Design PjBL projects so students must evaluate, critique, and improve AI-generated outputs rather than prohibiting AI use¶

Strategy · All strategies
Evidence · no claims cited

Description¶

Based on feedback that students increasingly use tools such as ChatGPT during project work, the authors recommend designing projects so that students must evaluate, critique, and improve AI-generated outputs rather than prohibiting such tools. Projects were revised to emphasize justification of decisions, comparison of alternative solutions, debugging, and assessments that reward reasoning and verification. The strategy treats AI assistance as an design opportunity to promote deep understanding rather than a threat to be banned.

Design Implications¶

Context¶

Requirements¶

• Assessments that reward reasoning and verification
• Project tasks requiring justification of decisions and comparison of alternative solutions

Constraints¶

• AI tools may also support instructors in drafting materials, but these still require instructor review for technical correctness and alignment with learning objectives

Target Learners¶

• undergraduate computing students

Target Learning Goals¶

• deep understanding, debugging and verification skills, higher-order evaluation of AI-generated solutions

Related Strategies¶

• Engage students in active learning assignments using generative AI with structured reflection and critique of AI outputs
• Embed AI as a learning partner and redesign assessments for AI-rich contexts
• Use MAPS-based reflection questions to prepare students to critique AI-generated physics solutions
• Five design principles for responsible implementation of AI-generated course readings
• Structure opportunities for students to evaluate AI outputs using locally grounded criteria of relevance, harm, and usefulness

Examples¶

-

Key Sources¶

• Suleiman, A. D., DeWaters, J., Shepherd, D. C., Korkmaz, T., Hussain, F., Liu, Y., & Hou, D. (2026). Adoption-Ready Project-Based Learning for Computing Education: The FORAP Framework and a Multi-Scale Project Portfolio. https://arxiv.org/abs/2607.11129. doi:10.48550/arxiv.2607.11129