Computerized Adaptive Trait Measurement: Problems and Prospects.
Abstract
This symposium consisted of four papers and the comments of two discussants. C. David Vale. Problem: Strategies of Branching through an Item Pool. This paper describes a variety of strategies for adapting tests to the trait level of each individual on the basis of the testee's responses to previously administered items. Based on data from computer simulations, the various strategies are compared. James R. McBride. Problem: Scoring Adaptive Tests. Several approaches to scoring adaptive tests are described. James B. Sympson. Problem: Evaluating the Results of Adaptive Testing. Six component elements of a testing procedure are described; it is suggested that proper evaluation of a testing procedure should be based on consideration of these elements as separable components. Nancy E. Betz. Prospects: New Types of Information and Psychological Implications. Several types of new information available from computerized adaptive measurement are described. Comments by the discussants, Robert L. Linn of the University of Illinois and R. Darrell Bock of the University of Chicago, include a discussion of some of the limitations of the research presented, some differing interpretations, and suggestions for future research in adaptive testing.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1975
- Accession Number
- ADA018675
Entities
People
- C. David Vale
- James B. Sympson
- James R McBride
- Nancy E. Betz
- R. Darrell Bock
Organizations
- University of Minnesota