Application of overbooking model to college admission at The Citadel

Abstract

The admission office in almost every four‐year college faces the challenge of determining the number of students to be admitted each year, because only a small percentage of the students enroll after being accepted (called yield rate). The national average yield rate in four‐year colleges and universities has declined. At The Citadel under normal circumstances, it is even more challenging since the cadets are required to stay on campus where the beds in the barracks are limited. An insufficient number of freshmen enrolled at The Citadel result in a loss of revenue, and too many freshmen enrolled results in shortage of beds. In this paper, we conduct a case study and apply the overbooking model in revenue management to find the optimal admission limit at The Citadel. We perform a computation experiment on the enrollment data to obtain the optimal admission limit for the academic year 2020–2021. Also, over the last five years, we have compared the admission limit obtained using the risk‐based policy to the admission limit obtained using the deterministic heuristic method, and with the actual admission number from the historical data. The results of this case study can be used by the admission office at The Citadel to measure its past performance, identify strategies, and make informed admission decisions in the near future.

A binomial decision tree to manage yield‐uncertainty in multi‐round academic admissions processes

“…They mention the problem of uncertainty around enrollment yield and suggest using waitlists to make up for shortfalls. Acknowledging the issue of yield uncertainty, Zhang and Pippins (2021) apply the overbooking model to the college admissions problem and perform a computational experiment to help make an optimal decision regarding sending out admission offers, which is somewhat similar to the goal of our paper. However, they do not consider multi-round admissions and do not take into account the possibility that an applicant's probability of accepting an offer is dependent on their rank.…”

“…Our work adds to the literature by contributing to an area of concern acknowledged by Maltz et al (2007), where they estimate enrollment likelihood of individual applicants and enrollment yield of the aggregate, and suggest using waitlists to handle uncertainty in both of these. We also add to the work done by Zhang and Pippins (2021), where they have applied an overbooking model to address yield uncertainty in college admissions. We address this problem in a multi-round setting by providing a framework to handle the uncertainties around offer acceptance and enrollment, and to recommend an appropriate number of admission offers in each round.…”