Clarifying the grading policy

I'm mulling a reply to this comment from Kirsty on the subject of appearance and mixed-sex friendship.  But with class in twenty-five minutes, it will have to wait.

Tony Vila and I have been having a bit of a disagreement about subjective grading (or "curved" grading) in the comments section below this post.  I wanted to clarify my grading policy (which might be helpful for students who read this blog).

Yes, I do grade students in comparison with one another.   This is done most simply with the multiple-choice quizzes I give in my Western Civ classes (I don't give such quizzes in the gender studies courses).  I take the highest grade on the quiz, whatever it may be, and declare that to be the A.  I simply go down from there in 10% increments.  My quizzes have 25 questions; usually the highest grade is a perfect score or near that.  If the highest score is low (it's been as low as 17 out of 25) I again apply my 10% rule.  It's not a perfect curve, because while someone is guaranteed an A, no one is guaranteed an F.  After all, if the high score is 25 and the low score is 21, 21 is 84% of 25.   It's never happened, but if it should, everyone would receive As and Bs.  The same thing would happen if the high score was 10 and the low score was 8.  (Also something that hasn't happened in over a decade of teaching this way.)

With essays, I continue to compare students to each other.  I always read midterms and papers twice: the first time to select the best essays in the class, the second time, to rank all of the others relative to those best essays.  Grades are assigned accordingly, but again, no one is guaranteed a failing grade -- even if there is always at least one A.  Of course, it's conceivable that my best paper could be of such poor quality that I couldn't justify giving it an A.  Thankfully, that hasn't happened to me in my years of teaching.  (Admittedly, I've been close: I've had a couple of sections where I've given only one A in a class of more than thirty students, but those occasions have been rare.)

Philosophically, I think grades serve to rank students relative to one another.   This ensures that they aren't trying to meet what may be my unrealistic standards, but are instead striving to match what their classmates -- who take the same exams under the same conditions -- have already proved possible to achieve.   

I'll surely catch some flak for this, but my grading policy is linked to, of all things, my experience running.   In 1999-2000, I was at my fastest as a runner.  I did lots of marathons and 10Ks.   I remember that in October '99, I ran my personal best time in a huge 10K on the Westside (38:49, for those who care).  I finished 27th male overall in a highly competitive field.  No medal for me.  Six months later, I ran in a much smaller 10K at the Rose Bowl -- and to my utter amazement, won the damn race despite running a 40:59, more than two minutes slower than my PR time.  I got a nice little medal and a ribbon, and even better, the sweet feel of tape hitting my chest!  (It was a small 10K, nothing fancy, and to be fair, there were only about forty or fifty people in the whole race.)  Whether or not I got a medal was contingent on two things: my performance, and the performance of the others who showed up that day.  I tell my students that my grading policy is similar, and that their results will be linked to the results of their classmates.   In this way, I encourage them not to do the "bare minimum", but to do the absolute maximum because surely some of the "competition" is willing to do just that.  I had this standard in place before 1999, of course, but these experiences on road courses helped clarify what I hope is the essential justice of the policy!

These students will apply for transfer, where their applications will be considered as part of a pool and where they will be analyzed subjectively.  They will apply for jobs where they will be directly compared to other candidates.  In virtually every other area of American public life, success is determined both by one's own performance and the performance of those around you.  It seems reasonable and fair and right to apply those same standards in a college environment.

Comments

[mercedes]

Hugo,
I feel the same as you do. I think it's only fair that those who put in the best effort be given the highest grades. ___I also agree with the comment you made in one of your classes about limiting your class load according to your schedule in order to compete effectively.__A very wise comment. I think knowing your limitations and scheduling appropriately can save a student a lot of stress.

[Emily H.]

I think you may well be right in a college environment (where grades are kept much more private, on the whole, and the social environment is less poisonous).

I had a bad time of it in high school when I really destroyed the curve on a few tests (in one case, I got a 97 when no one else got higher than an 80). No one else thought it was fair--and, you know, I didn't quite think it was fair either, because I was putting in hours of study on extra work, outside the bounds of the curriculum and the textbook, for my own personal benefit. I'm not really sure it was reasonable to expect everyone else in the class to come up to those standards.

[Hugo]

Emily, I assure you I never make known those who get the highest grades! But I do think that an A ought to mean something, and what it means to me is "best in show", as it were.

[Anne]

Well, I entirely disagree, I'm afraid. I don't approve of "curve" grading.

Students shouldn't be measured against each other. They should be measured against an objective standard...one like, "did they learn the material?"

[Hugo]

Anne, trust me. Any student who gets an A at the end of the class has learned the material. But if no student has reached a certain standard, the problem is less likely with them than it is with me -- and I need to change my standards. Competition seems to improve, rather than retard, actual learning.

[Tony Vila]

Look, are you arguing for relative grading on a deontological "what's right" scale, or on what practice makes the best results?

Deontological: I think the argument that a student's reward and what they are judged against should be independent of how other students in that class performed that year, is pretty strong. The ability for a teacher to objectively know what a C or an A is regardless of how many people got them, isn't much harder than the ability of a teacher to tell which papers are better anyway.

Pragmatic: I'm sure relative grading hasn't hurt your classroom and is fine for your situation. But it remains that relative grading is still a bad idea in many other places, and makes one of the key uses of grades (a signal to employers and other schools) suddenly collapse and have to be based on the avg of that school, class, and year.

[Hugo]

Tony, I'm definitely arguing the latter.