Lessons learned from the large-scale implementation of a social-emotional learning program in Brazilian primary schools

• In Event: Implementation findings from three different approaches to social emotional learning in Brazil early education

Proposal

A number of studies illustrate the promise of school-based programs focused on social-emotional learning (SEL) to have positive and enduring impacts on children’s social-emotional competencies (Taylor, Oberle, Durlak, & Weissberg, 2017). However, a related literature highlights the challenges of successfully implementing such programs at scale (Elias, Zins, Graczyk, & Weissberg, 2003). For example, SEL is often not viewed as a core component of the academic day and thus might be deprioritized as compared to instruction in traditional academic domains like mathematics and reading. Despite the proliferation of SEL programs to non-U.S. settings, the existing literature provides little insight into how unique challenges might threaten the success of social-emotional learning programs taking place outside the U.S. In the present study, we describe the implementation of a classroom-based SEL program called Programa Compasso (PC) in 47 primary schools in Rio de Janeiro, Brazil. In particular, we explore whether teacher- or school-level factors predict the likelihood of teachers implementing and appreciating PC lessons.

Data for the present study come from a randomized control trial of PC in 94 public primary schools in Rio de Janeiro. At the start of the study, schools were randomized to either receive the curriculum (n = 47) or to the control group (n = 47). Within each treatment school, we followed the implementation of one randomly selected third and fifth grade teacher for a total of 94 treatment teachers. Across the school year, teachers implementing the PC curriculum were expected to deliver 22 weekly, 30-minute lessons focused on promoting children’s social-emotional skills like self-regulation, executive function, emotion knowledge, and social problem solving. We collected data from 91 of the 94 treatment teachers (97%) during at least one of three times throughout the year: beginning, middle, and end. In the present study, we use information reported by treatment teachers primarily at the baseline to predict the number of PC lessons completed and teacher perception of the program. Table 1 describes our measures of implementation and key predictors.

The results of two-level regression models predicting PC implementation as a function of teacher-reported factors are presented in Table 2. These results suggest that a number of teacher and school factors seem to matter for implementation. First, these results illustrate that teachers at baseline who were already engaged in social-emotional related instruction (e.g., spoke to students about bullying) were more likely to complete a greater number of PC lessons than teachers who did not engage in these behaviors. Second, teachers with higher baseline self-efficacy were more likely to report that children had learned or were positively affected by PC than those with lower self-efficacy. Third, teachers in schools that were closed for more days due to violence implemented fewer lessons than teachers in less affected schools.

The final presentation will include updated analyses that account for missing data and explore additional measures of implementation. We will also discuss the implications of this work for efforts to scale-up school-based SEL programs in global contexts.

Authors

• Emily C. Hanno, Harvard Graduate School of Education

• Dana McCoy, Harvard Graduate School of Education

• Vladimir Pinheiro Ponczek, São Paulo School of Economics

• Christine Campos de Xavier Pinto, São Paulo School of Economics

• Gabriela Fonseca, São Paulo School of Economics

• Natália Zanetti Marchi, São Paulo School of Economics

• Renee Lamoreau, Harvard Graduate School of Education