I've been pretty quiet about educational matters of late, for the simple reason that I was too busy teaching to say much. The dust having settled a bit, though, I thought I would put some notes here about what I did this past term, and what worked.
I had two sections of the introductory Newtonian mechanics course in the Spring term; this was off the normal sequence for engineering majors (the engineers mostly take this in the Winter term of their first year), but this year we had yet another larger-than-expected engineering class, and needed to open another section. I picked up both of these, which meant I didn't need to coordinate with anybody else, and could experiment a bit.
One of the things I tried this term was making more use of the Direct Measurement Videos made by Peter Bohacek at Carleton. These are high-quality videos of various basic physics scenarios, with a frame-by-frame player and length scales put directly on the videos, so you can determine velocities and so forth by measuring positions and counting frames.
We did a few of these as in-class exercises, and I assigned a few others for homework. I was a little disappointed with the results from the homework problems-- a depressing number of students didn't recognize what they were supposed to do, even after I explained in detail in class-- but student comments about these were surprisingly positive. I'll do more with these videos in the future, possibly with more scaffolding in the statement of the homework questions.
I continued to experiment with standards-based-grading, using the same sort of leveled scheme I did the last couple of times I taught introductory E&M. I broke the content up into a series of subjects (Vectors, Kinematics, the Momentum Principle, etc.) and sorted the various skills associated with each into three levels (here's the full list (PDF)). I coded each assignment (homework, quizzes, exams) in terms of these standards, and then gave a grade of 0, 1, or 2 to each standard. At the end of the term, I averaged these, and weighted the scores so a 2/2 on all Level I standards would give a student a C, 2/2 on all Level II would raise it to a B, and at least some Level III standards would be needed to lift a student into the A range.
The main advantages of this are that it gives students somewhat better feedback on what areas they need to work on, and doesn't penalize students for bad scores early in the term, provided they figure out what's going on later. The main disadvantages are that there's a lot more tedious clerical work on the faculty side to make it all work, and it's non-standard enough that students are sometimes confused about their grades. To get around the latter problem, I handed out "if the course ended today" grade breakdowns after each of the midterm exams; that seems to have soothed nerves enough that I didn't get many complaints about the grading on the course comments.
As with the last time I taught this, a few years ago, I did whiteboard discussions in class. The room is set up so students sit in pairs (though there were a few groups of three), and each group has a whiteboard and markers. I use modified versions of the "clicker questions" found at the Colorado PER page-- tricky conceptual questions with multiple-choice answers, and generally no numbers. They work on these together with their partner, then we talk about the answers, "polling" the class by having them hold up the whiteboards.
This works pretty well by the various measures I have access to-- the class performance on exams was about what I usually expect, and the gains on the standard conceptual test that we use to track things was pretty good (not spectacular, but above the "traditional lecture" range). Student response was generally fairly positive-- there were the inevitable complaints that I didn't work enough example problems in class, but a fair number of comments saying that they found this more engaging and interactive.
I was a bit concerned that there would be problems with the group dynamics; some times in the past, I've switched groups up after a couple of weeks, so the same students weren't together for a whole term. This addresses the uncanny ability of confused students to find each other when making self-selected groups on the first day, but sometimes annoys students who just get comfortable with a particular partner, and then have to switch to a new partner. This term, I didn't do that because reasons, and just made an effort to keep a closer eye on the pairs of confused students so I could provide them more assistance. That seems to have worked out, but I'm not sure it wasn't just luck.
So, you know, a reasonably successful term. The negative comments I got were mostly things attributable at least in part to being department chair-- I wasn't able to offer many office hours, and I was slow getting homework graded and returned. Some of the latter was also due to this being the first time I've used SBG in the intro mechanics course, so I had to code the problems for the standards and write up new solutions to everything. Next time I do it, that part will go more quickly.
Of course, the next time I do this will be September 2016 at the earliest, as I'm on sabbatical next year. Calloo, callay, and all that.