Evaluations of interventions for children and youth typically show marked variability in their results. This project will explore that variability using the effect size estimates and associated descriptive variables from seven existing meta-analyses. These meta-analyses each include 58 to more than 500 evaluations of the effects of prevention and treatment programs on a range of youth outcomes. Lipsey and colleagues will further code community and organizational factors to shed light on other possible sources of variability not yet represented in these data sets. They will estimate how much of the variability in observed effects is due to the nature of the programs, their implementation, the settings, and the participants. They will also estimate how much of the variability is due to differences in the methods used to gather information, the study designs, the procedures used in analysis, and sampling error. Finally, the investigators will estimate how much of the observed differences are due to factors that cannot be accounted for by any of these sources. This three-step strategy will be applied to a broad range of settings, including after-school programs, school-based emotional and behavioral programs, interventions for juvenile offenders and youth at risk for delinquency, and family interventions. They will examine variation across participant samples, community contexts, organizational features, intervention components, doses, training, and methodologies. Understanding how these factors influence the variability of program effects will help researchers and practitioners identify the conditions under which program effects are particularly robust and the extent to which they can be generalized across different settings and applications.
This project will explore that variability using the effect size estimates and associated descriptive variables from seven existing meta-analyses.
