Demographic Interactions of Differential Item Functioning in Attitudinal Scales

• In Event: Consequences of Testing and Rasch-Based Differential Item Functioning Analysis in Validation Research

Abstract

In this study, we use Rasch trees to incorporate multiple demographic variables into an analysis of differential item functioning (DIF) with two sets of abortion attitude items. These two sets of items address abortion legality in eleven circumstances that might be interpreted differently by subgroups within the population. The use of Rasch trees allows for the investigation of demographic variable interaction effects with DIF. The results indicated that abortion identity interacted with race in regard to differences in item difficulty parameters. These differences occurred with abortion items that did not include a reference to a specific gestational age in pregnancy. However, there were no differences in subgroups when the items included a four month gestational age.

Authors

• Nana Amma Asamoah, University of Arkansas

• Ronna C. Turner, University of Arkansas

• Wen-Juo Lo, University of Arkansas

• Brandon Crawford, Indiana University - Bloomington

• Kristen Jozkowski, Indiana University

• Sara McClelland, University of Michigan