Raw data for “User Trust Formation in Mental-Health Virtual Humans among Socially Anxious College Students: A Mixed-Methods Study”

Published: 22 June 2026| Version 1 | DOI: 10.17632/6rwjgdgtdp.1
Contributors:
,
,
,
,

Description

This dataset contains research data and supplementary materials associated with the manuscript entitled “User Trust Formation in Mental-Health Virtual Humans among Socially Anxious College Students: A Mixed-Methods Study.” The study investigates how socially anxious college students form trust in mental-health virtual human systems. It adopts an exploratory sequential mixed-methods design integrating grounded theory, structural equation modeling (SEM), common method bias testing, design evaluation, predictive evaluation, and XGBoost-SHAP analysis. The dataset includes the following materials: Grounded Theory Research Text Data, including anonymized interview transcripts, literature content data, and online review data used for the qualitative coding process. Design Evaluation Sample Data, used to support the evaluation of system design-related factors. Formal Survey Sample Data, used for structural equation modeling and hypothesis testing. Predictive Evaluation Sample Data, used for predictive modeling and XGBoost-SHAP analysis. Raw Data for Common Method Bias Test, used to examine potential common method bias in the questionnaire data. These materials support the identification of trust-related factors, the validation of relationships among system design features, user-perceived factors, and user trust, and the predictive explanation of key variables influencing user trust in mental-health virtual human systems. All shared data have been anonymized or aggregated where necessary to protect participant confidentiality and privacy. The dataset does not include personally identifiable information. The materials are provided to support research transparency, methodological reproducibility, and secondary academic analysis.

Files

Categories

Artificial Intelligence, Mental Health, Human-Computer Interaction

Licence