"Prompting Behaviors as Pedagogical Tools in the Age of AI: Empirical Insights for Engineering Education"
收藏DataCite Commons2025-07-23 更新2026-05-03 收录
下载链接：
https://ieee-dataport.org/documents/prompting-behaviors-pedagogical-tools-age-ai-empirical-insights-engineering-education
下载链接
链接失效反馈官方服务：
资源简介：
"This dataset accompanies the study \"Prompting and Performance: An Empirical Study of ChatGPT-4o Use in Engineering Education,\" which investigated how specific prompting behaviors with large language models (LLMs) relate to academic performance. The dataset includes anonymized AI interaction logs (prompts, responses, and timestamps), written student submissions, assignment grades, and a set of computed behavioral and writing quality metrics (e.g., Structural Complexity, Content Richness, Query Efficiency, Prompt Refinement Depth, and AI-Driven Problem-Solving). Data were collected over a 16-week stratified randomized experiment involving 128 senior engineering students across four programs. All identifying information was removed, and data sharing complies with institutional ethics approval and privacy guidelines. The dataset is intended to support reproducibility, secondary analysis, and educational research on AI-assisted learning and prompting behavior."
提供机构：
IEEE DataPort
创建时间：
2025-07-23
