Dataset of SAG Learning Intervention Integrating Socratic Questioning, Agentic AI and Generative AI in Computing Literacy Course

A 12-week quasi-experiment was conducted to compare the effects of the SAG model and the pure GAI learning model on senior high students’ digital competence, critical thinking, AI usage behaviors, and classroom learning experience.Data Collection Period: March 2026 – July 2026 (a full semester, 12-week instructional intervention)Research Site: Public senior high school in Guangdong Province, ChinaParticipants: 145 Grade 10 students divided into two groupsSAG Experimental Group: 75 students GAI Control Group: 70 studentsCourse Context: Mandatory Computing Literacy course, 1 session per week (40 mins per session) in computer labs.

Curriculum covers information systems, networks, software, programming, cybersecurity, artificial intelligence, big data, etc.AI Tools Used: Agentic AI (WorkBuddy), Generative AI (Doubao)Research Design: Mixed-methods research (quantitative quasi-experiment + coding of AI interaction logs + semi-structured teacher interviews)Ethical Statement: The study was approved by the school ethics committee. Informed consent was obtained from all students, guardians, and the instructor.