The dataset contains individual-level responses from a stated preference study of Polish residents concerning microbiological water quality in the coastal waters of the Gulf of Gdańsk and the Vistula Lagoon. The study examines how information about Escherichia coli, Vibrio species, enterococci, and cyanobacteria affects knowledge, attitudes, recreational behavior, and willingness to pay for measures reducing infection risk. Respondents were randomly assigned to information treatments and completed discrete choice experiments on recreation site choice and on public policy programs.
Five data files in XLSX format are included, each a single worksheet with one row per respondent and variable names in the first row:
- The file s1 holds 3312 screened-in recreational users surveyed between May and August 2024.
- The file s1_rejected holds the 8942 respondents screened out at that stage, with demographic and screening items only.
- The file s2 holds 2588 responses to a short follow-up fielded between July and August 2024.
- The file s3 holds 1593 responses to the third wave fielded between July and September 2025.
- The file s4 holds 1111 responses from an independent general population sample fielded between September and October 2025, replicating the third wave.
Files s1, s2, and s3 cover the same panel of respondents and are merged on the variable ID. File s4 is a separate sample of different people whose identifiers form an independent sequence and must not be merged with the other files.
Four ODT files provide the full survey instruments in English, each serving as the codebook for its data file. Variable names appear in square brackets before each question and response codes follow the numbering of the answer options. Appendices document identifiers, weights, experimental condition assignments, the naming convention of the choice experiment design columns, and the variables withheld from this release.
Direct identifiers, open-ended answers and health data within the meaning of Article 9 GDPR were removed before publication. Affected values carry the string anonymized. See the README file for the full structure of the dataset, the anonymization record, and known data issues.
FUNDING AND ACKNOWLEDGEMENTS
--------------------------------------------------------------------------------
The authors acknowledge funding from the European Union's Horizon Europe research and innovation program under grant agreement no. 101057764 (BlueAdapt project). The views expressed are solely those of the authors and do not represent an official view of their organizations or granting authorities. KS, MC, and WZ gratefully acknowledge the support of the National Science Centre of Poland (Opus, 2023/49/B/HS4/01655; Preludium, 2023/49/N/HS4/02035; Preludium, 2024/53/N/HS4/04303). WZ gratefully acknowledges the support of the Foundation for Polish Science (FNP START). These grants did not fund the data collection reported here. They support related work by the authors.
