A Comparison of Reviewing and Removing Misfitting Items During Rasch Analysis

• In Event: 54.080 - Roundtable Session 22
  In Roundtable Session: 54.080-22 - Rasch Analysis

Abstract

Rasch measurement model requires data to fit the model. Within Rasch measurement model, the fit statistics indicate how accurately or predictably data fit the Rasch measurement model. The items with fit statistics outside the recommended ranged are identified as misfitting. Usually, misfitting items are removed from the analysis. In this study we use different approach of handling misfit data; reviewing the misfitting items. We explored two methods of handling misfitting data and explored any differences.

Authors

• Subin Hona, CRP, Incorporated

• Trent Haines, Morgan State University