UC Berkeley

Social-emotional competence (SEC) is a set of cognitive, affective, and behavioral skills that enable an individual to engage in healthy interactions with oneself and other human beings. SEC is a critical factor for individual well-being and may also contribute to creating a more safe and caring community. SEC can develop throughout a lifetime, but it has been studied to be particularly malleable during childhood and adolescence. School-based social and emotional learning (SEL) is an interventional approach to fostering SEC in children within a school setting. In the United States and many other countries, universal SEL targeting the entire student body has become a popular practice with the goal of promoting mental, behavioral, and academic well-being of young people. Experimental, quasi-experimental, and meta-analytic research has supported the overall benefits of universal SEL on student SEC development and other positive outcomes. However, the literature lacks sufficient evidence to answer the questions about the typical SEC developmental trajectory in the context of school-based SEL and the disparities therein across diverse student subgroups. Also, the question remains unanswered as to whether universal SEL works equally or differentially across diverse sociocultural subgroups, and the discussion of its potential impacts on subgroup disparities has been rare. Composed of three chapters each featuring a freestanding empirical study, this dissertation addresses important but under-investigated questions in the field: (1) What is a typical SEC developmental trajectory expected for school-aged children under a routine SEL practice setting? (2) To what extent do disparities exist in SEC developmental trajectories across diverse sociocultural subgroups? (3) What are the effects of universal SEL on subgroup disparities in SEC developmental trajectories? Chapter 1 examines how student SEC develops, on average, across elementary school years under a routine SEL practice condition. Although universal progress monitoring of student SEC has increasingly been adopted as part of regular educational practices, an evidence base has not yet been established on the typical SEC growth to expect within the school setting. In addition, longitudinal measurement invariance is a prerequisite when measuring student growth, but little evidence exists on the equivalence of SEC measurement across time, especially for the assessment tools that are currently widely used in practice. To address these gaps, this chapter examines how student SEC develops, on average, across elementary school years under a routine SEL practice condition, after first testing the longitudinal measurement invariance of a widely-used teacher-completed behavioral rating scale. The data analyzed in this chapter originate from six waves of teacher ratings of student SEC, collected from a districtwide SEL initiative for three consecutive years using the DESSA-Mini (N = 1,146; Grades K-2 at baseline). Using longitudinal confirmatory factor analysis methods, this chapter found no evidence for violation of measurement invariance across all six time points throughout the three consecutive years, suggesting that the same construct of SEC was measured across different seasons and grade levels. Then, using second-order latent growth modeling methods that did not impose any predetermined shape of growth, this chapter found that (a) student SEC increased within each year, (b) student SEC decreased over each summer by about a half of the yearly gain, and (c) the rate of yearly growth gradually decreased across years. Implications and limitations of these findings are discussed with suggestions for future research and practices. Chapter 2 examines the extent to which disparities exist in student SEC developmental trajectories across socioculturally classified subgroups. Although student gender, race, ethnicity, and socioeconomic status (SES) have conceptually been linked to differential and unequal social-emotional developmental experiences, empirical evidence on subgroup disparities in SEC developmental trajectories has been limited and inconclusive. Also, despite the field’s widespread concern about teacher bias in assessing SEC within diverse student bodies, little evidence is available on the measurement invariance of SEC progress monitoring tools across student subgroups, a prerequisite condition for examining subgroup disparities. The data analyzed in this chapter involve nine waves of teacher-rated student SEC, collected from a districtwide SEL initiative for three consecutive years (N = 5,452; Grades K-2 at baseline). Using multigroup confirmatory factor analysis methods, this chapter first tested measurement invariance of the DESSA-Mini by gender (male vs. female), race and ethnicity (Black vs. Hispanic vs. White), and SES (low-income vs. middle-to-high-income). No evidence for violation of measurement invariance across all the examined subgroups was found, suggesting that the same construct of SEC was measured across diverse student subgroups. Then, using second-order piecewise latent growth modeling methods, this chapter examined the subgroup disparities in the three-year growth trajectories of student SEC under a routine universal SEL practice condition. The results suggest that (a) male (vs. female), Black or Hispanic (vs. White), and lower-income (vs. higher-income) students started with a lower level of SEC, and (b) these gaps were either sustained across time or slightly widened. Implications and limitations of these findings are discussed with suggestions for future research and practices. Chapter 3 examines the universality of the intervention effects of universal SEL across diverse student subgroups, and the effects of universal SEL on existing subgroup disparities. Although previous SEL research has accumulated robust evidence of the average intervention effects of universal SEL, it remains unclear whether universal SEL works equally or differentially across diverse sociocultural subgroups of students. This has implications for possible subgroup disparities in student social-emotional competence (SEC) development. The data analyzed in this chapter came from student SEC progress monitoring collected during a one-year quasi-experimental study of a universal SEL program (N = 1,592; Grades K-2). Using a multigroup structural equation model by student subgroup, this chapter first examined whether the effects of universal SEL on student SEC growth differed across diverse student subgroups classified by gender, race, ethnicity, socioeconomic status, disability status, and English learner status. The results suggest that (a) the intervention effects were slightly larger for Black students, compared to White or other racial-ethnic subgroups, and (b) the effects were not different across other examined subgroups. Using a multigroup structural equation model by intervention condition, this chapter also examined whether the patterns of any subgroup disparities in SEC growth trajectories differed by intervention condition. The results suggest that (a) in the comparison condition, the SEC disparities between Black and White students tended to widen throughout the year, whereas in the intervention condition, Black students showed a similar rate of growth as their peers, and (b) no other subgroup disparities in SEC growth trajectories showed different patterns between the two conditions. Implications and limitations of these findings are discussed with suggestions for future research and practices. In summary, this dissertation contributes to the literature by providing evidence of (1) an average multi-year developmental trajectory of SEC among school-aged children, assessed with a measure tested to be longitudinally invariant, (2) gender, racial-ethnic, and socioeconomic disparities that exist in this SEC developmental trajectory, assessed with a measure tested to be invariant across subgroups, and (3) the potential of universal SEL for preventing some disparities from further widening, but not necessarily closing all the existing gaps. I hope the findings of this dissertation lead to and inform larger questions in the field, such as how to make sure SEL assessment tools are accurate and equitable, how to set SEL standards and benchmarks, and how to ensure school-based SEL truly benefits all students and produces more equitable outcomes.