Adaptive Dialog to Support Student Understanding of Climate Change Mechanism and Who is Most Impacted

Abstract

To support ninth graders to take advantage of the ideas, intuitions, and experiences that contribute to their understanding of climate change, we designed an NLP-based adaptive dialog and tested it in a week-long unit exploring Urban Heat Islands. The dialog’s guidance prompts were designed to elicit students’ ideas about climate change mechanisms. Students interacted with the adaptive dialog twice during the unit. We scored their initial and revised explanations (after engaging with the adaptive dialog each time) using a Knowledge Integration (KI) rubric. A repeated measures mixed ANOVA revealed that students who initially expressed descriptive ideas often rooted in experience made significantly greater gains during the dialog than those who initially expressed mechanistic ideas. The dialog supported all students to broaden the ideas they considered when exploring climate change. Further, a paired t-test revealed that students made overall gains in KI from pretest to posttest (d=.76).