Questionnaires are used for a variety of reasons – to gather information, to survey opinions or attitudes, to measure customer satisfaction, or to drive market research. But although sophisticated questionnaires usually require professional expertise to compile, administer, analyse, and interpret, you can construct simple instruments for surveys, interviews, and focus groups by following these guidelines…
1. Be clear about what you’re looking for.
Before you begin designing any questions, articulate clearly what need or problem you want to address using the information you intend to gather from the survey. Why are you doing the evaluation? What do you hope to accomplish? For example, questionnaires can measure attitudes, market trends, consumption patterns, beliefs or expectations about your services or products, the effect of competition, media, etc. Having a clear focus will help you frame your questions.
2. Select the appropriate method.
How will you administer your questionnaire – by post, telephone, fax, personal interview, the Web, e-mail? Each method has disadvantages and advantages relating to issues such as speed, cost, ability to reach a scattered sample, response rate, and interviewer bias. A combination of methods may be considered, such as using the post to elicit basic data, and following up with phone interviews to explore some issues in greater depth.
3. Select and define your sample.
To whom will the questionnaire be administered? Whose views are you seeking? Consider clients, customers and non-customers, age, sex, socio-economic groupings, and the sources of names, addresses and telephone numbers. Will you be sampling randomly or targeting a specific group or strata?
4. Know how you will handle the results.
Questionnaires can generate a great deal of paper and even more data. Determine in advance who will do the work, how the information will be sorted, how it will be analysed, how the responses will be used. By rightly focusing on these issues up-front, you will be forced to put more thought into the construction of the instrument.
5. Construct your questions carefully.
Consider the variety of ways questions can be posed – closed questions where respondents simply ‘tick a box’ or ‘circle a number’; open-ended questions where respondents are asked to write the answer in their own words; attitude questions where respondents are asked to mark their response on a 5-point Likert-type scale (ranging from ‘strongly agree’ to ‘strongly disagree’).
Consider the following advice:
- Begin with a couple of fact-based questions (e.g. demographic data), followed by general, easy questions to get the ball rolling. Move from the general to the specific.
- Sensitive questions should be asked towards the end of the questionnaire.
- Follow a logical sequence to avoid confusing the respondents.
- Limit the number of questions or the respondents may be put off answering.
- Attempt to get also the respondents’ comments or perspectives to supplement their ticked-box answers.
- Make sure you don’t put ideas in early questions which may influence respondents’ answers later.
- Consider concluding with a question which seeks the respondent’s view of the questionnaire itself and any suggestions for improvement.
- If using multiple-choice questions, make sure the choices are mutually exclusive and embrace the total range of answers. Ensure that respondents won’t have a clearly preferred answer that is not among your alternatives.
- As a general rule, keep it short, to the point, and easy to complete.
Questions to avoid are those that:
- are too difficult to answer
- assume the respondents have certain information – but they don’t
- do not provide enough information for the respondents to answer intelligently
- point respondents to the answer desired by the researcher
- require the respondents to put more effort into answering than they are willing to give
- are too confusing to answer, e.g. use of the word ‘not’ can cause confusion when a yes or no response is wanted.
- are emotion-laden, e.g. use of the phrase ‘our prompt, reliable service’
- use slang, cultural-specific, or technical words
- are, in reality, two different questions in one.
6. Pay attention to the questionnaire’s design.
An attractive layout for a printed questionnaire can have a major influence on the response rate. Focus on legibility of the typeface used, its size, generous use of white space, and, importantly, allow sufficient room for recording responses.
7. Give clear directions.
Include a brief explanation of the purpose of the questionnaire, how to complete the questionnaire, how to return the survey forms, and any details relating to confidentiality of the information or the respondent. And don’t say it will take 5 minutes when it will take 15. Finally, remember to thank the respondents for their time and assistance.
8. Conduct a pilot.
Test your draft questionnaire on a small group of staff or clients, to check on the clarity of the questions, their sequencing, the layout, the time taken to respond, and whether you get the type of data you really were seeking. Use this information to develop the final version of your questionnaire.