Invariance testing in diagnostic classification models
Abstract
This study examined the invariance of items (based on sex of individuals) on a diagnostic classification screener, Body Image Rating Scale (BIRS), for Body Dysmorphic Disorder (BDD). The purpose of this study was to develop a method for testing invariance in diagnostic classification models (DCMs). The development of a diagnostic classification model for BDD was vital; specifically, using MPlus, a loglinear cognitive diagnostic model (LCDM) was created. The investigation to construct a method for testing invariance in DCMs began using a method for testing invariance for confirmatory factor analysis (CFA), which was already established. Adapting the syntax for invariance testing for CFA to work for DCMs was based on the how CFA models are similar to DCMs in comparing the type of data and latent variable appropriate for the model. Using MPlus, syntax was created to perform invariance testing for the LCDM for BDD.
URI
http://purl.galileo.usg.edu/uga_etd/bozard_jennifer_l_201005_mahttp://hdl.handle.net/10724/26248