Abstract
BACKGROUND: Behavioral parenting programs have been exhaustively studied over the past five decades. We used this wealth of research to examine how estimates of parenting program effects have evolved over time, and if any time trends in effect estimates can be explained by trial, sample, or intervention characteristics. METHODS: We based our meta-analysis on a systematic search of 22 international and regional databases, gray literature, and 4 trial registries for randomized controlled trials of behavioral parenting programs. RESULTS: We identified 244 eligible trials (1,100 effect sizes; 28,916 families) from 36 countries. Parenting program effects initially reduced and then stabilized. More recent trials used more rigorous methods (e.g., more active control conditions and less risk of bias), samples that were generally older and included more girls, and evaluated interventions that on average had fewer sessions, were more often delivered by independent staff and made less use of time-out. However, none of these developments explained the initial reduction in effect size estimates during the first decades. CONCLUSIONS: Our findings suggest that estimates of parenting program effects are currently stable: Effect sizes are no longer reducing but there is also no evidence of increases over time. Experimentation with the content, delivery, and personalization of parenting programs is needed to identify ways to increase program effects.