Exam Quantity Analysis And Student's Ability Assessment Based On IRT Model

Item Response model is the most widely used education measurement model. It is using test scores, analyzing the questionnaire data by testers topic quality or the potential psychological characteristics (latent trait), mathematical model of its significance lies in can guide the project staffing, selection and test can be assessed subjects simultaneously in the potential capacity corresponding project.In this paper, first, the item response model, and then illustrates the introduction of the model parameters estimation and Gibbs sampling algorithms, including Gibbs sampling algorithms in the parameter estimation has dominated Gibbs sampling algorithms, this paper the implementation process, convergence conditions and the selection of the initial value are described. Finally, the model is applied in practice, to 110 students had a math test, collect data, its answer to the question through this model quality level and detailed analysis, also can find assess a different students to study mathematics potential ability.