How to design a survey / questionnaire
Surveys and questionnaires are a quick way to gather information and assess attitudes and opinions. However, don’t be fooled by their apparent simplicity – good design is essential. Here are our tips (although please note, we are not experts in qualitative research).
Does an appropriate validated questionnaire/survey already exist?
If not and you need to design your own, consider carefully what you’re trying to ask about and why. What information will you require to ultimately act on/present/publish the results? Then work backwards to ensure you collect all of this information in an appropriate format:
- free text?
- closed questions; single best answer?
- closed questions; multiple answers permitted (note that this will affect ‘totals’ at analysis)?
- graded scales?
Decide whether you will create a paper survey/questionnaire or use an online platform e.g. SurveyMonkey (a free account is available or your University may have a more comprehensive, paid-for account).
For paper-based surveys/questionnaires consider:
- the cost of printing and postage
- how respondents will return the survey/questionnaire (e.g. pre-addressed postage paid envelopes or secure deposit boxes)
- how you will record the survey / questionnaire results e.g. by hand onto a database
For online surveys/questionnaires consider:
- whether your target audience is likely to have access to a computer/the internet (if not, this may restrict participation)
- how you will access email addresses
- how you will securely store email addresses
Design the survey/questionnaire so that it is streamlined i.e.
- if you use an online platform: the responses given dictate subsequent questions so that respondents only answer questions which are relevant to them
- if you use a paper format: instructions are given for respondents to skip to the next relevant question
Pay attention to format:
- people often skim-read so ensure questions are concise and clear, leave no room for obfuscation
- consider including your institutional logo
Decide the minimum amount of subject-specific information you will need to analyse the results e.g. age, gender, location, training grade etc.
Carefully word the title of the survey/questionnaire and the introduction – an inflammatory title or lengthy/unclear introduction may put people off.
Test the survey/questionnaire (using a test account if designed using an online platform) – do the various pathways work? Pilot the survey on peers/colleagues/lay or patient representatives or PPI groups .
Consider whether you will offer an incentive to complete the survey/questionnaire – e.g. a prize draw.
Consider whether you need to apply for ethical approval to issue the survey/questionnaire (NB you may need University approval to distribute a survey/questionnaire to undergraduates).
Consider how you will promote and distribute the survey/questionnaire. Will you include a cover letter/email?
Decide how long you will keep the survey/questionnaire open for and whether you will issue reminders.
When analysing the results, consider what the uptake was and how representative the respondents were.
How will you distribute the results to ensure maximum impact?
Will you thank the respondents and send them a summary of the results?