One of the most important ways to determine whether respondents are interpreting questions as intended and whether the order of questions may influence responses is to conduct a pretest using a small sample of people from the survey population. The pretest is conducted using the same protocol and setting as the survey and is typically conducted once the questionnaire and procedures have been finalized.

For telephone surveys, interviewers call respondents as they would in the actual survey. Surveyors often listen to respondents as they complete the questionnaire to understand if there are problems with particular questions or with the order questions are asked. In addition, surveyors get feedback from interviewers about the questions and an estimate of how much time it will take people to respond to the questionnaire.

The Pew Research Center pretests all of its questionnaires, typically on the evening before a survey is scheduled to begin. The staff then meets the following day to discuss the pretest and make any changes to the questionnaire before the survey goes into the field. Information from pretesting is invaluable when making final decisions about the survey questionnaire.