The Sorority Match-Making Approach
At Cornell, homecoming weekend is a big opportunity for organizations to introduce themselves in a public, social setting to underclassmen whether it is active members being representatives or even alumni. Looking through the crowd, seeing interactions amongst students made me wonder if there was a way to model the connectedness of students and Greek life and I began searching if there had been any studies. Not only have there been many papers and articles discussing the networks of campuses across the globe, but more specifically there has been much analysis of the bidding system in sororities in American universities. The paper below from Susan Mongell and Alvin E. Roth dives into the real-world application of “market matching” that we have discussed in class through the lens of American sororities.
Paper: https://stanford.edu/~alroth/sorority.html
The American labor market for medical students consists of a newly graduating medical students competing for desired positions at certain hospitals and medical offices across the nation. It treats students and positions as a network in which preferences of the student and the position are considered to proceed in different matching sequences. This system has been implemented and advanced in many facets, specifically the Sorority Rush process in American universities. Susan Mongell and Alvin E. Roth have analyzed the implementation of the matching process called “The Preferential-Bidding System” and researched its efficiency on 4 different campuses through 21 rushes in two stages; formal rush and open bidding. Roth and Mongell were able to model the bidding system, develop new theorems for the algorithm, and compare to empirical observations. The model consisted of a set of sororities, a set of rushees, connections termed as preferences, and a quota limiting the number of connections a sorority can make. Some of these observations included the act of rushees showing preference to just one sorority, a term coined as “suiciding” because the likeness of their matching is uncommon, and yet are still matched to their desired sorority. Other observations show that some conditions affected the stability of the algorithm such as having sororities make more preferences than the quota. The study concluded that although the algorithm was effective, there are other items that must be taken into account before attempting to predict/confirm how matches are made in these sorority communities in American Universities. The most important being that the incentives provided across rushee and sorority affect the rationality of individual preferences made.
This discussion and analysis of the sorority market has taken our investigation of market matching to a level with greater depth. It shows how markets and graph theory consists of preconditions and assumptions that may be inapplicable in certain fields but crucial to describe others. It amplifies the idea of unity in theoretical applications amongst breadths and how creating arbitrary rules can impact the conclusion of each individual breadth.