MCMC Z-G

Author: Wei Wang, Philseok Lee, Seang-Hwane Joo, & Stephen Stark
Contact: computationalpsychology.org/resources/
Description: The MCMC Z-G computer program estimates the Zinnes Griggs (Z-G) item response theory model (Z-G IRT; Zinnes & Griggs, 1974). Contrast to the GGUM2004 (Roberts, Fang, Cui, & Y. Wang, 2006) and the MCMC GGUM (W. Wang, de la Torre, & Drasgow, 2014) that calibrate the single-stimulus response data, the MCMC Z-G calibrate response data from the forced-choice measurement formats, which potentially address the response style and bias problem that are typically caused by the popular single-stimulus measurement formats. Similar to the MCMC GGUM (Wang, et al., 2014), the MCMC Z-G computer program also employs the Metropolis-within-Gibbs algorithm based Markov chain Monte Carlo method (Hastings, 1970; Metropolis, A. Rosenbluth, M. Rosenbluth, A. Teller, & E. Teller, 1953) and estimates both the item and person parameters, and it can handle missing values in the data.
Analyses: Simulation/Resampling; Test Construction & Administration; CAT
System Requirements: Windows
Measurement Model: IRT/Rasch
License Type: Freeware
Documentation: User manual