Item Response Models for Time Limit Test

Abstract

In the first part of the talk, a speeded item response model is proposed. We consider the situation where examinees may retain harder items to a later test period in a time limit test. With such a strategy, examinees may not finish answering some of the harder items within the allocated time. In the proposed model, we try to describe such a mechanism by incorporating a speeded-effect term into the two-parameter logistic item response model. A Bayesian estimation procedure of the current model using Markov chain Monte Carlo is presented. The methodology is applied to physics examination data of the Department Required Test for college entrance in Taiwan for illustration. ??? ? In the second part of the talk, non-response is regarded as a response category. Based on the assumption that being not in the problem solving process leads to non-responses and is the only source of non-response, we find that the non-response model in Knott, Albanese, and Galbraith (1990) is broad enough for test speededness data with non-response. To relax the assumption, a tree model with 4 end nodes of not-all-distinct response categories is suggested for non-response modeling for the condition that there is certain probability of non-response for being not in the solving process and not knowing the answer. Penalized quasi-likelihood estimation is suggested. Simulations are conducted to demonstrate the advantage of considerations of the two non-response models. ???? If time permits, I will briefly introduce my other recent research: (i) Unifying Differential Item Functioning in categorical CFA and GRM under a discretization of a normal variant.(ii) Modeling online learning in cognitive diagnostic models. (iii) Modeling longitudinal progression of children post cochlear implant. Keywords : item response model, non-response, test speededness.