« Legal Scholarship Jams | Main | Assigning Cases Decided Mid-Semester? »

Monday, April 27, 2009

The Professorial Second Opinion: Should Law Professors be Able to Change Student Grades After Exam Conferences?

From the movie Clueless :
Mel: Cher, what's this all about

Cher: My report card?

Mel: The same semester?

Cher: Uh-huh.

Mel: What'd you do? Turn in some extra-credit reports?

Cher: No.

Mel: You take the mid-terms over?

Cher: Uh-uh.

Mel: You mean to tell me that you argued your way from a C+ to an A-?

Cher: Totally based on my powers of persuasion. You proud?

Mel: Honey, I couldn't be happier than if they were based on real grades.  
Many law schools preclude professors from changing student grades based upon "errors" that those professors made during the grading process. At my law school, John Marshall, professors cannot change grades based upon qualitative errors, such as failing to notice that a student addressed a hearsay exception on an Evidence exam, but they can change grades based upon computational errors, such as assigning 5 points to a student's discussion of a hearsay exception but failing to add those points to the student's final score (In this way, my school is like Federal Rules of Evidence 606(b), which precludes jurors from impeaching their verdicts based upon qualitative errors, such as failing to consider the plaintiff's comparative negligence, but allows them to impeach their verdicts based upon transcription errors, such as finding the plaintiff 10% negligent, but reducing his award by 40% based upon bad handwriting).

I suspect that there are two main reasons that many law schools generally preclude professors from changing student grades. First, blind grading rules the roost in law school, and if we routinely allowed professors to change grades, it would re-introduce the possibility of biased grading that it was designed to eradicate. Second, law students like to argue. This is why many of them came to law school in the first place. If law schools readily allowed professors to change grades, the number of exam conferences would greatly increase (which wouldn't necessarily a bad thing) as would the contentiousness of those conferences (which would necessarily be a bad thing).

These are certainly strong reasons to preclude grade changes, but it is no doubt extremely difficult for the professor, and, of course, the student, when the professor has to explain that the student deserved a B but got a C because of a mistake that the professor made. Now, maybe this doesn't happen very often (I've only heard of one such case in 1.5 years of teaching), meaning that the benefits a change would accomplish would not outweigh the burden the change would impose. But if the problem is more prevalent, what might be a way to allow for the correction of obvious grading mistakes without facilitating the problems addressed above?

In The Professional Responsibility of the Law Professor: Three Neglected Questions, 39 Vand. L. Rev. 275 (1986), Monroe H. Freedman hints that there should be "due process in grading" and notes that he uses the following review procedure (which apparently his school does not preclude):

The review procedure works as follows. If the student is not persuaded that the grade I have was a fair one, he or she can elect to have the grade reviewed by a committee of three students from the same class. I pick one of the three committee members, the student picks the second, and those two pick a third. I then explain to the committee how I arrived at the grade I gave. The student then explains to the committee why the grade should be higher. The committee then chooses between my grade and the one the student considers appropriate. The committee must choose one of the two; it is not permitted, for example, to split the difference between the two grades.

The committee uses whatever standards its members consider fair. I impose no criteria. The committees, however, have tended fairly consistently to review the challenged grade in the context of other grades given in the same exam. (I make it clear to them that I cannot lose: either they affirm my grade, or they validate my review procedure.) The process has taken an average of two or three hours of my time a year.

I could see a law school that currently precludes grade changes adopting this procedure because it would (partially) preclude claims of biased grading and, if Professor Freedman's experience is representative, it seems that it wouldn't (really) increase the number or contentiousness of exam conferences. Of course, such a procedure only addresses cases where there is a genuine disagreement between professor and student over grading, not the situation where professor and student agree that there was a grading error.
 
Professor Freedman found that his procedure did not produce the negative result of students always affirming his grades, but I would have to imagine that students would readily agree to a grade change when both professor and student agree that there was a grading error. Thus, such a procedure in such a circumstance would in effect be the professor changing the grade, which reintroduces the possibility (or at least the perception of the possibility) of biased grading (and if the students rejected the grade change, wouldn't they, in effect, be saying that the professor was biased?).

But this seems to me to only be a problem if the student reviewers engage in non-blind grading. In other words, if professor and student agree that there was a grading error, three students could be appointed to review the exam, and they could engage in "blind grading" in that they wouldn't know the student involved, the initial grade, or the proposed new grade. Of course, one might argue that without the input from professor and student that Freedman's procedure provides, the students wouldn't know how to "grade" the exam. I imagine that the solution would be to give the students 3 exams that received scores (1) about the same as the initial score, (2) about the same as the score the exam "should have gotten" without the error(s), and (3) somewhere in between the actual and "deserved" score. Using these guideposts, the students would assign a score to the exam that would or would not change the grade. Of course, this could be done by another professor as well, under what might be called a professorial second opinion. 

