Alternative System for Grading
The way we currently grade is broken. Universities, though I am primarily concerned with law schools, are incentivized to make their grading system as opaque as possible, reaping what rewards they can by maximizing the number of students whose grades look “good.” Though something like GPA is of course only meaningful when contextualized within the school, employers aren’t going to bother caring about the difference. They’ll glance at the number, look at class position if available, maybe look at individual grades if accessible, and rely on their gut from there. Nobody has the time to worry that much about differing class medians.
Trying to capture the variety of things we’d like within a single data point per class is hurting our ability to communicate most effectively. The different things a grading system should hope to communicate, as I see them, are:
Course material mastery/competency
Relative quality of students
There are some schools that feel the need to obfuscate differences between students as much as possible because the school is good enough to do so. Instead, they can focus on course by course content mastery. The system adopted by Yale Law, Stanford Law, Harvard Law (somewhat), and Berkeley Law all use some variation of the Honors/High Pass, Pass, Low Pass, Fail system. This communicates pretty effectively the 1. point.
Those students who demonstrate a true mastery of the material that goes beyond the expectations of the course can get an H/HP. Those students who demonstrated they understood the material well enough to earn credit and take more advanced material in the subject can get a P. Those who maybe deserve credit but shouldn’t take more advanced coursework in that area an LP. And students not deserving of credit could receive an F.
This can then be converted into a system allowing comparisons between students. Harvard Law does this by assigning grade point equivalents then determining class rank. Schools like Yale Law and Stanford Law do no such conversion. They let the transcript speak for itself.
Most schools can’t get away with the Yale Law and Stanford Law approach. Not only do schools want most of their students to get a job, schools want their best students to get the best jobs. And if the school in question is not at the caliber necessary to be able to get those jobs basically no matter the class placement, student differentiation is necessary so at least some of the graduates can get those more highly coveted jobs.
That’s why I think law schools should utilize a parallel grading scheme. Provide HP/P/LP/F grades based on the descriptors I provided or something similar. Then I recommend something slightly different as well. Rank students within each doctrinal course according to their raw scores. Provide a score equal to the sum of the number of students they outscored, plus one half the number of students they tied, divided by the total number of students in the class. Add up the scores from each doctrinal and calculate a GPA that then distributes students along a curve.
From here there are different options, based on the needs of the school. To start, I do not think the individual class scores should be part of the transcript. They imply a degree of accuracy in student assessment that I believe to be off base. Class assessments, outside of the level of material mastery already shared in the HP/P/LP/F system, are only really useful in aggregation. Instead, the class scores should only be aggregated in order to evaluate class position. Because schools want to be more or less revealing, they can choose how transparent they want to be. One method would be sharing exclusively class percentiles, another class position, yet another could be a hybrid of the two. The class percentiles or positions could be for the whole class or only the upper part of it. Here there is greater flexibility, which is useful and what schools already require.
Were law schools to adopt this, they would be able to accomplish a couple things. One is that they can better communicate course material mastery without having to worry about being able to differentiate students. Two is that they can control more carefully how much information they reveal about their class, and in a way that is arguably more accurate and less dependent on professorial whim. Law schools already rely on a curve, but the curve can still be tweaked depending on the professor. This form of assessment aggregation better recognizes what really matters: relative position.
Some, or many, may question the necessity of such a change. Though not necessary, I think the change would be useful and beneficial for many. There would be less of a possibility of a school gaming the system, or getting screwed by one’s school not gaming the system appropriately. Employers could compare and understand grades more easily. And there would be greater clarity to the students in how they are being assessed and compared.