This is the first in a series of 6 modules that introduces item response theory (IRT). In it the differences in IRT and classical approaches are introduced, specifically the differences in their scoring logics.
This is really helpful for me since I am a person needs conceptual understanding to move on before mathematical applications. Nice! I really like this module!
I'm looking forward to viewing the whole series. In Part 1, the response scale 0-4 (higher numbers less difficulty) didn't match the item scale you presented first 0-4 (higher numbers more difficulty). I will also send you an email. Please let me know if these comments are helpful. I'm working with some young researchers at UC Berkeley, very new to IRT and developing a gentle introduction to IRT for them.
Could IRT work with a scale that looks at psychological well being? I am working on a project that requires me to figure out how to adapt the Schwartz Outcome Scale-10 from English to Punjabi. Each item on this scale is given a Likert Rating scale from 0 to 6 and probes at an individual's overall well-being in the past two weeks. Each participant being administered this scale could be feeling "differently" than the others. So how does IRT work with more abstract, hard to quantitatively measure topics such as psychological well-being?
Dear Dr. Cook, In the first module, within the example you provided, Jean and Robin's responses on one of the three tasks is "Much difficulty"; however, in the scale you explain in the beginning does not define such a response; it rather has "Great difficulty" - is this what you mean by "Much difficulty"?
I'm currently in Dr. Richard Sudweek's Measurement Theory class at BYU. This video was a helpful introduction to the concept of Classical Test theory. I got out of it that True Score = Observed Score + Error, and that items can be measured differently. Suggestion, this intro video might be more engaging to new viewers if these two principles were taught in 6 minutes instead of 9.
I understand that you have administered different items to the test persons for the purpose of this demonstration, but why would I ever do that in real life? And if I administer the same items, why would I need IRT?
Dear Dr. Cook, In the first module, within the example you provided, both Jean and Robin's responses on one of the three tasks are "Much difficulty"; however, the scale you explain in the beginning does not define such a response; it rather has "Great difficulty" - is this what you mean by "Much difficulty"?