So, what do readers think? Should law professors be able to change student grades, and if so, under what circumstances? Or, as Richard Heny Seamon notes in Lightening and Enlightening Exam Conferences, 56 J.Legal Educ. 122 (2006), is it "arguably unfair...to change the grade of a student who requests an exam conference without re-scoring exams of students who do not request conferences but whose exams might contain clear scoring errors?" Of course, the best way to "solve" the issue is to avoid the issue to the greatest extent possible, and I'm sure that many professors follow the same technique as Seaman:

I try to mitigate this arguably unfair situation by automatically re-scoring all exams with scores that fall within two (and, for some courses, within three) points of the next higher grade cut-off. For example, if I have decided to award a grade of “A” to all exams that have scored between 96 and 100 points (where 100 points is the highest score), I will re-score all exams with scores of 94 and 95 points; and, if the point range for “A.” is 90-95 points, I'll re-score all exams with scores of 88 or 89 points, etc.

Posted by Evidence ProfBlogger on April 27, 2009 at 09:08 AM in Teaching Law | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c6a7953ef01157051da64970b

Listed below are links to weblogs that reference The Professorial Second Opinion: Should Law Professors be Able to Change Student Grades After Exam Conferences?:

Comments

I would be in favor of allowing professors to change grades given in genuine mistake with the caveat that no more than 1 such change will be made over the course of the student's law school career. Given the rarity of such errors, it is extremely unlikely that students will be subject to more than 1 over the course of their law school career. Limiting the changes to 1 solves the problem of students who feel like grades are negotiable and routinely get half their grades changed every semester just by arguing (and yes, I did know people like that in law school).

I was one of the unfortunate people who did experience an objective qualitative grading error during my time in law school. The professor acknowledged that the mistake was his, acknowledged that there was nothing I could have done to prevent it, and acknowledged that my grade would have been a full letter grade higher if he hadn't made the error yet refused to change the grade on the principle that doing so a) was unfair to any student on whose exam he might have made the same error but who didn't contact him, and b) would encourage too many non-meritorious student complaints from repeat complainers who always try to squeeze out an extra 1/3 of a letter grade on every exam, every semester.

While I understood his point, as someone who had never before in his entire life, birth through law school, quarreled with any grade before and was only doing so then because the error was flagrant, obvious, and significant, it felt extremely unfair. Needless to say, now that I'm grading my own exams, I do not share his approach. Nevertheless, I always wonder when a student comes in how many other professors the student has approached that semester and whether entertaining such appeals simply rewards certain repeat complainers rather than allowing students to double-check my work to find mistakes that I am absolutely sure I did make in the course of grading 100+ exams every semester.

Posted by: Anonymous | Apr 27, 2009 9:57:11 AM

Thanks, Anonymous. Your "rule of one" sounds like it could be an interesting compromise. It could, however, create a problem if there is a mistake in grading that would change a student's grade from, say, a B to a B+, and the student is uncertain whether to take his one change or save it for a situation down the road in case he uncovers a larger error (such as your full letter grade example). Of course, as you note, such errors are probably rare, so this probably wouldn't be that much of a problem. And, I suppose that we could let that student take the first grade change, and then, if the student uncovers a "worse" error later, the student could forfeit the earlier grade change for the new grade change.

Posted by: Colin Miller | Apr 27, 2009 10:24:25 AM

This is the policy at our school (I think): professors may make grade changes due to ministerial errors (typos, adding wrong, etc.) at any time. All other grade changes must be presented to and approved by the faculty.

If the student wants to contest a grade, there is a procedure that I am not entirely familiar with, but a committee of professors (and maybe a student) hear arguments from both sides, and render a decision.

Posted by: Michael Risch | Apr 27, 2009 2:57:38 PM

To correct calculation errors, yes. To change the substantive evaluation of the answers, no.

Allowing students to contest substantive grades will tend to favor those who are the noisiest and make the biggest stink: I don't think law schools should reward that. Of course, that puts a burden on professors to grade as accurately as possible to ensure that the grades are as fair as they can be.

Posted by: Orin Kerr | Apr 27, 2009 3:58:11 PM

I've always wondered what quality-control mechanisms were in place with the registrar. Does the professor or a second person review the grades that are input? Wouldn't it be pretty easy for a typist inputting the letter grades of 30-50 students to mistakenly read the wrong line, or worse, be bribed? Many professor don't give back exams, or even have objective criteria for grading, so how would a student even know if the wrong letter was input?

