You were right on this one. The conclusions here seem... weak, on a number of different axes (participation pool = undergraduates, topics of discussion, conversations lasting 7 minutes only, seemingly small effect sizes).