Survey data and analysis scripts on privacy–utility trade-offs (N = 1,275)

Anonymised responses from a survey of 1,275 Vietnamese users on the privacy–utility trade-off, collected via Google Forms between 11 and 20 April 2026, together with the questionnaire exactly as administered and the Python scripts that reproduce every statistic, table and figure in the accompanying paper. The instrument presented three simulated security scenarios (Low, Medium, High), each followed by the same seven data-type willingness items on a five-point scale, plus items on AI-risk concern, acceptable percentage of utility sacrifice, objective technical knowledge, and attitudes to responsibility and pricing. The deposit includes the five data-quality diagnostics reported in the paper: test–retest agreement across repeated administrations of identical items, exact-duplicate detection, Cronbach's alpha, goodness-of-fit against uniform marginals, and a cross-item consistency check. These diagnostics indicate the multi-item willingness scale did not measure a stable individual disposition in this administration; anyone reusing the data should read Section 3.2 of the paper first. No direct identifiers were collected. Google Forms was configured not to record e-mail addresses, all items are closed-ended, and the only non-response column is a submission timestamp, retained because the duplicate-record diagnostic depends on it.