Designing a questionnaire involves 10 main steps:
This involves getting acquainted with the subject, making a literature review, decide on objectives, formulate a hypothesis, and define the main information needed to test the hypothesis.
This steps determines how the information defined in step 1 should be analysed. The plan of analysis should contain the measures of association and the statistical tests that you intend to use. In addition, you should draw dummy tables with the information of interest. The plan of analysis will help you to determine which type of results you want to obtain. An example of a dummy table is shown below.
From the plan of analysis you can draw a list of the information you need to collect from participants. In this step you should determine the type and format of variables needed.
You can start now designing different parts of the questionnaire using this list of needed information.
Knowing the education and occupation level of the study population, ethnic or migration background, language knowledge and special sensitivities at this step is crucial at this stage. Please keep in mind that the questionnaire needs to be adapted to your study population. Please see "Format of Questions" section for more details.
You should start from easy, general and factual to difficult, particular or abstract questions. Please consider carefully where to place the most sensitive questions. They should be rather placed in the middle or towards the end of the questionnaire. Make sure, however, not to put the most important item last, since some people might not complete the interview.
Add instructions for the interviewers and definitions of key words for participants. Insure a smooth flow from one topic to the next one (ex. "and now I will ask you some questions about your own health..."). Insert jumps between questions if some questions are only targeted at a subgroup of the respondents.
Verify that each question answers to one of the objectives and all your objectives are covered by the questions asked. Delete questions that are not directly related to your objectives. Make sure that each question is clear, unambiguous, simple and short. Check the logical order and flow of the questions. Make sure the questionnaire is easy to read and has an clear layout. Please see the Hints to Design a good Questionnaire section for more details.
You should always conduct a pilot study among the intended population before starting the study. Please see the Piloting Questionnaires section for more details.
Depending on the results of the pilot study, you will need to amend the questionnaire before the main survey starts.