If it is really the case that everyone is around 1 to -1 and this person alone hits the double digits, yes I would worry. But I have known subjects who gave very large beta values. I once had a person I know participate in my experiment (they asked), and they tried really hard and had a very high behavioral score and very large Beta values compared to the rest of the sample. It can happen.
Of course you should very carefully check everything, and their raw data as well. See if the signal in that person is unusually high or they have a high variance, or maybe some task-correlated motion which can wildly inflate the Beta values.