The Construction and Validation of a Test for Measuring Critical Thinking of Students: Using Item-Response Theory (IRT)

This study aimed to build and validate the Critical Thinking Scale using item-response theory. For this purpose, a sample size (n=7478) was selected from Qom and Sheikh Mofid university students and the questionnaire was implemented with their cooperation. Three-parameter model was detected more suitable among various IRT models. Data was analyzed through three-parameter logistic model. Exploratory factor analysis was applied to evaluate the theory’s assumptions (one-dimensional and local independence). Factor analysis’s results confirmed the existence of a dominant factor and local independence. Item Information Function was used to evaluate the propriety of questions. Item Information Function examination showed that some of questions had very little information or the information function was not distributed over ability range. The rest of the item parameters were estimated again after removing inappropriate items. New estimations showed that the range of Discriminant Coefficient was .133 to 1.899 and difficulty coefficient -.990 to 2.780. In addition, coefficient conjecture was .0 to .0790. Eventually, the results showed that the Critical Thinking Scale with 7 components and 29 items has desirable properties to measure the structure.