Talking to Learn: Investigating the Relationship Between Classroom Discussion and Persuasive Writing

Abstract

Dialogic discourse emphasizes making meaning through discussion, and stands in contrast to recitation, where the teacher initiates a question that is followed by a student response and is then evaluated by the teacher (known as IRE). Though dialogic discussion is a desired medium through which students can develop various skills fundamental to their academic success such as reading comprehension and vocabulary development, it remains a rare occurrence in today’s classrooms. Moreover, the relationship between dialogic talk and outcomes such as writing quality is less understood. This dissertation explores the predictive association between features of whole class discussion and students’ writing quality in classrooms implementing debate-centered curricula. Findings from two empirical studies using the Low-inference Discourse Observation tool (LIDO) are highlighted. The LIDO is a unique instrument that captures various research-based indicators of discussion produced by students and teachers in whole class settings. The first paper analyzes teacher and student talk using audio recordings from 42 classrooms in grades four through seven, as well as corresponding student essays (n=471). Multilevel modeling shows that teacher-produced dialogic talk positively predicts students’ persuasive writing, while teacher- and student-produced low dialogic talk negatively predicts their writing. Contrary to hypothesized relationships, however, student talk coded as more dialogic using the LIDO was negatively predictive of persuasive writing. In this study, participation rate was also positively associated with students’ writing. The second paper analyzes video recorded student data from one eighth grade classroom (n=15), where students discuss four civics topics over eight class sessions. Findings are consistent with the previous study, as dialogic student measures were negatively associated with persuasive essays. The fourth chapter examines implications for classroom practice gleaned from these results, such as emphasizing contestable questions over semi-open or quiz-like questions. The thesis concludes with suggested approaches to move the field of discussion research forward.