CRII: Understanding Conceptual Transfer from Imperative Rule-Based Programming to AI/ML Data-Driven Computation Approaches

San Jose State University will explore how to improve computer programming education for introductory artificial intelligence (AI) and machine learning (ML). As computing education evolves, students must learn both traditional programming and newer approaches like AI and ML. These newer topics often require different ways of thinking, and many students struggle to transfer what they’ve learned in early traditional programming to these more complex, data-driven approaches. This project explores where students face challenges in making these transfers and develops teaching approaches that use real-world examples to build stronger connections to prior knowledge. By helping students relate new concepts to what they already know, the project supports deeper understanding, better engagement, and greater success in computing. The result will be a flexible teaching framework that educators can use to make advanced computing topics more accessible to a wide range of learners by building connections. This work supports NSF’s mission by advancing effective STEM education in a critical computing and AI field and ultimately preparing a strong technology workforce. This project investigates how undergraduate students navigate the conceptual transition from imperative, rule-based programming to data-driven paradigms such as AI and ML. The proposed work has the potential to advance knowledge about: (1) the specific programming concepts students struggle to transfer when moving between imperative and AI/ML approaches; (2) how contextualized, real-world examples can support students in bridging conceptual gaps across programming paradigms; (3) the design of a pedagogical framework that scaffolds conceptual transfer using these contextualized examples; and (4) the impact of such a framework on students’ understanding, confidence, and ability to apply prior programming knowledge in new contexts. The project will use a mixed-methods approach, beginning with qualitative exploration to inform the development of contextualized materials, followed by quantitative testing in undergraduate classrooms to evaluate learning outcomes and conceptual gains. This work contributes to computer science education by improving support for student learning as computing curricula increasingly include advanced topics in AI and data-driven methods. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.