Files
Abstract
Diagnostic Classification Model (DCMs) are psychometric models that enable the categorization of examinees based on their proficiency (proficient or non-proficient) in specific skills or attributes. The information from DCMs can aid educators in identifying specific areas where examinees require additional support and provide actionable feedback. This study examines the robustness of a recently developed DCM, the One-Parameter Log-Linear Cognitive Diagnostic Model (1-PLCDM), that prioritizes interpretability and ease of application. Through a simulation study, we demonstrate that the 1-PLCDM maintains robust classification accuracy and reliability when the assumptions of attribute independence and simple Q-matrix structure are violated.