You do not need to know R to use mirtverse. You need to install it once and type one line; after that, you can work with buttons or plain language.
Watch the workbench in action first (about 90 seconds, with captions):
Try it before you install anything
Open the live example dashboard. It shows the results for a 32-item test: Which items work, which need review, and how scores are distributed. It runs in your browser, and you can make one like it for your own test.
Your path in three steps
1. Install R, RStudio and mirtverse (once, about 10 minutes).
- Install R from CRAN and RStudio Desktop, both free.
- Open RStudio and paste this line into the Console (bottom left), then press Enter:
install.packages("mirtverse", repos = c("https://okanbulut.r-universe.dev", "https://cloud.r-project.org"))2. Open the workbench. Paste these two lines:
A window opens with eight steps: Load your data (CSV, Excel, SPSS), choose a model, check items, score people and download a report. The workbench guide walks through each step.
3. Or describe what you want in words.
mv_ask() writes the analysis for you, and you can read it
before it runs:
mv_ask("fit a Rasch model, check item fit and make a report", data = mydata)It works offline for common analyses; see the guide.
Coming from other software?
- Winsteps or other Rasch software: The workbench fits the Rasch model, item fit and a Wright map, and writes the R code so you can see the equivalent of each step.
-
SPSS or Excel: Load your
.sav,.xlsxor.csvfile directly in the workbench. -
ltm, eRm, TAM or irtQ in R:
mirtify()translates your code; see Moving to mirt.
Words you will see
The glossary explains each term (discrimination, information, DIF and so on) in plain language.