Begin with the decision the survey serves
Write down what you want to learn and how the answer will be used. Then decide whether a survey is appropriate or whether documents, interviews or direct observation would address the question more clearly. A list of interesting questions is not yet a research design. If people will participate, consult your supervisor and the relevant ethics process before collecting responses. This article offers planning questions, not a validated instrument or permission to conduct a study.
Ask one thing at a time
Consider the hypothetical item: how satisfied are you with teaching and administrative support? A respondent may have different experiences of those services, making one answer difficult to interpret. Split the topics and define the period or situation where needed. Avoid wording that assumes an experience occurred. If a respondent has never used a service, an appropriate route may be different from a satisfaction rating; the design should make that distinction explicit.
Test the response options
Check whether the available options match the question, cover relevant cases and avoid forcing a false answer. Define unfamiliar terms and use the same meaning consistently. A pilot with appropriate participants can help reveal misunderstandings, but a small pilot does not establish that an instrument is valid for every purpose. Follow any required permissions when adapting an existing questionnaire. Keep a record of changes so that later readers know which version participants actually saw.
Plan the limits of interpretation
Before collecting responses, identify who can be reached, who may be excluded and what the recruitment route could mean for the results. Responses from a convenient group should not automatically be presented as the views of every student. Record missing answers and relevant context rather than treating an incomplete response as proof of a particular opinion. Ask for methodological guidance when estimating a sample or making statistical claims; those decisions require more than a generic checklist.
Write the report before the launch
Draft the headings of the eventual report: purpose, participants, recruitment, questions, analysis and limitations. If you cannot explain how an item contributes to that report, reconsider including it. Collect only information justified by the approved purpose and arrangements. When reporting, show the wording needed to understand a result and distinguish description from explanation. A useful survey report lets readers see how the answers were produced, rather than presenting numbers detached from their questions.
A practical exercise
Draft three possible survey items on a fictional study-support service. For each item, write what each answer would allow you to say and one conclusion it would not justify. Include a route for a respondent with no experience of the service where appropriate. Next look for double questions, unclear periods and assumptions hidden in the wording. Do not distribute the draft as research until the relevant supervision and approvals are in place. The useful output is a set of explicit design questions, not a claim that you have created a reliable or validated questionnaire.
