Çeldiricilerin Yer Aldığı Q Matrisin Bilişsel Tanılamada ve Seçeneklerin Ağırlıklandırılmasında Kullanımının İncelenmesi

Abstract

The aim of the study was to develop a cognitive diagnostic test according to the MC-DINA model that utilizes distractors in cognitive diagnosis, to examine the characteristics of the items and the study group in the developed cognitive diagnostic test in comparison with the DINA and G-DINA models, and to determine the correlation between the scores obtained from different scoring methods according to the latent classes obtained when using EAP, MAP, and MLE prediction methods. Since distractors are of great importance in the ability of multiple-choice items to measure the desired behavior, MC-DINA model was used to obtain information about the skills of the students from their incorrect answers. In the study, the item parameters and latent class sizes obtained according to the DINA and G-DINA models were compared, and the choice probability parameters of the individuals in the latent classes associated with the item options were calculated for each option. The scoring methods used to determine the correlation between the scores obtained from different scoring methods according to the latent classes obtained when EAP, MAP, MLE estimation methods are 1-0 scoring and scoring by weighting the options differently. In weighting the options, the number of attributes with which the options were associated was used using the Q matrix. Due to the purpose of the research, the research was conducted with the study group since generalization was not made.