Building the Evidence Base for Social-Emotional Teaching Practices in Early Childhood

• In Event: Measuring Social-Emotional Teaching and Learning in Early Childhood Classrooms

Abstract

Despite agreement among scholars and practitioners about the importance of emotional competence for children’s school and life success, little guidance exists for specific evidence-based strategies that teachers can use to support children’s emotional development. In order to fill this gap, we systematically reviewed the literature linking aspects of emotion socialization to emotional competence (EC) in young children.

We organized our review using the framework of social and emotional teaching (SET) which describes modeling, teaching, and reacting behaviors that adults can use to convey to children norms and values about emotions in their worlds (defined in Table 1). A long theoretical literature (Gottman 1996, Denham, 1998) has linked such emotion socializing practices to children’s knowledge of and abilities to express and regulate their emotions (defined in Table 2). Although studies have attempted to empirically demonstrate these expectations, a formal synthesis is necessary to inform future research and practice.

The screening procedures and coding processes for this synthesis is summarized in Figure 1. The final collection of 29 studies were coded for study characteristics (design, sampling, sample demographics) and measures were coded by SET and EC construct. Then study-reported effect sizes captured and linked to measure codes and analysis type (e.g., bivariate correlation, multivariate regression).

The studies fell into two broad sets: (a) small-scale correlational studies using convenience samples of families to examine the emotion socialization conceptual framework (n = 24); and, (b) randomized control trials building on emotion socialization concepts in curricula implemented by teachers in preschool classrooms (n = 5). The former studies primarily included white, middle-class samples, and reported regression results only for significant associations, used few controls, and failed to report standard errors, t-statistics, or confidence intervals. Using Hunter and Schmidt’s (2014) approach we therefore meta-synthesized bivariate correlations from these studies using sample sizes to define weights (or impact effect sizes, weighting by the inverse of the variances for the classroom interventions).

Tables 1 and 2 list the measures of SET and EC used by each study. Regarding SET, many papers relied upon the same published parent self-report measures. Studies measuring SET based on observations generally followed unpublished study-derived procedures. For EC, many studies used the same direct assessment of children’s emotion knowledge. Expression and regulation were typically parent and teacher report (except for two studies with physiological measures of regulation). Figure 2 graphically exemplifies individual effect sizes from papers assessing SET-modeling and EC-knowledge. These effects varied considerably, were all small in size, and most were nonsignificant. The overall sample-size weighted effect was quite small (.07). In the full paper, we also formally test for heterogeneity of effects across studies and adjust for non-independence of multiple effects within studies.

This manuscript provides the first systematic review of empirical evidence linking SET to EC, a topic of considerable interest given contemporary focus on social and emotional learning across grades and the historical focus on these constructs in early childhood. We draw implications for current and future research and practice to support young children’s emotional competencies in classroom contexts.
 

Authors

• Rachel A. Gordon, University of Illinois at Chicago

• Xue Jiang, University of Illinois at Chicago

• Kate M. Zinsser, University of Illinois at Chicago