Premium
The Impact of Omitted Responses on the Accuracy of Ability Estimation in Item Response Theory
Author(s) -
Ayala R. J.,
Plake Barbara S.,
Impara James C.
Publication year - 2001
Publication title -
journal of educational measurement
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.917
H-Index - 47
eISSN - 1745-3984
pISSN - 0022-0655
DOI - 10.1111/j.1745-3984.2001.tb01124.x
Subject(s) - estimation , item response theory , statistics , psychology , monte carlo method , econometrics , contrast (vision) , test (biology) , maximum likelihood , psychometrics , computer science , mathematics , artificial intelligence , management , economics , paleontology , biology
Practitioners typically face situations in which examinees have not responded to all test items. This study investigated the effect on an examinee's ability estimate when an examinee is presented an item, has ample time to answer, but decides not to respond to the item. Three approaches to ability estimation (biweight estimation, expected a posteriori, and maximum likelihood estimation) were examined. A Monte Carlo study was performed and the effect of different levels of omissions on the simulee's ability estimates was determined. Results showed that the worst estimation occurred when omits were treated as incorrect. In contrast, substitution of 0.5 for omitted responses resulted in ability estimates that were almost as accurate as those using complete data. Implications for practitioners are discussed.