The R-Package GLMMRR. This package gives support to Generalized Linear Mixed Modeling (GLMM) for Binary Randomized Response Data. It extends the modeling possibilities of lme4, by including different link functions to deal with randomized response data. This includes Cauchit, Compl. Log-Log, Logistic, and Probit link functions for Bernoulli distributed RR data. RR Designs: Warner, Forced Response, Unrelated Question, Kuk, Crosswise, and Triangular. The package is listed on CRAN: GLMMRR.
The R package LNIRT. This is the package for log-normal response time IRT modeling for responses and response times. It is listed on CRAN: LNIRT . It allows the simultaneous analysis of responses and response times in an Item Response Theory (IRT) modelling framework. Parameter estimation is done with a MCMC algorithm. An RMarkdown document has been written to further support the use of the R-package (LNIRT RMarkdown).
LNIRT replaces the package CIRT (written by Rinke Klein Entink). For reference, see the paper by Fox, Klein Entink and Van der Linden (2007), “Modeling of Responses and Response Times with the Package cirt”, Journal of Statistical Software, link.
Questions or remarks about the LNIRT or GLMMRR package, please use our maintainer email address: firstname.lastname@example.org
The R package MLIRT (Version R 2.15.0) contains methods related to Bayesian Item Respons Modeling. The package is build under R-2.15.0. The package can also handle randomized response data. There is a 32 bits version, compiled under i386 architecture (including a win32 Fortan dll), and a 64 bits , compiled under Intel 64 architecture (including a x64 Fortan dll). It was not possible to construct one package for both architectures.
The mlirt R-package is not listed on CRAN.
Download: MLIRT R-Package 32 Bits (zip)
Download: MLIRT R-Package 64 Bits (zip)
The R package CIRT for R-3.0.0 is meant for the joint analysis of response data and response times on (educational) tests. There is a 32 bits version, compiled under i386 architecture (including a win32 Fortan dll), and a 64bits, compiled under Intel 64 architecture (including a x64 Fortan dll).
Download: CIRT Package 32 Bits (zip)
Download: CIRT Package 64 Bits (zip)
This is an example that can be followed to see how to use the software CIRT (from Rinke Klein Entink)
Download: CIRT Example (zip)
The R package MIXIS 1.0 contains methods related to Bayesian joint mixture modeling of survival and explanatory item response data using IRT. The package is build under R-2.13.0.
Download: Mixis Package 32 Bits (zip)