Joint Reflection Promotes Students’ Use of Evidence in Argumentive Writing

A basic component of argumentive writing is the coordination of claims with evidence bearing on them. Deep engagement in dialogic argumentation has been found to facilitate development of beginning students’ individual written argument. Despite progress in several respects following such engagement, in their argumentive writing middle-school writers frequently ignore evidence incongruent with their claims -- a violation of norms of skilled argument. The present research examines the effectiveness of engaging middle-school student dyads in joint meta-level reflection on the use of evidence in their argumentation, both anticipating its potential use and evaluating its actual use. A total of 54 Chinese 7th graders participated in a dialogic argument curriculum in 33 class sessions over four months. For each of three successive topics, evidence both congruent and incongruent with a dyads’ position on the topic was made available for their use. Half of the participants were assigned to an Evidence Reflection and Argument Practice (ER+AP) condition, in which in addition the dyad was prompted to discuss verbally and jointly complete reflection sheets regarding their evidence use. The other half of participants served in an Argument Practice (AP) condition, identical except for omission of the Evidence Reflection component. Analysis of participants’ individual written essays on the topic at the end of their engagement with each topic revealed superior performance on the part of the ER+AP group, with the reflection component enhancing their addressing evidence both congruent and incongruent with their claims. However, this happened only slowly. The superiority of the ER+AP group was most decisive by the last topic, when members of the ER+AP students also demonstrated an ability to connect two pieces of evidence serving conflicting argumentive functions. Fifty additional students participated in a control condition, included for the purpose of comparing their performance to that of the intervention students on a topic new to both groups. Both the ER+AP and AP intervention groups showed superior performance relative to the control group in including evidence congruent with their own position in their essays. Only the ER+AP group, however, showed superiority in addressing evidence incongruent with their position. Analysis of responses students provided to the evidence reflection sheets revealed developmental patterns over time, and explicated the underlying mechanism driving ER+AP students’ superior performance. Theories regarding the interiroization of cognition from inter- to intra-mental planes, as well as the supportive effects of meta-level engagement on transfer of skills, are invoked in accounting for the findings.