Our semester is into its third full week, and most of my time (as you know from checking my Twitter or Facebook feed) is being spent, it seems, on making screencasts for the MATLAB class. I feel like I’ve learned a great deal from a year’s worth of reflection on the first run of the class last spring, and it’s showing in the materials I’m producing and the work the students are giving back.

The whole idea of the inverted classroom has gotten a lot of attention in between the current version of the course and the inaugural run — the time period I think of as the “MATLAB offseason” — through my blogging, conference talks, and everyday conversations at my work. One of my associate deans, off of whom I’ve bounced a number of ideas about this course, related a conversation he recently had with someone about what I’m doing.

Associate Dean: So, Talbert is using this thing called the inverted classroom.

Other person: What’s that all about?

AD: He puts the lectures all online, and instead of lecturing in class he has them do group assignments on various kinds of problems.

OP: Doesn’t that double the amount of time students have to spend on the class?

I’ve never encountered that exact reaction before, although I did mention once that the biggest negative comment from students last year in the MATLAB course was that it took too much time relative to the credit load (1 credit). I liked how my associate dean put the answer:

AD: Well… think about it this way. You are still doing both lecture and “homework”. But which part of that are going to need the most amount of help on?

OP: OK, now I get it.

Exactly. Students are going to need a lot more guidance on the difficult task of assimilating information than they will need on the relatively easy — incredibly easy, in fact — task of receiving a transmission of information. Both phases of the game need to take place in some form, but assimilation is harder, and the probability of sinking massive amounts of time into work that goes nowhere is a lot higher, than in transmission.

I’ve seen some great examples of where the inverted classroom method has actually saved students possibly hours of fruitless labor in the last two weeks.

Today, for instance, we were doing a lab problem set on command line plotting. In one of the tasks, students are asked to produce a 1×2 subplot illustrating the behavior of a two-parameter family of functions. One team was stuck because their M-file wouldn’t execute properly even though their code looked correct. The problem: They used a dash (-) in the title, which causes MATLAB to think that the stuff preceding the dash is a variable name, which wasn’t in the workspace. It’s an innocent error but not one that students with just two weeks of MATLAB under their belts could easily debug themselves. Had they run into this problem outside of class, who knows how much time would have been wasted getting nowhere? But inside class, it was solved in the amount of time it took for them to raise their hands and for me to come over and look.

Another example from today: A team had entered this code:

x = linspace(0,10); y = 100 - exp(-2*x); axis([0 15 90 105]) plot(x,y)

They had entered the code without line 3 already but didn’t like the look of the plot, so they added the **axis** command to try and change the viewing window. But nothing changed. Why? To the trained eye, it’s simple — you have to have something plotted first before you can change the axis. So just reverse lines 3 and 4. But to the untrained eye, again, who knows how much time would be lost in trying to figure this out? Instead I was able to instruct them directly on this, at the conceptual level (*How is MATLAB thinking its way through your code?*) and they got it. (It wasn’t just me telling them, “You need to switch lines 3 and 4.”)

So above and beyond being more instructionally effective, I’m realizing — and I hope students are too — that the inverted classroom makes student time a lot more efficient, and there’s a much higher success-to-effort ratio than in the traditional mode of teaching.