However, students relatively infrequently exchanged reasoning with fully articulated warrants Exchange of Quality Reasoning level 3 claims logically connected to evidence (Table 4).
This measure emphasizes the value both of exchanging reasoning and of providing reasoning that logically connects evidence with a claim (warrants).
In addition, some "discussions" mainly involved group members listening to the dominant person in the group rather than exchanging reasoning, especially when correct answers were rewarded with more points (also previously discussed in James, 2006).
Discussions in which a high proportion of the discussion was spent exchanging reasoning or a high proportion was spent exchanging claims were no more likely to produce a high correct revote than were discussions in which a low proportion of time was spent on reasoning or claims.
We find it notable that exchanging reasoning does not guarantee that students will arrive at a correct answer: the fraction of each discussion spent explaining reasoning did not correlate with the percent of students at a given table who ultimately answered the question correctly (Table 5).
In contrast to introductory astronomy students (James and Willoughby, 2011), recorded volunteers in this upper-division course engaged in the type of discussion the instructor intended for almost all of the transcripts analyzed: that is, they exchanged reasoning related to the clicker question asked.
Discussions that involved an exchange of reasoning (levels 2 and 3) devoted a significantly higher percent of the discussion to reasoning and also had significantly more turns of talk, compared with level 0/1 discussions (Table 5).
We found evidence of coconstruction in more than three-quarters of these discussions, which exhibited either an exchange of reasoning (level 2) or an exchange of warrants (level 3).
In addition, peer coaches would undoubtedly benefit from more hands-on training, with particular attention paid to practicing how to scaffold learning with their students (Kulatunga and Lewis, 2013) and using role-playing strategies to practice encouraging exchange of reasoning.
We found that the majority of student discussions included exchanges of reasoning that used evidence and that many such exchanges resulted in students achieving the correct answer.
