The Constrained Diagnostic Classification Models for Polytomous Attributes

• In Event: AERA Poster Session 13
  In Poster Session: Quantitative and Qualitative Approaches to Assess Students' Cognitive Performance

Abstract

DCMs have been used to provide dichotomous feedback about students’ mastery and non-mastery levels. In educational contexts, further delineating mastery categories into 3, 4 or 5 levels may be useful for meaningfully grouping students to provide tailored instruction or interventions. The polytomous DCM (PDCM) classifies students into more than two mastery levels for each attribute. However, it often requires longer test lengths and larger sample sizes. This study proposed two constrained models, paDINA-1 and paDINA-2 that can reduce the number of item parameters. Through a simulation study, we investigated the sample size and test length requirement for the two constrained models. We conducted an empirical study to examine how the attribute classifications were different under the PDCM and constrained models.

Author

• Yu Bao, James Madison University