Student Evaluative Profiles of Generative AI in Higher Education: Trust, Ethical Acceptability, Perceived Learning Value, Familiarity, and Academic-Use Frequency — Dataset and Supplementary Materials

This record contains the dataset and supplementary materials supporting the manuscript “Student Evaluative Profiles of Generative AI in Higher Education: Trust, Ethical Acceptability, Perceived Learning Value, Familiarity, and Academic-Use Frequency.” The study examines how three evaluative dimensions of academic generative-AI use — Trust and Reliability, Ethical Acceptability, and Perceived Learning Value — combine within students, and whether the resulting evaluative profiles differ in self-reported GenAI familiarity and frequency of academic use. Profile formation is based only on the three evaluative dimensions; familiarity and academic-use frequency are examined subsequently as external characteristics. The public analytic dataset contains 173 undergraduate respondents who provided affirmative informed consent. Five survey records without affirmative consent were excluded from the public dataset and from all analyses. The repository includes:S1 — English version of the questionnaire, including response options and scoring structure;S2 — de-identified participant-level analytic dataset for the 173 consenting respondents;S3 — Pearson correlation matrix for the study indicators;S4 — internal-validity diagnostics for candidate k-means solutions (k = 2–5), including average silhouette width, Calinski–Harabasz, Davies–Bouldin, and cluster sizes;S5 — reproducible R analysis script covering data checks, composite construction, descriptive statistics, reliability, correlations, candidate cluster diagnostics, the two- and three-cluster solutions, and external profile characterisation by GenAI familiarity and academic-use frequency. The materials are provided to support transparency, reproducibility, and direct or conceptual replication across cohorts, disciplines, institutions, and stages of GenAI adoption.