A conversation on blind grading began over the weekend as comments to an unrelated post. I am moving the conversation to its own post so that this valuable discussion does not get lost.
The comments raised two different but related issues: (1) Is blind grading a myth? and (2) Why do professors grade participation? I'm going to address these questions in reverse order because the second question has the shorter answer.
Why do I grade participation? Grading participation is an administrative hassle, but this year I have chose to include participation in the final grade as a "carrot." The one thing that professors want is an active class. I want students to read the material, come to class, and talk about the material. I want volunteers and I want students to answer my questions when I call on them. I don't even care what your answer is. I am not in love with the sound of my own voice (for many reasons!) and I abhor that Ferris Bueller silence ("Anyone, Anyone?") I also hate it when I call on students and they read out of the book. So, I can try to achieve my classroom goals in many ways. First, I could have an attendance policy and reduce your grade if you don't show up. That's a stick. Instead, I use a carrot. I will raise your grade if you suit up and show up. In practice, it may work like a stick. If the whole class except one comes to class and participates, then effectively the no-show's grade is reduced because everyone else's grade benefits.
Is Blind Grading a Myth? No. I have been teaching since 1998, and I have never seen any evidence that any professor graded exams on any other basis than blind. Yet, this perception still continues among students. I'm not sure if there is anything that I can add to dispel that misperception, but I'll try. First, I think this perception is based on an assumption that the professor has either an intense like or dislike for another student that would trump that professor's own good judgment. To give a biased grade, a professor would have to be willing to put his or her own career at stake to either benefit or abuse a student to the tune of a few credit hours of a total of 90-100 credit hours. I can't imagine that the enormous cost would justify the small benefit. Absent an inappropriate romantic relationship, a professor has too many students over the course of a career to become so passionate about one student's situation to get involved in that manner.
Second, I also think that the elusive quality of grades, especially first-year grades, leads students to latch on to bias as a factor in grades. Unlike in undergraduate courses, grades in law school do not equate neatly to effort. As you sit in your first-year Torts class, you guess who will do well based on who seems to study the most. You gauge your own studying accordingly. However, the resulting grades surprise everyone. How did the professor grade these exams? As students struggle to make sense in a world that seems more arbitrary every minute, law students seem to subscribe to either a chaos theory of throwing the exams down the steps or a bias theory of nonanonymous grading. (We were convinced our Crim Law teacher randomly assigned grades.)
Because it is too unwieldy for professors to hold after-exam rap sessions where the professor explains why each student received each grade, there is a gap in the information. Neither students nor professors are very good at guessing who will get the top scores in their classes because most exams are written to grade application, not information-gathering, and the ability to apply knowledge is not readily recognizable until the exam. I know that last semester, I was surprised by the top-scorers in my Torts class.
Finally, once students are willing to consider that the professor was biased in the grading, students may point out that some high-scorers were students that the professor called on in class a lot or were talking to the professor after class. The cause and effect part of the scenario is not: Student A talks to Professor B after class a lot/Professor B gave Student A a high grade/Professor B plays favorites. The cause and effect part is probably: Student A really likes Torts and understands it/Student A wants to ask questions that are beyond the class discussion, so Student A talks to Professor B after class/Student A gets a high grade in Professor B's class/Student A has an aptitude for analyzing the rules and policy of Torts law.
Professors are human. We like students who like us. We call on students who raise their hands. We go back to students who give good answers and do the reading. We like to chat after class with students who are interested in the material. However, we do not take actions to ensure that those students will get good grades.
TrackBack URL for this entry:
Links to weblogs that reference Blind Grading:
1. Posted by jh on April 4, 2005 @ 8:31 | Permalink
Do you keep a record of class participation? If not, I think participation grades are wrong--just a way for profs to reward people who brown nose. If, on the other hand, the prof writes a check mark per comment or something, I think it is good.
2. Posted by phred on April 4, 2005 @ 8:32 | Permalink
I'm unclear on what you mean by 'blind' here. You argue that blind grading is a myth based on the costs of giving a biased grade. Grading can be biased with or without being blind (e.g., biased against bad handwriting), and it can be unbiased with or without being blind (e.g., where the professor doesn't even know the students' names). Is your argument that law professors don't circumvent exam anonymity because the only reason for doing so would be to bias a grade, and biasing a grade isn't worth it? Or is it just that by 'blind' grading, students just mean 'unbiased' grading.
3. Posted by Christine on April 4, 2005 @ 8:45 | Permalink
1. I do keep a running list on participation, and I think that is the norm.
2. I am using "blind" grading to mean "anonymous" grading, so when I then refer to "biased" grading, I refer to grading that would be biased in favor of a student's known identity. I think all good professors struggle with other types of conscious bias in grading law school exams -- bias against bad handwriting, bad spelling, incoherent sentences, and stream of consciousness organization. I tell my students ahead of time that I will try to stay with them and wade through their exam looking for substance, but I'm not sure how many profs are willing or able to do that. I had 43 Torts students last semester. The next time I teach it, I will have twice that. I'm not sure if I will have the patience to hunt for substance if the form becomes too much of an obstacle. (I only had one exam that I would describe that way.)
