LibDem press office is trying to spin its way out of last night’s 22 point thumping, selectively quoting three lines of YouGov pollster Peter Kellner’s methodology to suggest the results were unfair on Clegg:
YouGov chief admits last night's #NickvNigel poll sample was "skewed towards people who identified with UKIP" http://t.co/7kJFFn0uCg (1/3)
— Lib Dem Press Office (@LibDemPress) March 27, 2014
YouGov: "group was skewed towards older people, more middle class people, broadsheet readers & also people who identified with UKIP" (2/3)
— Lib Dem Press Office (@LibDemPress) March 27, 2014
YouGov: "We did NOT seek to correct the demographic balance of the sample" http://t.co/7kJFFn0uCg (3/3)
— Lib Dem Press Office (@LibDemPress) March 27, 2014
Yet if you look at Kellner’s reasoning in full rather than the heavily edited LibDem version, if YouGov had done what the LibDems wanted and not “skewed” the poll Farage would have actually beaten Clegg 65% to 28%:
“An alternative approach would have been NOT to have corrected the political skew among our original 3,000 sample. The argument for doing this is that any assessment of audience reaction should take the audience as it is – in this case, accepting that UKIP supporters were much more likely to watch or listen to the debate than supporters of other parties. Had we done this, I estimate that the verdict of the audience would have been Farage 65%, Clegg 28%. Those who prefer to cite this figure, rather than to adjust for the UKIP-rich nature of the audience, are of course free to do so.”
A spectacular trajectory of toys from pram.
