Checklist: Reporting on statistical surveys

Hem/Posts/Checklist: Reporting on statistical surveys

Press releases based on statistics and surveys are becoming more common. But how can it be ensured that the statistics are accurate and don’t just favour the person or persons who ordered them? Use the checklist here to avoid the risk of comparing apples to bananas.

Sweden’s Market Research Organisation, SMIF, in collaboration with the Swedish Statistical Society and the Swedish Survey Association, has produced the following checklist for journalists.

Understanding how a survey is conducted and what/who is behind it can be essential to allowing its actual results to be published. The purpose of the checklist below is for surveys to be presented in a clear and transparent way by journalists.

Checklist for journalists:

  • Who paid for the survey?
  • What was the purpose of the survey?
  • Who conducted the survey?
  • How many people participated in the survey? What was the response rate?
  • How were the interviewees recruited to this survey?
  • What population does the sample refer to?
  • When was the survey conducted?
  • How was the survey conducted – mobile phone, Internet, telephone, at home, on the high street?
  • What questions were asked?
    Were the questions clearly formulated?
    b. Were the questions leading? –
  • In what order were the questions asked – could the order have affected how the participants answered?
  • Are the results based on the responses from all of the participants or just some?
    If only a subset, how many and why?
  • Were the results weighted? –
    If weighted – based on which variables and why?