/blackboxstudyR

R package to fit basic IRT models to FBI Black Box Data.

Primary LanguageC++

blackboxstudyR

blackboxstudyR is an R package for fitting basic item response theory (IRT) models to FBI Black Box Data.

Installation

This package is not yet available on CRAN. Install the development version from Github with:

# install.packages("devtools")
devtools::install_github("aluby/blackboxstudyR")

As of April 2019, this package requires a specific version of rstantools in order to install (error is documented elsewhere). Install the required version with:

devtools::install_github(“stan-dev/rstantools”, ref = “c3c59fb1ef”)

Usage

Usage consists of three steps: (1) Score the data, (2) Format a data list, (3) Fit the IRT model.

im_scored = score_bb_data(TestResponses, "inconclusive_mcar")
im_data = irt_data_bb(TestResponses, im_scored)
im_model = fit_rasch(im_data)