The Moderation of Individual- and School-level SES on the Effect of the Good Behavior Game

• In Event: Understanding the Link Between Socioeconomic Status and Children’s Developmental Outcomes Across Diverse Sociocultural Contexts

Abstract

Background: The early elementary school years have been suggested as an important period and a setting for implementing universal school-based intervention programs aimed to prevent the development of children’s internalizing and externalizing problems. One of the most successful universal school-based intervention programs for the prevention of internalizing and externalizing problems is the Good Behavior Game (GBG). However, previous studies indicated that the GBG might be more effective among children with elevated levels of externalizing and internalizing problems (Bowman-Perrott et al., 2016). Children who grow up in families of less-educated parents have been shown to be at risk of having elevated levels of internalizing and externalizing problems. Similarly, schools with predominantly children of lower parental education backgrounds have been shown to report elevated levels of internalizing and externalizing problems in their students. Hence, a potential moderator of the effectiveness of the GBG might be parental education levels at the individual level or at the broader school level. This study, therefore, examined whether individual-level (children of less- or higher-educated parents) and school-level parental education (schools with a higher or lower percentage of children of less-educated parents) moderated the effect of GBG on the reduction of internalizing and externalizing problems in children followed from kindergarten to the end of second grade.

Method: Using a randomized controlled trial design (66% in GBG group, 34% in practice as usual group), 597 children (52% girls) were annually followed from kindergarten to second grade. The GBG intervention was implemented in the first and second grade of elementary school. Children’s internalizing and externalizing problems were assessed by teacher reports. Individual-level parental education level was based on the highest completed education level of the mother and father. School-level parental education was based on the per-school percentage score of all children from less-educated households. Main effect associations and cross- and between-level interactive associations were examined using multi-level latent growth models.

Results: Overall, there was a trend-like main effect of the GBG in reducing externalizing problems and a significant effect on reducing internalizing problems. However, the effect of internalizing problems was moderated by school-level education, suggesting the GBG to be more effective in decreasing internalizing problems in higher parental education schools than in lower parental education schools. No such interaction was found for externalizing problems (See Figure 1 and 2). Moreover, no individual-level moderation of parental education on the impact of GBG on reducing internalizing and externalizing problems were found (cross-level interactions).

Conclusion: GBG was borderline effective in reducing externalizing problems universally and effective in reducing internalizing problems for the sample in total. However, further exploration of this effect for the group in total revealed that the GBG was less effective in reducing internalizing problems in schools where most children have less-educated parents. Results suggest that schools with a high percentage of children with less-educated parents may need additional resources or intervention components to prevent mental health problems in students.

Authors

• Nil Horoz, Vrije Universiteit Amsterdam
  Presenting Author

• Marieke J.M. Buil, Rotterdam Universitty of Applied Sciences
  Non-Presenting Author

• Susanne Koot, Vrije Universiteit Amsterdam
  Non-Presenting Author

• Hans M Koot, Vrije Universiteit Amsterdam
  Non-Presenting Author

• Pol Van Lier, Vrije Universiteit Amsterdam
  Non-Presenting Author