I've made this point before, but Survation are known for often being much more robust with their clients about the type of questions they will allow than many other pollsters. That being the case, why oh why do they regularly break that practice by running propaganda polls for the anti-independence pressure group "Scotland in Union" that always contain breathtakingly loaded, leading, or misleading questions? I mean, are Survation being paid a push-poll premium or something? Reading through the datasets of the latest one, it's just absolutely laughable - there are a series of questions asking Yes voters from 2014 "who have changed their minds" about their reasoning, but there's just one small snag. As far as I can see, respondents were not actually asked at any point in the poll whether they had changed their minds - the standard independence question was never posed. So the questions supposedly asked to switchers will have in reality been asked to a subset of respondents who may or may not have changed their minds since 2014, and the results are thus based on a false premise and are essentially worthless. If I was Survation, I'd be feeling embarrassed bordering on degraded by this stuff. Other polling firms are sometimes criticised for being too accommodating to clients, but I can't think of any other leading firm that has agreed to drop its standards to quite this extent. Not in recent years, at any rate.
As ever, in place of an actual independence question, Scotland in Union have insisted upon a question about "remaining in the United Kingdom or leaving the United Kingdom". It always produces radically different results from genuine independence polls, presumably because a significant minority of respondents misinterpret the question as being about support for the monarchy. Nevertheless, support for "remaining in the United Kingdom" has dropped to 51% in the latest poll (before Don't Knows are excluded), which is lower than in six of the nine previous polls in the series, and in three of the last four.
If you've seen Paul Hutcheon's tweet claiming that "a new Survation poll" shows that only one-third of voters want an independence referendum next year, rest assured that it was only from this poll and thus the results are hopelessly tainted by being towards the end of a question sequence in which respondents are practically bludgeoned to death with unionist propaganda. The question Hutcheon is referring to is itself extremely loaded, asking if respondents want "another" referendum "on leaving the UK". In any case, Hutcheon is giving a false impression by not excluding undecideds. With Don't Knows stripped out, 40% managed to withstand the onslaught and indicate a desire for a referendum next year.
Pretty much all that is of any use from this absurd poll are the parliamentary voting intention questions asked at the start of the question sequence, and which show a broadly similar pattern to last night's YouGov and ComRes polls. However, it's noteworthy that the Greens have drawn level with the Tories on the Holyrood regional list (indeed on the raw numbers they've moved very slightly ahead of the Tories).
Scottish voting intentions for the next UK general election (Survation, 28th-29th September 2022)
Check the weighing. They split it up into age groups. Some of which are strikingly anti independence. One of those is the 16 - 24 year olds. The vast majority of those won't have had a vote in 2014. Some of them won't even have had a vote in recent elections. On the plus side it shows 8 years as being a generation.
ReplyDeleteAlso worth looking at some of the additional questions and ask if they were presented in a random order.
Hey could a REAL indendence suporter please self ID as Nicola Sturgeon for three months and then actually start doing something to get Scotland free from this horror show called the UK.
ReplyDeleteCould a REAL independence supporter also please self ID as the Britnat, British/Irish liar Skier, a liar since 2013, for a few months and point out the horror show called the WGD numpty shitshow.
ReplyDelete