Diagnostic Modeling of Skill Hierarchies and Cognitive Processes with MLTM-D

This chapter formally describes the multicomponent latent trait model for diagnosis (MLTM-D; Embretson S.E., Yang X, Psychometrika 78:14–36, 2013) and then provides examples of applications to diagnose broad and narrow skills, as well as measure processing complexity and attainment. MLTM-D can be applied to diagnose either skill mastery or cognitive processing capabilities of examinees. MLTM-D is readily applicable to diagnose hierarchically-structured skills or to assess cognitive processes with postulated sources of complexity. That is, MLTM-D is a multidimensional conjunctive model for item responses that are impacted by varying underlying components with specifiable sources of complexity. MLTM-D can be applied to assess both processing competencies of examinees and the impact of the postulated features on process difficulty.