CAJ | 학술논문

认知诊断以微观认知角度对被试做出准确评估与反馈的优势在心理与教育测量领域展现出巨大发展潜力，但要利用这种优势就必须确保Q矩阵的合理性。研究表明错误界定的Q矩阵会严重影响模型参数估计和被试分类准确性。本文从参数化与否出发介绍6种基于被试作答反应的Q矩阵估计方法的基本思想、实现步骤及应用情况，总结这些方法的特点并展望未来研究方向，为认知诊断研究及应用提供借鉴和基础。
인지진단이미관인지각도대피시주출준학평고여반궤적우세재심리여교육측량영역전현출거대발전잠력，단요이용저충우세취필수학보Q구진적합이성。연구표명착오계정적Q구진회엄중영향모형삼수고계화피시분류준학성。본문종삼수화여부출발개소6충기우피시작답반응적Q구진고계방법적기본사상、실현보취급응용정황，총결저사방법적특점병전망미래연구방향，위인지진단연구급응용제공차감화기출。
Cognitive Diagnosis Assessment(CDA)has become a huge developing potential in fields of educational and psychological measurement with advantages which could make accurate evaluation and feedback of participants from the view of microscopic cognition. However, it is necessary to ensure the rationality of the constructed test Q-matrix which indicates the set of skills that are required to answer a particular item correctly, before taking these advantages. It is widely recognized that a misspecification of the Q-matrix can negatively affect the estimation of the model parameters, which may then result in the misclassification of examinees. From the perspective of the parameterized or not, we introduces the basic ideas, implementation steps and applications study of six methods, which are based on the participants’response data. Last, four directions for future study are discussed to researchers and practitioners.