GAITA_proactive support_study

Description

This deidentified dataset contains 228 session-level observations from 57 university students who completed four online learning sessions under four proactive support modes of a generative AI teaching assistant: none, notification, suggestion, and intervention. It includes participant characteristics, experimental conditions, learning performance, interaction behaviors, and eye-tracking measures, including attention allocation, gaze entropy, and transitions among three areas of interest. The dataset supports research on proactive AI support, human–AI interaction, and technology-enhanced learning.

Institutions

• Tsinghua University

  Beijing, Beijing

Categories

Licence