In 2015, Mark has green lighted the $500 prize to your MS advisory account for champion in optimizing a figure of merit of your (materialized) admits. Over time we’ll tune the FoM. For now, I have a version 1 below. The general idea is a good student you admitted adds to your score and a bad/mediocre one subtracts. Every MS-PhD convert adds a bonus point. Good or bad is measured by GPA and a subjective evaluation (such as TA performance and exit exam). For now, Michele will grade everyone on a 100% scale.
FoM = Σ(GPA+SE-Th) + N_c
SE: subjective evaluation, initially weighs about 5-10% of GPA until we are better at it.
Th: threshold will be average score for the 1st cohort.
N_c: Number of converts
Notes:
- The FoM is NOT normalized to the number of students: We want more good students.
- The threshold is a fixed number so that we can see over time whether our admission is improving or not.
- The FoM can’t help you diagnose your criteria of admission. You should keep your own statistics if you want to debug and improve your algorithm.
- Yes, I understand luck is a factor.
UPDATE:
The first cohort’s average is 3.69 (They are evaluated as A, A-, etc) so there is limited precision in the threshold.