A novel bidirectional test of emotion socialization among high-risk families

• In Event: Social and Emotional Development Within the Family

Abstract

According to the parent emotion socialization (ES) model (Eisenberg et al., 1998), parents’ reactions to their children’s emotions influence children’s socioemotional and behavioral competence via children’s expression and regulation of emotion (Eisenberg et al., 1998). Despite the ES model’s theorizing of the immediate outcomes of parent ES, most studies of parent ES focus on distal, longer-term outcomes (Cui et al., 2022; Eisenberg, 2020). Examining the proximal effects of parent ES, while accounting for bidirectional exchanges between parent and child, is necessary to advance understanding of parent ES in real-time interactions. Further, ES models are traditionally evaluated among community samples, despite the potential of supportive ES to promote resilience even among youth exposed to early caregiving adversity (Milojevich et al., 2020).
Thus, our aims were two-fold. Among a high-risk sample enriched for early life adversity exposure, we evaluated the effects of parent and child ES during a challenging task on their own and each other’s subsequent positive and negative affect during a dyadic stress recovery task. Specifically, we hypothesized that (1a) parents’ supportiveness would positively predict subsequent child positive affect and negatively predict subsequent child negative affect, whereas (1b) parents’ nonsupportiveness would negatively predict subsequent child positive affect and positively predict subsequent child negative affect. Similarly, to test the hypothesized bidirectional ES model, we simultaneously evaluated the effects of child ES on parent affect.
We evaluated these hypotheses among a sample of 161 parent-child dyads (child ages 6 – 16 years; 48.4% female) that was enriched for exposure to early life adversity (50% were previously in institutional care or foster care prior to being adopted). All parent-child dyads participated in a 5-min conflict discussion task followed by a 5-min pleasant event activity planning task, a procedure designed to elicit dyadic stress and subsequent recovery. Trained coders rated parent and child supportiveness, nonsupportiveness (pressuring the other to agree with them), and the intensity and frequency of negative and positive affect using the FIMS (Holmbeck, 2007). Our hypotheses were evaluated using actor-partner interdependence models (Cook & Kenny, 2005) in a structural equation modeling framework, which enables estimation of interpersonal partner effects, after adjusting for the intrapersonal actor effects and concurrent covariation in parent and child behavior. Models also adjusted for potential confounds, including child age, biological sex, parent ethnicity, early life adversity exposure, and mode of data collection (in the laboratory v. remote).
Results offered mixed support for our hypotheses (Tables 1 & 2). Though there was no evidence that parent ES predicted child affect, there was some evidence of child-driven ES. Specifically, child nonsupportiveness positively predicted parent negative affect, Beta = 0.173, SE Est = 0.073, p = .018. The covariances between parent and child behavior during each task were significant in all models.
The present study adds to the limited designs in the ES literature that control for stability in individual behavior and test for bidirectional effects (e.g., except Cui et al., 2022). Results challenge dominant models of parent ES behaviors as applied in high-risk samples, highlighting instead child-driven effects.

Authors

• Jennifer Anne Somers, University of California - Los Angeles
  Presenting Author

• Kristen A. Chu, University of California - Los Angeles

• Fran Querdasi

• Naomi Gancz, UCLA

• Emily Towner, Cambridge University

• Bridget Callaghan, The University of California, Los Angeles