Grading on a Curve: A Prisoner’s Dilemma
When a class is graded on a curve, many students feel that it causes them to have to compete against other students for grades. What they may not realize is that they can use this curve to their advantage. Bryan Caplan, an Economics Professor at George Mason University, grades on a curve. He tells his class that if everyone were to study less, they would learn less information but would still attain the same grade. They just keep on studying because they hope that they will be outdoing their fellow classmates and don’t want to fail if their classmates do indeed study. As a result, he knows that collusion will ultimately fail and is not worried that they will be successful.
One of my professors here at Cornell offered our class of about 80 students the same proposition. He stated that he gives the two highest grades an A+ and then curves down with a normal letter scale from there. If everyone got the same two high grades then there would be nothing that he could do. He said that collusion has not worked in all his years teaching and is sure that it will not work with our class.
The root of this predicament lies in game theory. If no one were to do any work, then they would all get the same good grade and be better off (assuming that they value free time more than they do attaining knowledge). However, if one person were to “cheat” and actually do the work, then this person would get an A+ and everyone else would suffer. Thus, it would be better for someone in this group to do the work, as would be the thinking of everyone. Ultimately, everyone does the work and collusion fails.
This is an example of a Prisoner’s Dilemma: they know that they are better off if no one does any work, but out of fear of someone not “following the plan,” they themselves do not follow the plan. Thus, the Nash Equilibrium is that everyone ultimately studies. The greater the number of people who need to collude, the more difficult it is to coordinate and make sure that no one is “cheating.” Collusion is bound to fail and Professor Caplan knows that.
-DTM
http://econlog.econlib.org/archives/2007/02/collusion_in_th.html
[…] professors specifically make light of the incentives to grading on a scale. As described by one Cornell student: One of my professors here at Cornell offered our class of about 80 students the same proposition. […]