mostly off-topic title: elephant talk

L&A: Experimenting on students sounds kind of odd and vaguely wrong, but as a new teacher (2nd year), I feel like I am eager to try lots of new strategies in my classroom with only descriptive evidence and some outside research to guide my decisions. The idea of teaching a whole course with only one experimental variable different between control and experimental classes sounds absurd to someone whose curriculum and strategies are in constant revision. Of course, not all experiments are like the one I describe or the sentence combining example from L&A. I know that they used two instructors each teaching a control and a variable group, but didn’t these instructors see the effects of sentence combining and feel tempted to change other, controlled aspects of the experimental section’s course? It seems like a long-term experiment could face more problems from interactions. Another thought about O’Hare’s study: What are the 84 observation units mentioned at the top of page 157? Finally, I hope others were as relieved as I when L&A brushed aside most of the threats to validity by citing randomization. What about when randomization just isn’t possible? I’m looking forward to quasi-experiments; will that tell us?

We learned from Brodkey the difficulty of interpreting and analyzing conversation; the problems that arise in interpreting talk are much like those in writing. Like my classmates, I was intrigued by the "triplet" nature of those "tacit rules" in turn-taking that govern teacher-student talk. I would argue that I often ask students to ask a question about a specific subject, so the process is a little different. In any case, teachers create systems of conversation that students generally choose or learn to follow. I also sometimes feel like my conversations with students in my speech class follow the concern raised by Freedman and Katz. When I know a topic well, the conversation is different than if I am helping a student write an informative speech about which I know relatively little (I'm happy to say though that my kids usually are surprised to find that I know a lot more about a lot more topics than they do....hehe). Page 114 brings up another good point: Sometimes the best conversations end in disagreement or tension. Strong students thrive on this stuff when they feel safe. On the same page, the "functional taxonomy" of utterances reminded me of the challenge of coding questioning types in my observation for this class. To conclude, I want to ask, did I miss something? Wouldn't we benefit from a bit about identifying the characteristics of talk that produce successful outcomes? That would be really interesting!