Incentive Programs
Over the years I have given a great deal of thought to grading, incentives, and fairness. One anecdote will illustrate what started me on the road to thinking about how often our grading systems can slip into the red zone of unfairness was an incident when I was a TA in graduate school. The professor I was teaching for (mind you this is anthropology) maybe didn’t have the clearest grasp of basic math, like percentages, means, modes, and medians.
This professor had constructed a wildly illogical method for ‘smoothing’ out those pesky in-between grades, you know, the ‘so close to an A minus that you can taste it B plus’ and the ‘oops, one slip of a digit can send this guys off the map into F land’ kind of grades. His system, and sadly I can’t remember all of the gory mathematical details, had the effect of rewarding good students for doing well, and punishing bad students for doing poorly, that enhanced the bi-modal distribution of his grade distribution.
He is not alone in this, and I remember some of my thinking on this subject was sparked by an NEA Higher Education Advocate newsletter (http://www.nea.org/home/37810.htm) which reported that quite a few of our colleagues have somewhat sketchy grading practices (I’d love to hear some further examples from you in the comments section!)
An example drawn from my son’s elementary school can help illustrate what can happen. Recently, a teacher was trying to incentive effort by measuring performance on multiple choice exams, while not understanding some of the statistical effects such a practice would have:
If students improved their math scores by one percent, they were rewarded. Students who did not improve got extra work, and students who had already done well were also rewarded. Problems ensued because they had created a two-pronged incentive system that rewarded good students for doing well, and rewarded poor students for doing ‘better’ but punished the middle for not improving even when they have exceeded a statistically average outcome. I like to call this the, “you got a B but you should have gotten an A so I am giving you a C” rationale, courtesy of a former social sciences teacher of mine.
This system is arbitrary. Let’s see why:
If a student is struggling in class, and got say, a 59%, then brought it up to a 60%, they received a reward though they are still performing below the average; whereas the child who got an 80% and didn’t make any improvement got extra work. Statistically speaking, it’s much harder to close a smaller gap than a larger one. Also statistically speaking, a low performing student could still, through random chance, improve their grade by one point more easily than a higher performing one. To further ‘de-randomize’ the experiment, the teacher had actually removed the top scoring students from needing to improve. What if they had gone DOWN on the second try? They have luck on their side on multiple choice exams just as much as any other student. Regression to the mean practically stipulates that some of those pupils would have performed more poorly.
I appreciate the attempt to incentive effort – and believe me, I know how hard it is to do this sort of thing; I wrestle with it all the time with my own students. But I think seeking to reward ‘intent’ by measuring it with performance on a multiple choice exam, is a recipe for unhappiness. Here are some things I do in my classroom to when I want to incentivize either effort or performance, but do not wish to conflate the two:
For rewarding performance
If you get 100% on an exam, I give you a 5$ gift certificate (like to Starbucks). This is a straightforward outcome, you either achieve it or you don’t. I’ve given out two in the past five years.
For rewarding effort
I have a ‘maps’ assignment that we never get time to finish (a filler activity) so I send them home and say, ‘hey, you can work on this at home, or not, your choice.’ Then, in a future class, at the end of the session, I’ll give them 15 minutes to work on the maps, but those who did them at home get to turn them in and leave early. This rewards extra effort.
To promote effort
I will often reward everyone for trying whether they all ‘deserve’ it or not. Nothing succeeds quite as well as liberal praise. Further, I am a big believer that stickers and candy, judiciously applied, remain as great a motivator for eighteen as they do for eight.
