Item Parameter Estimation And Model Selection In DINO And DINA Models

Cognitive diagnosis model is a kind of psychological measurement model.It has enormous potential to provide abundant cognitive information,can be used to assess the strengths and weaknesses of students.Although many researchers more and more understand the cognitive diagnosis model,but the framework has not been fully used.There are two main limitations: first of all,compared with the traditional item response model,a cognitive diagnosis model,in some cases,a parameterized model is a novel,may be more complex,therefore,many researchers familiar with the properties of these models and their lack of.Second,the traditional item response model can be commercially available through the use of the soft analysis,cognitive diagnosis model is more widely used,however,has been impeded by the lack of commercially available software.In order to solve this problem,this article focuses on an easy to handle and can explain and apply a wide range of cognitive model of the diagnosis model-DINO model and DINA model.This paper mainly adopts marginal maximum likelihood estimation to estimate DINO model item parameters,and by using the rule of information to solve model selection problem between DINO model and DINA model.Finally,this paper introduced the data using MATLAB software simulation,and the simulation results are analyzed.