Posted by: Wonderer | Apr 27, 2009 4:30:18 PM

Orin, that's one approach. The other is to do what district courts, in combination with appellate courts, do: either describe the case as turning on an essentially unreviewable question of fact, or define the initial scope of discretion sufficiently broadly that only a difficult-to-show abuse of discretion can overturn a decision.

Of course, like you, I take the first approach of simply making sure my grades are highly accurate.

Posted by: Paul Horwitz | Apr 27, 2009 4:31:30 PM

To the commenters who have a policy of not correcting qualitative errors in grading--have you ever had to have a conversation with a student in which you had to explain to the student that although the mistake was yours, you are not willing to adjust their grade?

I admit my own experience with having been a student in one of those conversations has profoundly impacted my own grading today. With grades as important as they are to many of our students, especially in first year courses, and with the knowledge that the B- I accidentally gave a student who really deserved a B+ might mean the difference between getting a job and not getting a job, I am unwilling to let the student bear the brunt of what is entirely my mistake. Having said this, I have yet to change a student grade because of a qualitative grading error. When I pass back exams, I tell students that I welcome office visits to discuss their grades but warn them upfront that I have a policy of not changing grades unless grading mistakes were clear and unambiguous.

The mistake when I was a student was that the professor appeared to have confused my exam with a different exam and failed to award me credit for 1/3 of what I wrote in response to one issue spotter because he believed (incorrectly) that I had already discussed that group of issues in an earlier part of the essay. He wrote "do not repeat yourself" in the margin and not only awarded me 0 points for the section but subtracted 2 points for the supposed redundancy error. Like Michael's school, my school's policy for grade appeals involved a faculty committee, and I was advised by a supportive faculty member not on the committee that in general the faculty would be unlikely to override the judgment of a colleague barring truly extraordinary circumstances (like, a professor who vindictively gave a student an F). So I decided not to protest further.

I would not change a grade if the call were questionable or if a student was merely asking me to rethink a judgment call I made while grading. But if 2 pages of the exam stuck together so I didn't read one, if I blanked out on my 60th exam and didn't notice the student had discussed a key issue, or if there were some other clear, obvious error that was entirely my fault, I can't see flat-out refusing to change the student's grade.

Posted by: Anonymous | Apr 27, 2009 4:35:11 PM

Anonymous, your experience is what I would consider a "calculation error," assuming the professor agreed with you that he had confused someone else's earlier discussion with yours. I.e., it wasn't based on a judgement call, it was based on mechanical error on the grader's part in not keeping the exam answers separate. It sounds similar to the scenario where the prof wrongly gives you someone else's raw score for an essay question and bases the grade on that -- surely that's a mechanical error that could be corrected under most of the policies quoted above.

If the professor didn't agree that he must have been looking at a different exam, well, that sounds like a different matter to me, even if on reconsideration the comment in the margin looked erroneous.

Posted by: Bruce Boyden | Apr 27, 2009 5:05:51 PM

Do the policies that permit grade changes allow grades to go down as well as up? Has anyone that is tempted to change a grade ever really considered a downward change? It seems to me that the one-way-ratchet effect is the biggest problem, since a policy of liberal grade changing will result in many lottery ticket buyers who figure they have nothing to lose.

Posted by: TJ | Apr 27, 2009 5:18:09 PM

Thanks, Michael. How often does professors present their cases to the faculty?

Posted by: Colin Miller | Apr 27, 2009 6:32:12 PM

Thanks, Orin. Indeed, I knew a student in college who contested every grade. Even though he was only rarely successful, that still meant that he had several grades changed.

Posted by: Colin Miller | Apr 27, 2009 6:34:48 PM

Wonderer, I think that it's on the individual professor. I record the grade I give to each exam and then compare that list to the list given back to me by the administration. So far, I haven't come across an error, but I wouldn't be surprised if some are made.

Posted by: Colin Miller | Apr 27, 2009 6:37:07 PM

Thanks, Paul. I would guess that if professors tell students that it would be the very rare error that would result in a grade change, there wouldn't be that many students seeking grade changes. But maybe I am underestimating the average law student.

Posted by: Colin Miller | Apr 27, 2009 7:00:42 PM

Thanks, Anonymous. The errors you list in the list paragraph are exactly the types of errors that I would like to see law professor be able to correct. I just wonder whether we should put that power in the hands of professors or whether, even in these cases, professor should have to submit the issue to third parties.

