Program Profile: Keep Cool...Start at School

Study Title: School Based Intervention for Childhood Disruptive Behavior in Disadvantaged Settings: A Randomized Controlled Trial with and without Active Teacher Support (which is associated with Outcome 1 and Outcome 4)

Research Design

Liber and colleagues (2013) studied the intervention effects of a cognitive behavioral treatment (CBT) program, using a school-based, randomized controlled trial method. The “Keep Cool... Start at School” intervention was implemented during 3 consecutive school years at 17 elementary schools in low or low-to-middle socioeconomic status urban areas in the Netherlands. Students were randomly placed into the intervention group or waitlist control group based on grade level; both groups comprised of participants from grades 5–8. To be eligible for inclusion, students were screened by two teachers independently using the List Global Screening (LSG), a 6-item questionnaire that seeks to identify children that display symptoms of antisocial behavior. The scores from each teacher were combined to identify at-risk students, and these students were then selected to participate in the study. Of the students selected to participate (n = 280), parental consent was requested and obtained for 224 students. The sample was further confined to include only students whose IQ was 85 or higher. Students with an IQ below 85 were excluded from the study sample, as they were not expected to understand the cognitive content of the training administered. 

Sample

The sample consisted of 173 children between the ages of 8 and 12 years (the average age was 10), their parents, and their teachers.  The majority of the students were male (79 percent), low socioeconomic status (55 percent), and self-identified as non-Western immigrants (63 percent), including those who identified as Turkish, Moroccan, and Afghan. 

Students were randomly assigned to the intervention group (n = 70) or waitlist control group (n = 103). Students were assessed before treatment (pre-treatment) and after treatment (post-treatment). Data on pre-post changes between the intervention and waitlist control group was collected 1 to 2 weeks after the Keep Cool... Start at School intervention concluded for the intervention group. There was no statistically significant difference between students in the intervention group and students in the waitlist control group at the pre-treatment period. The students in the waitlist control group received the intervention after the waitlist control group received the pre-treatment assessment and after the intervention group received the post-treatment assessment. 

Data Collection/Outcome Measures

The outcomes examined were externalizing behavior and conduct disorder/oppositional defiant disorder (ODD). Externalizing behavior was a combined measure of the Peer Measure of Internalizing and Externalizing Behaviors (PMIEB) Externalizing scale, and the Teacher Report Form (TRF) Externalizing scale. The PMIEB is a 22-item peer-nomination scale evaluating psychopathology in children. Classmates are asked to select up to three of their classmates who best fit the description of each item (e.g., “sometimes fights”; “is shy or withdrawn”; “swears or uses bad language”), including externalizing behaviors (10 items). The TRF is a 113-item Likert scale used to assess behavior problems. Externalizing behavior was measured using 32 items from the TRF’s externalizing subscale. Conduct disorder/ODD was a combined measure of the TRF Conduct Disorder scale and Oppositional Defiant Disorder scale. Items from the TRF were used to measure conduct problems (13 items) and oppositional defiant problems (five items). Items are rated on a three-point Likert scale. 

The CrimeSolutions review of this study focused on the differences between the waitlist control group outcome measures reported pre-treatment and intervention group outcome measures reported post-treatment. 

Statistical Analysis 

For each outcome measure, the authors reported means, standard deviations, and pre-post effect-sizes (Cohen’s d) with confidence intervals for both the intervention condition and the waitlist control condition. The data were analyzed with multilevel modeling accounting for random effects of the trainer, as well as child-level and school-level variables. The authors did not conduct subgroup analyses. 

Citation:

Liber, Juliette M., Gerly M. De Boo, Hilde Huizenga, and Pier J. Prins. 2013. “School-Based Intervention for Childhood Disruptive Behavior in Disadvantaged Settings: A Randomized Controlled Trial With and Without Active Teacher Support.” Journal of Consulting and Clinical Psychology 81(6):975–987.

This study is a high-quality randomized controlled trial.

Study Title: The Role of Perspective Taking and Self-Control in a Preventive Intervention Targeting Childhood Disruptive Behavior (which is associated with Outcome 2 and Outcome 3)

Research Design/Sample

Nijhof and colleagues (2021) used the full study sample from the evaluation conducted by Liber and colleagues (2013) assessing the impact of “Keep Cool … Start at School” on perspective taking, self-control, and disruptive behavior. 

Children (n = 173) from 17 elementary schools in low or low-to-middle socioeconomic status urban areas in the Netherlands, their parents, and their teachers participated in the study from September 2008 through October 2011. Eligibility requirements and sample characteristics were the same as described in Liber and colleagues (2013). 

All children in the current study received the program “Keep Cool ... Start at School,” either in the first or second intervention wave. Children in the intervention condition (n = 70) received the treatment in intervention wave 1, and children in the waitlist condition (n = 103) received the treatment in intervention wave 2. Assessment took place prior to the beginning of the first intervention wave and at the end of the first intervention wave (14 weeks later) for children in both the intervention and waitlist conditions. The follow-up measurement for children in the intervention group was taken 15 weeks after the first intervention wave. The post-test measurement for children in the waitlist condition was taken 15 weeks after the first intervention wave. Children within the waitlist condition were also measured at follow-up 17 weeks after the post-test assessment. The CrimeSolutions review focused on outcomes reported at the post-treatment assessment for the treatment group and the post-waitlist/pre-treatment assessment for the control group.

There was a statistically significant difference in SES between the intervention condition and the waitlist condition at baseline. Specifically, the number of children in the intervention condition from low socioeconomic status backgrounds and from low-to-middle socioeconomic status backgrounds was approximately equal. However, the waitlist condition included a larger percentage of children from low socioeconomic status backgrounds compared with low-to-middle socioeconomic status backgrounds. In both conditions, the smallest percentage of children were from high socioeconomic status backgrounds. No statistically significant between-group differences were found for all other background variables (i.e., child’s gender, age, and ethnicity) and for baseline levels of perspective taking, self-control, and disruptive behavior.

Data Collection/Outcome Measures

Three outcomes were examined: 1) self-control, 2) parent-reported problem behavior, and 3) conduct disorder/oppositional defiant disorder (ODD). Teachers rated self-control using the Self-Control Rating Scale, which consists of 33 items (e.g., "Does the child stick to what he or she is doing until he or she is finished with it?"). Items are rated on a seven-point rating scale ranging from maximum self-control (1) to maximum impulsivity (7). Parent-reported problem behavior was measured using the Total Difficulties scale (20 items), which assesses students’ problem behaviors. Higher scores indicate more problem behavior. Conduct disorder/ODD was a combined measure of teacher-reported conduct disorder and teacher- reported oppositional defiant disorder. Conduct disorder/ODD was measured using the Disruptive Behavior Disorders Rating Scale (DBDRS) Oppositional Defiant Disorder scale (eight items) and the Condict Disorder scale (16 items). Items are rated on a four-point Likert scale ranging from not at all (0) to very much (3). 

Statistical Analysis

Means and standard deviations of all outcome measures were reported pre-treatment, post-treatment, and at the follow-up for the intervention condition and the waitlist control condition. The CrimeSolutions review focused on outcomes reported at the end of the first intervention wave. The data were analyzed with multilevel modeling. No subgroup analyses were conducted. 

Citation:

Nijhof, Karlijn, Lysanne W. Te Brinke, Urdur Njardvik, and Juliette M. Liber. 2021. "The Role of Perspective Taking and Self-Control in a Preventive Intervention Targeting Childhood Disruptive Behavior." Research on Child and Adolescent Psychopathology 49(5):657–670.

This study is a high-quality randomized controlled trial.