Dimensionality assessment of proximity-based data in unfolding model applications
Williams, Elizabeth Jane
MetadataShow full item record
Dimensionality assessment of measurement instruments in psychology is no easy task. Within the field of IRT, much attention has been given to the issue of dimensionality assessment for item responses that are consistent with cumulative models. This has resulted in several well-performing dimensionality assessment procedures for data that are a monotonically increasing function of the latent trait. However, there has been very little attention paid to dimensionality assessment for unfolding data. The development of a simple, yet accurate, dimensionality assessment tool for ideal-point responses is crucial to proper utilization of multidimensional unfolding IRT models like the Multidimensional Generalized Graded Unfolding Model (MGGUM; Roberts & Shim, 2010), and will increase the practicality of the MGGUM in common psychological research areas. The current study satisfies the need for a more robust dimensionality assessment tool for unfolding data, specifically when test data generated by the MGGUM have homogenous item structure.