Font Size: a A A

A Comparative Study Of Pretest Item -Calibration Methods In Computerized Adaptive Testing

Posted on:2008-09-23Degree:MasterType:Thesis
Country:ChinaCandidate:J GuoFull Text:PDF
GTID:2178360215479037Subject:Probability theory and mathematical statistics
Abstract/Summary:PDF Full Text Request
Based on the framework of item response theory(IRT), this paper compares and evaluates four pretest item-calibration/scaling methods in computerized adaptive testing(CAT): one EM cycle (OEM), multiple EM cycle (MEM), Stocking's Method A and Stocking's Method B. This paper describes their respective scope and weaknesses of each methods, it provides the basis for the investigation of four pretest item-calibration/scaling methods in computerized adaptive testing.This paper is divided into three chapters, the content is as follows:Chapter I simply introduces the model of the item response theory and some useful methods of parameter estimation.Chapter II introduces four item-calibration/scaling methods in computerized adaptive testing.Chapter III compares the four item-calibration/scaling methods in computerized adaptive testing and carries on the simulation to the real data according to Stocking's Method A.
Keywords/Search Tags:Computer Adaptive Test(CAT), Item Response Theory(IRT), One EM Cycle(OEM), Multiple EM Cycle(MEM), Stocking's Method A, Stocking's Method B, Joint Maximum Likelihood Esti-mation(JMLE), Marginal Maximum Likelihood Estimation(MMLE)
PDF Full Text Request
Related items