Tackling Missing Data from Omitted and Not-reached Items: A Model-based Approach

• In Event: Missing Data

Abstract

In low-stakes assessment, there is often a substantive amount of missing data observed from omitted and not-reached items. Missing item responses are often MNAR and lead to biased parameter estimates if not handled properly. This study evaluated and compared the accuracy of parameter estimates between classical approaches and model-based approaches.

Authors

• Dongwei Wang, UMass Amherst

• Craig Wells, UMass Amherst