YouGov’s IndyRef NO lead grows even though the firm’s tweaked its methodology to make it more favourable to YES

August 11th, 2014

Fieldwork straddled last Tuesday’s Darling-Salmond debate

There’s a new IndyRef poll out from YouGov – the firm which has generally been showing bigger NO leads and which has been critical of the way some other pollsters have been surveying this election. According to YouGov’s Anthony Wells on UKPR the firm has made adjustments to its approaches.

“There are two slight changes in YouGov’s Scottish methodology since the previous poll. The first is that the sample is extended to include 16 and 17 year olds – though this didn’t actually make any difference to the result.

The second is that YouGov have added an extra weighting variable, weighting according to people’s country of birth. For some reason raw samples seem to contain too many respondents who were born in England, and English born people are more likely to vote NO (Panelbase found the same, and also adopted place of birth as an extra weighting variable in their latest poll). This additional weight does makes a slight difference to final result, making the results slightly more “YES”. Under the old weighting scheme the results would have been YES 38%, NO 62%, a slight shift towards NO.

Including 16 and 17 year olds is clearly the right thing to do and it is a bit strange that it is only now, with five weeks to go, that they are being polled by YouGov.

Dealing with the issue if English born respondents seems sensible because the data has been showing that they have a very different view of the referendum and for some reason are being over-represented in samples.

This poll would have been much more useful if it had, like Survation for the Scottish Daily Mail, all been carried out after the debate. In fact just over half the fieldwork took place before.

I’m hoping that the coming week will see the latest ICM referendum polling.

Mike Smithson

Ranked in top 33 most influential over 50s on Twitter