Aggregate Survey Results for Patient Information Sources and Trust in Aesthetic Medicine
Description
This dataset contains aggregate results from a seven-item anonymous electronic survey of 50 adult patients at an aesthetic medicine practice in South Carolina during July 2026. Data are provided as one-way frequency tables with counts and percentages for every response option. The package was reconstructed from Tables 1–7 of the associated manuscript because the original response-level export was unavailable. No participant-level, simulated, imputed, or AI-generated research records are included. The files contain no direct identifiers, free-text responses, timestamps, or cross-tabulations. The package includes CSV and Excel data, a data dictionary, reconstructed survey documentation, validation results, and detailed provenance and limitations.
Files
Steps to reproduce
Open aggregate_survey_results.csv or the Aggregate Results sheet in aggregate_survey_results.xlsx. Group rows by question_id. For each of Q1–Q7, sum frequency_n to confirm a denominator of 50, sum percent to confirm 100%, and verify that percent = frequency_n / 50 × 100. Consult data_dictionary.csv, README.md, and provenance_scope_and_limitations.md for field definitions, provenance, and appropriate analytical limits. Because the original response-level export is unavailable, participant-level analyses and cross-tabulations cannot be reproduced from this package.
Institutions
- Walden UniversityMinnesota, Minneapolis