Students’ Statistical Thinking When Using Generative AI: A Descriptive Case Study

Generative artificial intelligence (AI) technologies are transforming the world of education, but their impact on students’ thinking and learning remains unclear. In this study, we investigated AI’s potential role in supporting statistical thinking by interviewing undergraduate students ([Formula: see text]) as they completed a graphing task using Rtutor.AI, an AI-powered tool that integrates ChatGPT with R. The analysis yielded five key themes that describe students’ statistical thinking while using Rtutor.AI. The first theme demonstrated how the iterative and intuitive nature of prompting within Rtutor.AI shaped participants’ approaches to problem-solving. The next three themes—“Building statistical understanding through a step-by-step process,” “Identifying key elements of a problem to create specific prompts,” and “Lowering barriers to completion of statistical tasks”—illustrated how Rtutor.AI facilitated statistical thinking in various ways. The fifth theme showed that students’ prior statistical knowledge influenced their ability to interpret and contextualize Rtutor.AI’s output, and that Rtutor.AI did not fully absolve students of the need to think statistically. Overall, these results highlight both the potential benefits and risks of incorporating AI into statistics classrooms and can serve as an empirical basis for future scholarship aimed at creating scaffolding around AI technologies to support statistics instruction.