A useful branch begins with a reason.
A support-service evaluation should ask whether the participant used the service before asking for a rating. A training needs survey may ask about a missed opportunity before asking what prevented attendance. These branches protect both the participant experience and the meaning of the data.
A long survey does not automatically need complicated flow. If every participant can answer the same short set, branching may add testing work without helping the research.
Two answers, two paths.
| Eligibility question | Next step |
|---|---|
| “In the last six months, was there training you wanted to attend but could not?” — Yes | Ask about the main barrier at the most recent missed training. |
| No or cannot remember | Skip the past-barrier question and ask about future participation conditions. |
The two questions are not interchangeable: one concerns an actual experience and the other a possible future barrier. Keep the distinction when analysing the responses.
Make the decisions before drawing the flow.
- Define the answer that controls the path and the question it affects.
- Decide what happens after “do not know”, refusal and missing answers.
- Keep eligibility, quota-full and completion routes distinct.
- Check that language versions follow the same underlying logic.
- Avoid making a skipped question a required question on another screen.
Pilot every route, not only the first one.
Walk through each answer combination that changes a path. Include someone with no relevant experience, someone who refuses and someone who is ineligible. Check on a phone as well as a desktop.
In Nuvopoll, the flow editor helps inspect the questionnaire’s paths. The SPSS export distinguishes −99 (not shown by flow), −98 (shown but unanswered) and −97 (declined). A skipped rating should not become a zero score.
Read results using the people who could answer.
A support rating applies to support users, not every customer. Report the base of that question and preserve the route that produced it. Conditional flow improves relevance; it does not correct recruitment bias.