S.L. 350.21 Art. 22 [ 3

Opinion polls 2.1. Opinion polls are a useful tool to inform the audience of voting intentions. However, they can also be used as a tool of manipulation of the viewers and listeners - especially if they are, for instance, broadcast on the last days before a general election, a European Union parliamentary election or a referendum. 2.2. A broadcaster who broadcasts the results of an opinion survey must broadcast the following information with the results of the survey:

i. the name of the sponsor/originator/originator of the survey; ii. the name of the individual or organization who conducted the survey; iii. the dates when the survey was conducted; iv. to the extent that the information is applicable to the survey, the number of individuals contacted for the survey and the percentage of those who refused to take part in the survey; v. to the extent that the information is applicable to the survey, the margin of error for the survey; vi. the exact wording of each question for which data are reported; S.L.350.21 BROADCAST ON RADIO AND TELEVISION SERVICES

vii. for each question for which the margin of error is greater than that reported, the margin of error for the question; viii. a mailing address or telephone number, indicating it as the address or telephone number at which the sponsor/ originator can be contacted to obtain a written report regarding the survey.

2.3. The person carrying out the survey should provide to the station a copy of a written report on the results of the survey, including the following information to the extent that it is applicable to the survey: i. the name and address of the sponsor / originator of the survey; ii. the name and address of the individual or organization who conducted the survey; iii. the dates when the survey was conducted; iv. the exact wording of each question for which data are reported; v. the method used to collect the information; vi. the population from which the sample was drawn; vii. the size of the initial sample and the number of individuals contacted for the survey; viii. the number and percentage of individuals contacted who answered the survey; ix. the number and percentage of individuals contacted who refused to take part in the survey; x. the method used to recalculate percentages when those who expressed no opinion or those who did not respond are omitted; xi. the times of any interviews; xii. the sampling method; xiii. the number of ineligible individuals contacted; xiv. any weighting factors or normalization procedures used; xv. the margin of error for the survey. 2.4. Further, programmes reporting polls are to ensure that they: i. never rely on the interpretation given to a poll’s result by the publication or organisation which commissioned it: look at the questions, the data, and the trend; ii. never use language which gives greater credibility to the polls than they merit - results never "prove" or even "show", they simply "suggest"; iii. always report the expected margin of error, and where the gap between the two leading contenders is within the combined margin of error say so; iv. always say if it was a telephone poll; v. never be tempted to single out one as more "newsworthy".

2.5. Time must be allowed in programmes for all the proper details to be given.