This was part of Teaching and Evaluating Data Communication At Scale

Supporting Student Writing and Reasoning in Data Science Using a Scalable Automated Feedback System

David Brown, Suguru Ishizaki, Michael Laudenbach, Carnegie-Mellon University

Wednesday, January 10, 2024

Abstract: This workshop will introduce a student-facing, interactive tool called DocuScope Write & Audit. The tool is web-based and integrated into the learning management system Canvas, where it can be made accessible to students though a Canvas assignment. Once the system is launched, students can copy and paste their text into a text editor and submit it for processing. The tool is designed to provide tailored, assignment-specific feedback to students as they work through the drafting process. Moreover, that feedback is automated, alleviating some of the challenges that instructors face when trying to teach and support writing in large classes, particularly in content courses where instructors are additionally pressed for time. During this workshop, we will provide an overview of Write & Audit’s design, demonstrate how it works using sample assignments from an introductory statistics and data science course, and briefly preview supplemental functions that take advantage of Large Language Models like ChatGPT.