AI Metaphor Framing and Learners' Epistemic Judgments: Anonymized Data, Materials, and Analysis Outputs

This anonymized dataset contains participant-level and trial-level data, task materials, a variable dictionary, analysis syntax, reproducible analysis outputs, and figures for a four-arm learning experiment on AI metaphor framing, knowledge authority judgment, critical distance, reliance calibration, and learning outcomes. The public package is prepared for double-anonymous review. It excludes names, student IDs, email addresses, contact records, institutional certificates, author names, affiliations, ORCID identifiers, local workflow files, submission-system records, Git metadata, and the original participant ID mapping.