Beyond trait models of self-criticism and self-compassion: Variability over domains and the search for signatures

Abstract

Research on risk and resilience variables such as self-criticism and self-compassion typically assumes that they are broad traits reflecting highly cross-situationally consistent dispositions. However, contemporary personality science conceptualizes contextual variability as a universal feature of personality characteristics, complementary to but not inconsistent with central tendencies, or traits (Fleeson & Jayawickreme 2015; Moskowitz & Fournier 2015). We adopted the person × situation interaction model of personality, distinguishing person differences in mean levels, normative effects of situations, and person × situation interactions, termed signatures. However, rather than sampling external situations, we modified measures of self-criticism, self-reassurance, and self-compassion to assess them in eight domains of the self. Study 1 estimated person, domain, and person × domain variance components in four young adult samples. Differences between persons (traits) accounted for considerable variance, ranging from 29% (self-compassion) to 46.5% (self-reassurance), and domain differences accounted for relatively small variance components. Signatures accounted for substantial additional variance, ranging from 13.7% (self-compassion) to 26% (self-criticism). In Study 2, multilevel modeling demonstrated that person, domain, and signatures were each significantly related to domain-specific negative and positive affect. These findings highlight the importance of expanding risk and resilience studies beyond traits to include normative domain effects and person-specific signatures.