Seminar in Psychometrics

COMPS seminar main page

DIF detection by comparing item response curves

Date and time: Tuesday, April 11, 2023 (2:00 PM CET)
Place ICS CAS room 318, Pod Vodárenskou věží 2, Prague 8, also on Zoom.

Differential item functioning (DIF) is a phenomenon when two respondents with the same underlying latent trait but from different social groups have different probabilities to endorse an item in multi-item measurement.
Many DIF detection methods are derived from comparison of regression curves of the reference and focal group, however, most of them are limited to detection of DIF caused either by difference in difficulty or discrimination parameters. Possible methodological gap can be filled by generalized logistic regression models (Drabinová & Martinková, 2017) implemented within the difNLR R package (Hladká & Martinková, 2020). These models offer a possibility to account for guessing or inattention when answering and moreover to test whether these item characteristics differ between the two groups. In this talk we will discuss several approaches to estimate item parameters (Hladká, Martinková, & Brabec, 2023).
Another approach which can detect such differences is a nonparametric comparison of regression curves. This method can be useful for example in situations when no specific true model is expected.

Drabinová, A., & Martinková, P. (2017). Detection of differential item functioning with nonlinear regression: A non-IRT approach accounting for guessing. Journal of Educational Measurement, 54(4), 498-517,

Hladká, A., & Martinková, P. (2020). difNLR: Generalized logistic regression models for DIF and DDF detection. The R Journal, 12(1), 300-323,

Hladká, A., Martinková, P., & Brabec, M. (2023). Parameter estimation in generalised logistic model with application to DIF detection. arXiv preprint.

Adéla Hladká
ICS CAS, Czech Republic

Adéla Hladká obtained her PhD in mathematical statistics from Charles University, Prague, Czech Republic. She is now a postdoc fellow at the Institute of Computer Science of the Czech Academy of Sciences. She is the developer of an open-source 'difNLR' R package for DIF detection using generalized logistic regression models. She is also a co-author of the 'ShinyItemAnalysis' R package and its interactive online application. The center of her interest is a detailed description of between-group differences with analysis of DIF.