Methods: A pretest-posttest multi-site cluster-randomized design (MSCRT; Spybrook & Raudenbush, 2009) was used across 127 early childhood teachers (Tx = 61 & Comparison = 66) and 319 preschool children (Tx = 170 & Comparison = 149). Randomization occurred at the classroom-level, with classrooms blocked by school and randomly assigned to either the treatment or control conditions. Two-level analysis was conducted for teacher outcome measures and three-level analysis was conducted for child outcomes. Effect sizes were calculated using a multi-level variant of Cohen’s d.
Results: BEST in CLASS teachers demonstrated significantly higher levels of classroom organization, emotional support, and instructional support in comparison to control teachers as measured by the Classroom Assessment Scoring System (CLASS; Pianta, LaParo, & Hamre, 2008). BEST in CLASS teachers’ also self-reported higher levels of self-efficacy in comparison to control teachers as measured by the Teachers’ Self-Efficacy Beliefs Scale (TEBS; Dellinger et al., 2008) and the Teachers’ Sense of Efficacy Scale (TSES; Tschannen-Moran & Hoy, 2001). BEST in CLASS children demonstrated significantly lower levels of teacher reported conflict, lower levels of challenging behaviors, and significantly higher social skills in comparison to control children as measured by the Student Teacher Relationship Scale (STRS; Pianta & Hamre, 2001), Social Skills Improvement System (SSIS; Gresham & Elliott, 2008) and Caregiver-Teacher Report Form (C-TRF; Achenbach & Rescorla, 2000). In addition, BEST in CLASS children demonstrated significantly higher levels of positive interactions with teachers and task orientation while levels of conflict decreased in comparison to control children as measured by the Individualized Classroom Assessment Scoring System (inCLASS; Downer, Booren, Lima, Luckner, & Pianta, 2010).