Study 1
Conroy and colleagues (2015) assessed the effects of the Behavioral, Emotional, and Social Training: Competent Learners Achieving School Success (BEST in CLASS) intervention. Participants were drawn from a larger randomized efficacy trial, conducted over 4 years (2011-2015), examining the impact of the BEST in CLASS intervention (Conroy and Sutherland 2011) in early childhood classrooms in two southeastern states and four school districts. The majority of the classrooms were located within federal- or state-funded early childhood programs in rural, urban, and suburban communities. Across all classroom sites curricula included the Creative Curriculum for Preschool, the High Scope Early Childhood Curriculum, and Second Step Social-Emotional Skills for Early Learning. There were no statistically significant differences in location, funding, and curricula across potential research sites.
Study participants included both early childhood teachers and young children. The teachers were selected to participate through randomization at each research site. Classrooms were randomly assigned to the intervention (BEST in CLASS) group or to the business-as-usual (BAU) group. The BAU teachers continued with their usual instructional activities such as large and small group instructional time, center-time activities, outdoor play, lunch, and snack time.
Child participants, nominated by teachers, were selected through a systematic screening process, which included the administration of the Early Screening Project (ESP) to determine Emotional Behavioral Disorders (EBDs) and the Battelle Developmental Inventory, Second Edition Screener (BDI-II) to rule out a cognitive developmental disability (Feil, Sevenson, and Walker 1989; Newborg 2005). To be eligible to participate, children 1) had to be enrolled in a federal- or state-funded early childhood classroom with a participating teacher, 2) had to be at elevated risk for EBD as indicated by the ESP, 3) could not demonstrate any cognitive development delays as indicated by the BDI-II, and 4) displayed problem behaviors in the early childhood setting.
An initial sample of 53 female early childhood teachers participated in the study. Of these, 26 participated in the BEST in CLASS intervention group and 27 in the BAU control group. Of the teachers in the intervention group, 7 had an associate’s degree, 12 a bachelor’s degree, and 7 a master’s degree, and an average of 15 years of teaching experience. In terms of race/ethnicity, 13 were white, 11 Black, 1 Hispanic, and 1 other. Of the teachers in the BAU group, 11 had an associate’s degree, 10 a bachelor’s degree, and 6 a master’s degree, and an average of 13 years of teaching experience. In terms of race, 15 were white, and 12 were Black. There were no statistically significant differences in the level of education and teaching experience between the intervention and BAU groups at baseline.
Of an initial sample of 130 preschool-aged children, 66 were eligible for the BEST in CLASS intervention group, and 64 participated in the BAU group. There were 45 boys and 21 girls in the BEST in CLASS intervention group. In regard to age, 49 were 4 years old and 7 were 5 years old. In terms of race/ethnicity, 51 were Black, 9 white, 2 Hispanic, 3 other, and 1 for whom race was not included. There were 38 boys and 26 girls in the BAU group. In terms of age, 47 children were 4 years old, 9 were 3 years old, 7 were 5 years old, and 1 was 2 years old. There were 46 Black children, 10 white, 3 Hispanic, and 5 of other race/ethnicity. The study did not provide information on whether or not there were statistically significant differences in demographic characteristics between children in the intervention and BAU groups.
The Teacher-Child Interaction Direct Observation System-Research Version 2.1 (TCIDOS-RV2.1) was used to observe and code all teacher and child behaviors (Conroy et al. 2015; Pianta 1993). Teacher behaviors were coded for the six instructional practices: 1) rules, 2) precorrection, 3) opportunity to respond, 4) behavior-specific praise, 5) instructive feedback, and 6) corrective feedback. Child behaviors were coded for four behaviors: 1) disruption or a verbalization, physical act, or gesture that either interrupts or has the potential to interrupt classroom instruction; 2) aggression or a behavior aimed at causing harm, pain, or personal injury; 3) defiance or a behavior that is challenging, noncompliant, confrontational, openly and boldly challenging, and resisting authority; and 4) engagement or participating appropriately or working on an assigned/approved activity. Interactions between teachers and children were coded as positive or negative. Mixed procedure and covariance analyses were used to estimate differences between study groups at pre- and posttest and at maintenance time points. The study authors did not conduct subgroup analyses.
