Reporting Guidelines
General Press Release No. 246 (iv) (July 2001)

Opinion Polls

Opinion testing and polls have become an increasingly important source of news stories in all the media. The Press Council has long provided guidelines on such reporting to help both the press and the public get the maximum information from any published poll.

Many papers and magazines do provide the sort of information about poll-taking that the Council advocates, often in the form of a panel alongside the main report. The Council appreciates that tabloid papers and others may find space considerations restricting the amount of background information that can be given, but clearly a bedrock of who conducted the poll among whom is essential. In case of polls with a marked political content, more information is needed.

The public needs to be able to judge properly the value of the poll being reported.

The Press Council believes that reports should as far as possible include the following details:

1. The identity of the sponsor (if any) of the survey.
2. The exact wording of the question(s) asked.
3. A definition of the population from which sample was drawn.
4. The sample size and method of sampling.
5. Which results were based on only part of the sample: e.g. men or women; adherents of particular political parties; and the base number from which percentages were derived.
6. Name of the organisation that carried out the survey.

Additionally, the following information may also be included:

1. How and where the interviews were carried out: in person, in homes, by telephone, by mail, in the street, or whatever.
2. Date when the interviews were carried out.
3. Who carried out the poll, e.g. trained interviewers, telephonists, reporters etc.

When reporting ring-in or internet polls (i.e. those where readers are given a number or address to register a vote), it should be made clear that, as the results have been generated by self-selected respondents, and not by proper statistical sampling, they are not necessarily representative of the whole population. A distinction in reporting should be made between 'phone-in polls', where people are invited to call in, and 'phone-out polls', where people are phoned and asked their opinion.

In reporting the results of phone-in or internet polls, expressions such as "most people" and "the public" should be avoided if likely to give a misleading impression that the poll results are representative of public opinion. Well-organised statistically related polls are clearly likely to be more reliable, and the conclusions drawn likely to be more accurate. The methodology may well need to be explained to convince the readership.

From time to time the question of opinion poll reporting on political outcomes is brought up, with disappointed political advocates calling for some limitation on media reporting at politically sensitive times. The Press Council is firmly against any such limitations; the public has a right to known and a right to speak and comment freely.

This GPR supersedes GPR 39 and GPR 83.

return to
Index of guidelines

[ return to top ]