difNLR

DIF and DDF Detection by Non-Linear Regression Models.

GHversion version cranlogs

Description

The difNLR package contains method for detection of differential item functioning (DIF) based on non-linear regression. Both uniform and non-uniform DIF effects can be detected when considering one focal group. The method also allows to test the difference in guessing or inattention parameters between reference and focal group. DIF detection method is based either on likelihood-ratio test, F-test, or Wald’s test of a submodel. Package also offers methods for detection of differential distractor functioning (DDF) based on multinomial log-linear regression model and newly methods for DIF detection among ordinal data via adjacent category logit and cumulative logit regression models.

Installation

The easiest way to get difNLR package is to install it from CRAN:

install.packages("difNLR")

Or you can get the newest development version from GitHub:

# install.packages("devtools")
devtools::install_github("adelahladka/difNLR")

Version

Current version on CRAN is 1.3.7. The newest development version available on GitHub is 1.3.7.

Reference

To cite difNLR package in publications, please, use:

Hladka, A. & Martinkova, P. (2020). difNLR: Generalized logistic regression models for DIF and DDF detection. The R Journal, 12(1), 300–323, doi: 10.32614/RJ-2020-014.

Drabinova, A. & Martinkova, 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, doi: 10.1111/jedm.12158.

Try online

You can try some functionalities of the difNLR package online using ShinyItemAnalysis application and package and its DIF/Fairness section.

Getting help

In case you find any bug or just need help with difNLR package, you can leave your message as an issue here or directly contact us at hladka@cs.cas.cz