Speaker
Description
Background: Good Behavior Game (GBG) is a school-based intervention designed to reduce conduct problems, while increasing on-task behavior and a positive classroom climate. Earlier studies have shown long-term effects in several outcomes, making GBG a promising method of universal prevention. This study evaluates the effectiveness of an adapted version of GBG in a pragmatic, cluster-randomized, parallel group, superiority trial.
Methods: Schools with K-3 students were eligible for recruitment. Five schools were recruited to either receive training in GBG or continue with business-as-usual. Schools were allocated using randomization, stratified by a sociodemographic variable. The outcomes included teacher-rated conduct problems in the classroom (primary outcome) and common school areas, observer-rated on-task behavior, along with both teacher- and observer-rated classroom climate. All 43 classrooms had teacher-rated measures, while a subset of 20 classrooms were randomized to also receive the observer-rated measures. Three points of measurement were used; at the start of school year (pre-intervention), at the middle of the school year (interim follow-up), and at the end of the school year (primary endpoint). All classrooms were included in the analysis using Bayesian mixed effects models.
Results: Conduct problems in the classroom, on-task behavior and classroom climate had effects in the hypothesized direction, favoring the GBG-group. However, there was no effect on conduct problems in common school areas over time. Furthermore, effects for all outcomes were uncertain, as highest posterior density intervals overlapped for the intervention and control group.
Discussion: While the effects are somewhat uncertain, taken together, this study lends tentative support for cross-cultural transportation and adaptation of GBG to a pragmatic context, without major external provision of resources or personnel.
| Conflict of interest | The authors have no conflicts of interest to disclose. |
|---|
