The effects of routing and scoring within a computer adaptive multi-stage framework

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Andrew Dallas (Creator)
Institution
The University of North Carolina at Greensboro (UNCG )
Web Site: http://library.uncg.edu/
Advisor
Richard Luecht

Abstract: This dissertation examined the overall effects of routing and scoring within a computer adaptive multi-stage framework (ca-MST). Testing in a ca-MST environment has become extremely popular in the testing industry. Testing companies enjoy its efficiency benefits as compared to traditionally linear testing and its quality-control features over computer adaptive testing (CAT). Test takers enjoy being able to go back and change responses in review time before being assigned to the next module. Lord (1980) outlined a few salient characteristics that should be investigated before the implementation of multi-stage testing. Of these characteristics, decisions on routing mechanisms have received the least attention. This dissertation varied both item pool characteristics such as the location of information, and ca-MST configuration characteristics such as the ca-MST configuration design (e.g., 1-3, 1-2-3, 1-2-3-4). The results from this study hope to show that number correct scoring can serve as a capable surrogate for IRT calibrations at each step and that even if three-parameter scoring models are used at the end that the number correct method will not misroute as compared to traditional methods.

Additional Information

Publication
Dissertation
Language: English
Date: 2014
Keywords
Computer Adaptive Multi-Stage Testing, Routing Methods
Subjects
Educational tests and measurements
Computer adaptive testing

Email this document to