Imputation Vis-a-Vis Research Standards in Clearinghouses

• Activity Number: 303 - Imputation Vis-a-Vis Research Standards in Clearinghouses
• Type: Invited
• Date/Time: Wednesday, August 11, 2021 : 3:30 PM to 5:20 PM
• Sponsor: Social Statistics Section
• Author(s): Rebecca R Andridge* and Elias Walsh* and Joshua Polanin* and Paul T von Hippel* and Joerg Drechsler* and Ryan Williams*
• Companies: The Ohio State University College of Public Health and Mathematica and American Institutes for Research and University of Texas, Austin and Institute for Employment Research (Germany) and
• Keywords: WWC; Randomized trials; Evaluation research; Sensitivity analysis

Abstract

The What Works Clearinghouse (WWC) is an influential force in American education research. Studies that are endorsed by the WWC as providing strong evidence are more influential. Although the most recent version of the WWC handbook (v4.1) is less opposed to the use of imputed outcome data than prior versions, it can still be difficult to get a high study rating from the WWC if the study relies on imputed outcomes. The panel will discuss attitudes and misconceptions about imputation across a range of applications, including the appropriateness of WWC standards and alternative approaches to judging the quality of impact estimates involving imputation.

We have two panelists who were involved in preparing the WWC handbook and three who are active in conducting evaluations with the aid of imputation. One panelist from Germany brings us an international perspective.

David Judkins of Abt Associates will chair the discussion.