Individual test item analysis (e.g., difficulty and discrimination)

[koenderks]

@vandenman just had a meeting with Dylan today discussing a new IRT module, which had this analysis in it. We figured that it was maybe more suitable for the reliability module.

@juliuspfadt @vandenman What are your thoughts on this analysis? Should we even offer this analysis to users? Is there anything that you would add or remove? Everything can still be changed of course.

[juliuspfadt]

hmm. Tough question. Of course, there is some overlap with the regular reliability analysis, and also with the descriptives module. My 2c:

• In my mind difficulty and discrimination are terms connected to IRT, and ability tests.
• Naming it CTT might be confusing to users: within the unidimensional reliability we have coefficients based on CTT, and some based on FA. Could be separating the analysis based on test-theories might be more appealing to psychometricians, not sure about the average user.

I think this could either be a section within the "unidimensional reliability analysis", that section would then be something like "individual item analysis", if I remember correctly, Andries van der Ark was also asking for something similar. Or the analysis itself would be named "individual item analysis", but then I wonder if leaving it inside IRT would not be better. Maybe @vandenman has better ideas?