Investigating Institutional Bias Through a Visualization Dashboard

Abstract

In this work, we present an interactive data visualization dashboard developed for the 2025 VAST Challenge Mini-Challenge 2, which investigates stakeholder engagement and potential bias during council meetings. The dashboard integrates matrix-based visualizations, barplots, trend lines, boxplots, and tooltips to support exploratory analysis. Through the TROUT (tourism) and FILAH (fishing) datasets, we identify systematic sentiment differences aligned with each group’s sectoral interests—particularly when discussions involve their own industries (e.g., large-vessel fishing or tourism)—providing evidence of potential bias among council members (COOTEFOO). Incorporating the JOURNALIST dataset, which combines both perspectives, reveals a moderation of sentiment extremes when a broader, more balanced view is introduced. Overall, this paper demonstrates how interactive dashboards can effectively expose and contextualize institutional bias, underscoring the utility of data-driven approaches for transparency.