A model for multiple‐choice problem selection

This paper proposes a new model for the alternative selection process for multiple‐choice test items. The model has the following features in the description: (1) it can cope with the real‐time data, and (2) it is based on the entropy model. The feature of each selection item can be represented using the estimated parameters. Furthermore, using the indices derived from the model parameters, the extent of “illusion” among the test items can be represented quantitatively. The method is applied to the actual data and the following points are discussed: (1) validity of the model, (2) analysis of the characteristics of the selection item, and (3) characteristics analysis of alternative selection process. The effectiveness of the method is demonstrated.