AI-Assisted Coding of Students’ Graph Literacy in Dynamic Data Environments

In contemporary data-rich societies, statistical literacy has become a foundational component of informed citizenship. While public data is increasingly presented through dynamic, multivariable visualizations, high-school mathematics and statistics curricula often remain focused on static, univariate graphs. This doctoral study investigates the development of graph literacy among high school students, grounded in the frameworks of Curcio (1987) and Shaughnessy (2007). The study employs a research instrument incorporating interactive dashboards, dynamic visualizations, and Excel-based datasets derived from authentic social-science data. This presentation focuses on a pre-intervention questionnaire designed to map students' graph literacy levels as they engage with complex data representations. Students' open-ended responses were analyzed using a computational coding protocol developed collaboratively with a large language model. The presentation discusses this AI-assisted coding protocol and its potential for large-scale qualitative analysis, as well as how developing it alongside an LLM may help operationalize existing literacy frameworks in dynamic, multivariable data environments.