Friday, January 26, 2007
Workload Equity and the Popularity Penalty
Since we have increasingly limited resources, fluctuating student demand for different courses, and a unionized faculty, one of my semesterly battles involves finding ways to balance teaching loads.
Some of that is done for me by the contract. The contract specifies the ratios of lab to lecture for load purposes (don't ask – it's incredibly complicated, based mostly on who was on the negotiating team that year), so I don't have to reinvent that wheel. The contract also has a provision by which 'independent study' courses are compensated at a fraction of a credit per student. Over the years, this provision has come to be used to run some very small classes.
Running small classes is a financial problem for the college, but it can't always be avoided. With a few exceptions, we generally require any given section to include 12 students in order to run. (Exceptions include nursing clinicals and a few courses with very specific lab needs.) But sometimes you get stuck. For example, some of our majors require two semesters of a language. If we ran, say, Russian 1 in the Fall with 15 students, we're pretty much honor-bound to run Russian 2 in the Spring, even if only 8 or 10 come back for it. In some of the smaller majors, the upper-level (that is, sophomore) level classes are graduation requirements, so we run at least one section of each, whatever the size.
Equity issues abound. The composition and speech courses, for example, carry much lower caps than do the psychology or history courses. We justify the discrepancy by pointing to the different amount (and nature) of grading; since composition courses usually require more student writing than history or psych courses, the argument goes, it's as much work to teach composition to 22 students as it is to teach psych to 35. (This is why 'writing across the curriculum' died on the vine. Importing composition-level grading expectations to psych and history would require drastic reductions in psych and history class sizes, which would be financially prohibitive for the college.) But there, too, the college has a long history, and most of the faculty have accepted that difference simply as a fact of life.
The judgment calls come with the tiny classes. What do you do when, say, four students sign up for a class, and it's the only section?
The per-student rate for running it as an independent study is low enough that most faculty won't bother for fewer than six or seven students, since they still have the same amount of time in class and prep work. Yes, the grading load is lighter, but depending on the class, that may or may not compensate adequately for the reduced pay. If the class was supposed to be part of a professor's regular load (as opposed to an overload), the professor has to pick up another class to compensate, which makes for some very cranky faculty.
There's also the “I saw with my own eyes” problem. Last semester one sharp-eyed professor cornered me, asking why her classes were stuffed at 35 students a pop when just down the hall she saw a colleague teaching (a different course to) a group of five. I assured her that the colleague was only getting the independent study rate, and that we only ran the class because the students needed it to graduate, but I could see the explanation didn't really satisfy her. In a way, I couldn't blame her. It's harder to learn 35 names than 5. Multiply that out by five sections per term, and the differences become astronomical. (Of course, nobody gets 5 sections of 5 students, but emotional impressions are strong.)
The other side of the equity dilemma is the popularity penalty. Professors who become popular with the students find their sections perennially full, which increases their grading load. Professors whom students avoid like a bad smell wind up with fewer students, and therefore lighter grading loads, for the same salary.
Our student turnover is high enough, and ship is tight enough, that we don't have too much of the popularity penalty problem. I saw it at Proprietary U, though, since the student grapevine there was more fully developed. There was one professor in particular who achieved a sort of rock-star status among the students, which meant that his classes were always stuffed to overflowing. He complained regularly that he was effectively penalized for teaching well, which was pretty much true, since the increased grading load was uncompensated. Out of a desire to even out the workload a bit, I tried scheduling him one semester in the death valley time slot, where sections go to die. It didn't help. Students still poured in, only with more of an attitude. That experiment lasted exactly one semester.
From an administrative point of view, the popularity penalty is hard to address. I like to hear that students like their classes, and the soulless bean-counter in me can do the math well enough to realize that the books balance better when the classes are full. That said, there's something perverse about the better teachers having to do more work for free, and the dregs getting a free ride. (At Flagship State, where I got my doctorate, I knew people who openly admitted using the first week of class to scare students away so they wouldn't have to grade as much. That always bothered me.) Professors who earned reputations as ogres, for whatever reason, usually enjoyed much smaller sections, and therefore lighter grading loads. Other than tight control of logistics, such as we have here, I'm not really sure how to work around this.
Has anyone out there found a fair and intelligent way to offset or prevent the popularity penalty?
It may be perverse, but it's the inevitable byproduct of a union contract coupled with tenure. Free market-based outcomes will be optimal, all others will have perverse elements.
I think you've answered your post of last week on documenting good teaching.
Also, did the NY Times take a cue from you yesterday in writing about affirmative action today?
This cap also keeps good instructors from getting ground to dust with endless large classes.
Other ways to address the issue:
If an instructor has this problem consistently, this should affect the courses that the person gets assigned and/or things like merit pay. Everybody has a class once in a while where this happens, but if it's happening constantly it's a problem.
Another issue is when courses are being offered (if you know that intro to bio never has more than 10 students at 7 AM, you've got to stop offering it at 7AM, no matter what the professor's preference) and whether courses should be required if over time the enrollments remain low.
