Define the need before the questionnaire.
Write a concrete research question: “Which spreadsheet tasks make work difficult for the people we want to train, and what prevents them from attending?” This asks about experience and access rather than simply demand for a course.
Define the people whose views matter. Include those who have not attended previous activities if their barriers are part of the question. A convenient list of former participants may miss the people most affected.
A practical sequence.
- Name the decision
Will the evidence change content, delivery, timing, access support or whether training is the right response at all?
- Choose the target group
Describe who is included, how you can reach them and who the recruitment route may miss.
- Ask about actual experience
Use a time frame and specific tasks. Separate a recent difficulty from a future preference.
- Ask about participation conditions
Explore language, time, cost, transport, equipment and accessibility without presuming the same obstacle for everyone.
- Pilot the wording and flow
Check the draft with people from the target group. Test options, mobile screens and the paths for participants with no relevant experience.
- Read the answer in context
Show usable response counts and distributions. Compare meaningful groups and read open answers alongside the closed questions.
- Turn evidence into a decision
Explain the planned response, the need it addresses and what remains uncertain. Share an appropriate summary with participants or partners.
An example that changes the project plan.
Participants may report that filtering spreadsheet data is difficult. They may also say that live sessions are impossible during working hours. The resulting decision has two parts: teach a specific task and change how the activity is offered.
Some may prefer a reusable file or access to a support person rather than training. The survey should leave room for that answer. A course preference alone does not demonstrate a skill deficit.
What the result cannot establish.
A self-reported confidence score is not a tested skill. An open-link sample is not automatically representative. Stated interest is not confirmed attendance, and a high priority does not by itself prove that training will solve the difficulty.
Set it up in Nuvopoll.
Describe the target group and the decision to the AI survey generator, or start with the 12-question training needs example. Review the questions and routes in the editor.
Provide appropriate language versions, recruit through suitable channels and prepare a live report that shows the participant base and limitations. Keep personal-data collection proportionate to the study.