Friday, August 01, 2008

Problematic Students

One thing they don't warn you about in graduate school -- unless some places have changed their "teaching preparation" classes to be somewhat more useful -- is that, every once in a while, you'll get a student who is, shall we politely say, "problematic". This is the student that takes up 80% of the time you spend interacting with students that semester, and in a negative way.

I've probably seen a few more of these students than the average, because I've allowed my course to be offered through the Harvard extension school. There have definitely been many cases there of students who just enter the class insufficiently prepared, and most of them quickly drop the class. But occasionally there's one who misunderstands and thinks it's our fault (mine and the TAs) that they're failing a class that they may not have had the necessary background for to begin with. (I've recently had to deal with such a student, which brought up this line of thinking.)

For sheer annoyance value, though, my most problematic student was a Harvard student. He or she (let's use "he" from hereon) got a warning from me partway through the semester because he failed to turn in an assignment. I told him he had done fine on the assignments he had turned in, but if he didn't turn in one or more future assignments, his grade would suffer, and he could even fail the class. He said he'd understood.

After the midterm, he did not turn in another problem set. Which would be fine, except that he then made a rather large issue out of failing the class. He insisted on knowing the exact formula I used to assign grades, going over every question on the midterm and final with me, and so on. In short, he refused to take responsibility for the outcome, which is the hallmark of a problematic student.

I'm curious if other teachers have had similar experiences, and what advice they might have in dealing with such students. (My advice -- catch these students early, and document by e-mail what they have been told regarding their performance! And try to spend more time with more positive students.)


Anonymous said...

I tell students exactly what formula I use at the beginning of semester. And if I am going to give grades from 4 point e.g., I make maximum possible 6 points, and clearly state that this is a generous marking system. I give them different ways to earn points, so if some one does not like doing homeworks, he/she can give a talk on a topic I assign him/her.
Yet there are still problematic students. I just tell them that they have not done the work required to get the points, and that is the final word.

Anonymous said...

I always announce a very precise breakdown of percentages for the various components that go into the final grade, but never announce the grade curve, in advance or after grades have been assigned. So people can use the percentages to make rational allocation of effort. But all I say about the letter grades is the truth, that I am going to give each student the grade that, in my best judgment, he or she deserves. I sometimes implicitly use the first derivative as well as the weighted average in coming to those judgments (in a course where there is a homework assignment due almost every week and they count).
The "personal judgment" (or perhaps "professional judgment") standard of course does not satisfy the student, but at Harvard anyway, it's the end of the story; there is no challenging a professor's judgment on grading, unless the professor has done something to signal that he or she is being illegally discriminatory.
Though this may solve the grading problem, it doesn't solve the educational problem What the student really needs is not a cogent argument about his grade, but someone to sit down with him and tell him to grow up, that however well it may have worked in high school, it isn't going to work in real life to challenge every adverse decision on technical grounds. And college is the time when students should learn to shift their coping mechanisms towards the ones more useful in adult company, where mutual respect and good faith will generally take you farther than constant quibbling and quarreling. This fellow deserves to be told that he is developing a reputation that will not serve him well. You have tenure now, Michael; you can do it! :)

Michael Mitzenmacher said...

Harry -- for what it's worth, I follow (naturally :) ) your grading approach -- giving a breakdown of percentages, and reserving the right to use my own judgment in assigning final grades. That seems to be what, depending on your interpretation, was either bothering the student, or providing the student the opportunity to bother me even though it was clear what the grade should be. I agree with you, however, that it's the right approach to grading.

The student is, I believe, long gone now, so I can't pass on your wisdom to him -- though I believe that at the time, I, and his resident dean, and several other people, tried to give him the message about growing up. The truly problematic students, though, like this one, don't seem to get the message. I'm convinced that's because there's something psychologically deeper going on, that is what leads to their aberrant behavior. If the student doesn't get the message, my advice is to pass on the information as soon as possible to the right advisor/administrator. In such cases, I'm convinced the student likely needs more help that I can give.

Michael Mitzenmacher said...

That should have ended "than I can give", not "that I can give."

Dave said...

I want all students at a university to be required to take a "101" course. It could have a different title in each department, but the goal would be the same. They are given an unreasonably difficult homework assignment and exam. Any student who complains, fails, and any student who does not complain, passes.

Obviously this requires secrecy and so is not a good idea as stated. But I want to use some modification of this basic idea to somehow communicate to students (since many high schools are apparently not communicating this as effectively as my high school) that no matter what their major, and no matter how far along in their courses, their attitude and their behavior in responding to challenges will make a bigger difference than any other factor in determining how much their tuition investment will pay off.

Anonymous said...

Wow, amazing you use the same grading method! I wonder how that happened? ;)

We are all fighting societal forces on this that are bigger than ourselves. One reason so many graduates of good colleges think that all disagreements should be settled through litigation is that they learn it from hyperactive, competitive parents while they are in school. Colleges often reinforce the message, because they are reluctant to tell students that they are behaving badly. I don't doubt you tried with this fellow, and that it wasn't enough. But I also hope you've had students to whom it comes as a relief to be told that they are not doing themselves any favor by being so litigious. I have.

Anonymous said...

dave's good suggestion reminds me of a similar suggestion I put forward in a talk at Harvard's Morning Prayer service a couple of years ago. I gave this homily at a moment when Harvard was wrapping up an endless, aimless, chaotic revision of its graduation requirements.

Anonymous said...

But all I say about the letter grades is the truth, that I am going to give each student the grade that, in my best judgment, he or she deserves.

This doesn't seem fair. A student should be given some feedback (before the drop date) as to what their letter grade is likely to be. Otherwise, what is to stop someone from giving the top student an A and the rest C's (or some other arbitrary assignment of letter grades)?

