Date
Publisher
arXiv
This exploratory case study investigated two contrasting pedagogical
approaches for LCA-assisted programming with five novice high school students
preparing for a WeChat Mini Program competition. In Phase 1, students used LCAs
to generate code from abstract specifications (From-Scratch approach),
achieving only 20% MVP completion. In Phase 2, students adapted existing
Minimal Functional Units (MFUs), small, functional code examples, using LCAs,
achieving 100% MVP completion. Analysis revealed that the MFU-based approach
succeeded by aligning with LCA strengths in pattern modification rather than de
novo generation, while providing cognitive scaffolds that enabled students to
navigate complex development tasks. The study introduces a dual-scaffolding
model combining technical support (MFUs) with pedagogical guidance (structured
prompting strategies), demonstrating that effective LCA integration depends
less on AI capabilities than on instructional design. These findings offer
practical guidance for educators seeking to transform AI tools from sources of
frustration into productive learning partners in programming education.
Study at a glance
- Question: How does adapting existing functional code examples compare to coding from scratch when high schoolers use AI agents?
- Setting and data: Five private high school students in grades 10-12, Shanghai, China, April 2025 across two three-hour evening coding sessions.
- Method: Exploratory case study comparing two sequential coding sessions using instructor field notes.
Why it matters
Novice coders using AI agents often struggle to create projects from scratch, face debugging difficulties, and fail to gain learning satisfaction or self-efficacy without proper pedagogical support. Robust empirical evidence for effective classroom integration remains limited.
Key findings
- Students achieved a 100% Minimum Viable Product completion rate when adapting functional code examples with AI, compared to only 20% when generating code from scratch.
- In the adapting phase, students required fewer debugging requests, completed features faster, and reported higher confidence and more optimistic attitudes.
- Students shifted from single, abstract AI prompts in the from-scratch phase to multi-step analytical prompts when adapting existing code modules.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
