AI Health Assistant Communication Strategies and Persuasion Outcomes

367 Chinese university students, 84.5% with myopia, participated in a quasi-experimental study comparing three AI communication archetypes. The data, authored by Chaoyuan Zuo and last updated on 2026-05-20, measures outcomes across Information Processing, Relational Engagement, and Action Empowerment dimensions. It supports analysis of how AI communication style influences health persuasion.

Use Cases

• Compare the effectiveness of empathetic versus authoritative AI communication on relationship quality based on the HEAT framework dimensions.
• Analyze the predictive power of AI relationship quality versus knowledge acquisition on behavioral intention based on hierarchical regression results.
• Investigate the differential impact of gain-framed, loss-framed, and balanced-framed messaging in a health education context.

Strengths

• Dataset is based on a study with 367 participants, providing a substantive sample size for analysis.
• Includes specific statistical results, such as effect sizes (d_z > 1.0 for knowledge gains) and regression coefficients (β = 0.346 for relationship quality).
• Released under a permissive CC-BY-4.0 license, allowing for broad reuse and sharing.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (36.9 KB), indicating limited scope, likely containing summary results rather than raw participant-level data.

Provenance

Source

Chaoyuan Zuo via figshare.

Collection Method

Pre-post quasi-experimental design where participants interacted with one of three AI health assistant archetypes.

Freshness

Last updated 2026-05-20 05:51:47; freshness should be verified.

Geography

Study conducted with Chinese university students.