Laboratory for latent variable modeling


FLIRT

R package FLIRT for Flexible item response theory modeling

Source file: flirt_1.15.tar.gz (64bit Windows)

Source file: flirt.x32_1.15.tar.gz (32bit Windows)

Manual: flirt.pdf
R script: test_flirt_version1.15_shared.R
Vignette: flirt_vignette.pdf
Readme: readme.pdf

Please read the readme.pdf for important notes on installation and usage.

Note

The new version 1.15 is based on a newer version of MCR (Matlab 2014a).
To utilize this new version, users need to download/install the new MCR.
(Versions below 1.15 need MCR for Matlab 2012b)

To cite the package, please use

Jeon, M.,  Rijmen, F. & Rabe-Hesketh, S. (2014). Flexible item response theory modeling with FLIRT. Applied Psychological Methods, 38, 404-405.

Modeling Features:

One parameter logistic (1PL) and two parameter logistic (2PL) models
Multi-dimensional models (between-item and within-item models)
Bifactor models
Person covariates (latent regression)
Item covariates (linear logistic test model)
Multiple group analysis
Differential item functioning analysis
Polytomous responses (partial credit models, rating scale models, graded response models)
Combinations of above

Estimation features:

Efficient ML estimation with a modified EM algorithm using graphical model theory
Adaptive quadrature

Version 1.15 New Features:

Second-order models (and its extensions)
Three parameter logistic (3PL) model (basic model)
Item response curves for unidimensional 1PL, 2PL, and 3PL models
Item/test information functions for unidimensional 1PL, 2PL, and 3PL models

New Examples:

Longitudinal modeling with FLIRT: R_script
Item response tree modeling: example_trees.pdf / R_script
Plotting examples: R_script


For any questions, contact jeon.117@osu.edu