Surveys and Questionnaires

Surveys are used to gather views and comments from a sample of local people or service users on a range of issues. You can use surveys to find out what people think about a particular service, including proposed changes, and how services can be improved. A questionnaire-based survey can be used in a variety of formats, for example postal, online, face-to-face and telephone interviews.

Download a print versionPDF document

Please see also the section on Electronic Questionnaires.

How to do it

Planning

  • First, ask yourself: What do I want to find out? What do I do with the information? Is this the best way to collect it?
  • Before starting any survey make sure you have the necessary level of expertise, either yourself or available within your organisation, to undertake this kind of activity. NHS Boards' Clinical Governance or Research and Development Departments may offer advice on surveys and developing questionnaires.
  • Identify a clear and specific objective. This may focus on a particular geographical area, client group, or one aspect of a specialist service, or it might be more general.
  • Check whose approval you need to undertake the survey. Is it for audit, verification or research purposes? If it is for research purposes you must submit the project to the Research and Development Department in your NHS Board. If it is for audit or service evaluation purposes, this is not necessary.
  • Ask yourself: Is a questionnaire the best way to collect the information I need to address my objective? If you do not know (or cannot guess) the kinds of answers you are likely to get or how you will use these answers, a survey may not be the best tool, in which case interviews or focus groups may better suit your needs.
  • Have you developed a plan for your survey? You should decide on a timeline for the project, ensuring you have included time to: get permission to undertake it (when necessary); pilot it; send it out and wait for responses; follow up non-responders; analyse it; write up the findings.  The survey plan should also include any funding you may need.
  • Find out a little about the survey's subject. This is known as a literature search and may involve referring to relevant documents, policies and guidelines.
  • Find out whether a similar survey has been carried out locally or a similar piece of work is already planned or even underway elsewhere. A literature search may provide some of this information. It is perfectly acceptable to refer to work from other authoritative sources, or even to base the design of your work on previous work, as you may wish to compare your findings with those of others. You could also refer to their plan and tools. Any work that you use will require permission to be granted for its use and any extract or references should be clearly sourced.
  • Think about the type of survey that will be appropriate – postal, online, face-to-face or telephone interview.
  • How many people are you going to survey? If you are not going to survey everyone (e.g. all service users or everyone involved in a service change consultation) how are you going to identify your sample? Decide on the people you would like to collect information from (this is known as your sample) and try to ensure that it is representative of the population you are studying.
  • Consider whether your survey has any ethical implications, for example will you be obtaining people’s personal details? You should refer to Section 5 regarding when to seek ethical approval.
  • Consider the type of questions you would like to ask:
    • Qualitative, known as 'open-ended', questions allow the respondent to write his or her own reply to a question.
    • Quantitative, known as 'closed', questions require the respondent to choose from a set list of responses, e.g. by ticking yes/no boxes.
  • Think about how the information will be analysed and seek advice on the use of spreadsheets and statistical packages which can be used to carry out analyses.
  • You must include information on who is carrying out the survey, why and how the information will be used. Contact details of the person conducting the survey must be included and participants should be offered the opportunity to request a copy of any report produced.
  • Be clear about your Data Protection Act (1998) responsibilities. Once information has been recorded and analysed, you should destroy any that identifies respondents unless they have given you permission to keep their information on file, where it should be kept securely. To offer reassurance to respondents, information about data protection and confidentiality could be put in the covering letter and included in the preamble of face-to-face and telephone interviews. Refer to your Board's Data Protection and information governance policies for further information.

Developing Questions

  • Involve patients, carers and service users in designing the questions.
  • Aim to keep the questionnaire short and try not to go over one or two pages.
  • Try to keep questions short and simple, preferably with 25 words or fewer. This will also help ensure that the information will be simple to analyse.
  • Ask one question at a time. The question "Are staff helpful and courteous?" is actually two questions in one. Staff may be helpful but not necessarily courteous.
  • Group questions so that themes flow through in a logical way.
  • Avoid using leading questions which force or imply certain types of answer. An example of this is: "Do you think that access to services is limited because of staff attitudes?"
  • It is easier to analyse responses if you use categories. Use tick scales such as Agree, Don't know, Disagree; or Last Week, Last Month, Last Year.
  • Once you have produced a draft questionnaire, conduct a small pilot 'trial survey' to pick up any mistakes or ambiguities in the questionnaire before the survey is commenced. Revise the questionnaire.
  • Include a preamble to explain the purpose of the survey.
  • If you have decided on a postal or online questionnaire:
    • Provide details of how to return the questionnaire, the date by which it should be returned and how people can request assistance to complete the questionnaire.  Ask if they would like to receive the questionnaire in another format. This information should be in the preamble and should be repeated on the questionnaire itself.
    • Consider the relative merits of postal or online questionnaires. Electronic surveys are good if all your potential respondents have easy and regular access to a computer. If they don't, a postal survey might be better. The online survey services are good for short, straightforward surveys. For complex analyses, e.g. those involving many subgroups, other methods may be more useful.
    • Questions that ask people to specify their age, ethnic group, gender, geographic area and religion should be added towards the end and should clearly state that they are not mandatory, i.e. respondents may choose not to answer them.

