Active Learning Experiment: The Halfway Point

Tuesday was the last day of the fifth week of classes (out of ten; for reasons that passeth all understanding, we started on Wednesday, so all the week-based deadlines fall on Tuesday). Accordingly, it seems like a decent time for an update on the active learning stuff I’ve been doing in my classes.

Each class has had one exam at this point, and a handful of labs. In the regular intro class, we’re in Chapter 5 of Matter and Interactions, about to do curving motion, and in the integrated math-physics class, which only does half a term of physics, we’re just dealing with non-constant forces, though we skipped ahead to do the simple harmonic oscillator on Monday, so I can keep the two labs in sync.

All in all, I think it’s going pretty well. In both classes, the exams were pretty solid– average scores around 80%, which is higher than I usually like (though lower than the students would like…), but not by a whole lot. Both distributions were more or less normal– a cluster of scores right around the average, with a few higher, a few lower, and only a couple of really bad grades. More importantly, they mostly got the idea of how to do the problems, with most of the lost points coming from genuinely tricky conceptual questions (in one class, absolutely everyone fell for the question whose answer was “there is not enough information to answer”– I throw a few of those in every term, just to keep people honest, but they almost always decide that it’s always wrong). So that’s encouraging.

In terms of classroom dynamics, I think I’m getting a better feel for this. I still have trouble getting people to speak up in favor of wrong answers, so I’ve started offering fallacious arguments myself in an effort to get things going. That sorta-kinda works, but could be better.

I still haven’t been able to stop approving right answers, which a lot of the stuff I’m reading says is the way to go. The few times I’ve tried, I’ve gotten students practically begging me to say which answer is right, so I usually end up saying that the correct argument offered by one of the students is the right one.

As far as the discussions go, I need to shake up the groups, because some of them have started to fall into ruts. I know the students a little better at this point, at least, so I can break up the groups that just don’t speak, which should help. They’ve mostly got the hang of it, though, and with a few questions, I’ve had students very passionately arguing about what’s going on, which is cool. They’re not always passionately arguing for the right side, alas, but that’s one of the risks.

The curriculum continues to be a bit of a problem. The Matter and Interactions approach puts the standard topics in a different order, which makes it a little tricky to use clicker questions from the really excellent collection of questions I got from Colorado. The WebAssign package for M&I includes a set of clicker questions, but they really kind of suck. Far too many of them are calculational in nature, involving punching numbers into a calculator, and I continue to be astonished by how that stops a class dead. Anything involving a calculation absolutely crushes discussion, and takes forever. I can do the math out on a chalkboard faster than they can put the numbers into a calculator.

The best questions, it will surprise nobody who is familiar with this stuff to hear, are entirely conceptual. I got a great discussion out of the classic question asking them to compare the tension in a rope with one end tied to a wall supporting a known mass over a pulley to the tension in a rope between two pulleys with identical masses on either end. That really got everybody talking. They didn’t converge on an answer, alas, and didn’t believe the right answer until I did the demonstration at the front of the room, but it was a really active class.

I also got a good discussion going with a question I got from a colleague a while ago, which is: “A notebook sits on a chair, with a textbook on top of the notebook, and a water bottle on top of the textbook. Which objects exert an upward force on the water bottle?” They overwhelmingly wanted to say that all the objects lower down exert a force on the bottle. I tried to go all Socratic on the groups taking that view, asking whether I needed to include a force from the floor (yes, they said), and a force from the foundation (sure), and a force from the ground underneath the building (yep)… There didn’t appear to be any number of layers that they found disturbing enough to question the idea– evidently, it’s turtles all the way down.

(I had thought that one would be a slam dunk, since M&I introduces normal forces with a microscopic ball-and-spring picture of matter, emphasizing the need for contact to create normal forces. Didn’t help a bit.)

The text-message polling system is still working okay, though it can be kind of sluggish. This contributes to the problem of needing to approve an answer– one of the ways to get everything to be student-generated is to re-poll a question after soliciting arguments for the various answers, but it can take a couple of minutes just to tally all the votes, which makes that kind of impractical.

I’ve also worked in a couple of items from other sources– one clip from Mythbusters, and a little video of SteelyKid and me going down a slide (the question was “Who will finish faster?” which I used to illustrate a discussion of sliding friction, in spite of the fact that M&I goes to great lengths to avoid talking about objects on inclines). Those are fun, and have been fairly well received.

I’m enjoying the experiment, and it seems to be going all right. It can be a little hard to tell, though– I’m tempted to do a sort of mid-term course comment sheet, but I don’t know if I can spare the time (I’m a couple of classes behind the nominal schedule, though I can make up a little of that, and there was a little slack built in). I don’t think they hate me, though.

