Could be a study to see how much response such an obviously fraudulent e-mail gets, sent from that same university conducting the study.

I admit that it's a fringe case, but it could be a thing.