Combining Automated Scoring and Teacher Guidance to Improve Students' Science Learning

• In Event: Poster Session 11
  In Poster Session: Division H Poster Session: Assessment in Schools

Abstract

We investigate how alerts based on automated scoring of student-generated work can support teachers to guide struggling students’ to develop more robust scientific ideas. In Study 1, we examine how the combination of automated guidance and teacher alerts effects student learning, compared to multiple rounds of automated guidance only. In Study 2, we compare two types of teacher alerts based on different possible automatically scored indicators of student science learning. We draw on evidence from pre/post test and embedded assessments. The studies demonstrate a combination of automated guidance plus the teacher alert is more effective than automated guidance alone for low-scoring students. Short essays serve as an optimal prompt to structure teacher-student discussions on science reasoning.

Authors

• Libby F. Gerard, University of California - Berkeley

• Jonathan Michael Vitale, University of California - Berkeley

• Dermot Francis Donnelly, California State University - Fresno

• Marcia Linn, University of California - Berkeley