Study 2
Sutherland and colleagues (2018a) drew participants from the earlier study of the BEST in CLASS intervention (Conroy and Sutherland 2011) to examine whether training and practice-based coaching were associated with teacher reports of child behavioral outcomes.
The overall sample consisted of 185 teachers and 462 children. There were 92 teachers in the intervention group and 93 in the BAU group. The majority were female (99 percent) and had, on average, 12.09 years of prior teaching experience. In terms of race/ethnicity, 48 percent were Black, 47 percent white, 3 percent Hispanic, 1 percent Asian/Pacific Islander, and 1 percent other. Of this group, 30 percent had an associate’s degree, 39 percent a bachelor’s degree, 26 percent a master’s degree, and 5 percent another degree. The study did not provide information on whether or not there were statistically significant differences in demographic characteristics between teachers in the intervention and BAU groups.
There were 230 children in the intervention group and 232 in the BAU group. Overall, 65 percent of child participants were boys and had an average age of 4.32 years at the beginning of the study. Sixty-six percent were Black, 17 percent white, 5 percent Hispanic, and 12 percent other race/ethnicity. The study did not provide information on whether or not there were statistically significant differences in demographic characteristics between children in the intervention and BAU groups.
Data from observations occurred at baseline (pretest) and at the end of the intervention (posttest). The Caregiver Teacher Report Form (CTRF) was used to measure child externalizing problems and problem behavior of both groups at pre- and posttest (Achenbach and Rescorla 2000). The Social Skills Improvement System-Rating Scale (SSIS) was used to measure children’s social skills and problem behaviors (Gresham and Elliott 2008). Multilevel structural equation modeling was used to analyze the data. The study authors did not conduct subgroup analyses.
Study 3
Sutherland and colleagues (2018b) drew participants from the earlier study of the BEST in CLASS intervention (Conroy and Sutherland 2011) to assess the effect of the intervention on child problem behavior, teacher-child interactions, and teacher-child relationships.
A total of 185 teachers and 465 children participated in this study. There were 92 teachers in the intervention group and 93 in the BAU group. In the intervention group, 99 percent were female and 1 percent male. In terms of race/ethnicity, 48 percent were white, 45 percent Black, 3 percent Hispanic, 1 percent Asian, and 2 percent other. Of this group, 30 percent had an associate’s degree, 40 percent a bachelor’s degree, 25 percent a master’s degree, and 4 percent another degree. In the BAU group, 97 percent were female and 3 percent were male. In terms of race/ethnicity, 51 percent were Black, 45 percent white, 2 percent Hispanic, 1 percent Asian, and 1 percent other. Of the BAU group, 30 percent had an associate’s degree, 38 percent a bachelor’s degree, 27 percent a master’s degree, and 5 percent another degree. There were no statistically significant differences in demographics between teachers in the intervention and BAU groups at baseline.
There were 231 children in the intervention group and 234 in the BAU group. Sixty-four percent of children in the intervention group were boys, and 36 percent were girls. The average age was 4.29 years at the beginning of the study. Of the children in the intervention group, 66 percent were Black, 16 percent white, 5 percent Hispanic, and 7 percent other race/ethnicity. Sixty-six percent of children in the BAU group were boys and 34 percent girls. The average age was 4.36 years at the beginning of the study. In terms of race/ethnicity, 67 percent were Black,18 percent white, 4 percent Hispanic, and 7 percent other. There were no statistically significant differences in demographics between children in the intervention and BAU groups at baseline.
This study also used the TCIDOS-RV2.1, CTRF, and SSIS instruments to measure student-teacher relationships, problem behaviors, and social skills in children. In addition, an Individualized Classroom Assessment Scoring System (inCLASS) was used to assess children’s interactions with teachers, peers, and tasks in the classroom within the following three domains: 1) teacher interactions or positive engagement with teacher, teacher communication, and teacher conflict; 2) peer interactions or peer sociability, communication, assertiveness, and conflict; and 3) task orientation or engagement within tasks, self-reliance, and behavior control (Downer et al. 2010). A multilevel model was used to analyze data. The study authors did not conduct subgroup analyses.
Please note that across the three BEST in CLASS studies, there are variations in sample sizes for intervention group teachers and children. The authors attribute these variations in sample sizes to the fact that each study examined different outcome measures, which resulted in different sample sizes.
