The Application of Multiple Choice Model (MCM) in Item Analysis and the Comparison of its Fit and Information with Three Parametric model (3PM): Case Study: The Biology Test of University's Entrance Exam 2012

The main objective of the present study is to apply multiple-choice model (MCM) for analyzing multiple choice items and comparing its fit and information with the three parametric model (3PM). To this end, among all the participants in biology test (contained 50 four-choice items) of University’s entrance exam in the field of Natural Sciences in 2012(1391), 5000 participants were randomly selected as the sample by SPSS software. In order to analyze the test and its items based on the multiple choice model, the assumptions of unidimensionality were firstly investigated by NOHARM software. Then, the test items were analyzed by MULTILOG program. Three-parameter model, which is the closest one to multiple-choice model in terms of parametric structure, was selected as the most appropriate binary model for comparing its information and fitting with multiple-choice model. The results indicated the efficacy of multiple-choice model for analyzing distractors. In comparing the fit of 3PM with MCM, the results showed better fitness of 3PM to the whole test than multiple-choice model. In addition, results indicated the appropriateness of multiple-choice model in accuracy of estimating the ability of subjects with lower-than-average ability, while 3PM estimates the ability of higher than average subjects more accurately. Accordingly, it is recommended that MCM model can be applied for farther knowledge about items’ distractors performance along with other methods.