My wife and I went to listen to a panel discussion yesterday about polls and polling held at the Scottish Parliament as part of the Festival of Politics.
The context is the referendum on the 18th September on Scottish devolution.
Here are a couple of snippets that came up during the discussion.
Mark Diffey of IPSO Mori pointed out that the problem with polling people on mobile phones is that there is no geographical tag on a mobile number.
So if you want to ask people about the Scottish referendum and want to ask only Scottish people in Scotland, you can’t poll them by their mobile phone numbers. That’s because the person could be anywhere in the UK.
He also mentioned that the referendum polls show that one third of No voters think the YES campaign is running a better campaign.
Mike Smithson of the blog political betting said that people who opted to vote by postal vote tended to return their vote early. That meant they had cemented themselves into a YES or NO position and were not susceptible to late swings in the campaigns.
Tom Costley of TNS Polling countered that the majority of voters who had a strong opinion one way or the other, did not change their votes – so it was irrelevant that they denied themselves the opportunity to change their vote because they would not anyway.
I am not sure how reliable that is. People tend to behave in conformity with their declared intent – so once they have fixed their position they are likely to say that they wouldn’t change it anyway even if they could.
Mark Diffey again – he said that 16-24 age group were more volatile, i.e. more changeable in the views.