Assessing the Performance of Automated Short Answer Scoring to Support Integrated Engineering and Science Understanding

• In Event: 53.082 - Poster Session 10
  In Poster Session: 53.082-3 - Special Topics in Engineering and Computer Science: Engineering Design, Computational Thinking, and Assessment

Abstract

The Next Generation Science Standards emphasize connections between engineering and science in K-12 classrooms. However, K-12 students often have difficulty understanding and applying science ideas during engineering design. Providing automated guidance to students during design projects has the potential to help students understand and make connections among science and engineering ideas. This paper reports on the performance of an open-source automated scoring system using student responses from a middle school integrated science and engineering unit within an online engineering learning environment. Results indicate that the automated scorer is comparable in agreement to the highest performing automated graders. Results point to the potential for use of automated guidance to help students develop and apply science understanding in engineering design projects.

Authors

• Jim Bywater, University of Virginia

• Xiang Fu, Hofstra University

• Jennifer L. Chiu, University of Virginia