Paul De Boeck: Pervasive DIF

March 12, 2024

Tuesday, March 12, 2024

2:00 - 4:00 PM (PDT) at Berkeley Way West 1212 and via Zoom

Open to GSE faculty, students, and community.

Request a zoom link from convenors@bear.berkeley.edu

Abstract:

Most Differential Item Functioning (DIF) detection methods minimize DIF in favor of group mean differences and would not detect DIF as a possible source of group mean differences. We believe that pervasive DIF can go unnoticed whereas it may indeed cause group mean differences. We present a method to deal with DIF that possibly pervades a large part of the test. For illustrative purposes, we use real data on a neuropsychological test (Boston Naming Test) applied to two groups: Caucasian and African American respondents.

About the speaker:

Paul De Boeck obtained a Ph.D. degree in psychology from KU Leuven, a Dutch-speaking university in Belgium, with a focus on mathematical psychology and personality psychology. Until 2009, he held a position as a professor of psychological assessment at KU Leuven. During his tenure at the university, he served as the Chair of the Department of Psychology, Chair of the University Research Council, Vice-President for Research, and as a member of the Executive Committee and Board of Directors.

From 2009 to 2012, he was affiliated with the University of Amsterdam as a professor of psychological methods. Since 2012, he has been associated with Ohio State University (Columbus, OH) as a professor of quantitative psychology, with a courtesy affiliation at KU Leuven. In August 2023, he became an Emeritus Professor at OSU.

He served as the President of the Psychometric Society in 2008 and was honored with the 2021 Career Award from the Psychometric Society for Lifetime Achievement. Additionally, he was the founding editor of the Applied Research and Case Studies (ARCS) section of Psychometrika.