Flannery and colleagues (2003) evaluated PeaceBuilders using a randomized design with treatment and control groups. Eight elementary schools (grades K–5) in Pima County, Ariz., were selected from two large school districts, based on their high rates of juvenile arrests, suspensions, and expulsions. The total sample consisted of 4,128 students. Of these, 51 percent were Hispanic, 28 percent white, 13 percent American Indian, 6 percent African American, and 1.5 percent Asian American.
The study analyzed data for 2 full school years, denoted as year 1 and year 2. All participating schools remained in the study through the first 2 intervention years. Before baseline data collection, the eight project schools were matched into four pairs, primarily on the basis of geographic proximity. Four schools were then randomly assigned as PeaceBuilders Immediate intervention schools and began the program immediately following baseline data collection in fall 1994. These four schools implemented the PeaceBuilders program for the entire school year of year 1 (fall and spring). The remaining schools in the control group were assigned to the PeaceBuilders Delayed program, where implementation of the PeaceBuilders program was delayed until year 2 in 1995, after 1 year of baseline data collection. Thus in year 1, the four treatment schools received the PeaceBuilders intervention, while the four control schools did not receive any PeaceBuilders intervention. In year 2, both the treatment and control schools received the PeaceBuilders intervention. Teachers at the intervention schools were trained to implement the PeaceBuilders program during the school day. Teachers at control schools neither received training nor implemented PeaceBuilders until after the 1st year. Student-level data was used for the analyses.
The two primary outcome measures for the study were social competence and aggressive behavior, as measured by teacher reports and self-reports by students. Outcomes for self-reported behaviors were measured as two separate groups, for grades K–2 and grades 3–5. The questionnaires were modified to be made easier for the grade K–2 group; students in these grades were asked to respond “yes” or “no” (or close variations) to items, while students in grades 3–5 were asked to rate items on a point scale. Also, face-to-face interviews were conducted with students in grades K–2 to obtain responses, while students in grades 3-5 and teachers filled out bubble scan sheets to provide responses.
Data were collected at four points:
- Baseline (fall 1994)
- Time 2 (spring of year 1, 1995)
- Time 3 (fall of year 2, 1995)
- Time 4 (spring of year 2, 1996)
To evaluate the effect of the intervention in its 1st year, differences in outcome measures from baseline data collection (fall 1994) and spring of year 1 (spring 1995) were compared for the treatment and control groups. This was the only period when a true control group was used. Time 3 and time 4 data were used to assess whether or not the effects of the program could be sustained.
Social competence was measured based on teacher reports of students’ social competence, using the Walker–McConnell Scale of Social Competence and School Adjustment. This instrument has three subscales: School Adjustment, Peer-Preferred Behaviors, and Teacher-Preferred Behaviors, and include items on students’ cooperation, self-control, and social behavior. Teachers rated the frequency of such behaviors on a Likert scale ranging from “never” (1) to “frequently” (5).
Prosocial behavior for students in grades 3–5 was measured using a 16-item instrument designed for the study, which included items regarding empathy, caring, helpfulness, and support of others. Students reported the frequency of their own behaviors on a three-point scale (1=no, 2=a little, 3=a lot). Prosocial behavior for students in grades K–2 was assessed based on responses to six statements. Students reported the frequency of their own behaviors by responding “yes,” “sometimes,” or “no, not really.”
Peace-building behavior for students in grades 3–5 was assessed based on their responses to three peace-building statements, including “I helped build peace at school,” “I told other kids they were peace builders,” and “I earned rewards for peace building” on a three-point scale (1=no, 2=a little, 3=a lot). Peace-building behavior for students in grades K–2 was assessed based on their responses to four items regarding peace building. Students reported the frequency of their own behaviors by responding either “yes” or “no” to these statements.
Aggressive behavior was measured using teachers’ reports of aggressive behavior by students, using items adapted from the aggressive behavior subscale of Achenbach’s Teacher Report Form. Using this instrument, teachers rated aggressive child behavior on a three-point scale between 0 and 2 (0=not true, 1=somewhat true, 2=very true).