ISSN 1671-3710
CN 11-4766/R
主办:中国科学院心理研究所
出版:科学出版社

Advances in Psychological Science ›› 2016, Vol. 24 ›› Issue (12): 1946-1960.doi: 10.3724/SP.J.1042.2016.01946

• Research Methods • Previous Articles    

Model-data fit test methods and statistics for cognitive diagnostic models

CHEN Fu1; XIN Tao2; LIU Yanlou3; LIU Tuo4; TIAN Wei2   

  1. (1 School of Psychology, Beijing Normal University, Beijing 100875, China) (2 Collaborative Innovation Center of Assessment towards Basic Education Quality, Beijing Normal University, Beijing 100875, China) (3 China Academy of Big Data for Education, Qufu Normal University, Qufu 273165, China) (4 School of Education Science, Tianjin Normal University, Tianjin 300384, China)
  • Received:2015-12-30 Online:2016-12-15 Published:2016-12-15
  • Contact: XIN Tao, E-mail: xintao@bnu.edu.cn

Abstract:

Cognitive diagnosis models (CDMs) are one family of psychometric models developed primarily for assessing examinees’ mastery and non-mastery of skills or attributes. The extent to which the cognitive diagnosis models’ hypotheses fit the real response patterns substantially influences the validity of cognitive diagnostic evaluation. The model-data fit tests in cognitive diagnostic theory are typically based on the item fit tests, the overall fit tests and the person fit tests. Comprehensive and structural introduction and evaluation of the model-data fit test methods and related statistics are intended to aid the future cognitive diagnostic studies and practice. Directions of future researches on model-data fit for cognitive diagnosis models include further inspection of the evaluation and comparison of fit test statistics under more complex simulation conditions, the improvement of the existing fit statistics and the development of novel fit statistics.

Key words: cognitive diagnostic models, test of model-data fit, item fit, overall fit, person fit