Saturday, February 21, 2009

Giving an A for effort

A recent NY Times article laments the fact that students these days have come to expect an A for effort.

Thinking about it, I'm actually ok with the idea of A for effort (there's the Marxist in me again), Assuming of course we could measure effort accurately, and they really put in sufficient effort. From a contract theory point of view, arguably grades are designed just to incentivize effort, and if they really put in maximum effort then an A makes sense.

Also, from a Rawlsian distributive justice point of view, it also makes sense. There's an interesting paper by John Roemer on how in a fair Rawlsian economy wages should be based only on effort. and things like innate ability and privleged background should all be subtracted out.

The only reason not to give the A for effort is if we believe our job is to provide accurate signals for employers or if it would be unfair to other students. Though perhaps the ability to put in effort is the only dimension that employers really care about. That is far more useful to them, than the ability to write essays about the fall of carthage or to derive Legrangians for maximization problems. There is evidence (that gladwell for example popularizes in his latest book) that innate ability doesn't matter too much, and that people we call geniuses like Mozart, are people who just had a low cost of effort. They only became geniuses after 10 years of hard effort and practice.


James Lin said...

What? Employers don't care about effort. They care about people getting stuff done. A nincompoop who tries hard but can't accomplish tasks is useless.

Who says everything needs to be fair? Why shouldn't people with innate abilities have advantages? They should be rewarded with better reproductive opportunities.

HoBs said...

"They care about people getting stuff done"

they do care about getting stuff done. but the ability to get stuff done in a class (even a comp sci class) may have nothing to do with what they care about. like when was the last time you had to prove the halting problem? but someone able to put in effort to learn the halting problem, could put in effort to learn whatever new task the employer had.

yeah, things don't have to be fair, but it would be nice if they were. so rawls' argued that the society we live in, should be the one we would choose if before we were born, we had to pick the society, but didn't know who we would be born as.

i think that's reasonable.

ok, maybe there is a (potentially immoral) eugenic argument, but i really don't think there's much evolution going on anymore anyway. darwin said you need evolutionary pressure to get survival of the fittest, and we live in a world, at least injavascript:void(0) the US, where pretty much anyone can survive.

James Lin said...

Effort expended and the ability to learn are separate, though.

Someone who learns things easily without spending much effort is much more valuable to an employer than someone who learns things only with a lot of work.

From the standpoint that grades are a useful metric to employers, it's true that they aren't representative of "ability to learn". I think turning them into a measurement of effort spent is the wrong, though; you still need the "did this person actually learn the material" component.