Child and Family Center Barcelona Conference Room
2:30-3:30 pm
November 12 , 2004
John Light
Child & Family Center
"Average Causal Effect (CACE) Analysis of the Project Alliance Family Intervention: Findings, Methodology, and Implications"
Abstract: Complier Average Causal Effect (CACE) analysis is a statistical method that can be used to estimate the effect of a randomized intervention trial in situations where the intervention involves encouragement to expose one’s self to certain experiences, but actual exposure is voluntary and not universal. CACE is an alternative to historically-standard Intent-To-Treat (ITT) analysis, where intervention and control groups are compared regardless of treatment group members’ level of exposure to the intervention. The discussion will center on the underlying logic of CACE, model assumptions and identification requirements, and implications for improving effect size estimates. An example outcome analysis is presented using data from CFC’s Project Alliance, an ongoing longitudinal randomized intervention trial which can be thought of as a type of “encouragement” design in the sense described above.
Jo, Booil (2002). Estimation of intervention effects with noncompliance: Alternative model specifications. Journal of Educational and Behavioral Statistics, 27: 385-409.
