I suppose a next step would be to perform a brain scan before someone starts using Facebook, then several over the course of their subsequent use. The researchers would obviously need to do their best to control for confounding factors, but it seems like a reasonably interesting first study could be done this way.