TitleImpact of Item Drift on Candidate Ability Estimation
Publication TypeConference Paper
Year of Publication2011
AuthorsHagge, S, Woo, A, Dickison, P
Conference NameAnnual Conference of the International Association for Computerized Adaptive Testing
Date Published10/2011
Keywordsitem drift
Abstract

For large operational pools, candidate ability estimates appear robust to item drift, especially under conditions that may represent ‘normal’ amounts of drift. Even with ‘extreme’ conditions of drift (e.g., 20% of items drifting 1.00 logits), decision consistency was still high.