A Hierarchical Model for Accuracy and Choice on Standardized Tests
psychometrics
journal
Summary
This paper presents a hierarchical framework that jointly models both the accuracy of test-taker responses and which items they choose on standardized tests, providing new theoretical results on when allowing choice improves score precision. An “answer two, choose one” design is introduced and validated experimentally, with the methodology released as the cIRT R package.
My contributions
(Your role on this paper, to be filled in by operator.)