CANCELLED: Storytelling with Explainable Machine Learning in R for Civic Engagement

This short course explores how explainable machine learning (XML) can be combined with storytelling to strengthen civic engagement and democratic participation. Using R exclusively, participants will learn how to transform complex models into narratives that are interpretable, transparent, and impactful for non-technical audiences such as policymakers, educators, journalists, and community organizations. The course introduces XML methods, including interpretable models, LIME, SHAP, and DALEX, and demonstrates how they can be integrated with state-of-the-art visualization and communication tools in R (ggplot2, ggtext, and ModelStudio). Real-world and simulated civic datasets (e.g., voter participation, gender representation, housing and economic inequality) will ground the exercises and illustrate how explainability enhances trust in public-facing analytics. Participants will engage in both guided group work and individual practice to build and evaluate models, interpret their predictions, and create civic-focused narratives. Emphasis will be placed on balancing predictive performance with fairness, transparency, and accountability. By the end of the course, participants will be able to design understandable machine learning workflows in R, extract and visualize meaningful insights, and communicate them through compelling civic stories that build public trust, strengthen transparency, and promote responsible decision-making.