Files in this item



application/pdf9712323.pdf (4MB)Restricted to U of Illinois
(no description provided)PDF


Title:Applications of computational statistics in cognitive diagnosis and IRT modeling
Author(s):Jiang, Hai
Doctoral Committee Chair(s):Stout, William F.
Department / Program:Statistics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Psychology, Psychometrics
Abstract:The identifiability and estimability of the parameters for the Unified Cognitive/IRT Model are studies. A calibration procedure for the Unified Model is then proposed. This procedure uses the marginal maximum likelihood estimation approach and utilizes the EM algorithm. It differs from other calibration procedures for IRT models such as BILOG in that we use Genetic Algorithm in the maximization (M) Step of the EM algorithm. Procedures for classifying examinees are also proposed. A simulation study shows that our calibration procedure works remarkably well for a wide variety of model settings.
A new regression correction used to adjust for the Type I error inflating and estimation biasing influence of group target ability differences on the DIF detection procedure SIBTEST is proposed. This new regression correction uses a piecewise linear regression of the true on observed matching subtest scores. A realistic simulation study of the new approach shows that when there is a clear group ability distributional difference, the new approach displays improved SIBTEST Type I error performance, and when there is no group ability distributional difference, its Type I error rate is comparable to the current SIBTEST. A power study indicates that the new approach has on average similar power as the current SIBTEST. It is thus concluded that the new version of SIBTEST seems appropriately robust against sizable Type I error inflation while retaining other desirable features of the current version.
Issue Date:1996
Rights Information:Copyright 1996 Jiang, Hai
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9712323
OCLC Identifier:(UMI)AAI9712323

This item appears in the following Collection(s)

Item Statistics