Questionnaire Survey on Perceived Flood Risk

# Data description

Anonymized survey data and questionnaire for a study of household flood experience, perceived pluvial flood risk, and stated support for local flood-risk reduction measures in the Copenhagen area. The dataset contains 2,565 respondent-level observations from two survey versions and has been generalized and suppressed to reduce re-identification risk before public sharing.

This upload contains anonymized respondent-level survey data and the survey instrument used in a study of household perceptions of pluvial flood risk, flood experience, and willingness to support local flood-risk reduction measures in the Copenhagen area.

The dataset is provided as a comma-separated values file:

- `survey_share_anonymized.csv`

The survey instrument is provided as:

- `Endelig_survey_(regn).pdf`

## Contents

`survey_share_anonymized.csv` contains 2,565 anonymized survey responses. The file combines two questionnaire versions that differed in the framing of the flood-risk reduction scenario:

- `25pct_reduction`: 1,301 respondents
- `50pct_reduction`: 1,264 respondents

Each row represents one respondent. `respondent_id` is a synthetic identifier created for data sharing and does not correspond to any original personal identifier.

## Substantive coverage

The survey includes variables on:

- prior experience with flooding from heavy rainfall
- perceived flood risk and information seeking
- attitudes toward public flood protection
- stated voting responses to hypothetical flood-risk reduction programs with different payment levels
- self-reported household adaptation measures
- health and stress-related outcomes
- basic demographic and household characteristics

The questionnaire text and response options are documented in `Endelig_survey_(regn).pdf`.

## File structure

The anonymized data file retains the original survey variable names used in the research workflow. Important added variables are:

- `respondent_id`: synthetic respondent identifier
- `scenario_group`: indicates whether the respondent was in the 25% or 50% risk-reduction scenario version
- `birth_cohort`: grouped version of birth year
- `movein_period`: grouped version of year of moving into the current dwelling
- `household_adults_group`: grouped household adult count
- `household_children_group`: grouped household child count
- `income_group`: grouped household income band
- `s_54_grouped`: grouped version of reported sick days
- `s_55_grouped`: grouped version of contact frequency with a general practitioner

The file also includes original coded survey items and derived analysis variables such as bid levels and flood-experience indicators used in the empirical analysis.

## Anonymization

The shared file is an anonymized version prepared for public release. Direct identifiers and high-risk indirect identifiers were removed. In particular, the shared dataset excludes:

- personal identifier fields
- open-text responses
- exact birth year
- exact move-in year
- an additional property-risk field judged to increase re-identification risk

To reduce disclosure risk further, selected demographic variables were generalized into grouped categories. Rare combinations of quasi-identifiers were then suppressed. Under the applied rule, there are no remaining nonblank combinations with fewer than 5 respondents for the anonymity key based on:

- `birth_cohort`
- `movein_period`
- `household_adults_group`
- `household_children_group`
- `income_group`
- `flood_dummy`

## Reuse note

Users should consult the questionnaire PDF together with the dataset, as the variable names in the CSV follow the original survey coding rather than descriptive English labels. Interpretation of individual items should therefore be based on the questionnaire and, where relevant, the accompanying study documentation.