Choosing an Appropriate Cognitive Diagnostic Model for Reading Comprehension Tests: A Case Study of the Graduate Entrance Exam of the English language

Cognitive Diagnostic Assessment is a type of educational assessment that serves to identification and diagnosis of the learning disabilities through the application of psychometric models. Presence of numerous models, makes it crucial choose an appropriate Cognitive Diagnostic Model for data analysis. The purpose of this study was to determine a fitting model for reading comprehension tests. The examined data in this study were included responses from 3,000 participants in the Graduate Entrance Exam for English language field, which were selected randomly out of 16044 people. After Creation of matrix Q, two general models (G-DINA and LCDM), two compensatory models (ACDM and DINO) and two non-compensatory models (RRUM and DINA) were considered for comparison. In terms of Chi-squared maximum Fit Index, none of the models were fitted; but in contrast the results of SRMSR index showed that the fit of all data were acceptable with the data. A comparison of Relative Fit Indices cleared that G-DINA and ACDM models have a stronger fit and DINO and DINA have a poorer fit in comparison to the other models. Also, investigating the Question Fit index showed that only G-DINA and LCDM have a fit with all questions.