Bias and Information of Bayesian Adaptive Testing.
Abstract
The present study was designed to further investigate the nature of the bias and the information characteristics of Owen's Bayesian adaptive testing strategy and to examine possible causes of the bias. Factors investigated included (1) the effects of item discrimination, (2) the effects of fixed vs. variable test length, and (3) the effect of an accurate prior theta estimate. Monte carlo simulation of Owen's adaptive test was used.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1983
- Accession Number
- ADA129280
Entities
People
- David J. Weiss
- James R McBride
Organizations
- University of Minnesota