Grade retention is a policy in which students who have not mastered the skills of a particular grade are required to repeat that grade. It is an educational policy used to combat social promotion though there is little rigorous evidence about the causal effects (positive or negative) of retention on student outcomes. Moreover, little is known about the effects of compensatory policies that many districts now operate in tandem with grade retention policies. This project will adapt and refine an innovative statistical approach (Classification Trees) for use in causal estimation for observational studies, using it to investigate the effects of grade retention on students' cognitive and behavioral outcomes. The project also will use an instrumental variables strategy to estimate effects of retention. Importantly, effects will be allowed to vary depending on the mix of compensatory programs available to children to better understand the implications of the findings within specific school contexts.
Grade retention is an area that is ripe for further investigation given the lack of rigorous causal analyses, the improbability of a randomized experiment to assess its effect, the need to examine the effects of retention in conjunction with other compensatory policies, and the policy importance of the topic. In addition, refinement of classification tree methods for use in causal inference will be useful for many researchers trying to answer causal questions in the absence of randomized experiments. This award was supported as part of the Fiscal Year 2005 Mathematical Sciences priority area special competition on Mathematical Social and Behavioral Sciences (MSBS).