Dataset for "Correcting or Confirming? A Pilot Study of LLM Responses to Simulated Student Pressure in Higher Education"
This dataset accompanies the study “Correcting or Confirming? A Pilot Study of LLM Responses to Simulated Student Pressure in Higher Education”. The workbook documents 60 conversations and 120 model responses collected from ChatGPT, Gemini, and Claude. The protocol comprises ten English-language academic scenarios, two...