difNLR: Generalized Logistic Regression Models for DIF and DDF Detection

0532886 - ÚI 2021 RIV AT eng J - Journal Article
Hladká, Adéla - Martinková, Patrícia
difNLR: Generalized Logistic Regression Models for DIF and DDF Detection.
R Journal. Roč. 12, č. 1 (2020), s. 300-323. ISSN 2073-4859
Institutional support: RVO:67985807
Keywords : differential item functioning * differential distractor functioning * group disparities * generalized logistic regression
OECD category: Statistics and probability
Impact factor: 3.984, year: 2020
Method of publishing: Open access

Differential item functioning (DIF) and differential distractor functioning (DDF) are important topics in psychometrics, pointing to potential unfairness in items with respect to minorities or different social groups. Various methods have been proposed to detect these issues. The difNLR R package extends DIF methods currently provided in other packages by offering approaches based on generalized logistic regression models that account for possible guessing or inattention, and by pro viding methods to detect DIF and DDF among ordinal and nominal data. In the current paper, we describe implementation of the main functions of the difNLR package, from data generation, through the model fitting and hypothesis testing, to graphical representation of the results. Finally, we provide a real data example to bring the concepts together.
Permanent Link: http://hdl.handle.net/11104/0311264