Asking the right questions in quantitative surveys

A quantitative survey questionnaire is only as good as the questions that go into it – if the questions are ambiguous, leading or presented in the wrong order then the quality of the output will suffer.

We use our years of experience in market research to build questionnaires that are easy to understand, avoid creating preconceptions and have a logical flow to them. We start by understanding the key objectives of the survey, identify the key questions that need to be asked in order to address these objectives and then build the rest of the survey around these questions.

We program most quantitative surveys in-house in the UK and our clients have direct access to our programming team. We find that this greatly speeds up the development and review process, and means that we can often make minor changes and improvements in a matter of minutes.

Once a questionnaire is fully tested and signed off, we pilot it with a subset of the sample (typically 10%) and review the responses before proceeding with a full launch.

Quantitative market research is almost always based on a structured questionnaire made up of yes/no questions, multiple choice questions, and open-ended questions. Besides ensuring that the questions are clear and comprehensive, close attention should be paid to the ordering of the questions:

  • opening questions should be easy to answer and make the respondents feel at ease
  • questions should be grouped in a logical order, leading naturally from one topic to another, without introducing any bias
  • the format of the questions should be varied to reduce the risk of boredom (e.g. if there are a lot of multiple choice questions, consider breaking them up with a few open-ended questions)
  • potentially sensitive questions (e.g. income level) should be left to the end

If there are multiple routes through the questionnaire (e.g. if the next question asked depends on the response to the previous question), then each route should be checked to ensure that the logical flow is maintained for all respondents.