Over breakfast today I told two classmates I expected the torts exam to be the most difficult, crim law in the middle, and civ pro the best chance of a decent grade. They said "uh, everyone thinks they'll do well on the civ pro exam."
That is true. Everyone enjoyed the civ pro professor so naturally we all think we'll do great on that exam. But thanks to the curve, 20% of us, even if we are all civ pro geniuses, will nevertheless get C's. And on the torts exam, though we may all dread it, 20% of us will nevertheless get A's.
So in fact, the exam people feel generally most confident about will be the one that's most likely to produce a grade below expectations. And vice versa. Man, I hadn't thought I'd get a better grade in torts than civ pro but I've just convinced myself it could well happen.
03 Dec 04
I love the website. I'd like to talk to you about it.
As for the grades--you are correct. What is more appalling is that if you were compared against last year's class or next year's class, you grades might improve or worsen. So to say you were a genius for finishing near the top of your class is a false sense of security and frankly a false way to evaluate people who are competing for jobs and career chances that determine many of the future opportunities facing them.
In my opinion, an A is a A, not it's an A for this year but next year it might be a B.
Congratulations. Well done.
I imagine there's a lot of law students who would agree with you, especially since the level of admissions competition has ticked up noticeably in the last few years. Bigger pool of applicants + same number of seats = curves are getting meaner everywhere.
The LSAT has an elegant solution -- they curve the results, but use 10 years of results. So if everyone who takes the test the day you do is a super genius, you won't be disproportionately penalized.
Ultimately the origins of the scaled grading system must have something to do with the hiring habits of law firms. It doesn't matter how many A's you give out, certain firms just want to see the top 10% (or whatever) and a gentler curve doesn't change the overall rankings, just what GPA is nominally assigned to them.
Statistically I think the best reason to have a widely distributed curve is to create contrast, and reduce the role of rounding errors. The shorter the curve gets, the coarser it becomes, and the more difficult it is to numerically distinguish an outstanding student from a good one, or a good one from a mediocre one.
Posted by: MB at January 12, 2005 05:02 PM