Automated Support to Scaffold Students' Short- and Long-form STEM Writing

Explore automated analysis techniques for assessing student writing in STEM fields during this 59-minute lecture by Professor Becky Passonneau from Penn State University. Discover how automated support can provide alternatives to multiple-choice questions and enable more comprehensive evaluation of students' reasoning skills through long-form writing. Learn about methods focused on analyzing specific ideas expressed by students to support formative assessment and provide targeted feedback. Examine research from the PSU NLP lab on automated assessment of various forms of student writing, including middle school physics essays and college physics lab reports. Gain insights into novel approaches like recurrent relation networks incorporating contrastive learning for short answer assessment. Understand the potential of these techniques to scaffold students' abilities to reason and communicate effectively in STEM subjects.