Zhang, Yudong's dataset from Harvard Dataverse contains data from three scenario-based experiments investigating user tolerance toward AI errors. The experiments cover commercial shopping, academic tutoring, and medical consultation domains. The dataset was last updated on May 12, 2026.
Use Cases
- Compare error tolerance mechanisms between commercial, academic, and medical AI contexts based on the described scenario experiments.
- Model user acceptance thresholds for AI errors based on task criticality implied by the described domains.
- Design AI system interfaces that adapt to user error expectations based on the described experimental findings.
Strengths
- Data originates from three distinct scenario-based experiments, suggesting a multi-context design.
- Experiments cover three domains: commercial shopping, academic tutoring, and medical consultation.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- Harvard Dataverse
- Collection Method
- Scenario-based experiments
- Freshness
- Last updated 2026-05-12 01:07:31; freshness should be verified.