Through the Looking Glass: Insights into Visualization Pedagogy through Sentiment Analysis of Peer Review Text

TL;DR

This paper uses sentiment analysis of peer review texts in visualization courses to gain insights into student engagement, concept mastery, and course trends, providing practical recommendations for educators.

Contribution

It introduces a novel application of sentiment analysis to peer review data in visualization education, offering empirical insights and pedagogical recommendations.

Findings

Peer review enhances student engagement and concept reinforcement.

Sentiment analysis reveals focus areas and engagement disparities.

Insights inform targeted teaching strategies and peer review improvements.

Abstract

Peer review is a widely utilized feedback mechanism for engaging students. As a pedagogical method, it has been shown to improve educational outcomes, but we have found limited empirical measurement of peer review in visualization courses. In addition to increasing engagement, peer review provides diverse feedback and reinforces recently-learned course concepts through critical evaluation of others' work. We discuss the construction and application of peer review in two visualization courses from different colleges at the University of South Florida. We then analyze student projects and peer review text via sentiment analysis to infer insights for visualization educators, including the focus of course content, engagement across student groups, student mastery of concepts, course trends over time, and expert intervention effectiveness. Finally, we provide suggestions for adapting peer review to other visualization courses to engage students and increase instructor understanding of the peer review process.