This is far and away the most work I’ve had to put into teaching in a long time– probably 2-3 hours of prep for every hour in class. Of course, I’ll be able to re-use a lot of the material I’m generating if I teach this way again, but it’s really a grind. Which is why there’s been so little blogging here, despite a lot of interesting physics news, and why you shouldn’t expect any big increase until mid-November when I’m done with these classes.

1. #1 Joss Ives
October 12, 2011

Chad, I’m still not great at getting the intro students to discuss their reasoning for incorrect answers. But when I have had success it has been when I ask them to tell me “what their group thought” or “why a person might choose x” as opposed to asking them why they (the individual) chose a certain answer. That shift in responsibility from “looking dumb because I had the wrong idea” to being able to discuss “what our group thought” seems to really make a big difference.

October 12, 2011

I was writing this somewhat hastily last night, expecting SteelyKid to demand bedtime stories at any moment, so I didn’t get to everything. I’ve adopted a similar approach to trying to elicit reasons, though– I usually ask “What’s a reason why somebody might choose option A?” and that occasionally works. Sometimes, it’s a student who I know went another way (from listening in on their discussion) who steps up to give the wrong answer, which is kind of interesting.

3. #3 David Speyer
October 12, 2011

Re breaking up groups: In my calculus class, I’ve been bringing a deck of cards to class, with values from ace to 8. I hand them out when students walk in. Cards with the same rank are working together that day. Extremely rapid sorting into groups of 4, and the students seem to find it mildly amusing.

I am also amazed at the slowness of routine calculations (using calculators). Has anyone found a good way to deal with this?

4. #4 ann connolly
October 13, 2011

Hi Chad–don’t print this; it is just for you from a fan of your blog and your book. As a former teacher of English, I am fascinated by your thoughts about improving teaching and learning. Why not want EVERY student to get 100% on a test? Why be sad that so many earned 80%.

5. #5 J
October 13, 2011

Thanks for these helpful updates, Chad. I have a naive question: what’s so wrong with telling the class what the right answer is, after a full conversation that has not necessarily converged to the correct place? Physics is a subject full of right answers, after all. Don’t you want to make sure they go out knowing the correct ones?

I suspect that you feel this way to some extent, which is why you’re finding yourself ratifying the correct answers. But can you say a few words about why a lot of the stuff you’re reading says that’s a bad idea?

6. #6 Kaleberg
October 15, 2011

I’m glad this approach is working well for you and your students. I was always a listen to the lecture guy. I’d hate having to come up with alternate hypotheses. That would be like doing problem sets without being able to look at the text book, an occasional nightmare I have. It’s one thing if no one knows the answer or there may be many answers as there are in engineering courses, but the whole point of bothering to attend a class was to learn stuff that someone else knew and I didn’t, not to figure it out for myself.

Maybe it was my low self esteem. I was always sure that my professors knew some things that I didn’t.

October 15, 2011

Ann, if everyone get’s 100% the class is far too easy. Additionally, having the average be, say 70% allows the test sufficient dynamic range to gauge the student’s mastery of the material, which is the point of spending time on the test in the first place.

October 15, 2011

Why not want EVERY student to get 100% on a test? Why be sad that so many earned 80%.

If every student in the class got 100% because they had really mastered 100% of the material, that’d be fine. The nature of the subject, though, is that basically nobody will truly master 100% of the material the first time they see it. When lots of students get really high scores, the most likely explanation is that I didn’t gauge the difficulty of the test properly, and could’ve asked them harder questions.

What’s so wrong with telling the class what the right answer is, after a full conversation that has not necessarily converged to the correct place? Physics is a subject full of right answers, after all. Don’t you want to make sure they go out knowing the correct ones?

The idea is that they retain the correct answer better when they have arrived at the answers themselves, rather than having me tell them what the right answers are. The point is to make them go through the process of determining the answer, not just memorize facts that I tell them.

9. #9 JM
October 15, 2011

Thank you for writing this behind-the-scenes perspective.

It sounds as if a sticky point is student participation that involves problem solving skills, as opposed to conceptual discussions, that could touch on students’ feelings of competence, or lack there of, which in turn might expose the students to feeling embarrassed, especially if speculating about wrong answers.

Do you have a student TA who could be like a confederate in the class, someone more like a peer who could present fallacious arguments or be more like a model of the type of discussion you want?

And, if you think part of the difficulty is time-constraints, such as the amount of time it takes to do calculations on the clicker, could you create an on-line blog or forum for class discussion?

Lastly, the comments about grading curves and how difficult exams are reminds me that, based on my undergraduate son’s report, the second mid-term invariably is harder than the first one, at least for math and physics courses.