![[PDF]](/pb-assets/Icons/adobe-pdf-icon-1498649896243.png){kind=icon}
Abstract
There are many item response theory software packages designed for users. Here, the authors introduce an environment tailored to method development and simulation. Implementations of a selection of classic algorithms are available as well as some recently developed methods. Source code is developed in public repositories on GitHub; your collaboration is welcome.
While open-source software packages for item response theory (IRT) are available (e.g., Chalmers, 2012), the software tends to be designed for users as opposed to methodological researchers. This is a small, but important distinction. For instance, the latter group may have more demands for modularity and greater need for high performance to efficiently run large-scale simulations. Here, the authors introduce OpenMx (Pritikin et al., 2015), rpf (Pritikin, 2020), and affiliated packages that, together, they regard as a modular IRT methodology development environment for R (R Core Team, 2020). Performance critical parts are written in C++. An optimized implementation of marginal maximum likelihood by quadrature (Bock & Aitkin, 1981) can fit a model with 1,536 dichotomous items and 2,500 responses in about 12 s on an Intel i7-8550U CPU running at 1.80 GHz.
A selection of popular fit diagnostic tests are available, including the S-X2 statistic (Orlando & Thissen, 2000), a pairwise test of local dependence (Chen & Thissen, 1997), a multinomial fit test (Bartholomew & Tzamourani, 1999), and a sum-score expected a posteriori test (Li & Cai, 2018). The menu of features is organized in the style of à la carte as opposed to table d’hôte. Each diagnostic can be run separately to facilitate simulation studies. Since the initial publication of rpf, a suite of monotonic polynomial models have been contributed (e.g., Falk, 2020; Falk & Cai, 2016).
As with all IRT packages, additional methods and features are under development, yet users are encouraged to contribute or develop additional modules. There is already enough modeling flexibility to combine IRT with a kernel smoother (Hunter & Bard, 2018), and a point-and-click user interface for model construction is available from the ifaTools package (Pritikin & Schmidt, 2016). The software runs on all major operating systems. Source code is developed in public repositories on GitHub.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by National Institute of Health (Grants R01-DA018673 and R25-DA026119, Principal Investigator [PI] Neale), and Fonds de recherche du Quebec—Nature et technologies (2019-NC-255344).
ORCID iDs
Joshua N. Pritikin 
https://orcid.org/0000-0002-9862-5484
Carl F. Falk https://orcid.org/0000-0002-4788-7206
Supplemental Material
Supplementary material is available for this article online.
References
|
Bartholomew, D. J., Tzamourani, P. (1999). The goodness-of-fit of latent trait models in attitude measurement. Sociological Methods & Research, 27(4), 525–546. Google Scholar | SAGE Journals | ISI | |
|
Bock, R. D., Aitkin, M. (1981). Marginal maximum likelihood estimation of item parameters: Application of an EM algorithm. Psychometrika, 46, 443–459. Google Scholar | Crossref | ISI | |
|
Chalmers, R. P. (2012). mirt: A multidimensional item response theory package for the R environment. Journal of Statistical Software, 48(6), 1–29. Google Scholar | Crossref | ISI | |
|
Chen, W.-H., Thissen, D. (1997). Local dependence indexes for item pairs using item response theory. Journal of Educational and Behavioral Statistics, 22(3), 265–289. Google Scholar | SAGE Journals | ISI | |
|
Falk, C. F. (2020). The monotonic polynomial graded response model: Implementation and a comparative study. Applied Psychological Measurement. https://doi.org/10.1177/0146621620909897 Google Scholar | |
|
Falk, C. F., Cai, L. (2016). Maximum marginal likelihood estimation of a monotonic polynomial generalized partial credit model with applications to multiple group analysis. Psychometrika, 81, 434–460. https://doi.org/10.1007/s11336-014-9428-7 Google Scholar | |
|
Hunter, M. D., Bard, D. E. (2018, July 10–13). Linking scales without common items [Talk presentation]. The Annual International Meeting of the Psychometric Society, New York, NY. Google Scholar | |
|
Li, Z., Cai, L. (2018). Summed score likelihood–based indices for testing latent variable distribution fit in item response theory. Educational and Psychological Measurement, 78(5), 857–886. https://doi.org/10.1177/0013164417717024 Google Scholar | |
|
Orlando, M., Thissen, D. (2000). Likelihood-based item-fit indices for dichotomous item response theory models. Applied Psychological Measurement, 24(1), 50–64. Google Scholar | SAGE Journals | ISI | |
|
Pritikin, J. N. (2020). rpf: Response probability functions [Computer software]. https://CRAN.R-project.org/package=rpf Google Scholar | |
|
Pritikin, J. N., Hunter, M. D., Boker, S. M. (2015). Modular open-source software for Item Factor Analysis. Educational and Psychological Measurement, 75(3), 458–474. https://doi.org/10.1177/0013164414554615 Google Scholar | |
|
Pritikin, J. N., Schmidt, K. M. (2016). Model builder for Item Factor Analysis with OpenMx. R Journal, 8(1), 182–203. Google Scholar | Crossref | |
|
R Core Team . (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org Google Scholar |
