Font Size: a A A

Exploration And Achievement Of A Technology For Computer Adaptive Test

Posted on:2012-04-24Degree:MasterType:Thesis
Country:ChinaCandidate:W LiuFull Text:PDF
GTID:2178330332989942Subject:Education Technology
Abstract/Summary:PDF Full Text Request
Examination is an important part of teaching activities while the examination has also changed from traditional written form to computer-mediated form. In order to improve the quality and efficiency of examination and provide better service for teaching, how to make full use of computer in computer-mediated form has drawn the thinking and attention of many researchers. In computer adaptive testing (CAT) based on item response theory (IRT), computer not only act as a "medium", but also plays the role as "decision maker" , which is favored by many researchers and enable computer adaptive testing (CAT) to be widely used in examination.In this article, literature research and experimental research have been mainly adopted. By literature research, first, compare the advantages and disadvantages of classical test theory (CTT) with those of item response theory (IRT), only to find that item response theory has the characters of item parameter invariance and capacity parameter invariance, which provides powerful theoretical support for the research and development of computer adaptive testing. Second, give systematic introduction to item response theory. By experimental research, explore the key technology of computer adaptive testing and starting from the fitting testing of data and model, apply Monte Carlo method and standard residual error method to determine suitable logistic model. In this article, 3-parameter logistic will be finally selected. Then, make use of Bayes method to complete the determination of project parameter and organize these items to complete preliminary construction of examination pool; in the following, it is the core links of CAT, that is, select items adaptive to examination ability to determine the capacity of examination, which make use of test information function and item information function to select the most adaptive examination ability project and use Bayes maximum a posteriori method to finally determine its parameter; finally, give overall evaluation test to the quality. In this article, after the exploration of CAT technology, combining with the public course and with the students of 2007 in Shandong Normal University as the study object, CAT system based on Modern Education Technology is developed to make educational technology teachers in college and universities to be free from tedious examination.
Keywords/Search Tags:CAT, IRT, Bays parameter estimation, Fitting testing
PDF Full Text Request
Related items