The Triumvirate of Teaching and Work Life Balance

Most courses have a series of learning outcomes for students. Once you have done the course (and, I assume, gotten a B or some reasonably high mark), then you know how to accomplish the learning outcomes.

Some may break those learning outcomes down to smaller units per module.

For instructors, it occurs to me there are three objectives to balance (at least):

1. How much effort it takes to teach the topic
2. How much students appreciate the topic and teaching choices
3. How much students learn after being taught

Number 2 is not ever considered in pedagogy, but is the ONLY thing that matters from a management point of view. It maps directly to the things RateMyProf and course evaluations measure. Therefore from a pragmatic point of view, a prof should only care about 1 and 2.

Number 3 is what pedagogy is all about: how much are students learning? This is what is referred to when we look at things governments fund us to do. They want more “skilled workers”. Students in the short-term don’t really care much about this stuff. And course evaluations don’t really test for this. Teachers who are really good at 3 often end up getting punished for 2 (learning things is hard and not fun!)

Number 1 is often ignored, too, but can be the difference between having a good term and a shitty term. For any given topic, there are many ways to think of teaching that topic. Take data flow diagrams:

- we could lecture off a set of slides showing DFDs
- We could use a textbook reference and just skim the topic
- We could create a detailed case study and show DFDs by construction
- And for each of these, we could come up with different teaching strategies: whiteboard, live coding/drawing, interviewing experts etc.

As someone with limited time, one goal has to be minimize #1. My contention is that it is easy to go for perfection in 3 and absolutely devastate yourself in #1.

What I try to do is:

• Destroy with fire any plan that maximizes 1, but minimizes 2 and 3. If students aren’t learning, aren’t enjoying it, and it is a lot of work, you should NEVER do it. And I bet you would be surprised how often this case happens. For example, in my third year software design course I spent a ton of time (increasing 1) ingesting the Play framework, learning how it worked, so that students could use it in the project. But it was a huge pain for the students to work with (hurting 2), they wanted to use React instead (hurting 2), and most of the learning was about the Play framework itself, rather than software design concepts (hurting 3). I won’t be doing that again.
• Ruthlessly assess how important #3 is for your career. I haven’t seen anyone whose teaching packet evaluates 3. And yet it seems like we all talk about how it is the only goal. I really hope researchers improve on this. For tenure, for example, I seriously doubt anyone is looking at this. The closest we come is peer evaluation, but as Mark Guzdial wrote, this is often the blind leading the blind.
• Teaching awards seem to be about 2: winners tell jokes, they dress nice, they are male, they seem knowledgeable.
• Given what we think we know about how software is built, I am going to guess that some teachers are really effective (either for 2 or 3) for minimal effort, and others put in orders of magnitude more time on 1, but have little extra to show for 2 or 3. I believe there is a non-linear, diminishing returns model for teaching effort; you might do 10 hours of prep for a lecture and not have much more to show for (3) then if you had done 1-2 hours.
• There is a sunk cost/amortization problem here, too. If you teach a course the first time, you may have a lot of 1 to pay off. Subsequent offerings might greatly reduce 1 and allow you to focus on 3 (or 2) to a greater extent. But I’m not sure how much this is true. Things move fast in software courses, especially in 2nd year and above, some costs simply don’t amortize (marking), and we often try to improve the course year to year. Plus, we might not get to teach that course more than a few times.

I want to be clear that I am not endorsing a focus on teaching for “show” instead of long-term learning. 3 is clearly the goal. But the reward structure does not reflect this. We should figure out how to ensure that teaching is measured against outcomes on 3, and not on 2 (2 is also horribly biased!).

More importantly, there seems to be embarrassingly little data on how to minimize 1 and maximize 3. I think this is a problem. We have a lot of info on (for CS1 at least) how to best teach linked lists, such as using Parsons problems. But frankly, my job involves teaching 40% of my time. I may not be able to dedicate the time required to prepare Parsons problems for the course. So a “cost/benefit” (1 vs 3) analysis would be very useful to help me maximize the teaching effectiveness for unit of teaching effort.