USF's VisualStats: AI-driven learning for statistical data visualization
The future of AI in education is being built in real classrooms. At the USF College of Education, Assistant Professor Bo Pei received national recognition for that work, earning the 2025 Emerging Learning Technology Award from the Association for Educational Communications and Technology's Division of Emerging Learning Technologies (DELT) for his AI-driven platform, VisualStats.
VisualStats supports how students learn statistical data visualization through interactive dialogue. As Pei put it, "VisualStats is not just another tech tool. It represents a new vision for how students can learn and master statistical data visualization through interactive dialogues in an increasingly data-driven world."
Why this matters for education, science, and research
Data visualization turns raw data into charts, graphics, and maps so people can quickly spot trends and patterns. It's central to science, government, software, and finance, yet many students struggle to go from numbers to clear visuals that actually explain what's going on.
"It's a core skill because even the most intuitive analysis means nothing if people can't make sense of the findings," Pei said. VisualStats aims to give every student the confidence to explore, analyze, and communicate with data by pairing data literacy, statistical reasoning, and visualization into one learning experience. For a quick primer on the concept, see data visualization.
From clicking menus to learning through conversation
Most tools force students to guess their way through unfamiliar interfaces. VisualStats flips that. Learners ask questions, get immediate feedback, and move step-by-step through a guided conversation that links statistical choices to the visuals they produce.
"It acts like a knowledgeable peer, encouraging students to ask questions through a series of low-stakes conversations that help them understand the rationale from the data to the visualizations, not just the physical images."
Integrated scaffolding that builds real competence
Pei's research focuses on AI in education and how to support meaningful progress, not just button-clicking. VisualStats includes built-in scaffolding that helps students choose appropriate charts, justify statistical assumptions, and explain their decisions in plain language.
The result is a tighter connection between method and message. Students learn to communicate insights clearly, and instructors get a window into where reasoning breaks down.
Piloted in USF courses, built for scale
Pei has piloted early versions of VisualStats in his courses to study learning patterns and tune the embedded supports. The next phase includes teacher-facing features to monitor behaviors, spot gaps in understanding, and generate targeted interventions.
The long-term plan is a scalable platform that can support classrooms nationwide while giving faculty the analytics they need to improve instruction. Recognition from AECT signals how this work is setting a standard for AI-supported learning design.
What this means for your courses and labs
- Move beyond "make a chart" tasks. Ask students to justify why a visual fits the data and audience.
- Use AI-guided dialogues to lower the stakes and increase repetition where it matters: selecting, interpreting, and explaining.
- Leverage logs and feedback to detect misconceptions early (e.g., misuse of scales, inappropriate chart types, weak variable reasoning).
- Target support to specific stumbling blocks with just-in-time prompts, examples, and micro-assessments.
- Assess communication, not just correctness-students should be able to defend their choices.
Where it's headed
As development continues, expect richer analytics for instructors, smarter prompts for students, and tighter alignment with learning goals across disciplines. The aim is simple: make statistical visualization teachable, coachable, and repeatable at scale.
For teams building AI fluency
If you're planning faculty or TA upskilling around AI and data, explore curated options by role at Complete AI Training. Pairing pedagogy with practical AI skills sets the stage for strong outcomes when tools like VisualStats enter your curriculum.
Your membership also unlocks:
