This Article Full Text (OnlineFirst PDF) All Versions of this Article: 0013164407313370v1 68/5/742    most recent 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 Finch, W. H. Articles by French, B. F. Search for Related Content Social Bookmarking                         What's this?

Anomalous Type I Error Rates for Identifying One Type of Differential Item Functioning in the Presence of the Other

^* To whom correspondence should be addressed. E-mail: whfinch{at}bsu.edu.

	Abstract

A number of statistical methods exist for the detection of differential item functioning (DIF). The performance of DIF methods has been widely studied and generally found to be effective in the detection of both uniform and nonuniform DIF. Anecdotal reports suggest that these techniques may too often incorrectly detect the presence of one type of DIF in the presence of the other type (Type I error). The purposes of this simulation study are to ascertain whether these observations are in fact accurate and, if so, to gain some understanding as to the cause of the inflated Type I error. Results do support that the Type I error rates for detecting one type of DIF in the presence of the other are inflated for most common DIF detection techniques. Discussion focuses on potential causes of these results.

First published on February 6, 2008, doi:10.1177/0013164407313370

Educational and Psychological Measurement 2008;68:742.

A more recent version of this article appeared on October 1, 2008


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