Anonymous said...

I do give midterm grades, just before the deadline for getting out of the course. They're part of the regular Harvard protocol -- though not everyone turns them in early enough so students can use them to make a decision. They don't become part of the permanent record.

Michael Mitzenmacher said...

Anonymous #8: Nothing stops a professor from assigning grades like that -- except professionalism and courtesy to students.

I give an approximate breakdown of where the "lines" are between grades, so students should have a rough idea of where they are, but they are meant to be fluid and approximate -- that's my use of judgment. It doesn't follow that my grading is capricious, and that I'll just assign C's arbitrarily. It does follow that whether you get a B or a B+ may depend on things like whether you've been improving over time, whether you've done something exceptional (like get a great score on the final, or offer interesting insights in class), and other judgment calls that I do not specify as a formula.

However, you shouldn't mistake "not proscribed by an exact mathematical formula" as "arbitrary" when it comes to grades.

If a student asks at any time what their approximate grade is, I will give it to them. I've found that most times students who ask are doing better than they expected... they think the curve is actually harder than it is. (They tend to forget that I don't curve starting from 100% of the points-- historically, nobody gets 100% in my class...)

Anonymous said...

One time a student asked me how they were doing in the course; I told them they were headed for a D but might be able to pass the class if they did well on the final.

When they (didn't do well enough on the final and) failed the class, they complained to the department head that I had misled them.

I never made that mistake again! Now, when a student in that situation asks me how they are doing, I say they are headed toward a D and while it is theoretically possible they can turn it around, most students in their situation don't. This is one of the ways I learned to stop trying to be "nice" to students in my courses...

Unknown said...

Fun discussion. Three tidbits.

First, I used the Harry Lewis grading system (or, more correctly, the Lewis-Denenberg-Thielens grading system devised for AS/CS11) at Stanford in the 1990s in an advanced undergrad course. The biggest issue I ran into was students who complained that they'd worked hard on the assignment and deserved a better grade. A combination of pointing out that what matters is the final work product, not the time invested -- combined with being able to show them the written grading standards I'd given the TAs (so they saw the grade was not arbitrary) did the trick.

Second, many of my students were remote (joining the lecture through the excellent Stanford Instructional Television system, which allowed them to ask questions and participate, to some degree, in class discussions) and their work being funded by their company, subject to achieving a good grade. I had students appeal for higher final grades because it would save them thousands of dollars. My approach, instead, was to ask to talk to their HR person and seek to convince the company to reimburse on a sliding scale (on the grounds that a student who got a C+ or B- [usually where companies stopped reimbursing] did, in fact, learn something).

Last, it appears that Michael has not yet hit the true problem student -- the perpetual grad student. When he hits that problem, drop me a note -- while at Stanford, I was forced to develop a process that, once put into effect, seems to successfully get students out the door in about a year to 18 months while using only 15 minutes a week of my time.

Anonymous said...

craig, I'd be interested to hear your strategy for perpetual graduate students. Can you share it?

Unknown said...

Perpetual grad student solution.

First, pre-requisite is they have a thesis topic. Most departments have a way to ensure that you've got a topic within a modest number of years or you've exited with a Masters degree.

So... the student's job is to stop by your office at a fixed time every Friday with a printout of the *new* pages (written in required thesis format) they wrote that week. [Phone call with emailed PDF also works]. The requirement is that the pages be *new* (not revised) and for the current chapter in progress (no handing me stuff and saying "I'll use this in chapter 7" when they are writing chapter 2). Minimum number of new pages is 5.

Failure to deliver 5 pages means they have to explain, in great detail, why you do not have 5 new pages in your hands and what they will do next week to deliver 10 pages (5 from this week plus next week's 5). Your job, as advisor, is to remove every possible impediment in the upcoming week. I've demanded a student give me his girlfriend's phone number so I could cancel their three-day weekend together. I've called colleagues and told them that the student was not going to be helping install those new lab servers next week. I've recruited the student's co-author to go give a talk. Nicely enough, my experience suggests you only have to do something like this once to make the point.

The goal is to make absolutely clear in word and deed to the student that her or his job is to finish the dissertation and do it in feasible increments. At the end of a year, they have a thesis draft and, usually, with some editing, can be out the door about four months after that.

Two gotchas I've learned. First, every so often 5 pages is too much. I've had students doing prior literature work be chagrined that they'd boiled down a weeks' reading to only 2 or 3 pages. Those folks (and similar good work) get a pass for the week. Second, the process can reveal a student who actually isn't ready to do a dissertation (or has a dissertation topic that proves unsound). But now you know why the student is hanging around and can try to address the core problem.

Anonymous said...

Wow! So does that mean that you guarantee your student funding? Since otherwise there would surely be impediments to him handing in the required pages each week ... Sounds like you are a great adviser.

Unknown said...

Stanford used to guarantee funding (may still) and, in my experience, most students had plenty of time each week for writing. 5 pages isn't that onerous (thesis formats are invariably double spaced, so we're talking 2.5 pages of text). Yes some work often has to be done (writing a little simulation, reading some papers, sitting in a quiet room thinking through a theorem) but, week in and week out, 5 pages isn't onerous and a good mechanism for working with students who, for whatever reason, can't make progress on their own.

(As a counterpoint -- I had a student who wrote roughly a third of his dissertation on a plane flight to India).

And thanks for the kudos!

Anonymous said...

>(As a counterpoint -- I had a student who >wrote roughly a third of his dissertation >on a plane flight to India).

I am sure your student is brilliant. But given how long flights to India are, somehow I can believe this! :)