Air Pollution Visualizations for Promoting Data Literacy with Middle Schoolers and the Public

Air pollution is a data-rich problem space, with vast networks of sensors constantly capturing and reporting measurements. The measured pollutants are superficially familiar to many people, but few understand the specific sources, acceptable levels, and negative effects of any individual pollutant. In addition, coverage of air quality in formal school curricula is limited. This project seeks to deepen the public's understanding of air quality data through a public kiosk and to develop middle schoolers' understanding of air quality data through a youth summer camp. The project will investigate how air quality data interactions (via these two concurrent designs: the public kiosk and the youth summer camp) can be designed to support learners' personal agency in data investigations, visualizations, and communications as well as how these experiences help non-experts learn about their environment. This project expands public engagement with air quality data in ways that extend beyond the simplified metric of the Air Quality Index (AQI). Because this work has the potential to lessen the knowledge gap between the scientific community and the general public it could be an important starting point for enacting meaningful change in mitigating air pollution.

This four-year interdisciplinary effort draws on expertise in the learning sciences, visualization, middle and high school STEM education, and teacher learning and professional development to create two design spaces for informal learning aimed toward advancing engagement and self-efficacy in reasoning with scientific data. This project will advance knowledge in human-data interaction (HDI) design and support environmental data literacy through air quality visualizations. The Data Visualization and the Environment (DVE) summer camp will be designed in collaboration with local middle school science educators and will invite middle school students from the Atlanta area to Georgia Tech's campus for a sponsored 2-week program during which they will engage with science educators and visualization experts to collect, visualize, and display environmental data. Recruitment for these funded camps will prioritize underserved and underrepresented groups in science. By working first with air quality data and then a self-selected environmental dataset of interest to them, students will participate in the scientific practices of generating and presenting evidence-based arguments about their local environment to investigate projects of relevance to their own community, culture, and interests. The project will concurrently develop and iterate the Contextualized Air Quality Timelines (ContAQT) platform as a public data display in Midtown Atlanta breaking down current and historical local multipollutant air quality data. The project will answer the following research questions: (1) How can data interactions in informal settings support learning about data interpretations and the environment? (2) How can data interactions be designed to support learners' opportunities to investigate questions of personal interest? and (3) What design features support data exploration by non-expert learners? Through design-based research the team will investigate how changes to the interface and interaction impact data fluency indicators such as statistical reasoning talk (identifying outliers, reasoning about trends, etc.), connections to prior knowledge, and asking questions. Using qualitative research methods, such as conjecture mapping, the project will investigate how data interactions in brief, free-choice public kiosk interactions (ContAQT) as well as extended data investigations (summer camp) can assist learners in asking questions, making connections to existing knowledge, and proposing and evaluating claims. This work will advance knowledge in the following areas: 1) best practices for framing scientific data to be accessible to lay audiences; 2) understanding of how specific design decisions in visualization interfaces and interactions can influence data interactions and learning talk; 3) design of summer camp activities to support middle school students in developing data fluency and increasing self-efficacy in science; 4) synthesizing knowledge from science educators in best practices for engaging female, Black, and/or Latine learners in data practices.