Questionnaires to gather information and ideas

All these factors are significant:

• the design
• the sample selection
• the way the questions are asked
• the way the respondent interprets the questions
• the way the reply is recorded
• the analysis of the responses.

Before you start, clarify what it is you need to know and why. Check that a questionnaire is the most effective way of getting the information you want. Involve research and evaluation staff from the earliest planning stage.

There are two basic methods for administering a questionnaire: self-administered, or interviews undertaken face-to-face or by telephone.

Use a proper sampling technique to select your questionnaire respondents so that they are representative of the relevant population.

Test the questionnaire first. When you pilot a questionnaire with people similar to your survey population, this helps ensure the clarity and success of the questionnaire.

Before finalising a questionnaire, decide how you are going to process and record the responses.

Wording of questions is crucial

Ask only those questions that are absolutely necessary. If the questionnaire is too long, people will be reluctant to answer it.

Make sure the questions fit the task. If you are trying to find out how effective a policy has been, look at the original aims for the policy and break those down into indicators that would measure success in achieving those aims.

When deciding what questions to ask, you should be constantly asking, ‘How will I use this information?’ This helps to ensure every question is relevant and necessary.

Are you likely to get the most useful response by using multiple-choice questions or open questions? Tick-box responses are quicker to answer and easier to process.

Make sure the questions follow a logical sequence, as this will affect how they are answered.

More tips for wording questions successfully

• Use filter questions wherever possible (eg, ‘If no, go to question ...’). This will save the respondent's time.
• Keep the wording simple. Avoid jargon, technical terms and abbreviations.
• Avoid leading questions, complex questions and negative questions. Ask simple questions, each about a single point.
• Be clear and consistent about the meaning of concepts. Words such as ‘income’ or ‘employment’ may need to be defined.
• Make sure the response categories do not overlap. A common overlap occurs in age categories, eg, 10-15, 15-20 years.
• Make sure that your response categories cover the full range of possibilities. Always provide a catch-all option such as ‘other (please state)’, ‘not applicable’ or ‘don't know’.

Questionnaire instructions

Every questionnaire, whether self-administered or administered by interview, should contain clear instructions on how it is to be filled out. Go through the questionnaire with the people who will be administering it, to ensure that they understand each question and the reasons for asking.

Keep a record of the number of questionnaires administered or sent out, and the number of responses received. If the survey population is made up of groups with distinct characteristics (eg, urban/rural, different ages or ethnicities) you may want to know the response rate from each group. Your recording system will need to make that possible. The response rate is fundamental to the weight or reliability that can be attributed to the results.