Embodied Conversational Agents for Role-Playing Alcohol Peer Pressure in Virtual Reality

A study by Simon Langener, uploaded to figshare in 2026, evaluating the use of Embodied Conversational Agents (ECAs) in Immersive Virtual Reality to simulate peer pressure to drink alcohol. The dataset, 120.1 MB in size, contains results from a repeated measures experiment with twenty patients with Mild to Borderline Intellectual Disability and Alcohol Use Disorder. It assesses the ECA's persuasiveness and effects on perception, emotional state, and coping behavior using actor-recorded dialogues in dominant-friendly and dominant-hostile styles.

Use Cases

• Training machine learning models to recognize and classify peer pressure dialogue styles based on the described dominant-friendly and dominant-hostile scripts.
• Analyzing the correlation between virtual agent behavior and patient-reported outcomes like alcohol craving and anxiety scores.
• Studying the impact of ECA design features, such as paralanguage and group dynamics, on user perception and engagement in therapeutic role-plays.

Strengths

• Dataset is based on a repeated measures experiment with twenty patients, providing a controlled study design.
• The ECA system was co-designed with therapists, suggesting clinical relevance.
• The 120.1 MB file size indicates the dataset likely contains detailed documentation or supplementary materials.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The sample size of twenty patients may limit the generalizability of findings.

Provenance

Source

figshare

Collection Method

Data was generated from a repeated measures experiment evaluating an Embodied Conversational Agent in Immersive Virtual Reality.

Freshness

Last updated 2026-04-20 10:31:37; freshness should be verified.