Ranking Frats and Sororities

In my TIL for the day (this I learned), I stumbled across this site that ranks fraternities and sororities on campuses. Go here for UGA (where I teach) rankings.

For those keeping score (i.e., any Greek):

UGA Fraternities

1. Sigma Chi
2. Phi Gamma Delta
3. Pi Kappa Phi (my frat, but at different school)
4. Sigma Nu
5. Kappa Sigma

UGA Sororities

1. Pi Beta Phi
2. Delta Zeta
3. Delta Gamma
4. Alpha Gamma Delta
5. Gamma Phi Beta

Me being me, I peeked under the hood at the methodology. For sororities, they ask you score a sorority on (I'm not making these up) hotness/attractiveness, popularity/dominance, classiness, campus/community involvement, fun/social life, and sisterhood. For frats it's the same except, of course, brotherhood replaces sisterhood. There's a funny joke in there, somewhere, but I'm not gonna go digging for it. Also they ask you to rank what tier the frat/sorority belongs in, though that doesn't play a part in the overall rankings.

My favorite part of the evaulation process is this:

Please describe why you either like or dislike this chapter. DO NOT post a person's name or malicious attacks. Before you post imagine you are reading this about you or your organization. Keep it Positive.

If you click on an individual fraternity or sorority you can see the score breakdown and, best of all, the comments. The "keep it positive" request, not quite followed. For example, clicking on Pi Kappa Phi fraternity you see, among the comments, this:

Easily the biggest self rankers on this site. Around campus these guys are known as the guys who couldnt get a bid from any of the top 5. Middle tier at best

Our budding methodologist above is on to something -- self rating. If we assume all the Greek orgs will self rate at roughly the same amount, then any differences may, theoretically, be due to others ranking them. So the cream would rise to the top. Of course there's no systematic reason this would be the case. My hunch is some orgs go out of their way to either self rate or get others to rate them, meaning there is so much measurement bias in the data that the lists and their ranks are, essentially, useless.