Would be interesting to see some other stats for people in the study... anything around mental health, number of sexual partners, the length of relationships, academic performance, personality types, etc... Something that explained why people make relationships the way they do.<p>Looking at the graphs I think I would fall into the compartmentalized option, I had friends I would go to classes with, friends outside of classes, friends in my fraternity... it was rare that they all got together in the same place.<p>My goals were to get good grades, have people to party with on the weekends (on demand), and then have some people who I was actually close with -- the guys I played poker with. Post college I kept up with a few people from each group I guess -- certainly wasn't exactly as I planned though. People float in and out over time...