Author: Lou Blouin
Publisher: University of Michigan-Dearborn News
Publication Year: 2022
Summary: The following article describes the ethical issues inherent in data-driven education, particularly in the context of digital learning. Blouin notes that student privacy is often overlooked, particularly when instructors require students to install and use third-party software that tracks user data. In this case, the institution is not at risk, while the students are left vulnerable to misuse of their data. For example, third-party educational software has sold student data to advertising groups that would promote higher education opportunities solely to white students. More broadly, though, the article describes the risks that come with defining and collecting “engagement metrics” in classroom settings. When any kind of metric is defined to measure engagement, Blouin points out that analysts may forget the importance of other factors that were not encoded as variables. As a result, when using engagement metrics to judge subjects, organizations may fail to capture the importance of hard-to-measure qualitative features. Engagement metrics are also prone to misuse, and have been used to unfairly penalize students due to low metrics alone.