ABSTRACT
This chapter provides some background on multistage testing (MST) using diagnostic models. Diagnostic classification models (DCM; e.g., DiBello, Stout, and Roussos 1995; Rupp, Templin, and Henson 2010; von Davier 2005) are constrained latent class models (Maris 1999; von Davier 2005, 2009) that are intended to enable simultaneous measurement of several (mastery type) latent dichotomies. To date, there are few examples of how to apply computer adaptive testing (CAT) principles when estimating proficiency using diagnostic classification models (we will use the acronym CD-CAT for these types of applications). To our knowledge, no attempts have been made to apply diagnostic models to the selection of stages in MST.
