Enhancing Student Feedback Using Predictive Models in Visual Literacy Courses

TL;DR

This study addresses the durability and predictive value of peer review in visualization education by applying a Naïve Bayes classifier to five years of student feedback. It analyzes both parts‑of‑speech usage and visual peer‑review rubric scores to predict feedback characteristics, finding nouns to be the dominant POS and the lie factor to align strongly with noun usage. The NB model achieved high accuracy (∼95% overall with CV) and demonstrated the rubric’s utility in forecasting educational directions, though PPV/NPV indicated areas for refinement. The work suggests a practical, data‑driven approach to improving visual literacy courses, including rubric design and course content, facilitated by predictive analytics. These insights hold potential for scalable, instructional decision support in higher‑education visualization programs.

Abstract

Peer review is a popular feedback mechanism in higher education that actively engages students and provides researchers with a means to assess student engagement. However, there is little empirical support for the durability of peer review, particularly when using data predictive modeling to analyze student comments. This study uses Naïve Bayes modeling to analyze peer review data obtained from an undergraduate visual literacy course over five years. We expand on the research of Friedman and Rosen and Beasley et al. by focusing on the Naïve Bayes model of students' remarks. Our findings highlight the utility of Naïve Bayes modeling, particularly in the analysis of student comments based on parts of speech, where nouns emerged as the prominent category. Additionally, when examining students' comments using the visual peer review rubric, the lie factor emerged as the predominant factor. Comparing Naïve Bayes model to Beasley's approach, we found both help instructors map directions taken in the class, but the Naïve Bayes model provides a more specific outline for forecasting with a more detailed framework for identifying core topics within the course, enhancing the forecasting of educational directions. Through the application of the Holdout Method and $\mathrm{k}$-fold cross-validation with continuity correction, we have validated the model's predictive accuracy, underscoring its effectiveness in offering deep insights into peer review mechanisms. Our study findings suggest that using predictive modeling to assess student comments can provide a new way to better serve the students' classroom comments on their visual peer work. This can benefit courses by inspiring changes to course content, reinforcement of course content, modification of projects, or modifications to the rubric itself.

Figures (6)

• Figure 1: Student map-making project and visual peer review rubric the student evaluate the work.
• Figure 2: Example of a student's comments on their peer's work.
• Figure 3: Summary of the word counts from 2017 to 2022.
• Figure 4: Summary of parts of speech from students' comments.
• Figure 5: Summary of the correlation matrix score on parts of speech (nouns, verbs, adverb (adv), and adjective (adj)).