Abstract
Logistic regression provides a flexible framework for detecting various types of differential item functioning (DIF). Previous efforts extended the framework by using item response theory (IRT) based trait scores, and by employing an iterative process using group-specific item parameters to account for DIF in the trait scores, analogous to purification approaches used in other DIF detection frameworks. The current investigation advances the technique by developing a computational platform integrating both statistical and IRT procedures into a single program. Furthermore, a Monte Carlo simulation approach was incorporated to derive empirical criteria for various DIF statistics and effect size measures. For purposes of illustration, the procedure was applied to data from a questionnaire of anxiety symptoms for detecting DIF associated with age from the Patient-Reported Outcomes Measurement Information System.
Highlights
Standardized tests and questionnaires are used in many settings, including education, psychology, business, and medicine
Previous efforts extended the logistic regression differential item functioning (DIF) technique into a framework known as difwithpar (Crane et al 2006) by using item response theory (IRT) based trait estimates and employing an iterative process of accounting for DIF in the trait estimate with the use of group-specific IRT item parameters for items identified with DIF (Crane et al 2006, 2007b,c, 2004)
Detecting DIF: Flag DIF items based on the detection criterion ( “Chisqr”, “R2”, or “Beta”) and a corresponding flagging criterion specified by the user; for criterion = “Chisqr” an item is flagged if any one of the three likelihood ratio χ2 statistics is significant
Summary
Standardized tests and questionnaires are used in many settings, including education, psychology, business, and medicine. Previous efforts extended the logistic regression DIF technique into a framework known as difwithpar (Crane et al 2006) by using IRT based trait estimates and employing an iterative process of accounting for DIF in the trait estimate with the use of group-specific IRT item parameters for items identified with DIF (Crane et al 2006, 2007b,c, 2004) This framework has been found to be facile at accounting for multiple sources of DIF and paying specific attention to DIF impact. Several values may be used for flagging criteria in analyzing a single dataset, resulting in varying numbers of items identified with DIF, but fairly consistent DIF impact for individuals and groups across different values for the flagging criteria (Crane et al 2007b). The resulting R package lordif is available from the Comprehensive R Archive Network at http://CRAN.R-project.org/package=lordif
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
