full-randomized PNT will be a bit trickier due to random test order
100-item IRT we can't do currently, since we don't have data. Perhaps if 30 and 60 work though, that's good enough. Or we should just exclude the 100 item?
reminder for self: 30-item IRT tests whether ability estimates and next items are working similarly to the old java app. walker and full pnt tests are checking that responses entered into the app are recorded and output correctly in the downloaded data
Questions/thoughts:
reminder for self: 30-item IRT tests whether ability estimates and next items are working similarly to the old java app. walker and full pnt tests are checking that responses entered into the app are recorded and output correctly in the downloaded data