The Sensitivity to Threat and Affiliative Reward Scale (STARS) in Children and Adults

• In Event: Poster Session 13
  In Poster Session: PS 13 Topic - Developmental Psychopathology

Abstract

Background: Callous-Unemotional (CU) traits (i.e., lack of empathy, prosociality, and guilt) identify children at risk for severe antisocial behavior and aggression (Frick et al., 2014). CU traits are conceptualized as a developmental precursor of the affective features of psychopathy, which form a distinct factor from harmful lifestyle and antisocial behaviors (Frick & Marsee, 2018) A recent theoretical model proposed that CU traits arise from deficits in two psychobiological dimensions of temperament: sensitivity to threat (i.e., fear in response to social and nonsocial cues of threat) and affiliative reward (i.e., motivation to seek out and get pleasure from social and physical closeness with others) (Waller & Wagner, 2019). To assess variability in these dimensions across the lifespan, we developed the parent- and self-reported Sensitivity to Threat and Affiliative Reward Scale (STARS). Methods: In the current study, we tested the psychometric properties, factor structure, and construct validity of the STARS in three independent samples. Sample 1 included children aged 3-10 years old (N=303, 50.5% female), Sample 2 was children aged 7-10 years old (N=53, 54.7% female), and Sample 3 (N=314, 48.7% female) was adults aged 19-73 years old. To establish the most parsimonious set of items that showed homotopic continuity for both sensitivity to threat and affiliation across the lifespan, we conducted separate item refinement analyses for children using the parent-reported STARS and for adults using the self-reported STARS. Results: In Samples 1 and 3 we used Differential Item Functioning (DIF) and Item Response Theory (IRT) techniques to identify the best-performing set of items from an initial 64-item pool. In children, DIF and IRT analyses of the parent-reported STARS items resulted in a revised set of 28 items that functioned equivalently for regardless of age (early vs. middle-to-late childhood) and gender (Table 1). In adults, the DIF and IRT analyses resulted in partially overlapping set of 40 items that functioned equivalently regardless of gender or race (Table 2). In both samples, confirmatory Factor Analysis (CFA) indicated adequate fit for a 2-factor solution consistent with the theorized separate dimensions of threat sensitivity and sensitivity to affiliation (Waller & Wagner, 2019). In Sample 1 the STARS dimensions showed convergent validity in relation to parent-reported behavioral inhibition and activation and construct validity in relations to CU traits, such that reduced sensitivity to threat (β=-.15, p<.01) and reduced sensitivity to affiliation (β=-.39, p<.001) were independently related to higher CU traits in 3-10-year-old children. Likewise, in Sample 2, both reduced sensitivity to threat (β=-.43, p<.05) and reduced sensitivity to affiliation (β=-.43, p<.05) were uniquely related to higher CU traits in 3-10-year-old children. Finally, in Sample 3, reduced sensitivity to affiliation among adults was uniquely related to the affective facet of psychopathy (β=-.08, p<.001), but not to other features of psychopathy. Discussion: Findings from three independent samples support the psychometric and structural properties of the parent- and self-reported STARS to assess variability in sensitivity to threat and affiliation, with implications for understanding risk for CU traits and the affective features of psychopathy across the lifespan.

Authors

• Samantha Conte Perlstein, University of Pennsylvania
  Presenting Author

• Nicholas James Wagner, Boston University
  Non-Presenting Author

• Rebecca Waller, University of Pennsylvania
  Non-Presenting Author