The Influence of Student Characteristics on Cognitive-Behavioral Intervention Effectiveness

Introduction: Student problem behavior can have a negative effect on the overall school environment. Among the possible causes of problem behaviors are cognitive processing deficits and distortions that can be addressed by school personnel through the use of cognitive behavioral interventions (CBI).  In recent years, CBI research has moved from the determination of its efficacy to the extension and refinement of CBIs to better meet the needs of a diverse school population.  There are several limitations in the current school-based CBI literature: a lack of (a) focus on the effectiveness of CBI for culturally diverse and low income students, (b) studies that examine CBI’s effectiveness in changing internalizing and externalizing behaviors simultaneously, and (c) studies on if and how student characteristics such as anger and rejection status at pretest play a role in intervention outcomes for students. Our study addressed these limitations by examining whether student characteristics (i.e., anger control, internalizing and/or externalizing behavior, peer status, socio-economic status (SES), ethnicity) were associated with how Tools for Getting Along (TFGA), a universally delivered CBI for 4^th and 5^th grades, affected outcomes related to behavior and social problem-solving.

Methods: Using extant data from a randomized control trial of the TFGA intervention, we conducted mediation and moderation analyses using structural equation modeling (Preacher et al., 2011) to investigate the effects of baseline levels of anger control, internalizing and externalizing behavior, peer social status, and demographic variables on the efficacy of TFGA for increasing positive social problem solving and reducing negative social problem solving, externalizing behavior, and aggression.

Results/Conclusions: The analysis revealed that TFGA participants qualifying for SES had more externalizing behavior (β = -0.076, p < .05) but less reactive aggression (β = 0.069, p < .05) than control participants. In addition, treatment participants with low externalizing behavior at pretest had greater positive social problem solving at posttest compared to the control group (β = 0.014, p < .05). Peer social status, race, and baseline levels of anger control and internalizing behavior did not moderate the effect of TFGA on the outcome variables of interest. TFGA did not have a direct effect on anger control and as a result, we did not find evidence that anger control mediated the effect of TFGA on outcome measures.  Our presentation will include a discussion of possible intervention refinements to improve CBI effectiveness with certain student populations and propose future research directions.