Socio-emotional Skills and the Socioeconomic Achievement Gap

Abstract

Empirical evidence suggests children’s socio-emotional skills—an important determinant of school achievement—vary according to socioeconomic family background. This study assesses the degree to which differences in socio-emotional skills contribute to the achievement gap between socioeconomically advantaged and disadvantaged children. We used data on 74 countries from the 2018 Programme for International Student Assessment, which contains an extensive set of psychological measures, including growth mindset, self-efficacy, and work mastery. We developed three conceptual scenarios to analyze the role of socio-emotional skills in learning inequality: simple accumulation, multiplicative accumulation, and compensatory accumulation. Our findings are in line with the simple accumulation scenario: Socioeconomically advantaged children have somewhat higher levels of socio-emotional skills than their disadvantaged peers, but the effect of these skills on academic performance is largely similar in both groups. Using a counterfactual decomposition method, we show that the measured socio-emotional skills explain no more than 8.8 percent of the socioeconomic achievement gap. Based on these findings, we argue that initiatives to promote social and emotional learning are unlikely to substantially reduce educational inequality.

Declines in social–emotional skills in college students during the COVID-19 pandemic

“…We conducted a total of six one-way analyses of covariance (ANCOVAs) to assess whether mean scores of SSI measures were significantly different between the pre-lockdown and post-lockdown groups, controlling for gender, race, and socioeconomic status. We included these variables as covariates given prior research showing that social and emotional skills vary by gender, race, and socioeconomic status ( de Ridder et al, 2012 ; Kuo et al, 2020 ; Gruijters et al, 2024 ). To estimate the SSI measures’ internal consistencies, Cronbach’s alpha coefficients were generated for pre- and post-lockdown groups.…”