Login or signup to connect with paper authors and to register for specific Author Connect sessions (if available).

LEAP: A Method for Programming Education
Nada Hashmi, Zhi Li, Salvatore Parise, Ganesan Shankaranarayanan
This research introduces LEAP (Learning Enabled by AI for Programming), a novel pedagogical approach integrating Generative AI into programming education. Addressing persistent challenges of high dropout rates and learner frustration, LEAP strategically incorporates tools like ChatGPT and GitHub Copilot to provide personalized, real-time assistance. Applying Robins et al.'s (2003) framework, we categorize student-GenAI interactions across knowledge, strategies, and mental models during design, generation, and evaluation phases. Analysis of over 1,500 prompts from graduate business students reveals a fundamental shift in learning dynamics—from syntax-focused instruction (reduced from 50% to 20%) toward enhanced conceptual understanding and problem-solving (each expanded to 40%). While students effectively leverage GenAI for knowledge acquisition and debugging, our findings highlight the continued necessity for explicit instruction in strategic problem decomposition and independent troubleshooting. LEAP demonstrates how thoughtfully integrated GenAI can transform programming education while preparing students for AI-augmented professional environments.

AuthorConnect Sessions

No sessions scheduled yet