Hi all,

I am analysing a task and wanting to apply SDT(signal detection theory) on the analysis. Here's the description of the task: the task has two variables: match/mismatch and labels. So when it's match condition and response also shows match, it is a "hit" (when it's match but response is mismatch, it's a "miss"). On the other hand, when it's mismatch and response is mismatch, it's a "correct rejection" (when it's mismatch but response is match, it's a "false alarm").

Since the task is relatively simple and there's a ceiling effect (accuracy equals to 1), I found this might lead to some problems when calculating d' of SDT. I was wondering whether there are some alternative methods I can use to calculate d'? (if there's a paper that I can refer to, it would be great). Thanks!