Rules for publishing public opinion research
How a survey was carried out, and the context in which it was carried out, are an inseparable part of it. Research agencies recommend that the media publish the passport together with a link to fuller information. The media should also check a study’s credibility before publishing and discuss any interpretation of the results with its authors. That avoids needless errors and misinterpretation.
The passport
The public opinion research passport is the list of information the provider gives when a survey is published.
- the nature of the output
- the agency and the client
- the number of respondents
- the data collection method
- the date of collection / fieldwork period
It is also good practice to give the public a link to the fuller information in the research report (see the passport).
The most common mistakes the media make when interpreting research
- Quoting results without a source (who carried the study out, and how). See the credibility test.
- Not checking where the research came from or how reputable the provider is. See the credibility test.
- Interpreting the results beyond what was measured or beyond the purpose of the study (claiming, for instance, that it is a “prediction”).
- Confusing a straw poll with serious research. On sample surveys versus straw polls, see here.
- Generalising results from a specific target group to the general population (or to a wider target group). On sampling and representativeness, see here.
- Misreading the wording of the question (or of the answer options).
- Telling only part of the story (trust in public figure “X” has fallen — when in fact the change is within the margin of error, or the comparison with others is missing).
- Comparing results of studies that are not methodologically comparable (a different sample, a different method — a straw poll against quota sampling). On comparisons, see here.
- Using data that is out of date or has been superseded. See Fieldwork period.