Now, I hate administrators getting involved with this kind of stuff, which I really think should be left to individual departments to manage. But I'll say this: I'm a "popular" teacher, and while I'm fine with teaching the maximum in theory (it's what I plan for when I design the syllabus), I am pissed off in practice I feel like I'm being punished for being good at what I do. No, it's not my job to worry about what's happening with Prof. X, and I try not to do so, but after a while, it begins to feel like one is being used, and that makes it really hard to be an effective teacher.
I disagree. This is yet another way to measure professor popularity. I would be surprised if the correlation between enrollment and teaching effectiveness was stronger than the correlation other measures (e.g., student evaluations) and teaching effectiveness.
Where I work, we've tried to solve the problem of small but necessary classes like Russian II or Calculus III by offering "overlay" classes. Teachers have Russian I and Russian II students in the same classroom at the same time. It's more work and far from ideal, but teachers are willing to do it because they seldom get to teach the advanced sections--and because it's necessary.
Philip
Drone @ Funhouse U
Interestingly, many of the most popular classes in those days were taught by department chairs.
But from the point of view of students, do we realy want the less popular teachers teaching more students? If less popular teachers are also less skilled, are they the ones we want to be teaching the core courses?
It seems to me that would amount to saving the good teachers for students in the major, and letting the rest of the students, the ones who are majoring in something else but who are taking some required courses, get the bad teachers. In this case, most students, who only take one or two classes in the department, would get the bad teachers, and the students who are majoring in the department's field would repeatedly get the good teachers. That's one way to make the department look bad to everyone outside it.
Hire graders for sections which exceed a certain size -- or for instructors whose combined section size exceeds a certain number. You don't even have to pay much, because they can be Federal Work-Study jobs. This is a massive win-win; students get on-campus work which contributes to campus life, and professors get to hang out with smart and interesting students as a perk of being good.
In addition, I don't know if you have student-nominated Teacher of the Year awards (or some such), but if they came with a few thou as prizes, that'd calm a lot of people down.
Heck, assign your parking spaces by who's teaching the most students in their department.
It's true that markets often come with incentives built in, but a modicum of effort can replicate them.
After a couple of semesters I decided to push the boundaries of my popularity. I offered popular courses at unpopular times (but times I liked..), I made my courses more difficult and added unpopular things like presentations.
My success has been mixed -- sometimes my intro class at 4:00 has 30, this semester it is full... sometimes my early morning killer-subject is full and other times it has 30 students. Either way, I'm teaching when I want to teach -- and if they really want me rather than my colleague, they'll figure out how to get into my class.
I've had colleagues stress my ability to say no to these requests, but let me tell you, just having a disproportionate burden of having to say no much more often can be taxing. Even fielding the requests and having to go through a slightly more in-depth process of deciding who to accept and who to say no to takes more time. And I worry about how my saying no so often will come across - to both students and my senior department members.
What you're saying is that popularity (as measured by enrollment?) correlates with student judgment of teaching effectiveness. If student judgment is a good method, then all we need to do is turn to student evaluations for documentation. Part of the "new correspondent"'s concern in the post you referred to is over the validity of student evaluations.
Given that the lack of a truly good measure, I'd make use of both teaching evaluations and enrollment, but I don't know how much enrollment will really help this particular correspondent. For one, he is a TA, which has two consequences: (1) He's not the only instructor involved in student's decision making (2) It's likely that the students won't know he is teaching the section when they enroll (in my experience, TA assignments are generally made, or at least finalized, *after* regular enrollment closes, partly because number students taking a course can affect the number of TAs involved). Further, he mentioned that it was a 2-year appointment, which may not have left enough time for student lore to have developed.
Just my 2c
As the TA who sent the plea for help, your comments are right on in this case. As a TA at my large public university, we are often not assigned sections until the week before a course starts. And two years is not enough time to build an enrollment legacy. Further, as TA I have no control of what course I am assigned to on a quarter-by-quarter basis. It may be a nice sequential series of courses, or it may be a random set of which no student would take all courses.
So while in priciple Kimmitt, enrollment may speak louder than evaluations, for TAs (and new faculty), there has not been enough time for enrollment to be directly related to teaching. Also, enrollment numbers are closely tied to the course itself (is a required intro stats course led by a more popular/better teacher than the advanced Russian course because the enrollment is higher?).
As several posters noted here, being popular has to do with more than just being a good teacher. What if the teacher is super-nice? Or a super-easy grader? Or super-hot to look at (it happens...)? Clearly, all these overlap to get popularity. My original post was about how evaluations can tease these apart, or at least tease out the "good teacher" part. Or rather, do can they tease it apart?
All of that aside, if pay can't be linked to teaching merit, I like Kimmitt's suggestion for non-traditional rewards like the good parking space, or having a laptop guarenteed for my use in class rather than having to fight the other teachers....
<< Home

