Thursday, October 24, 2013


What if We Never Cancelled Classes?

I have to admit liking this idea a lot, even though I’m having trouble imagining how it would work.  Henry Ford Community College, in Michigan, has announced that starting next year, it will refrain from cancelling classes for low enrollment.  (It will retain the right to cancel classes for lack of faculty to teach them, which seems fair.)  The idea is to ensure students that once they’re in, they’re in.  

Apparently, the mechanism will involve posting a minimal schedule first, and then adding sections as needed.  That way, students won’t have classes cancelled out from under them.  They may not get optimal schedules, but the ones they get, they can keep.  (The article refers to adjuncts being “bumped off” from sections.  Maybe it’s a regional thing, but to me, “bumped off” has a more sinister meaning.)

I understand the appeal to students (and to faculty) of knowing that once a class is on the books, it’s running.  For students who have complicated lives, shifting days and times at the last minute can be a real hardship.  For faculty, the prep time that goes into a class that never runs is essentially wasted.  I get that.  I also see the academic appeal of the occasional unusually small class.  

And it would solve a persistent, nagging problem.  Cancelling underenrolled classes is a painful and imperfect process.  On the one hand, you want to do it as early as possible, in order to give students and faculty time to make whatever adjustments they need to make.  On the other, you want to do it as late as you can, to preserve the possibility for ninth-inning rallies in enrollment.  In either case, you’re basing decisions on best-guess predictions, which are necessarily imperfect.  I’ve been told by the occasional angry professor that his class “would have made it” if it had been given a chance.  There’s literally no way to know the truth of that one way or the other, which makes the discussion frustrating for both sides.

To make matters worse, we’ve found historically that when you cancel a class with, say, ten students in it, five of them don’t re-enroll in something else.  Students have scheduling constraints of their own; some will roll with the punch, but some won’t.  And a student who might cut a college some slack the first time it happens may walk away in disgust the third time it happens.

That said, though, the painful fiscal truth is that the college could not run a bevy of small sections without going under.  Enrollment pays bills.  Start dividing the same salaries by many fewer tuitions, and things get ugly fast.  So if the possibility of just throwing everything out there and running it is off the table, and you don’t want to go with late cancellations, you have to start with pretty minimal offerings and work your way up.  I presume that’s why HFCC isn’t going with a “run everything” model, but is instead going with “we’ll add it as we need it.”

The “add it as we need it” model seems to lend itself to other painful issues.  As things stand now, late staffing for a few sections causes issues with book ordering; if late staffing became more normal, I could see the bookstore issues mushrooming quickly.  The faculty problem of the “wasted prep” would be replaced by the faculty problem of the “instant prep”; it’s not obvious to me that that’s better educationally.  Adjunct recruitment would get much harder, since I assume that full-timers would have to be staffed first.  They would probably have to go with “tiers” of likelihood of backup classes: the most likely, the next most likely, and so on.  But getting good adjuncts is tough already, without trying to get someone to commit to a “fairly likely” section to run.  I foresee some defections from the ranks, as adjuncts rationally decide to take more concrete offers from other places rather than vaporous ones from HFCC.  Even room scheduling would have to be rethought.

Even as I struggle to comprehend the operational issues, though, I have to admit liking the concept and the spirit behind it.  I’ll be following this one with interest.  Henry Ford revolutionized the production process; maybe Henry Ford Community College will revolutionize the scheduling process.  I’m rooting for them.

Our campus began doing this for discussion sections of large lecture courses: they would schedule, but hide, a bunch of sections, and only as the first sections filled would the additional sections be revealed. It kept enrollment in sections balanced, and meant that we had as many as we needed but not too many extra. Since our campus is severely space challenged, this has been a boon.
My college has been doing that for ages, with varying degrees of success. It works great if (1) you have good institutional data on the historical enrollments and retention rates in sequences and (2) overall college enrollments remain predictable. We ran into trouble when we had big increases or decreases in enrollment.

I was astounded when I saw some of the data behind our scheduling. The statistics for things like the fraction of students who transfer in the middle of a sequence is really stable over many years. The number who enroll in the fall is much less predictable.

Where it really worked well for us was when we promised that we will teach what might look like niche classes during specific semesters or times (e.g. at night) and not cancel if the enrollment is low. An example from the math side might be differential equations, a class many used to take after transfer. It turned out that once it was predictable, students were advised to make long-term plans that included it and enrollment grew.
One possible long-term benefit is that if there are too many bookstore problems, programs might begin shifting to more e-books. Tablets are now cheap enough that they aren't really any worse then textbooks, especially if they are used to replace multiple courses. Newer e-reader software even supports attaching notes, so its possible to annotate the material.
Once upon a time, railroads had the best older passenger cars in reserve to strengthen formations at times of peak ridership, such as Thanksgiving or spring break or a sporting event. Susan's approach with the quiz sections is in that spirit.

For individual courses, there's a federal mandate that gets in the way. Our bookstore gets after professors to get the text orders in very early (such as the middle of October for classes beginning in January). Their impatience isn't with Fed Ex or the Postal Service, it's because of some requirement that the book requirements be made available to students well in advance.

So the requirements are met, the book lists are on the web site, the students can decide whether the text required for Differential Equations is too expensive or too hard, and the section doesn't make.
In California, part of the reason for lead time in textbook orders is so that they can be optically scanned so that they can be read through special readers that convert text to speech, magnify the text (for those with partial sight) or converted to braille.
Apropos ebooks, there's an article in the latest Scientific American about the difference in learning between paper and electronic books. Turns out paper has some advantages in recall, probably because of the contextual cues people use to recall information (eg. 1/2 inch into book) that are missing or not as effective (scroll bar) in an ebook.

Effect is significant for older folks, who grew up with paper, but a also noticeable in current students, who are much more used to electronic reading.
Followup on institutional data.

We are well aware of the many problems of adding a section late, especially if it is filled with last-minute enrollees. One thing that we do is keep really good data on the rate of enrollment during the various registration and orientation periods over many years. That is used to help guide decisions on when to open up "hidden" sections, so they can be available as early as possible.

I was glad to see you have data on whether students whose small class gets cancelled will enroll elsewhere. That is also a financial issue, as you note. Do you provide direct assistance to place those students in a replacement class? Collecting data about when that is problematical might tell you what day/time/class combos should not be cancelled (or even scheduled in the first place!) and which can.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?