Thursday, March 31, 2011

Shared labs

Last year I tackled a common problem in our Modern Physics lab: too many students, not enough copies of the equipment for them all to do the famous labs of the early 20th century. My solution was to have them all get a little time on the lab/equipment but not let anyone do the whole lab outright.

Here's how it worked. In the first week they break into four groups, one for each experiment (Franck-Hertz, Millikan Oil Drop, e/m measurement, angular momentum study (the TeachSpin one with the magnet in the cue ball)). In that week they were tasked with writing the theory and set-up/procedure section of the eventual write up. That's it. They were given manuals for all the equipment and, of course, they had their text books from class to help with the theory section.

In the second week they come in and are expected to only do data collection. On a different apparatus. For a different experiment. The only thing they are allowed to use is the write up from the previous group from the previous week.

In the third week they analyze data. For a different experiment. From a different lab group.

Then the Frankenstein lab reports are mashed together and turned in. It worked pretty well last year and I'm between weeks 1 and 2 this year.

At first I told them that the last group would get the full grade for that lab report. That didn't go over very well. We ended up deciding that I would grade each group and their contributions, trying to be careful not to penalize a group for some other group's poor work.

Spreading it over several weeks was done because some of this equipment is finicky and the analysis can be tricky. Typically I don't like students to have to do too much outside of our three hours per week together so the three weeks is about the right amount of time is you consider all the analysis and all the writing being done in class. What was cool, though, about the three week spread was how the groups pushed each other. If a group knew that it was going to have to take Millikan data, they would pressure the set-up group to make sure to give them all the details they would need. Last year Millikan gave us some fits and it was cool how the data collection group worked with the setup group to figure out the glitches.

I see a lot of value to the way this is set up. In three weeks they get exposed to different aspects of three different labs. They understand the value of a carefully written setup section and the value of carefully organized data. I do it because I don't have enough equipment to do it other ways but I have to say I like this solution.

4 comments:

  1. Andy-

    I appreciate you sharing this strategy. It not only addresses the supply issue, I think it helps students focus on quality when they are accountable to their peers. This allows for students to realize the importance of the details in their work. Grading the individual groups (rather than the entire report) is necessary because gives you a truer picture of where each groups' strengths and weakness are in relation to the standards.
    This post is sending my thoughts in a million directions as to the possible applications for this strategy. Thank you so much for sharing!

    ReplyDelete
  2. Andy,

    I like this idea and see how there can be some really valuable lessons to be learned, especially between weeks 1 and 2.

    One thing I am wondering is how this idea could be modified in the upper years when we start getting them to take more of their own initiative and develop their own procedures. Or how it could fit into the more realistic real-world experimental cycle of take initial data, do a quick analysis, revise procedures, take more data,...

    One thing you could try is something that I want to do in my own lab courses in the near future and that is to have "group meetings" with two or more groups, where these group meetings would be similar to the research group meetings that we would have (or continue to) experience as a grad or co-op student. This could be done in your fourth week. A research group would be you and one member from each other group and you would have two meetings (or more if group size > 2). At the meeting, the student would give a quick informal presentation consisting of a very brief bit of background and a couple of tables or figures showing the most important results. This would allow the groups that put time into those projects in weeks 1 and 2 to get a small sense of closure on those experiments and let the other group that never saw that lab get some hazy idea of what went down in that other experiment. Since you are trying to keep most of their work in lab time, the informal presentations based on just a couple of key results seems like it would work.

    ReplyDelete
  3. Ok, back from NSF panel review (wow, is that a lot of work!) so now I can start responding to some things.

    @Deana, thanks for the support, I'm glad this got you thinking. It's funny that just writing it got me to think a little more concretely about how it all works.

    @Joss I see a lot of value in both your proposals (the adding a revise part into the sequence and the group meeting idea). The problem with the first, I think, is probably time. To do it right you'd have to let yet another group take a look and do some change. Once cool thing you could do is spread a lab, like Millikan, let's say, out over the semester and have different groups do it at different times. Each one could make revisions to the previous. That could also work with the group meeting idea since some groups would be working on revising what another group did weeks ago. Hmmm. . .

    ReplyDelete
  4. You could also have it set up so that once "good" data has come in from the final procedure (so after an analysis week), one or two subsequent groups could take more data using those final procedures so that you have multiple data sets for the final analysis.

    I'm pretty excited to try out the group meeting idea this year. The idea originally came from Martin Madsen of Wabash College when I was talking to him at a poster session at the Advanced Labs conference that preceded the 2009 AAPT summer meeting. The thought is that it could make the lab feel more like an authentic research experience and it gives the lab instructor the opportunity to assess students throughout the process of the lab instead of only after the lab has been completed. I think that it could be a very nice form of formative assessment.

    ReplyDelete