Abstract:
|
Under rolling enrollment, individuals or entities choose when to enroll in an intervention. This staggered enrollment pattern presents challenges for a rigorous impact evaluation and can even introduce bias if the evaluation design does not account for selection pressures that may lead specific entities to enroll at specific times. In this talk, I describe solutions to some of the common challenges, focusing on comparison group selection and the optimal matching tool GroupMatch. As an example, we will look at matching results from the evaluation of the Medicare Care Choices Model, which provides supportive and palliative services to Medicare beneficiaries at the end of life.
|