Tuesday, September 22, 2009


Building for Non-Compliance

On my campus, we're discussing a change that's generating some feedback along the lines of "well, if everyone just did what they were supposed to do, that wouldn't be necessary."

Back in my younger days, arguments like that sometimes persuaded me. They appeal to my visceral distaste for rules that say one thing and mean another. Speed limits are an example. Most highways have posted limits that are more like opening bids. Most of the time, you can safely go about 10 mph over the limit; the unofficial limit is somewhat higher than the official one, though you're entirely sure by exactly how much. In my perfect world, I'd much prefer that the signs say what they mean and mean what they say. Instead of posting 55 when we mean 65, post 65 and enforce it. Take the guesswork out of it, and make one rulebook for everybody.

That would be great. But it's just not the reality of how people live and work. And systems that rely on perfect behavior are doomed to fail.

(For a close-to-home example, of course, take the 21 drinking age. The idea was that an age of 18 allowed high schoolers to buy beer, so raising it to 21 would get it out of the high schools. It was never really about colleges, although it applies to them. If you ever want to see prominent people fidget, ask a collection of residential college presidents what they do about underaged drinking on their campuses.)

Rules or expedients made for the masses generally undershoot what really conscientious people do on their own. Any experienced driver can tell you that speeds that may be perfectly reasonable on dry, sunny days may be irresponsibly excessive on dark and stormy nights. (It was a dark and stormy night...) So in practice, many limits seem to be set for dark and stormy nights, with an informal understanding that they'll stretch a bit on nice days. It's annoying when you guess wrong, but we haven't come up with a better system.

One alternative would be to do away with speed limits altogether. If there are no rules, then you don't have to guess what the rules are. But most of us suspect that some people would overestimate their own capacities quite badly, with deadly consequences for others. So we put imperfect rules in place on the theory that their admitted imperfection is less bad than relying on the good graces and judgment of everybody on the road.

That's basically what's happening here. We're discussing ways of communicating with students in the event of class cancellations for instructor absence. (The H1N1 scare was the prompt, but it would apply to absences for any reason.) A few folks have opined that if all affected professors just contacted their own students, no college-wide policy or practice would be necessary.

Well, yeah, but they won't. They haven't yet, and I don't see that changing. So we need a backup plan.

That's not to deny for a moment that some professors go above and beyond. In fact, I'd guess that the majority do all that could reasonably be expected. But there's a non-trivial number who do only what they're compelled to. When the costs of that attitude fall on the students, and they aren't trivial, it becomes fair to ask what else should be done.

I'm increasingly convinced that 'ideal' ideas are a dime a dozen. The ideas with real value are the ones that can survive heterogeneous behavior and compliance over time. Those ideas almost always fall short of the best behavior, but they have the unique virtue of being useful. The best individual performers may find the rules a bit underwhelming, and I salute them for that. They're right. But asking everyone to be perfect (or civic-minded, or virtuous, or altruistic, or...) just doesn't work.

Seeing beauty in the sustainable, mediocre idea is the administrative aesthetic. There's something a little bit sad in that, but there it is.

It's not sad at all, DD. Most successful systems involving semi-independent actors are sustainable but less perfect than they could be. (As an evolution geek, I am perfectly comfortable with that.)

There are people who ignore path dependence, human frailty, and the fact that not everyone cares as much about an issue as they do. They're called "ideologues". Fortunately, most of them never manage to get into a position of real power, but it's scary when they do.
"A few folks have opined that if all affected professors just contacted their own students, ..."


Thanks for that bit of CC admin humor. The only way it would have been funnier is if you told us who made that observation.

How? By e-mail? ROTFLMAO. By calling all 100 (or more) of them using the phone number they gave the college? Riiiight.

All with the result that the Dean wonders why only nine kids showed up for a large class after she paid an adjunct to cover the class?

PS - The only time faculty contact all of their students to cancel class is when they want to cut out early without taking any leave time, and they often get caught because some student didn't come to class or read e-mail for a few weeks, then decides to show up.
Ditto to what Dictyranger said, with the addition that what you're describing is quite familiar to those designing non-people systems as well; power system engineers, computer scientists interested in networks, etc.

A system that needs to be reliable has to be robust against deviations from `correct' behaviour. The only real difference is that there's a lot less dissapointment (or ideological dissonance) when the component flaking out is an aging switch somewhere, rather than a presumptively responsible adult in a position of responsibility.
What the hell? Letting someone know that you're not going to show up for work is "going above and beyond?" No, that's a condition of employment for most people. Why shouldn't it be for professors?

