Growth Trajectories of Social-Emotional Competence in K-2 Students: Toolbox Universal and Differential Intervention Effects

Background and Purpose: Social Emotional Learning (SEL) programs have demonstrated the capacity to enhance protective factors to prevent behavioral problems (Kam, Greenberg, & Kusche, 2004; Kim, Gloppen, Rhew, Oesterle, & Hawkins, 2015). Less work has been done to disaggregate the growth of student protective factors in responses to SEL programs by student characteristics (e.g., class, gender). Prior studies have found that social-emotional competence, as measured by the DESSA-Mini (LeBuffe, Naglieri, & Shapiro, 2011), grew overall among K-2 students in response to the PATHS curriculum, but trajectories differed based on gender and grade (Shapiro et al., 2017). The current paper examines: 1) To what extent social-emotional competence changed over time among K-2 students in response to a different SEL program; 2) To what extent the change in students’ social-emotional competence varied by individual student characteristics.

Methods: Social-emotional competence was assessed among students (n=1,625) at three time points (October, December, & May) using the DESSA-Mini. Within a single school district, 4 schools were assigned to implement TOOLBOX and 2 schools were assigned to the practice-as-usual condition. We first  assessed the growth in social-emotional competence over time by intervention condition. Then, we examined how student characteristics affected the rate of growth in students’ social-emotional competence among students at TOOLBOX schools. Growth curve modeling techniques were employed.

Findings: Students’ social-emotional competence (SEC) increased over time across conditions, averaging an increase of 2.65 T-Score points each year. At baseline, female (b=4.62, p<.001) and older (b=1.05, p<.001) students had significantly higher SEC relative to their counterparts; students receiving special education services (b=-8.76, p<.001) and free lunch (b=-1.17, p<.05) had significantly lower SEC relative to their counterparts. Controlling for student individual characteristics, there was no difference in baseline SEC by intervention condition, but the rate of growth was significantly higher for TOOLBOX students than comparison students by 1.88 T-Score points (p<.001). Differential growth at TOOLBOX schools was also examined. Students’ growth rate was not significantly different by gender, age, special education placement, or poverty status, while native English speakers (b=1.98, p<.01) had significantly more growth in SEC.

Conclusion and Implications: Youth across all intervention conditions experienced significant growth in SEC, but students in the TOOLBOX condition experienced higher rates of growth than students in the comparison schools. Within the TOOLBOX condition, the rate of growth in SEC was not different by sociodemographic characteristics except primary language. These results provide some preliminary evidence for the universal effectiveness of TOOLBOX in promoting SEC, including effectiveness among special education and low income students. Programming may benefit from modifications for English Language Learners. Limitations of the study will be discussed.