Surveys are a key component of many research and evaluation projects, and are a common method for gathering information.

Surveys focus on quantitative data such as ‘how many?’ and ‘how often?’. Survey questions are structured to give you answers that you can easily analyse, using mostly closed questions such as scales or multiple choice answers. They are best suited for feedback and outcome data, but can be used for all types of data collection.




  • Reaching a large number of people
  • Providing data that is easily analysed
  • Allowing people to respond anonymously
  • Comparing before and after (‘pre and post’)


  • Persuading people to respond can be tricky, and there is a risk of questionnaire and form-filling overload
  • Difficult for people with literacy and numeracy issues, learning difficulties, or those for whom English is a second language
User data

Best Practice

You can often use - or borrow from - existing questionnaires designed by professional researchers, which will make your survey more robust. Read more about using validated tools here.

If you decide to design your own survey questions, the following tips will help you get the best data from them:

  • Be brief: remove any unnecessary words and try to avoid more than 20 words per question.
  • Use simple language: avoid complicated words or confusing language.
  • Be short: Keep the overall length of the survey to a maximum of 15 questions (5 minutes for online surveys, 10 minutes for paper/phone, 15 minutes for face-to-face). This is likely to give you a better response rate.
  • Be objective: avoid leading questions, eg. ‘how satisfied are you with the service?’
  • Be specific: avoid words that are open to interpretation, eg. use ‘daily’ or ‘weekly’ rather than ‘often’ or ‘usually’.
  • Ask one thing at a time, eg. avoid questions such “Did you find the session helpful and interesting?”--ask in two separate questions.
  • Watch out for double negatives, eg. “Do you agree or disagree that you no longer need support?”
  • Use your common sense: will the respondent understand the question?
  • Pilot the questions: Ask a small group of respondents to complete the survey before rolling it out to the wider group - check how the pilot group understood the questions and tweak them based on their responses.

For more detailed guidance on writing your own questionnaire, see this guidance by NCVO. Also explore this interactive module on survey data and collecting longitudinal data by UK Data Service

coffee-break-and-working left facing.png

Using online surveys

People can complete your questionnaire on paper or face to face, but online surveys are a good option and many will also do simple analysis of the results. There are lots of free online survey tools available - this guide by M&E Consulting gives an overview of the pros and cons of some of the options available.

Online surveys don’t suit everyone; surveys can also be administered in person or on the phone. Often you’ll use different media for the survey at the beginning and at the end of the project. If people struggle to complete the survey, can you provide support or is it a problem with the survey design? If they aren’t willing, try to gain their trust before asking them to respond to a survey, keep it short, and consider if it’s appropriate to use incentives.

Don't want to do another survey?

If you just want feedback, here are some alternative methods you could try.


Data Diagnostic

The data diagnostic asks 10 multiple choice questions about what your programme or service is, how it works and who it targets. It then provides a tailored report that discusses what kind of data you should consider collecting and how.

Take data Diagnostic

Next steps