Optimism and Experimentation

Generalizing from my own approach to learning technology, I think it is natural to see possibilities that I might want to try, focus on the upside if the experiment works approximately according to plan so those possibilities are realized, and use that as the motivation for taking a different approach. I’ve argued before that there needs to be experimentation in teaching to keep it fresh and to energize the students and this focus on the possibilities is one way to make sure that things get tried.

However, we know that experiments fail and the frequency of failure is somewhat in proportion to the risk in the experiment. Consequently, it is possible to go through a “boom and bust” emotional cycle about the teaching and when the bust phase is upon us to get shy about further experimentation. That will not do, but it too seems natural.

So the question arises how to sustain the optimism or, alternatively, how to drive a little out of control so there is some excitement but not so out of control that there surely will be a wreck. I think there are two key skills/functions that the instructor must have to make this work.

First, the instructor must be able to trouble shoot on the experiments. So a good deal of time is spent in understanding why the experiment failed and then in trying to fix that. This is not time waste at all but rather critical time for the teaching approach. This will challenge the instructor’s time management skills, but the key point is to identify whether a minor error has caused the problem or if there is a fundamental issue that needs to be confronted.

In my own teaching this semester where I’m using the linked spreadsheet system I’ve come up with, it has worked so far quite well on my own computer for doing the Just In Time Teaching that I designed it for. But, at present, when I transfer the file that does the aggregation so the students can also see the results, the responses are cut off after 255 characters, as I reported in an earlier post. I’m not yet sure why so I don’t know how to fix the problem. I can envision a possible work around indeed I can envision several of them that might give a short term resolution for this semester. But each mean the general approach is not sufficiently robust that it might offer a preferred alternative to using a discussion board or a survey in an LMS. I’d like to see the linked spreadsheet approach work, so I’m going to spend some more time on looking for the root cause of the problem – march to victory rather than declare defeat. But in the meantime, I can post html pages of the student submissions that, because the class is small, is not too hard to generate.

Second, the instructor must be able to come up with plan B in the event the experiment is declared a failure. Plan B must be a good path for the students even if it was not the announced path at the beginning of the semester. The students must be able to endure the experiment just as the instructor endures. So they shouldn’t feel they are being jacked around in the process. This can only happen if the course has an open and flexible structure. The more scaffolding and prior setup that the instructor has put in place, the harder it is for there to be change in midstream. It may break things further down the road in the course.
So the experimentation approach to teaching is in conflict with the highly structured approach. And while I’ve cast the experiment in terms of the technology in my particular case, there can be experiments in what the students read, the type of assignments they are required to do, the nature of the in-class activity, etc. Then the question becomes whether the two can be reconciled and if not what to do about it.

As I watch others on campus in their teaching, it does seem as if there are two camps – the structuralists and the experimentalists. The former argue for clear objectives in every lesson so the student can readily assess whether those objectives have been met and so the student can have clear expectations of what to get out of that class. The latter argue for a sequence of questions where the resolution of one begats the posing of the next and where that pattern is largely unknowable ahead of time.

Call me a wishy-washy liberal on this one; I try to craft a middle ground. I have a path in mind for the course and try to articulate it to the students. But it is somewhat general with not too much detail. Experiments are either deviations from the path or variations on a theme. I try to apply a discipline that irrespective of how the experiment turns out, we’re back on the path when it is over. So there are lots of small experiments, rather than a few big ones. This makes sense to me and is a way for me to keep my sanity and engagement with the class.

Does it constitute a general approach? I don’t know.