A grade warrior in Chapel Hill, North Carolina, is trying to restore meaning to the grades given. All of us--students, faculty, administration, and anyone who plans to hire any of us in the future--owe him a vote of thanks. Here is the article.
In this post, I would only like to reflect on my own practice. Grades are crude, but they are the best we can do when we take on eighty students or more at a time. It would be much better to say that Bonnie (here used fictitiously but in fact one of my favorite students from my fifty years of teaching) writes descriptively with real ease and clarity, but whose analytical writing gets abstract and hard to follow; that she does not pick up new material quickly, but manages to master it all by applying herself; is very resilient in her work; shows amazing insight into complex issues, often leading her peers to look at things in a new way. That would be a useful array of things to know about Bonnie, wouldn't it?
Alas, we don't do that. We say A, B, C, D, or possibly P, and then those other letters, which include a word starting with f-. Really. You can look it up.
Using A--F, the question is whether it reflects absolute judgments or relative ones. In other words, when I know that Bonnie has received a B, what do I know about her work? Using the relative criterion, we would know that Bonnie's work was not as good as the A students, but better than the C students. Back before my time, some universities used pre-selected percentages so that A meant the top X% of the distribution. If comparison within the class is the most important thing to know, you can hardly do better than that.
The absolute criterion is most often used to measure mastery of knowledge or of skills. Bonnie--oh, let's go to Scott--knows or is able to do 80% of the things I have been trying to teach him. That might be more than anyone else in the class or less than everyone else in the class. We don't know. This is a very useful criterion for outsiders if the things that are measured are clear and similar to the outsider's use of them. Otherwise, not so much.
Each of these systems has its difficulties. I use a hybrid notion, which, of course, has its own difficulties. I use an absolute distribution for final grades. There are 900 points available and if you fall in the highest nontile--900 to 800--you get an A of some sort. My students could all get A's or none get A's. One of my most treasured memories of teaching at PSU is the woman who challenged me on that distribution on the first day of a new class. "Do you mean," she said, "that every student here could get an F?" I said, "Yes, or every student could get an A." She stood up and walked out, and in doing so, probably saved both of us a lot of grief.
For individual assignments or tests, I specify in advance what needs to be in the paper or, if it is a test, on the synthesis essay. The material needs to be formatted properly, and developed in an orderly way, and to treat the essential concepts correctly, and to cover the range of material and so on. You do all those things and no more, you will probably get a C on that assignment or test. If you do less, likely a D; if you do more--more on "more" in the next post--likely a B or an A. Complete failures at one or more of the criteria will produce a grade of F.
That combination gives me the final distribution I want--taking into account all kinds of student skills like brute memory are regular and careful completion of the assignments. It also gives me the comparative measurement I want on those narrow skills for which the criteria can be made entirely clear to the class and can be measured with reliable judgment, if not with complete precision.
Next post: What is "more?"
No comments:
Post a Comment