At my last research update I was looking at studying cohesion in teacher talk by analyzing how teachers use known-information questions in sequence. I had the idea that sequences of known-information questions allow teachers to construct cohesive talk while also involving students.
To test out this idea, I had to do a short pilot analysis. I had a hard time finding a data set — this is for a class project for this semester, so I don’t have time to collect my own data, much less get a data collection process approved by the IRB. I knew of a couple of shorter clips available online, but if I wanted to eventually build this up into a genre analysis, I would need recordings of full class sessions. So here’s what I ended up doing:
- All searches start with Google, I guess. In this case, googling didn’t lead me anywhere useful.
- I asked around at work. At a meeting, it came up that someone had been designated program area leader for K-12 research, so I talked to her. She sent me along to a couple of other people, who didn’t have any data to share but offered to put me in touch with teachers if I wanted to collect my own.
- In the course of working on my lit review, I learned of a professor who did a large-scale study in the ’90s collecting exactly the sort of data that I need. I mentioned her in passing to a professor in my department, and he said that he knows her and offered to put us in touch. So I emailed him, he emailed her, and eventually word got back to me that she wasn’t free to share her data with anyone, possibly due to the terms of her IRB.
- I emailed the subject specialist at the Georgetown library, and she got in touch with the National Library of Education for me, but while I was waiting to hear back from them, my deadline was approaching and I had to go with what I had.
- So I went ahead using one of the shorter clips that I already knew about, thinking, this is OK for a pilot analysis and we’ll see where it goes.
And then, going through the clip that I had, I realized that the teachers in this recording didn’t actually use known-information questions, or at least not in the way I had expected.
I’ll admit it: I panicked a little bit at this point. After all of that, I had found a data set I could use — not a perfect data set, but something usable — and it turned out not to have any examples of what I was looking for. Was I back to square one?
Then I took a step back and thought about my goals for this project, and I realized that I could make this work for me. Why was I interested in known-information questions? Aside from the fact that they’re so prominent in the existing research, what I really wanted to study was how teachers create cohesion. And even though I wasn’t seeing known-information questions, I was still looking at a 12-minute section of a cohesive text. I just had to find a better way to analyze it. So, looking at the questions that were present (discussion questions like “Can you paraphrase what she said?” rather than display questions like “What do you get when you multiply six by nine?”), and keeping in mind my goal of combining systemic functional and discourse analytic methods, I decided to analyze each teacher question using these two approaches:
- Goffman’s idea of reach: where else in the discourse is this utterance connected? Is it pointing back toward something that someone just said, or said some time ago? Or is it pointing forward to a newly introduced topic?
- Halliday’s idea of lexical reference: how many different things are actually being talked about here? What words are being used to talk about those things? How often is this referred to, as compared to that?
Preliminary results are interesting: questions used for topic initiation tend to be forward-reaching and lexically rich; questions used for topic continuation tend to be backward-reaching and more lexically restricted. And there’s a third class of question that doesn’t reach forward or backward but outward, referring to charts and graphs that are physically present in the classroom but not linguistically present in the discourse. (So now I’m reading O’Halloran on multimodality; thanks, Jodie.)
Of course, now that I’ve come this far, I’ve heard back from the Georgetown subject specialist that the National Library of Education actually knew of something kind of perfect; now I have to decide whether to start over familiarizing myself with a new data set, or keep going with what I have.