What I've understood from IRT is that it's able to predict how hard or easy an item is and how likely an applicant is to get an item right.
The question is: Can it predict how unlikely an applicant is to get it wrong too? I mean, can miss marks and time pressure distort the scale?
