I write a lot about the effective use of clickers and peer instruction, so I was excited at AAPT to see a talk with some interesting results on this educational technique. Judy Vondruska (South Dakota State University) spoke about the “Influence of previous subject experience on interactions during peer instruction.” She was using clickers like a pro, including peer discussion, but she consistently found that about 10-15% of students indicated that they didn’t find those discussions to be beneficial. She found tht a lot of those people were those without high school physics. She wondered if there were some issues that were being masked by her lumping all clicker outcome data together – were the discussions actually less beneficial for some groups, in terms of coming up with the right answer?
So, she looked across several different questions (I missed how many) with 46 students, some of whom had had high school physics and some who did not. Students were in pairs where both had had high school physics, both had not, or one had had high school physics and the other had not. She then looked at whether the pairs had the correct answer to the clicker question before discussion and after discussion. The most interesting categories (and those comprising the most students) are those who got the correct answer both before and after discussion, and those who got the incorrect answer before discussion and changed to the correct answer after discussion.
Here are the results:
- Both students have HS physics – 50% correct before discussion – 28% switch after discussion = ~80% correct by the end of the question
- One student has HS physics – 36% correct before discussion – 21% switch after discussion= 57% correct by the end
- Neither student has HS physics – 30% correct before discussion – 9% switch after discussion = 39% correct by the end.
As you can see, the students without HS physics not only don’t get as many questions correct, they are much less likely to change their answer due to discussion. These students were also less likely to indicate that they would like to work with that same partner in the future.
So, discussion may be less useful for students with less background preparation.
This was very interesting to me in light of some previous research by Jenny Knight. She found that, especially for difficult questions where few students had the right answer before discussion, student groups were able to put together the right answer even though none of them had the right answer in advance. So, I would be interested to see Judy’s results broken down by question difficulty. However, Knight et al. also found some nuanced results when looking at weak vs. strong students. In a majors course, the weak students have large learning gains when they talk to their peers. In a nonmajors course, however, the weaker students did NOT gain very much from discussing with one another. Hypothesized Knight, “One likely reason for this difference is that nonmajors were less inclined to regard their peers as learning resources.”
So, this seems to be a real phenomenon, though it’s not clear whether it’s in majors and non-majors classes. I’ve seen this happen in my own nonmajors courses; it’s really hard to get good discussion going among the students on clicker questions, unless they already know their stuff. They seem shy, and just not that interested. So, the question remains, how can we make peer discussions among weaker students more valuable, and encourage students to participate in them? I wonder if this relates to a fixed-growth mindset, where these weaker students might feel like either they know it or they don’t (fixed mindset), and so they just need to get the answer from an authority figure (the instructor), instead of recognizing that the process of reasoning through the ideas can help them improve (a growth mindset).