Abstract

Data visualization literacy has become crucial in STEM education, particularly in supporting students in becoming producers rather than consumers of graphical representations. Yet, secondary students face an accessibility valley where entry-level tools are too limited in their design capabilities while more advanced programming environments are too cumbersome to navigate. This paper introduces the framework for AI-enhanced Literacy In Visualization Education (AILIVE), which leverages large language models like ChatGPT to democratize data visualization in secondary classrooms.

Grounded in constructionist learning theory, our framework addresses visualization education's central paradox: students need to create sophisticated visualizations to develop representational competence and match their creative vision, but lack technical skills for implementation. The framework includes five design principles (meaningful data context, student agency, communication intent, iterative refinement, and collaborative discourse) implemented through three phases: preparation, investigation, and synthesis/communication.

We demonstrate AILIVE through the “Snackdown Challenge,” a hypothetical activity in which students use ChatGPT to visualize data about snack characteristics and class preference. Natural language interaction enables students to focus on statistical reasoning and communicating meaning through graphs rather than technical barriers, transforming learning from procedural exercises to authentic investigation. This approach develops data visualization literacy essential for 21st-century STEM participation while maintaining focus on conceptual understanding of graphical features over technical proficiency.

Keywords

data visualization literacy, artificial intelligence, ChatGPT, constructionism, secondary education, STEM education

How to Cite