Study
In an independent study conducted by Boyle and Hassett–Walker (2008), matched pairs of schools were randomly assigned to either ICPS instruction or control status. The student sample included 226 students who were assigned to one of three groups: the 2-year ICPS group (n = 96) received ICPS instruction for 2 consecutive years; the 1-year ICPS group (n = 106) received ICPS instruction for 1 year only in kindergarten or 1st grade; and the control group (n = 24) received no ICPS instruction.
More than 80 percent of participants in each group were Hispanic. African American students made up 9 percent of the 2-year instruction group and 5 percent of the 1-year instruction group. None of the control group participants were African American. A small percentage of participants in each group were white. Each group had more females than males. Additionally, more than 90 percent of children in all three groups were receiving a discounted lunch, a measure used as an indicator of income level.
Chi-square analyses revealed no significant differences among the groups in race/ethnicity, socioeconomic status, or gender. Two behavior ratings scales were employed: the Preschool Social Behavior Scale (PSBS) and the Hahnemann Behavior Rating Scale (HBRS). The PSBS is a 16-item Likert behavior rating scale that measures aggression and prosocial behavior. Teachers used the PSBS to rate children on items such as whether they are good at sharing or they shove other children. The HBRS is an 11-item Likert scale that measures aggression, impulsivity, passivity, and prosocial behavior. Teachers used to HBRS to rate children on items such as whether they are easily angered or overly timid. Analysis was conducted using multiple analysis of covariance and analysis of covariance techniques. No subgroup analysis was conducted.
Study
Kumpfer and colleagues (2002) conducted a quasi-experimental evaluation of ICPS. A total of 655 children were recruited from the 1st grade classes of 12 schools in two Rocky Mountain districts. Classrooms were randomly assigned to one of four conditions; an ICPS–only treatment group (n = 256), an ICPS combined with the full Strengthening Families program treatment group (n = 56), an ICPS combined with only the parent module of the Strengthening Families program treatment group (n = 21), and a no-treatment control group (n = 322). The CrimeSolutions review of this study examined the outcomes of the ICPS–only treatment group and compared them with the outcomes of the control group. The overall sample was 87 percent Caucasian and 7.6 percent Hispanic, with an average Hollingshead index position indicating a largely middle-class sample. The majority of the sample (53 percent) was female; 6 percent were enrolled in special education and 3.2 percent had a documented learning disability.
The study authors designed constructs for measuring school bonding, parenting skills, social competence, and self-regulation by deriving various indices from questionnaires administered to children, parents, and teachers. Questionnaires were administered at the beginning and end of the intervention year. The results show scores on each item and the study estimates effect sizes and mean differences controlling for group differences in the baseline measurements. No subgroup analysis was conducted.
Study
Shure and Spivack (1982) conducted a quasi-experimental evaluation of an interpersonal cognitive problem-solving intervention, which would later be called I Can Problem Solve (ICPS), followed low-income, inner-city African American nursery school and kindergarten children to evaluate them over a 2-year period. Subjects in the year 1 evaluation included 113 children (47 boys and 66 girls) trained in ICPS, and 106 control children (50 boys and 56 girls). In year 2 of the evaluation, the 131 children were divided into four groups: a twice trained group (n = 39); a trained in nursery school group (n = 30); a trained in kindergarten group (n = 35); and a never-trained control group (n = 27). The CrimeSolutions review of this study focused on the outcomes of the year 1 evaluation. At baseline the average age of the sample was 4 years and 3 months, with equality of the groups in age distribution, IQ, ICPS test scores, and behavioral characteristics. On average, treatment children were trained for 12 full weeks in formal, scripted sessions implemented by their teacher throughout the school day.
Researchers collected information on interpersonal cognitive problem-solving measures each year immediately before and following the 12-week intervention, for a total of four assessments. The measures were used to test children’s abilities to problem solve and understand consequences, and to evaluate teacher-rated interpersonal behavior in the classroom. Data was analyzed using analysis of variance. No subgroup analysis was conducted.