Our poll in Scotland for The Times: Clarification

On 19 October 2012, the Scottish Daily Mail coverage of our recent poll in Scotland for The Times stated that “Ipsos polls are traditionally highly favourable towards the SNP because they only include Scots who are certain to vote” and that Ipsos’s “methodology of surveying people definitely committed to voting favours the SNP”.

1. The exclusion from our figures of those who are not certain to vote applies only to the “headline” figures from questions on voting intention (including questions on voting in a referendum). This is to take account of likely differences in turnout. We ask all respondents to rate their likelihood of voting on a scale of 1 to 10. Only those who state that they are certain to vote — 10 out of 10 — are included in our headline voting intention figures. This is because we have found that, through testing this methodology, this filter appears to give the best estimate of the voting intentions of those who will turn out to vote. Nevertheless, our polls are conducted among a representative cross-section of adults, either on a GB-wide or Scottish-wide basis, and this includes people committed to voting as well as those who have no intention of casting a vote. We invariably also publish figures showing the voting intentions of the whole sample, including those who may not vote. Full data from all our polls is available on our website within 24 hours of publication. Such transparency allows public access to the data so that the effect of the headline figure compared to the full sample can be assessed.

 

2. Although taking account of turnout figures affects each party’s share of the vote when compared to the whole sample, we believe that using this filter makes our polling more accurate as a measure of voting intention. Our experience shows that reporting in this way will more accurately reflect the result in a referendum held now both in terms of who would vote and how they would vote. The accuracy and quality of our polling has been acknowledged widely. Most recently we were recognised for conducting the most accurate poll ahead of the 2011 Scottish Parliamentary elections  and we are rightly proud of our record.

Full details of our recent poll can be found here.

In the lead-up to the 2014 independence referendum and beyond we will continue to conduct polls of the utmost rigour, providing full transparency of results and guaranteeing the levels of high quality with which have been associated over many years.