Interview Dataset on Middle School Students' Experiences with Large Language Models (LLMs)

Data were collected between May and June 2025 to investigate the current status, practical logic, psychological perceptions, and impacts of LLMs on students in real-life academic and daily scenarios. The interviews cover topics including usage motivation, academic application scenarios, human-AI collaboration strategies, cognitive evaluation of LLMs, emotional experiences, and the moderating effects of social environments such as family, school, and peers.

The raw audio recordings were transcribed, resulting in over 80,000 Chinese characters of verbatim text, which has been structurally pre-processed into approximately 1,200 independent analytical units. This dataset provides rich qualitative material for understanding the interaction between digital natives and artificial intelligence, offering significant value for research in adolescent developmental psychology, educational psychology, human-computer interaction, and AI literacy.