6j/​?​view_​only=​251c110274964d72​927a9cbca669c7c7​.

Analyses were conducted in R (R Core Team, 2020). First, we calculated descriptive statistics and intercorrelations among all constructs. Further, we tested potential gender differences on all constructs. We also provide 95% confidence intervals around the Pearson correlation coefficients to compare the associations of actual:ideal and actual:ought discrepancies. In addition, we calculated total scores of the actual personality traits and correlated them with the other variables to have an indication of whether actual:ideal and actual:ought discrepancies are confounded with the actual personality values. Then, we tested the association of actual:ideal and actual:ought with the outcomes again while adjusting for actual personality scores to examine whether they still predict unique variance in the outcome variables. To further examine the relationship between self-discrepancy and depression, anxiety, and psychological well-being, we tested a pathway model that incorporated concealing, adjusting, tolerating, and self-efficacy as potential mediators of this relationship. To test differential effects of the different discrepancies on the outcomes, we tested these models separately for actual:ideal and actual:ought discrepancies, and for the three outcomes (depression, anxiety, and psychological well-being). Accordingly, we had six different single step multiple mediator models that we estimated with the lavaan package in R (Rosseel, 2012). To test whether indirect effects in our model were statistically significantly different from zero, we generated nonparametric confidence intervals using a bootstrap resampling procedure (Hayes, 2015). We generated 10,000 bootstrap resamples to estimate 95% confidence intervals (CIs) for the assessment of the indirect effect (Hayes, 2015). Given that gender was unequally distributed in our sample and gender differences were expected in our sample, we adjusted for gender in our pathway models. Last, we tested moderation effects of actual:ideal and actual:ought discrepancies with concealing, tolerating, adjusting, and self-efficacy on depression, anxiety, and well-being.

Table 1 depicts descriptive statistics of all constructs along with descriptive statistics separately for males and females. Females reported higher actual:ideal and actual:ought discrepancies. They further reported more depressive and anxiety symptoms and lower values on psychological well-being, concealing, adjusting, tolerating, and self-efficacy.

Actual:ideal and actual:ought discrepancies were highly interrelated, p < .001. (Table 2). Both discrepancies were significantly positively associated with depressive and anxiety symptoms, and negatively associated with well-being, the adjusting and tolerating affective styles, and self-efficacy, all ps < .001. No significant association with the concealing affective style emerged for both actual:ideal, p = .07 and actual:ought discrepancies, p = .12.

All associations were descriptively stronger for actual:ideal discrepancy associations compared to actual:ought discrepancy associations, but all confidence intervals were overlapping. However, for depression, well-being, adjusting, and self-efficacy, the actual:ought discrepancy confidence intervals did not contain the point estimates of the actual:ideal discrepancy associations.

When examining the three outcome variables, we observed that both types of discrepancies had descriptively the highest association with psychological well-being followed by depression, and then anxiety. All confidence intervals were yet overlapping, but the psychological well-being confidence intervals did not contain the anxiety point estimate.

Actual personality scores were highly related to both discrepancy scores and also significantly associated with all outcome variables, all ps < .001 (Table 2). When adjusting for actual personality scores (see Table 3 for multiple regression models), actual:ideal discrepancies were still significantly associated with depression, anxiety, well-being, the adjusting affective style, and self-efficacy. After adjusting for actual personality scores, actual:ought discrepancies were significantly associated with depression, and the affective style adjusting, while the other associations dropped below significance.

Overall, only two significant moderation effects occurred (Fig. 4). The first moderation was found for actual:ideal discrepancy and adjusting on depression, B = .07, p = .022, R² = .24. The second moderation effect emerged for the relationship between actual:ideal discrepancy and concealing with well-being as dependent variable, B = − .08, p = .006, R² = .21. All other moderation effects were non-significant, all ps ˃ .135.

Consistent with previous work, self-discrepancies emerged as transdiagnostic variables (Mason et al., 2019). Accordingly, self-discrepancies may be a viable intervention target in light of their empirical and theoretical relevance for mental health. Indeed, a reduction of self-discrepancies was associated with less depressive and anxiety symptoms in previous work (Watson et al., 2014). In therapy, it is advisable to focus on dysfunctional cognitions surrounding the fulfillment of obligations and expectations. In this regard, focusing on adjustment as emotion regulation strategy may be fruitful. Further, knowledge about actual:ideal discrepancies may help to articulate and move towards realistic and valued goals with the client. Such interventions could be further informed by findings about emotion regulation and self-efficacy, which are both also transdiagnosticially important (Bandura, 1977; Gross et al., 2019). Yet, future research first needs to systematically examine whether and how these findings can be implemented into intervention efforts. Given that our sample consisted of healthy participants, research in clinical populations is also warranted to gain more insights into pathological processes of self-discrepancies in individuals suffering from mental disorders.

The cross-sectional study design does not allow for any causal interpretation. While our models were based on theoretical considerations, it is also feasible that the directionality may be different. For instance, emotion regulation strategies can also show bidirectional relationships with emotional disorders (Aldao et al., 2010; Cludius et al., 2020). Further, other unmeasured variables may contribute to these relationships. Gender was unequally distributed as the majority of our sample was female. Although we statistically adjusted for gender in our analyses, studies with more balanced gender distributions are needed. Moreover, because of the number of statistical tests to detect moderation effects, these findings need to be interpreted with caution. In our analyses, we relied on nomothetic measures. Established measures of self-discrepancies such as the integrated self-discrepancy index use combinations of idiographic and nomothetic methods to assess self-discrepancies holistically (Hardin & Lakin, 2009). It was argued and theoretically expected that idiographic methods would elucidate the personally most relevant domains for individuals, which are prone to severe emotional reactions when discrepancies are detected. However, a meta-analysis suggests that effects of discrepancies on emotional reactions were stronger for nomothetic measures compared to idiographic measures (Mason et al., 2019). In this vein, it is also important to note that by using simple difference scores, all items contributed equally to the discrepancy scores. Further studies may remedy this limitation by asking for the subjective importance of the respective item and weighing scores accordingly or modeling discrepancies with different statistical methods (for an overview of limitations of difference scores see Shanock et al., 2010).

A further potential limitation of the self-discrepancy measure is the presentation of the ideal and the ought questions next to each other. This may have elicited response patterns that hamper participants’ ability to differentiate between the respective ideal and the ought question. The latter may have been perceived as subset of the former, and participants may have striven for consistency in their responses. This may, in part, account for the similar relationships of the two types of discrepancy. In addition, further studies could disentangle whether aspects of the ought-self relate to expectations and obligation perceived by themselves or by others. This could have relevant clinical implication as this may impact the underlying motivation to change the ought-self or to address cognitions implicated in actual:ought discrepancies. In the present study, we also focused on the core of individuals’ personality so that the attributes are likely to be highly valued and personally relevant while simultaneously capturing a wide range of attributes that may not have come up when using idiographic measures. As a result of this approach, internal consistencies of the discrepancy scales were below recommended cut-offs. However, to capture the relevant breadth of individual’s personality, the items needed to represent the construct broadly to increase their personal meaning for participants. This is in line with previous research reporting lower levels of internal consistency for short scales of personality (Rammstedt et al., 2021). It is also important to consider varying cultural backgrounds as the self is embedded in the sociocultural background of an individual (Cheung, 1997). For example, in collectivistic cultures, a higher importance may be assigned to the ought-self relative to the ideal-self compared to individualistic cultures, and vice versa (Cheung, 1997). Disentangling cultural specifics of self-discrepancies and resulting emotions in future studies may be relevant for a more thorough understanding of putative mechanisms.