By Ariadni Papana (Cleveland State University)
Information
BASE-AI-Spark is an instructional model designed to guide the integration of Generative AI into statistics teaching while connecting students’ disciplines and curiosity to the statistical concepts being taught. The model extends the BASE-on-AI framework introduced at USCOTS 2025 and includes three components: a syllabus AI-statement, AI-modified labs, and AI-discovery activities. The syllabus AI statement establishes expectations for responsible AI use and communication in the course. AI-modified labs serve as the primary learning space for developing statistical and technical skills, enriched with AI activities that support conceptual understanding. The new element in BASE-AI-Spark is the AI-discovery component which introduces opportunities for exploration, allowing students to investigate how statistics connects to their own disciplines, research interests and future careers. While the AI-modified labs provide structured opportunities to learn statistical concepts, the discovery component emphasizes curiosity and exploration, making the learning experience more engaging. Because lectures alone may leave some concepts unclear or misunderstood, BASE-AI-Spark offers a structured yet exploratory framework that integrates AI responsibly and helps students appreciate the role of statistics in their own fields. This model was piloted in an undergraduate introductory applied statistics course delivered in an online asynchronous format with students from a wide range of majors, including Biology, Psychology, Data Science, Health-related programs, Arts & Sciences, CCP, and Education pathways. This multidisciplinary setting motivated the need to help students connect statistics to their own academic interests. In the AI-discovery activity, students identified their major, selected a discipline-related course/topic, chose a statistical method learned in class, used a guided-AI prompt to connect the topic to the method, and reflected on what they learned. The model is designed for a wide range of instructional settings, including research universities, four-year institutions, community colleges, and AP Statistics courses, and can be adapted to online, hybrid, or in-person formats. It is primarily intended for undergraduate statistics courses across disciplines but may also extend to graduate-level courses with diverse student populations. Evidence is based on classroom observations, student feedback, and preliminary survey data. AI-modified labs have been implemented and appear to support student learning, while short surveys (e.g., Mentimeter) provide insight into students’ backgrounds and perspectives on AI. The AI-discovery component is newer and continues to be refined based on these early observations.