Exploring new methods to detect person misfit in CAT

Item scores that do not fit an assumed item response theory model may cause the latent trait value to be inaccurately estimated. Several person-fit statistics for detecting nonfitting response behavior for paper-and-pencil tests have been proposed. In the context of computerized adaptive testing, the use of person-fit analysis is hardly explored. Because it has been shown that the distribution of existing person-fit statistics is not applicable in a computer adaptive test (CAT), new person-fit statistics are proposed, and critical values for these statistics are derived from existing statistical theory. The theoretical and empirical distributions are compared, and a power study is performed.