Opinion polls are an important source and subject of news stories. Newspapers and broadcasters often commission their own polls to provide information about readers’ views or voters' intentions.
Opinion polls can be an effective tool in measuring what people think about issues, political parties and candidates. However, the public needs to be able to judge properly the value of polls. This requires that publications provide sufficient background information to prevent results from being misconstrued. Publications should take special care to do so in election periods.
Reports should not use language that overstates the possible interpretations or meaning of polls. Writers of opinion pieces and editorials should take care when referencing or interpreting polls to provide, where possible, context or information to enable readers to locate the poll results.1
Space considerations may restrict the amount of background information that can be provided about a poll, but background information on at least a number of important details is desirable This can be placed in the main body of an article, a footnote, another section that may be read separately or, if online, via a hyperlink.
Editors should take reasonable steps to ensure that reports about previously unpublished opinion poll results include, or have been written taking into account, at least the following matters:
- the name of the organisation that carried out the poll
- the identity of any sponsor or funder2
- the exact wording of the questions asked;
- a definition of the population from which the sample was drawn;
- the sample size and method of sampling3
- the dates when the interviews were carried out.
Publications are also encouraged to consider including the following matters where possible:
- how the interviews were carried out (in person, by telephone, by mail, online, etc); and
- the margin of error.
Editors and reporters should carefully evaluate whether to report online surveys, having regard to their scope and methodology.4
They should be cautious of open-access online polls where the sample size and the exact questions asked are unknown and the results have been generated by self-selecting respondents.5
Reports should not imply that the views of panels or focus groups or vox pops and straw polls represent the views of an entire population or the electorate at large.
References
- Adjudication No. 1636 (April 2015).
- Adjudication No. 1383 (January 2008)
- Adjudication No. 1636 (April 2015); Adjudication No. 1383 (January 2008).
- Adjudication No. 1383 (January 2008).
Relevant Adjudications
Adjudication No. 1636: Complainant/The Sunday Mail (April 2015)
- Issue: poll referenced in an opinion piece – question of whether it would be interpreted by readers as fact
- “The text of the print and online material included the statement that ‘[o]ur Galaxy Poll today clearly shows that Queenslanders are embracing asset sales rather than reduce government services or increase taxes’”
- “The material did not provide any statistical results, methodology or other details of the poll. There was no reference to any place where such details might be found. An article on page 8 of the print version did not provide these details. A separate online article on the same day said the survey involved 800 Queenslanders and found ‘38 percent of peopled believed asset sales were the best option to reduce debt, compared to 21 per cent for increased taxes and 24
- per cent for reduced services’.”
- “In this case, the unexplained and unqualified reference to the poll results relating to asset sales, government services and taxes was not distinguishable as the publication’s opinion about the meaning of the poll. The material was likely to be read as a statement of fact.”
- “The failure to indicate where detail of the poll findings and methodology could be found and the fact that the detail which was provided elsewhere did not enable readers to ascertain whether the statement in the material was opinion or fact, meant that relevant facts were not disclosed.”
- “Accordingly, the Council considers that the print material on page 54 and the online editorial were in breach of General Principle 6. It recommends that publications consider the Council’s Advisory Guideline on opinion polls which it issued in 2001.”
- General Principle 6 at the time was as follows:
- “Publications are free to advocate their own views and publish the by-lined opinions of others, as long as readers can recognise what is fact and what is opinion. Relevant facts should not be misrepresented or suppressed…”