The Role of Instructional Guidance in Generative AI-Assisted Learning: Empirical Evidence from Construction Engineering Education

2026-06-06 · Source: cs.AI updates on arXiv.org · Field: Education & Learning — Educational Technology (EdTech), Artificial Intelligence & Machine Learning, Academic Research & Higher Education · Depth: Expert, quick

Summary

A study published on arXiv:2606.05509 investigates the impact of instructional guidance on generative AI-assisted learning, specifically within construction engineering education. The research introduces a five-step prompting framework, grounded in Generative Learning Theory (GLT), designed to structure student-AI interaction during review activities. A controlled experiment compared three learning conditions: traditional slide-based learning, unprompted AI-supported learning, and prompted AI-supported learning. Performance was assessed via multiple-choice and open-ended tasks, alongside user experience measured by the User Experience Questionnaire (UEQ). Findings indicate that the prompted condition significantly improved open-ended scores by approximately 2 or 3 points on an 18-point scale (p < 0.01), particularly for tasks requiring explanation and reasoning. No significant differences were found in multiple-choice performance, and unprompted AI learning was comparable to slide-based methods, underscoring the critical role of structured interaction.

Key takeaway

For educators or instructional designers developing generative AI-assisted learning modules, you should prioritize integrating explicit instructional guidance. Unprompted AI interaction offers no significant advantage over traditional slide-based learning for tasks requiring explanation and reasoning. Instead, implement structured prompting frameworks, like the five-step GLT-based model, to significantly improve student performance on complex cognitive tasks and ensure your AI tools genuinely enhance deeper learning outcomes.

Key insights

Instructional guidance, like a five-step prompting framework, significantly enhances generative AI's effectiveness for deeper learning tasks.

Principles

• AI learning effectiveness depends on structured interaction.
• Guidance improves explanation and reasoning tasks.
• Generative Learning Theory informs AI prompting.

Method

A five-step prompting framework, grounded in Generative Learning Theory, guides learner interaction with generative AI during review activities to foster deeper cognitive processes.

In practice

• Implement structured prompting for AI-assisted review.
• Focus guidance on tasks needing explanation.
• Design prompts using GLT principles.

Topics

• Generative AI
• Instructional Guidance
• Construction Engineering Education
• Learning Performance
• Prompt Engineering
• Human-Computer Interaction

Best for: AI Scientist, Research Scientist, Domain Expert

Related on AIssential

• The Magic of Prompting: How to Talk to AI to Get What You Actually Want — Effective LLM interaction requires structured, specific prompts that provide context and guide the AI's response.
• ChatGPT Can Write My Essay, but It Can’t Do My Thinking — Instant AI content generation risks diminishing critical thinking and deep learning by bypassing essential intellectual struggle.
• Beyond Prompting: How Engineers Should Architect AI Code Generation — AI integration in software development shifts from prompting to designing controllable, continuous production processes and workflows.
• 500 Blog Posts To Learn About Ai — AI's rapid evolution spans foundational algorithms to complex applications, driving significant technological and societal shifts.
• 4 Steps to Master AI Prompting: From Quick Answers to Expert-Level Automation — Mastering AI prompting evolves through structured steps, from quick answers to automated, data-integrated systems.
• Promptions helps make AI prompting more precise with dynamic UI controls — Dynamic prompt refinement controls improve user interaction and reduce effort in generative AI comprehension tasks.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.AI updates on arXiv.org.