This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Yan, D. Articles by Stocking, M. Search for Related Content Social Bookmarking                         What's this?

Adaptive Testing With Regression Trees in the Presence of Multidimensionality

Educational Testing Service

It is unrealistic to suppose that standard item response theory (IRT) models will be appropriate for all the new and currently considered computer-based tests. In addition to developing new models, we also need to give attention to the possibility of constructing and analyzing new tests without the aid of strong models. Computerized adaptive testing currently relies heavily on IRT. Alternative, empirically based, nonparametric adaptive testing algorithms exist, but their properties are little known. This article introduces a nonparametric, tree-based algorithm for adaptive testing and shows that it may be superior to conventional, IRT-based adaptive testing in cases where the IRT assumptions are not satisfied. In particular, it shows that the tree-based approach clearly outperformed (one-dimensional) IRT when the pool was strongly two-dimensional.

Key Words: computerized adaptive testing • item response theory (IRT) • regression tree

Journal of Educational and Behavioral Statistics, Vol. 29, No. 3, 293-316 (2004)
DOI: 10.3102/10769986029003293


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?