Efficiency of explanatory Cognitive diagnostic Model under missing-at-random condition

[Speaker] Hung, Su Pin:1
[Co-author] Huang, Hung Yu:2
1:National Cheng Kung University (Taiwan), 2:University of Taipei (Taiwan)

A large number of cognitive diagnostic models (CDMs), have been developed and applied to analyze international academic surveys. Studies attempted to incorporate covariates into the simplest CDMs to enhancing the explanatory effect to the mastery profiles (Ayers, Rabe-Hesketh & Nugent, 2013; Park & Lee, 2014). Hung (2015) incorporated covariates into a generalized CDM-log-linear cognitive diagnosis model (LCDM; Henson, Templin & Willse, 2009). The present study aimed to further and deepen this effort by examining the efficiency of the new explanatory CDM under missing at random condition which was often occurred in large scale assessments. The performance of the estimator is evaluated via simulation, and the authors demonstrate how to apply the new model using actual data from large scale assessments. The result indicated the proposed model can improve the estimation of attribute mastery profile and reduced the standard errors under missing-at-random condition.
Advanced Search