Next Steps

  • When looking at the information you have collected (known as data analysis), look to see whether respondents have left any questions blank. Identify areas where respondents have indicated that improvement is needed and where the majority of respondents have indicated 'unsatisfactory'. Likewise, identify areas that are doing well, which may be indicated by consistently high marks. Look for a unified feeling about a certain topic among respondents of similar age, ethnic group, gender, geographic area and religion.
  • Produce a report (there is a section in this toolkit specifically on preparing a report of findings)and evaluate how well the survey went. Remember to include any problems you experienced while undertaking your survey, e.g. problems with your sample; or if you realised a question had been poorly written; or a low response rate.
  • Develop an action plan and provide feedback to the people who took part.

Pros

  • Information can be collected from a large group of people.
  • A representative sample of the population may be reached.
  • All respondents are asked the same questions.
  • Postal and online questionnaires can be completed in peoples' own time.
  • Face-to-face interviews, using questionnaires, are more flexible to the needs and difficulties of the chosen sample and allow the interviewer to 'probe' for further information.
  • Telephone interviews are faster and cheaper than face-to-face interviews.

Cons

  • If this tool is incorrectly used, the correct information will not be gathered.
  • When using a postal or online questionnaire, there is no opportunity to clarify what a question means.
  • Postal or online questionnaires usually have a low response rate.
  • There is a risk of excluding people with language and literacy issues.
  • Telephone interviews are more intrusive, likely to have a lower response and are limited to respondents having telephones.
  • Face-to-face interviews using questionnaires can be lengthy and costly.

Resources

  • It is expensive to commission an external organisation to conduct a large survey. It is less expensive if you are able to manage the survey yourself, although this needs a significant allocation of staff time.
  • Stationery and stamped addressed envelopes can be costly. Stamped addressed envelopes, however, may help to increase the response rate.
  • Telephone or face-to-face interviews may have costs associated with telephone calls, room hire and interviewers' pay. Face-to-face interviews require skilled interviewers.
  • Web-based services will be relatively inexpensive. Use of other software packages may incur further costs in terms of some combination of software costs, consultancy and/or staff time.
  • You may wish to offer incentives to respondents, if appropriate.

Top Tips

  • Give your questionnaire a short and meaningful title.
  • It is easy to ask too many questions, some of which you may not need. Always try to reduce the number of questions asked to a minimum.
  • In developing questions based on a set of options, it is sometimes advised that survey designers should set out an even number of options because otherwise there is a tendency for respondents to tick the middle box.
  • Use everyday language when devising questions and ensure your language caters for all levels of literacy. For some respondents, English may not be their first language.
  • Take care when covering sensitive issues. Try to explain why you need to know that information.
  • Try to use a font size of 14, ideally Arial or Comic Sans, and leave space for responses. Print questionnaires on white or yellow paper, with black text.
  • Include a stamped, self-addressed or pre-paid envelope to encourage the return of completed questionnaires.

Sources and Further Information

This information has been informed by the work of Link opens in a new windowLoughborough University Library, the Link opens in a new windowAssociation of Community Health Councils for England and Wales and Oppenheim. Their work is gratefully acknowledged.

  • Association of Community Health Councils for England and Wales (1996) Health News Briefing, Guidelines on Community Health Council Surveys and Research.
  • Link opens in a new windowFife Council's note on designing a survey or questionnaire
  • Link opens in a new windowHow to Analyze Survey Results
  • Oppenheim, A.N. (1992) Questionnaire Design, Interviewing and Attitude Measurement. London: Continuum.
  • Saris, Willem E. & Gallhofer, Irmtraud N. (2007) Design, evaluation, and analysis of questionnaires for survey research. New Jersey: Wiley-Blackwell