 Two Simple Approaches to Overcome a Problem With the Mantel-Haenszel Statistic: Comments on Wang, Bradlow, Wainer, and Muller (2008)Sandip Sinharay and Neil J. Dorans Journal of Educational and Behavioral Statistics, 2010, vol. 35, issue 4, 474-488 Abstract: The Mantel-Haenszel (MH) procedure (Mantel and Haenszel) is a popular method for estimating and testing a common two-factor association parameter in a 2 × 2 × K table. Holland and Holland and Thayer described how to use the procedure to detect differential item functioning (DIF) for tests with dichotomously scored items. Wang, Bradlow, Wainer, and Muller showed that the MH procedure unexpectedly often found DIF where there was none for very easy items. They showed that their suggested Bayesian procedure for DIF detection did not have the same problem. We discuss a simpler solution of the problem—the use of the standardized difference in proportion correct (STD P-DIF) procedure in conjugation with the MH procedure—that has been used by DIF analysts since 1988. We also show, using results from real and simulated data, that it is possible to overcome the problem using the empirical Bayes (EB) procedure of Zwick, Thayer, and Lewis, which also is computationally simpler than the procedure suggested by Wang et al. Keywords: empirical Bayes; Mantel-Haenszel statistic; prior distribution (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:sae:jedbes:v:35:y:2010:i:4:p:474-488 Access Statistics for this article More articles in Journal of Educational and Behavioral Statistics |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.