Comparison of Two Controlled Trials Designed to Reduce Aggressive Behavior in Primary School Children in China: Impact of the COVID Pandemic on Program Processes and Outcomes

Background and Purpose: In childhood, cognitive skill in processing social information is strongly related to long-term social adjustment. Indeed, recent research suggests that social information-processing skills can be used, in part, to conceptualize psychopathology as normal child development has gone away (Dodge, 2024). At the core of this assertion is the idea that the processes that govern psychopathological behavioral development are the same as those that govern normal behavioral development. This study describes the outcomes of two controlled trials of a social information-processing intervention, the Let’s Be Friends (LBF) program, which was designed to reduce aggressive behavior in primary school children in rural China. In addition, we discuss the impact of the COVID pandemic on study processes and outcomes. One study was completed before the epidemic, and the second study was completed during the epidemic.

Methods: The LBF program is a product of approximately 2 decades of collaboration between Sino-U.S. researchers to adapt a successful U.S.-based program—Making Choices: Social Problem-Solving for Children (MC)—for Chinese children. The work on LBF is translational, with systematic cultural adaptation. A blocked cluster design with random sampling was used in each trial. Schools in each trial were matched on Mahalanobis distances calculated from eight covariates. Combining the two trials and adjusting for differences between the studies, we estimated outcomes across all treated (n=698) and control (n=710) group children; and, additionally, we compared Trial 1 treated children with Trial 2 treated children.

Results: Across the combined studies, significant differences on children’s social skills and behavior favored the LBF condition. However, the outcomes for Trial 2 demonstrated less robust behavioral effects, suggesting that the program was weakened by pandemic-related implementation and measurement challenges. On all four dimensions of the SLA (i.e., encoding, interpretation – hostile attribution, goal formulation, and response decision), the treatment group showed statistically significant and beneficial difference-in-differences (DID) effects compared to the control group: of 20 outcome measures, the treatment group demonstrated 10 significant and beneficial DIDs. The teachers’ ratings using CCC also suggested that the classroom behavior of children in the combined LBF treatment groups improved significantly when compared to the behavior of children in the control group: of 10 outcome measures, 7 showed statistically significant DIDs. The study also confirmed that the treated children of Trial 1 showed more positive and beneficial changes on the SLA and CCC measures than the treated children of Trial 2 – the measurement and implementation were compromised by the pandemic.

Conclusions and Implications: Through the creation of culturally sensitive activities, the study showed potential for increasing social-emotional competence, disrupting aggressive behavior, and reducing hostile attributional bias in Chinese primary school children. But environmental disruptions, such as the COVID pandemic, create major challenges in implementing school-based programs. Perhaps the most important lesson learned from the comparative study is that school-based intervention during crises is feasible. It can be done. However, it requires the adaptation of program content, flexible program delivery timelines, close communication and coordination with school administrators and other stakeholders, and careful selection of research instruments.