Study
The Conduct Problems Prevention Research Group (1999) evaluated Promoting Alternative THinking Strategies (PATHS®) using 48 elementary schools located across the country including Durham, S.C.; Nashville, Tennessee; Seattle, Washington; and central Pennsylvania. Schools were divided into matched sets based on size, achievement levels, poverty, and ethnic diversity. The sets were then randomly assigned to intervention or comparison conditions. The study was conducted in three successive years, resulting in three cohorts of first graders. For each cohort, 12 elementary schools from each of the four geographic areas were included. A total of 6,715 children distributed across 378 classrooms (198 intervention and 180 matched comparisons) participated in the study. Groups were equivalent in percentage of children who received free or reduced lunch, percentage of ethnic minority children, and academic achievement scores.
While teachers in the comparison group pursued their usual lesson plans, teachers in the intervention classrooms delivered up to 57 lessons of the PATHS® curriculum emphasizing self-control, emotional awareness of peer relations, and solving problems.
The evaluators then assessed the impact of the PATHS® curriculum on students’ social competence using outcome measures, including:
- Teacher reports of student social competence
- Individual sociometric interviews with all children who had parental consent
- The Authority Acceptance and Cognitive Concentration subscales of the Teacher Observation of Classroom Adaptation—Revised
- The Social Health Profile
- Sociometric interviews of children’s ratings of their peers on aggression, hyperactivity/disruptive behavior, and prosocial behavior
- Teachers’ ratings of children on how liked they were by their peers
- The Classroom Rating Form completed during 30-minute observations by impartial raters on items such as child’s level of disruption, ability to handle classroom transitions, ability to follow rules, and ability to express feelings.
The data was analyzed using hierarchical linear modeling (HLM) models, with classroom as the second level in a mixed model design. The unit of analysis was the classroom, rather than the student. Thus, the analysis occurs at the level of implementation. The authors did not conduct subgroup analyses.
Study
This study—conducted by Kam, Greenberg, and Kusche (2004)—used a randomized control research design to examine the long-term effectiveness of the Promoting Alternative THinking Strategies (PATHS®) curriculum for special education students. The sample consisted of 133 students with disabilities: 97 boys, 36 girls; 88 were white, 27 were African American, and 18 were of other ethnicities. The average age of the children at baseline was 8years and 8 months. Fifty-three children had learning disabilities, 23 had mild mental retardation, 31 had emotional and behavioral disorders, 21 had physical disabilities or health impairment, and 5 had multiple handicaps. The classrooms were mixed age (grades 1–3), and each contained children with a variety of disability classifications. The intervention group had higher mean internalizing behavior, which was controlled in the analyses. However, the study did not specify the number of students assigned to receive the intervention condition and the number of students assigned to receive the control condition.
Eighteen teachers from seven elementary schools in Seattle, Washington, volunteered to participate in the study and were randomly assigned to teach either the intervention or the comparison condition. Teachers in the intervention classrooms implemented a 1-year model of the PATHS® curriculum, while teachers in the comparison group pursued their traditional lesson plans. A battery of sociometric tests and teacher reports was then used to assess students’ long-term emotional development.
Study outcomes consisted of changes in students’ externalizing and internalizing behaviors, social competence, self-reported depression, affective vocabulary, and problem-solving skills. Data for all variables was collected at baseline (time 1) and for 3 successive years (times 2, 3, 4). Measures used were the Child Behavior Checklist, the Beck Depression Inventory for Children, the Social Competence subscale of the Teacher–Child rating scale, the Kusche Affective Interview, and the Social Problem-Solving Interview, which generates a total effectiveness score.
Missing data in the follow-up years ranged from 6 percent to 48 percent on different variables, but the statistical procedure used in this study, Hierarchical Linear Modeling, manages missing data. The authors did not conduct subgroup analyses.
Study
This randomized clinical trial conducted by Domitrovich, Cortes, and Greenberg (2007) examined the impact of the PATHS® curriculum on social and emotional skills building among children in two regional Head Start programs in central Pennsylvania. The study used a mixed block design in which randomization took place at the building level. Within each of the two regional Head Start programs, blocks were created that contained at least two matched classrooms with similar neighborhood population density. Six classrooms (three pairs) were drawn from 15 eligible classrooms in one program and then randomly assigned to the intervention or the comparison group. Ten classrooms (five pairs) were drawn from among 13 eligible classrooms in the other program and randomly assigned to the intervention (n = 120) or the comparison (n = 126) conditions. The final sample was 246, of whom 120 boys were and 126 were girls. Forty-seven percent were African American, 38 percent white, and 10 percent Hispanic. Mean age for the children at baseline was 51 months.
Groups differed at baseline on age, ethnicity, and disability. These variables were controlled for using analysis of covariance. ANCOVAs were performed on each outcome, testing for main effects, site, group, and interaction of site by group. Attrition rate was 18 percent. Children who left the program tended to be higher functioning. The posttest evaluation occurred at the end of the school year.
Outcomes were affective development, social competence, and mental health. Measures used were a revised version of the Recognition of Emotion Concepts subtest from the Kusche Emotional Inventory, the Assessment of Children’s Emotions Scale, the Denham Puppet Interview assessing affective perspective-taking skills, two measures of inhibitory control, the Attention Sustained subtest from the Leiter Revised Assessment Battery on visual spatial memory and attention, the problem-solving portion of the Challenging Situations Task, the Preschool and Kindergarten Behavior Scale (both for teachers and for parents), and the Peabody Picture Vocabulary Test. Parents also filled out the Head Start Competence Scale, measuring children’s social and emotional skills that reflect interpersonal relationships and emotion regulation. The authors did not conduct subgroup analyses.