Here's an example of the sort of thing I'm looking for
Is obnubilate a word? YES/NO
How confident are you in your answer on a scale of 1 to 5?
Rate your confidence that obnubilate is a word from 1 to 10, where 0 means "very confident it is not a word" and 10 means "very confident it is a word".
The datasets don't need to relate to word/nonword studies, and could be any kind of scenario where there is a ground truth.
It would be best if there were at least 100 participants giving at least 100 responses, but this is not strictly necessary.
It would also be best if each participant was responding to the same question set in the same order.