Imagine I have 300 items in an item bank.
I am administering a 50-item test which will be comprised of items randomly selected from the 300-item bank.
Ultimately, across all my students and the various forms of the exam that are randomly generated, 175 of the items are administered.
What will my item analysis look like? Will I get item-level data for all 175 items? How will Canvas report reliability when items are not the same across all test forms?