Posted by: Colin Miller | Apr 27, 2009 7:02:42 PM

Thanks. Bruce. I agree that a common sense view of the errors listed by Anonymous would be to list them as calculation errors that schools like my school would allow to be corrected. But I wonder whether anyone ever come across such an issue and whether their school has seen it the same way.

Posted by: Colin Miller | Apr 27, 2009 7:04:44 PM

Thanks, TJ. That raises an interesting option. If a school allowed it, a professor could inform a student at the start of an exam conference that the student has two options: (1) the student can just get feedback, without the possibility of a grade change, or (b) the professor can seek a higher grade but will have to take the consequences if it turns out that the student deserved a lower grade. I wonder students would prefer this approach to a no grade change approach.

Posted by: Colin Miller | Apr 27, 2009 7:07:03 PM

Anonymous writes:

************
To the commenters who have a policy of not correcting qualitative errors in grading--have you ever had to have a conversation with a student in which you had to explain to the student that although the mistake was yours, you are not willing to adjust their grade?
**************

I have only had one mistake pointed out to me: I once made a math error in which I added .5 extra (out of 130) to student's raw score on a midterm exam. The student pointed it out to me, and I told him I would not deduct the .5 (which was vanishingly unlikely to change his final grade anyway). With that said, I agree with Bruce that what you're describing sounds like a calculation error.

More broadly, I spend a lot of time grading; I would estimate that I spend about a full hour of intense uninterrupted time grading a three-hour exam. I often end up regrading questions if I feel I am off, and if something seems fishy, like a page missing, I call the registrar and confirm that I have the full exam. I do that because I feel a duty to students that I give their exams the most accurate grade I can. One consequence of that is that -- at least so far -- I have not been in the unhappy circumstance of reading over a student's exam and feeling that the grade was inaccurate.

Perhaps the most stressful part of the process is entering in the final grades, especially with a large class. We have these long strings of numbers, something like 10 digits in all, and we need to match the digits to the raw score when we enter them into the computer.I am always paranoid that I will enter a score for one 10 digit number into the computer correctly. ( I could never let a secretary do this, as my secretary is pretty error prone when it comes to this sort of thing.) So I end up checking, double checking, triple checking, and quadruple checking the entering of the numbers. It's pretty stressful. (Why the school can't just use numbers from 1 to the number of students in the class is unclear.)

Posted by: Orin Kerr | Apr 28, 2009 12:35:56 AM

Colin -

Maybe one grade a year gets changed.

Posted by: Michael Risch | Apr 28, 2009 7:36:51 AM

I conKerr with Orin's first comment completely. I would not want to disadvantage students who are not constantly angling, like Colin's acquaintance, for any edge they can get by contesting everything and constantly making a stink.

I would add, speaking for myself, that I see grading as such an inherently subjective exercise (like a jury's determination of witness credibility) that having an appeals process seems silly.

For myself, the best and only way to deal with the issue is to try very, very hard to grade fairly and to get it right the first time. I'm not sure I spend as much time on each exam as Orin does, but, like Orin, I allow grading to cause me some stress. In fact, I'd go further and say that, in my judgment, professors have an obligation to get stressed out entering grades, making calculations, and so forth. I do several checks to make sure I haven't screwed anything up. But when I press send, I consider it done.

Posted by: Professor Anon E. Mous | Apr 29, 2009 8:51:29 AM

Law school grading is far from fair even if grading is blind.

At UT Austin, I gained a 95 (highest grade) on Hamilton's Property mid-term. But then our Property class was handed off for the second half of the year to "acclaimed" SJ, who proceeded to announce that, if one's score on his final were better than one's mid-term score, only the final score would count, else the two would be averaged. That was, of course, totally unfair to me and to many others.

When I got no satisfaction after complaining to the dean, I just quit attending SJ's classes and never even attended a single class of his vaunted Trusts, Wills and Estates. Didn't keep me from passing it, however, then graduating and pursuing my preferred career in computer engineering. I would have quit altogether, if my three years of law school hadn't been virtually free! Now thinking it over, I should have just sat for first year classes and skipped the rest, except maybe for Evidence and the very interesting seminars.

And my fellow student on law review reported sleeping with at least one of our current professors. Morally speaking, law school seems to put you right on the Group-W bench.

I have chosen to work a yearly average of 19 weeks per year since then, and I laugh now when I read of law grads fighting for jobs requiring 2500 billable hours! What choice do they have, not having any ability in math, computer technology or hard science?

Posted by: jimbino | Apr 29, 2009 10:42:27 AM

Post a comment