AI Literacy Intervention Experiments on Hallucination and Privacy Risks

Two experiments with 304 participants on AI hallucination and 314 participants on privacy risks test three AI literacy message strategies. The dataset likely contains results from these experiments, including mediation analyses on norms, self-efficacy, and perceived severity. It was authored by Chunsik Lee and last updated on May 5, 2026.

Use Cases

• Analyzing the effectiveness of norm-focused messages compared to nudges and factual information based on the described experiments
• Studying mediators like injunctive norms and self-efficacy across risk domains based on mediation analyses
• Investigating the relationship between perceived severity and protective intentions for AI hallucination based on findings

Strengths

• Includes results from two distinct experiments with participant counts of 304 and 314
• Mediation analyses provide insights into psychological mechanisms behind behavioral intentions

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment

Provenance

Source

figshare

Collection Method

Likely gathered from controlled experiments testing message strategies.

Freshness

Last updated 2026-05-05 15:36:47; freshness should be verified