A Friendly Workflow and Reporting Layer for mirt • mirtverse

mirtverse is a friendly workflow and reporting layer for the mirt package. mirt estimates item response theory (IRT) models; mirtverse helps you choose a model, understand the results, make publication-ready figures, and share everything, with or without writing R code.

About mirt

mirt, developed and maintained by Dr. R. Philip Chalmers, is one of the most comprehensive open-source packages for item response theory. It was introduced in the Journal of Statistical Software (Chalmers, 2012) and has been developed continuously since, with version 1.48 released on CRAN in October 2026. It is widely used in research on educational and psychological measurement.

mirt can:

Estimate a wide range of models: unidimensional and multidimensional, exploratory and confirmatory, bifactor and two-tier, with more than 30 item types (Rasch, 2PL to 4PL, graded response, (generalized) partial credit, nominal, ideal-point, nonparametric spline items and more), plus user-defined items.
Handle complex designs: multiple-group analysis and measurement invariance, mixed-effects and explanatory IRT (mixedmirt()), and latent class and diagnostic classification models (mdirt()).
Use several estimation engines: EM with fixed or quasi-Monte Carlo quadrature, and the Metropolis-Hastings Robbins-Monro (MHRM) algorithm for high-dimensional models, with several standard-error methods.
Check and use the models: item and person fit, limited-information fit (M2), local dependence, DIF and differential test functioning, and ability estimation (EAP, MAP, ML, WLE, plausible values).

mirtverse leaves all of this untouched and adds the workflow, interpretation and reporting tools around it.

What mirtverse adds

Feature	Functions
Point-and-click workbench with R code export	`launch_workbench()`
Analysis scripts written from a plain-language request	`mv_ask()`, `mv_ask_setup()`, `run_script()`
Model selection assistant	`select_model()`
Differential item functioning in one call	`dif()`, `gg_dif()`
Plain-language interpretation	`interpret()`, `interpret_prompt()`
Tidy outputs	`tidy()`, `glance()`, `augment()`, `item_table()`
ggplot2 graphics, including Wright maps	`gg_trace()`, `gg_info()`, `gg_wright()`, `gg_itemfit()`, `gg_q3()`, `gg_scores()`, `gg_empirical()`, `autoplot()`
Item health cards	`item_card()`, `item_cards()`
One-call HTML report	`mv_report()`
Results dashboard to share with non-R users, as a Shiny app or one offline HTML file	`mv_dashboard()`, `export_dashboard()`
Response-time modelling and rapid-guessing detection	`fit_rt()`, `sim_rt()`

Installation

Install a prebuilt version from r-universe (no Rtools or GitHub token needed):

install.packages("mirtverse", repos = c("https://okanbulut.r-universe.dev", "https://cloud.r-project.org"))

Or install the development version from GitHub:

# install.packages("remotes")
remotes::install_github("okanbulut/mirtverse")

A five-minute tour

library(mirtverse)

data <- expand.table(LSAT7)

sel <- select_model(data)   # compares Rasch, 2PL and 3PL and recommends one
mod <- sel$best             # an ordinary mirt model

interpret(mod)              # what the results mean, in plain language
tidy(mod)                   # parameters as a data frame
item_table(mod)             # item statistics, fit and flags in one table

gg_wright(mod)              # person-item map
item_card(mod, "Item.2")    # one-page item review

mv_report(sel, file = "report.html")   # everything in one HTML file
mv_dashboard(mod, title = "LSAT results")   # interactive dashboard to share

Prefer clicking? Run launch_workbench(). Prefer describing the analysis in words? mv_ask("fit a 2PL, check DIF by gender and make a report", data = mydata) writes the whole script for you.

Relationship to mirt

mirtverse depends on mirt and never changes how models are estimated. Every function accepts a model created with mirt::mirt(), so you can move between the two packages at any point. Please cite mirt (Chalmers, 2012) whenever you use mirtverse.

Chalmers, R. P. (2012). mirt: A multidimensional item response theory package for the R environment. Journal of Statistical Software, 48(6), 1-29. https://doi.org/10.18637/jss.v048.i06

Citation

If you use mirtverse, please cite it together with mirt. Run citation("mirtverse") in R, or use:

Bulut, O. (2026). mirtverse: A friendly workflow and reporting layer for ‘mirt’ (Version 0.1.0) [R package]. Zenodo. https://doi.org/10.5281/zenodo.23719385