Some good stuff here, but... (Score 1)

This is cross sectional research, so any implication of causality is tenuous at best. Also, though the author goes to some lengths to talk about the validity of the sample, it's small (~600). The longitudinal treatment of this topic, as well as the use of assessment scales not based on DSM/substance abuse models of "addiction" will move this inquiry even further forward. As far as the usefulness of this kind of thing - empowering the consumer to identify behavioral excesses and react to them is a good way to go. If you like this kind of stuff, check out Nick Yee's work and my own at ProjectMassive.com.