AI赋能材料科学实验教学：基于“三阶段–六步走”框架的准实验研究

AI-Enhanced Experimental Teaching in Materials Science: A Quasi-Experimental Study Based on the “Three-Stage Six-Step” Framework

摘要: 在新工科建设与人工智能(AI)技术快速发展的背景下，《材料科学研究综合实验》课程亟待更新教学内容、方法与设备。本研究构建了“三阶段–六步走”教学改革方案，即课前开发AI生成资源与虚拟项目情境，课中运用机器学习预测与VR/AR场景解决问题，课后基于学习痕迹构建精准评价体系，旨在提升学生的工程实践能力和创新精神。采用准实验研究设计，在复合材料工程专业76名本科生中开展试点，随机分为实验班(n = 38)与对照班(n = 38)。通过工程实践能力量表、创新精神评价量规、学习痕迹分析与半结构化访谈收集数据，运用协方差分析(ANCOVA)与质性编码进行统计分析。结果显示，实验班工程实践能力、创新精神、学习体验和学习效果均得到了有效提升。该框架为AI赋能材料科学实验教学提供了可验证的实践模型，对培养复合型工程技术人才具有重要参考价值。

Abstract: Under the background of emerging engineering education and rapid development of artificial intelligence (AI) technologies, the course “Comprehensive Experiments in Material Science Research” urgently requires updates to its teaching content, methodologies, and equipment. This study developed a “three-stage six-step” instructional reform framework: pre-class AI-generated resource development with virtual project scenarios, during-class application of machine learning prediction and VR/AR problem-solving environments, and post-class construction of precise evaluation systems based on learning trace analysis, aiming to enhance students’ engineering practical capabilities and innovative thinking. A quasi-experimental design was implemented with 76 undergraduate students majoring in composite materials engineering, randomly assigned into experimental group (n = 38) and control group (n = 38). Data were collected through Engineering Practice Ability Scale, Innovation Spirit Assessment Rubric, learning trace analysis, and semi-structured interviews, followed by statistical analysis using Analysis of Covariance (ANCOVA) and qualitative coding. Results demonstrated significant improvements in engineering practice ability, innovation spirit, learning experiences, and academic performance among the experimental group. This framework provides an empirically validated model for AI-enhanced material science laboratory instruction and offers valuable references for cultivating interdisciplinary engineering talents.