It's not hard. Professor notifies department, department notifies students. That way someone can physically post a sign up for whoever didn't get the email.
When I was going to night school back in the dark ages, I was not happy to drive 38 miles to my commuter school to find my 3 hour class canceled because the professor (and he was a tenured professor, not an instructor or an adjunct) had been in a bar fight.

My current employer gives us the option of providing a cell phone number to receive text messages regarding weather closings and crime alerts. If I were also a student, I could theoretically receive notices of individual class cancellations this way.

-I'm accounting as fast as I can
We have recently developed a system for communicating with everyone, through automated emails and phone calls. The system snags email addresses and hone numbers from the official records and places calls in case of (campus-wide) emergencies. It's not intended for, or adaptable to, individual class cancellations...but I would hope that most of us would recognize some responsibility in those cases.
If I read DD correctly, no one is suggesting that professors don't contact *anyone* if they're cancelling class. The suggestion is to have instructors contact *students*. Big difference.

CCPhysicist is right - this is near impossible (and, I would guess, for unionized instructors could also represent an un-negotiated addition to workloads). A mass email to all students, posting something on the school's website ASAP, and a written notice on the classroom door should be sufficient.
You could do worse than quoting Madison: "If men were angels, no government would be necessary."

To the specific issue: if the prof is in the hospital with H1N1 perhaps expecting them to call students is not the best idea. A decent rule needs to take into account those sorts of likely quirks.
I'm a prof at an SLAC,and my son has swine flu, so I have to stay home to care for him (he is 2) all week. All I did was use the "email all" tool on Blackboard to notify my two large classes of students and had my administrator put a note on my door also. I guess I just don't see what the problem is. it ain't hard. It maybe took 2-3 minutes of my time.

I am probably an "idealogue," (:)) but if I were a Dean, I might risk being an unpopular one by at least occasionally mentioning the types of easy responsibilities that faculty have. Am I just being naive?
As a few commenters have already said, mass-emailing the class, posting the same message to the course website, and calling someone at the school to post a sign is not at all difficult, even when you're quite ill.

But getting all students to actually provide a current email address and then check it before class? Crazy talk. Relying on the admin's office to actually post the sign they're supposed to post (because we're supposed to call them for the express reason of them posting a sign)? Not gonna happen. Expecting the students who didn't check their email to read the signs the when the admin's office can be bothered to post them? Ain't happenin.'

I'm sure there are profs who don't notify their students, but communication is a two-way street.
What Tree of Knowledge said. I'd also add that the need for an institution-wide plan-of-action - as opposed to the traditional means of notifying students of an absence, such as sending an email to the entire class plus contacting the office to make sure a sign gets put on the door - becomes paramount if there is the likelihood of mass cancellations across a department/university.

I don't think the issue is one individual instructor coming down with an illness so much as what happens when many instructors all are ill on the same day? And what if the department staff who would be responsible for posting notices on doors are also out sick? How do we notify students of the absences, and how do we institutionalize a protocol for that possibility? Further, what if a class needs to be canceled not because the instructor is ill but because so many students in the course are ill that one can't conduct business as usual? The university isn't closed, but if in a class, if a majority of students have H1N1, is it really reasonable to hold class as scheduled and to charge those students with absences, to follow the course policies as written for when there isn't a pandemic afoot?

The "email your students" option also assumes that all students *have an internet connection at home* - let alone that they check campus email regularly from it - and at my institution, this is just not the case.

In other words, obviously a worker should be responsible to notify someone if he/she will be out sick. BUT, and I think this is the key, how does this work in *out of the ordinary* circumstances? Leaving it up to individual instructors isn't a policy. And yes, out of the ordinary circumstances can mean broader policies are necessary.
That would be great. But it's just not the reality of how people live and work. And systems that rely on perfect behavior are doomed to fail.


I deal with this shit all the fucking time with some delusional washed-up senior faculty in my department. Their answer to everyfuckingthing that involves people not doing what is expected or desired is to make a rule. Then when they make the rule, and behavior still doesn't meet expectations or desires, their answer is nothing more than an indignant whine: "But I *told* them what they are supposed to do!!" Fucking morons.

Making a rule and telling people to obey it is pretty much the single least effective way available to get people to do something. The way to get people to do something is to create a system of reinforcers and incentives that make people *want* to do the thing you want them to do.

When the answer of a manager or administrator to a failure of action is, "But I *told* them to do it", you know you are dealing with a complete incompetent.
"When the answer of a manager or administrator to a failure of action is, 'But I *told* them to do it', you know you are dealing with a complete incompetent."

ha ha ha...The people who say "But I *told* them" are probably the same people who think teaching is really easy...you just have to *tell* them and they learn, right?
Post a Comment

<< Home

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