And at some point, form and substance merge. You could say that I have a "bias" against superficial analysis, and I would admit that I have a "bias" for well-reasoned answers.
4. Posted by D. Sgruntilt Wonelle on April 4, 2005 @ 8:53 | Permalink
So while you're yakking it up after class, day after day, sharing subtleties of Torts beyond the class discussion with the non-favorite favorites, we students in percentile 1-50 stand angrily by trying to get some remedial help, and feeling further lost by the substance of your musings with your non-favorite who winds up with the A. Your conscience should be clear, since the non-favorites pay so much more for the privileged of sitting in your class than the rest of us, and you have no obligation to make sure we all get a minimum level of instruction. Nice ivory tower there, pal.
5. Posted by Joshua on April 4, 2005 @ 9:04 | Permalink
I'm just a passer-by, neither law prof nor student, but I would think that the "minimum level of instruction" that the professor is obliged to provide would be covered by the class, and perhaps office hours. Maybe the disgruntled should try asking questions during the actual class, rather than pinning their entire hopes on the few minutes of informal yakking afterwards?
6. Posted by Bob Dobalina on April 4, 2005 @ 9:10 | Permalink
Don't fellow students hate "gunners?"
7. Posted by Kate on April 4, 2005 @ 9:10 | Permalink
Sometimes there are students who hijack the class and WILL NOT shut up when all you had was one or two short questions. These students think they know everything there is to know about, say, Torts, and will ramble on and on about things that are a chapter or two ahead. So you wait till after class to ask the simple question and there is another one, going on and on about something he read in a magazine article and compares it to something amazing he read in a hornbook. Classes are not that long.
8. Posted by Scott Moss on April 4, 2005 @ 9:10 | Permalink
Dear D. Sgruntilt Wonelle:
Perhaps the reason you're in percentile 1-50 is not some professorial conpsiracy against you, but the simple fact that you can neither write a coherent parahgraph nor respond in a civil manner to a reasoned discussion.
9. Posted by DBL on April 4, 2005 @ 9:12 | Permalink
One way to help students understand how exams are graded is to make available, after the exam, the best student exam paper in the class for all to read. Your average students will marvel at all the issues they missed, at the subtleties they glided by, and will begin to perceive that grading is not purely random.
10. Posted by Scott Moss on April 4, 2005 @ 9:32 | Permalink
In the prior thread that turned into a discussion of blind grading, I at one point posted that as I applied Marquette's blind brading system, “we grade blind, turn in the grades for each exam, and then get a sheet back with those grades attached to names. We then can change the students' grades based on participation, attendance, or other relevant factors (in my view, 'factors I mentioned in the syllabus')."
And then an anonymous poster responded, “So, you people admit that exams are not really graded blindly. Grading things up or down with a 'specific justification' means they are not graded blindly.. . . Yes. I think professors act in bad faith. I am not saying that they are sitting around evilly cackling about how to harm or help students, but they are usually quite up front about 'the type' of person they like to see succeed. Students often try to pretend to be that type of person and usually it works, especially if that student has had years of practice.”
This is an interesting response that bears response.
(1) This clearly IS blind grading of exams, which eliminates the real possibility that professors will be biased in assessing how well you answered the exam questions -- biased based on which students professors expect to do well, biased based on race or gender, whatever.
(2) I can't see how it's "bad faith" when "professors ... usually quite up front about 'the type' of person they like to see succeed." I think the ideal grading system is (a) blind grading of exams plus (b) full, transparent disclosure of what else affects your grade. That's why I'll go up or down from the blind exam grade only for factors I mention in my syllabus.
(3) I think it's 100% educationally legitimate for professors to have a "type" of student who will get a boost for participation. Obviously a bad or vain or stupid professor can choose an inappropriate "type" to grade up, like a preference for suck-ups or a gender bias or whatever. But I don't think I'd be grading students' academic performance accurately if I said, "OK, two students got a BC on my exam, and one of them consistently gave cogent analyses in class, whereas the other never spoke a word -- but I have to give the same BC grade to both." I think a higher level of educational performance -- legal reasoning, ability to form and respond to arguments, etc. -- was shown by the student who had the BC exam but all semester showed a strong ability to comprehend and apply case law. That student earned a B rather than a BC in my book.
(4) In law, there's always a tension between hard-and-fast "rules" and murkier "standards" (as my students know). I respect professors who go by blind grading with no participation adjustments; that "rule" has the advantage of clarity/transparency -- but it's a mistake to think that automatically means it's the a superior system. The "standard" I apply, i.e., modifying blind grading (for only a small number of students, incidentally) based on fully disclosed considerations, is slightly less transparent but has the advantage of, in my view, more accurately assessing eductaional performance. The trick with "standards" is to make them as transparent and predictable as possible. I go up or down from the exam grade for only a small number of students; I disclose the limited number of reasons for an up-or-down adjustment in my syllabus; I never go up more than one grade notch; I won't adjust downward without very good reason, and I won't go down more than one notch without very very very good reason.
I think that most of my students would guess right if I asked (though of course I won't really ask) whether they thought they'd get a plus factor, which is to say I think my system maximizes accuracy of assessment without sacrificing all that much transparency.