Hi @Josh , thanks for the reply!
I paused the study because I was worried about the average dropping below the minimum, so I haven’t let any more people take it. When I paused it, 250/500 had already taken it, so I’m worried the average won’t change. Further, I looked at my Group 1 data, and there were hardly any outliers in terms of time–only a few people taking longer than 20 minutes and no one taking anywhere near the max amount of time, whereas in Group 2, at least half are taking >20 minutes with several taking the maximum amount of time.
My main concern is with the average compensation dropping below the minimum and having to pay Group 2 more (which lowers the sample size I can afford). If the problem was common to both groups, I’d say I must’ve underestimated how long the study would take and take the hit. But I have 500 people in Group 1 who indicate otherwise. Does Prolific have a way of investigating something like this, or should I just plan to stop my study early if my Group 2 participants are being “underpaid”?