Differences between low and high achievers in whole-classroom dialogue participation quality

Academically productive dialogue activities in classrooms aspire to be inclusive, such that students of all capabilities participate meaningfully and equally in discussions. We empirically examine the extent to which this aspiration is achieved regarding students of different prior achievement levels. Low- ( N = 33), mid- ( N = 90), and high-achieving ( N = 22) upper elementary students participated in whole-classroom discussions around texts, facilitated by six highly motivated, trained teachers. Based on a quantitative ethnography approach, we coded 5975 separate speech turns and compared aggregated differences across the three subgroups using Epistemic Network Analysis (ENA). Whereas no differences were found in amount of speech turns, statistically significant differences were found in the quality of participation: Low-achievers' discussion participation was characterized by recitation-style talk, reduced network connectivity, and repetitive loops, compared to higher connectivity and more complex argumentative reasoning in their high-achieving classmates' network model. A follow-up qualitative analysis uncovered some of the social dynamics behind this inferior participation pattern, and highlighted the potentially stymieing behavior of peers. The findings underscore the need for more empirical research that takes into account differences in quality of student participation in dialogic activities and the reasons behind it. • Dialogic teaching strives for equitable participation in reasoned discussions. • These aspirations were tested in upper elementary whole classroom discussions. • Number of speech turns in discussions were similar for low- and high-achievers. • Epistemic network analyses revealed differences in discussion participation quality. • Low-achievers tended to participate in recitation-style dialogue spells.