A safety survey should measure a defined decision
Begin with what the group will decide from the results: where to conduct a walk, which maintenance issue to prioritize, when a service is needed, or whether a change improved use. A broad question such as “Is the neighborhood safe?” combines different places, times, experiences, and meanings into one answer.
Separate experience from explanation
Ask about a place, time, and observable condition
Useful questions can ask whether a person used a route, encountered a failed light, could find help, avoided a location at a certain time, or experienced a specific event. Ask separately why they made a choice. Do not present a list that assumes strangers, youth, unhoused people, or another identity group causes danger.
When asking about incidents, state the period clearly and offer “prefer not to answer.” Avoid requesting names, exact addresses, or detailed victim information unless the project has a justified, protected process for those data.
Design the sample before reading the result
- Identify who uses the place and who is usually missed by online outreach.
- Offer accessible formats, languages, and non-digital options where feasible.
- Record how and when responses were collected.
- Do not describe a convenience sample as representative of every resident.
- Report response count, missing answers, and important limitations.
The CPTED neighborhood walk guide can test reported conditions onsite. The program evaluation guide explains baseline and outcome measures.
Publish findings without stigmatizing a place
Group results enough to protect privacy, especially for small areas or rare events. Report strengths and useful activity as well as problems. Distinguish perception, direct experience, administrative records, and site observations instead of combining them as one crime measure. Explain what action follows, who owns it, and when residents will hear the result.