Experimental Data on Virtual Agent Response Type and Loneliness

Binbin Sun's dataset from figshare, last updated March 18, 2026, contains experimental data on consumer perceptions of virtual agents. The data, approximately 1.0 MB in size, was generated from three scenario-based, single-factor between-subjects experiments. It examines how scripted versus natural response types influence usage willingness, mediated by trust and moderated by loneliness type.

Use Cases

• Analyzing the main effect of virtual agent response type on user willingness based on the described experimental manipulation.
• Investigating the mediating roles of cognitive and affective trust in human-AI interaction as outlined in the study methods.
• Modeling the moderating effect of state versus trait loneliness on user preferences for agent behavioral realism.
• Testing conditional mediation pathways where loneliness type influences trust and subsequent usage willingness.

Strengths

• Data is from three controlled, single-factor between-subjects experiments, suggesting a structured design.
• Analysis was conducted using SPSS 25.0 and PROCESS 4.1, indicating a quantitative, process-oriented approach.
• Dataset is shared under a CC-BY-4.0 license, permitting broad reuse with attribution.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• The 1.0 MB file size suggests a relatively small-scale experimental dataset.

Provenance

Source

Binbin Sun via figshare.

Collection Method

Collected via scenario-based experimental approach with three single-factor between-subjects experiments.

Freshness

Last updated 2026-03-18 07:46:07; freshness should be verified.