Impact of a Universal Social-Emotional Learning Curriculum over Two Years: Examining Group Differences in a Developmental Context

Introduction: This study examined impacts on social-emotional and academic indices following two years (four data points) of implementation of an elementary universal SEL program (Second Step). Few school-based SEL interventions assess outcomes over multiple years, because of complexities involved, but doing so is important for both enhancing ecological validity and program validity. That is, most schools adopt curricula over multiple years, and in many cases, change in behavior may not be sufficiently robust to detect after only one year of implementation.

Methods: This study utilized a randomized-controlled wait-list design, including 61 schools, 321 teachers, and 8,941 students across two states. The sample was very ethnically diverse, and between 50-80% of students received free/reduced price lunch. Teachers reported on students’ social-emotional competencies, and brief academic probes were obtained for reading and math fluency. In addition, classroom observations were conducted to assess academic engagement and disruptive behavior. In order to maximize generalizability and reduce misspecification, this study utilized a information-theoretic approach in which we 1) proposed eight models to estimate potential patterns of growth for the four assessments across two years 2) assessed the best fitting model for all outcomes, and then 3) assessed differences between intervention and control groups. 

Results: Data suggest the standard curvilinear growth assumption did not fit a single outcome. Most outcomes fit a pattern in which students improved only during instructional time in school; though, a few, such as empathy and skills for learning, showed a pattern of early acceleration then flattening (in Year 2). Intervention effects were found on several important distal outcomes including hyperactivity, conduct problems, emotional symptoms and reading achievement. 

Conclusion: Findings overall, provide validation of Second Step within parameters of a large-scale, rigorous multi-state trial. Discussion emphasizes the importance of developmentally-sensitive methods for yielding valid conclusions, and elaborates on how this approach provides context for interpreting longitudinal outcomes in school-based RCTs. That is, findings help inform schools of skills that may be particularly responsive to intervention, such as hyperactivity, and helps elucidate when the most robust changes occur.