Using a Nonparametric Unfolding Item Response Theory Approach for Attitude Measurement

• In Event: Rasch Models: Considerations and Novel Uses

Abstract

This study describes the use of nonparametric unfolding IRT models for attitude measurement. We first provide a theoretical framework for nonparametric unfolding models. This framework discusses nonparametric IRT models as an intermediate step between deterministic and parametric models in examining unfolding data. Next, we conduct empirical analyses related to attitudes towards censorship. Data were collected through a Qualtrics survey on Amazon Mechanical Turk (N=735). The Multiple Unidimensional Unfolding model (MUDFold) and Generalized Graded Unfolding model (GGUM) were used to analyze the data. The purpose of the study is to introduce the application of nonparametric and parametric unfolding models in combination to analyze unfolding data. Nonparametric models provide useful approaches to diagnose model misfit that can complement parametric models.

Authors

• Ye Yuan, University of Georgia
  Presenting Author

• George Engelhard, University of Georgia
  Non-Presenting Author