May 21, 2013
In a recent conversation with someone in the early stages of preparing for medical school, I was struck by the parallels in the considerations of the medical school admissions process and those for Deal Scoring. It’s no secret that med school admissions are highly competitive, but the formula for admission and the weighting of the various factors considered at each institution is enigmatic.
Finding predictive measures for who will be successful in school and then later on in the medical profession isn’t straight forward for a number of reasons. To start with, book learning and bedside manner are certainly different skills, as are research and straight memorization/ cramming for exams. Similarly, there are many metrics related to a price agreement which can be considered in Deal Scoring.
Historically, GPA was considered as a summary metric around applicant quality, weighting achievement in each course into one number which could be compared across candidates throughout the country. When you dig into that metric though, you are faced with disparities in institutional quality and questions of grade inflation which confound that variable and highlight the ways it can be gamed.
At Columbia in particular, the science courses are graded on a forced normal curve, tightly centered around a B, with only 10% of the class eligible for an A. When grades are converted back to a four point scale, the difference between the 89th and 90th percentile translates to the difference between a 3.3 and 3.6 GPA. GPA is comprised of numbers which are not on a smooth curve, but a step function which forces students in to the A and B classifications where such stark distinction was not apparent before. Letter grades are the equivalent of customer classification dimensions rather than numerical measures. There are so many quality med school applicants that some universities find it expedient to categorically ignore applications from those individuals who don’t fall into the A grade classification. Then, within that group of applicants additional vetting occurs. More qualitative criteria include a review of reference letters and the evaluation of an applicant’s published research and work experience.
In pricing, total revenue is often the first measure considered when evaluating a deal, but the quality of that revenue (profitability) can be as different as an Ivy League education versus an unaccredited online degree. To better control for variation in school quality, standardized tests (the MCAT) were developed and administered as a new requirement for admissions.
As with the academic admissions process, deal approvals can begin algorithmically with numerical scoring, but most of them also involve manual approval and judgment calls to incorporate more nuanced trade-offs such as how to handle strategic customers. The goal of scoring should be to simplify the information into metrics which can be legitimately compared across customers, markets and products and bring a lens of objectivity to the deal approval process. The workflow or escalation process is an acknowledgement that there is a subjective element to deal scoring and admissions which highlights the importance of knowledgeable managers and med school admissions officers.