Import results from the calibration and store with items
What is Item Response Theory?
Item response theory (IRT) is a psychometric paradigm used to design, develop, assemble, deliver, score, and analyze assessments. It offers many advantages over its predecessor, classical test theory, by implementing more sophisticated mathematical modeling. Most large-scale assessments that you can think of, from professional certifications up to worldwide assessments like PISA, utilize IRT.
The basic building block of IRT is the item response function, which describes the probability of a given response as a function of a person’s true standing on a latent trait or ability. As you would expect, students of higher ability have a higher probability of a correct response, but this probability of course cannot exceed 1.0.
IRT uses these functions to evaluate item performance, but also to solve other important measurement problems: how to assemble equivalent test forms, how to evaluate score precision, how to design adaptive tests, how to equate test scores across different years, and more.
- More powerful test assembly with TIF and CSEM, including parallel form construction
- Better description of item performance (difficulty, discrimination, and guessing) and model fit
- More precise scoring – learn more about the theory behind it with this free software download
- Examinees and items are placed on the same scale (important for adaptive testing
- Greatly improves linking and equating of scores across forms or years
- Classical test theory tries to force a linear model on something that is known to be nonlinear
If you just want to use IRT to analyze and evaluate an existing test, you need specially designed software for that purpose. We recommend Xcalibre, as it has a number of advantages over competitive systems, including:
- Scientifically validated results
- Professionally designed user interface
- No command code
- Output directly into narrated Word documents and spreadsheets
- Extensive graphical analysis.