Working with clients to produce shorter surveys

Shorter questionnaires produce better quality data than long questionnaires. I imagine most research agencies agree with this statement but it seems that few are following through on it and producing questionnaires that don’t overburden respondents.

A barrier to producing shorter surveys can be a desire to please the end client. Understandably the client wants to get as much out of the research project as they can. For some it seems that the more questions are included in the survey the more comfortable they feel that they will get useful information and the better value for money they believe they are getting.

So what can we do to work with clients to ensure our surveys are as short as possible? Here are a few tips:

  • Ensure the research objectives are focussed. Having a conversation as early as possible in the project cycle in order to firm up and clarify objectives makes it easier to exclude “nice to have” questions during the design phase.
  • Explain the consequences of over-long questionnaires. Let them know that 20 minutes is around the maximum length for an online questionnaire before respondent concentration wanes and data quality suffers.
  • Ask your client to share what they already know and aim not to replicate existing knowledge. As well as ensuring that you avoid duplication, reviewing past research can provide valuable input into the design of your new survey.
  • For each question discuss what the resulting data might tell you and how it will help to meet the research objectives. Cut any questions that aren’t aligned to objectives.
  • Check that your survey doesn't ask the same thing more than once. For example in a recent survey we carried out a brand association exercise using the adjectives “modern” and “innovative”. The brands associated with both adjectives were the same. Respondents interpreted these words in the same way. We only needed to include one of them in the survey.
  • Provide estimates of questionnaire length for draft questionnaire versions. This helps to focus attention on the need to cut questions.

From my experience, as long as you justify the reasons for avoiding lengthy questionnaires clients understand and respect your point of view and make compromises. Respondents, agencies and end clients can all benefit from shorter surveys. Let’s try to make sure we all do.

1 comment on “Working with clients to produce shorter surveys

  1. Asantha Attanayake

    They say it is more effective to ask the same question a couple of times (maybe twice at least) using different words many questions apart so that the data will be more reliable.
    For instance, to check learner confidence level to use a language (objective) two questions as follow can be asked, a couple of questions apart. (e.g., Q1. Your confidence level to speak with your teacher in L2 – highly improved, somewhat improved, Cannot say, No much improved, Not improved at all. Q7. You are scared to speak in L2 with your teacher – Strongly agree, Somewhat agree, Cannot say, Somewhat disagree, Strongly disagree.)

    What is you take in this respect?
    Thank you

Submit comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Recent Posts