Tuesday, December 01, 2015


Cohorts and Critical Mass

This one is best expressed as a word problem.  And no, it doesn’t involve trains heading in opposite directions.

You can’t afford consistently to run classes below fifteen students.  You build a tightly-constructed two-year program on a “cohort” model: in other words, on the assumption that students will move through as a bloc.  How many students do you need at the outset to make the program viable?

  1. it’s only money
  2. 15
  3. 25
  4. 30+

The correct answer depends on institutional context.  In an amply-funded institution, or with a really killer grant behind it, “a” could be right.  In a residential college full of high achievers who attend full-time and have an on-time graduation rate over 90, b might be correct.  In a community college, though, even c would be optimistic.  I’d go with d.


Voluntary, informal cohorts fly apart quickly.  

Let’s say you get lucky and 30 students sign up for the first attempt.  Congratulations!  But eight of them need developmental classes, and another five bring some odd mix of transfer credits.  Now your first-semester classes range from 17 to 30.  

Assume that three students fail or drop out from the primary cohort.  (That’s pretty optimistic.)  Assume that another two drop out from the developmental classes.  At the second semester, you’ve got the “catchup” first semester classes running at six each, with the second-semester classes running at 14 to maybe 27, if you’re lucky.  (With transfer credits, the max is probably closer to 25.)  

Repeat, then repeat again.  By the end, you’ve got class sizes you can’t support, but you’re locked in because students need them to graduate.  Between the classes you planned to run and the catchup classes for the ones who either started or fell behind, you’re swimming in single digits.  And even if you closed the door behind the first cohort, you have an ethical obligation to the first cohort to give it a reasonable shot to finish.

Yes, there are exceptions.  Some employer-based programs hold together much better than that, because there are immediate economic rewards and penalties.  Selective programs also do better, because they’ve screened out the still-developmental and other high-risk groups at the outset.  Programs with captive populations -- high schools or prisons -- prevent attrition by preventing leaving.  But for programs built to take all comers and to last two years or more, this is a broadly accurate picture.

In the absence of either a grant or some other third-party payer (i.e. an employer), it’s easy for a program like this to start circling the drain.  And that’s assuming a relatively healthy sized starting group.  Start with 20 instead of 30, and things get uglier faster.

The folks who want community colleges to respond immediately when the economic winds shift often don’t understand this.  Absent some sort of funding cushion, the nightmare scenario is that you start with 15 to 20.  Turning away that many looks heartless and/or stupid; starting with that many virtually guarantees sustained losses.  

If we want community colleges to be more entrepreneurial and to take more risks like these, they need the funding to do it.  In the private sector, that would be so obvious as to seem tautological: investment requires capital.  But in the public sector, it’s considered suspect.  It shouldn’t be.  Here, too, investment requires capital.  We just don’t call it that.

If we could afford answer A, we could do nearly anything.  Even answer B would allow far more responsiveness to local conditions than we can manage now.  As long as we’re stuck with “maybe C but probably D,” well, some needs will go unmet.  I’d like to change the direction of that train.

Non sequitur is not quite the right term, but "tightly constructed program" and allowing students who need developmental classes to "sign up" just doesn't make sense. It has to be one or the other. It's OK to have students "express interest" in that program, to build up a mailing list or get them ready for the first class that starts next year, but you don't have a cohort if many of them are not prepared to start the program because they never should have graduated from HS.

You certainly don't run a nursing cohort that way!

To be clear, I'm not saying you have to be really selective like you are in a nursing program, but you do have to have college-level students if it is a college-level program. Perhaps what you need to do is have just-in-time remediation like they do in machinist programs: teach the fraction arithmetic and digital programming while they are learning other things. Think that through, with the incoming student body in mind, and then your program will be "tightly constructed".

That said, the workforce-oriented programs that worked well at my college in response to a community need were ones that had outside funding so they could run with a small cohort. That was easier to do back when there were lots of jobs programs to address the Great Recession, however. Not so easy now.
Yeah, you may need to move from "informal cohort" to "formal cohort you specifically apply to be in" to make this kind of thing work. The only cohort-based program I was ever in was a graduate teaching program designed to produce secondary teachers, and was very much a thing where you applied to be in the cohort for a given 15-month cycle but needed to meet certain criteria to get in. Students not in the cohort could also sign up for the classes, but the cohort students were the only ones who were particularly guaranteed to get through the program and get their license on schedule. (People who were already teachers would sometimes take individual classes from our program for professional development, and some of the classes were cross-listed with an Ed. D. program, so they weren't all cohort-exclusive.)

If you state "to be considered a part of Program, you need to apply and have met the follow checklist of [developmental classes] and to remain in the cohort you must pass [classes required for program] in the following quarters (since that's the only time they'll be offered in this two year cycle)" upfront, and then try to have options for student who leak out of your cohort to get some other kind of degree or lesser related credential, that's probably the way around this. Of course, that doesn't get those "leaky" students the credential they want in a timely fashion, but if the alternative is not offering the program at all, it's not like they'd otherwise be served better.
That is a really good idea, Anonymous.

Sort of a "reverse stackable credential", where you earn a lower (but still useful) credential after failing to meet the standards of a more exclusive one.
You need to get adjuncts to teach all these classes and lower their pay per student. For teaching a class of 10, for example, the adjunct should be paid between $500-$1000 for the semester, or less if you are in a low-cost rural area. Get rid of as much full-time faculty as you can. Raise the tuition for students in low classes, who are getting more attention.

To paraphrase someone in the 1992 Clinton campaign, it's the money, stupid.
Post a Comment

<< Home

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