What is the impact of giving students guidance on their infrequent ideas in an AI Dialog?

• In Event: AERA Poster Session Sunday 7:45 am
  In Poster Session: Equitable Participation and Discourse in Science Teaching and Learning

Abstract

Students benefit when their ideas are taken up as sensemaking resources. When students have unique ideas about a science topic, their ideas might get recognized in science classes. We investigated the impact of adaptive dialogs that leverage AI idea detection models to provide students with guidance based on the ideas in their science explanations. We examine the impact of guiding students’ infrequent ideas, ideas likely held by only one or two students in a class of 35. We found that students given guidance on infrequent ideas benefit as much or more than students guided on more frequent ideas. Further, many students with infrequent ideas reported feeling understood and like the guidance helped them notice more of their own ideas.

Authors

• Allison Bradford, University of California - Berkeley

• Libby F. Gerard, University of California - Berkeley

• Weiying Li, University of California - Berkeley

• Marcia Linn, University of California - Berkeley