Reports are coming through Twitter of a bad poll for the SNP:
Labour - 42%
SNP - 20%
Tories - 19%
Lib Dems - 14%
I'm not even going to bother calculating the seat breakdown if that was the final result. (Oh, ok then, Labour - 43 (+2), Lib Dems - 7 (-4), SNP - 6 (-1), Tory - 3(+2))
I say this with no joy and no detail regarding the above, but what was I saying about a battle for second?
The SNP needs to get on the front foot and an independence referendum, well played as it was today, isn't going to cut it for the campaign to come.
(Story here, no mention of a sample size. And not sure why Iain Gray gets the 42% against him given it's a Westminster poll...)
SNP Tax
2 minutes ago
28 comments:
The ComRes poll last week showed the SNP back in an outright lead over Labour.
The polls seem to be all over the place as the phoney election campaign get's under way.
I think the SNP holding back is the right thing.
No need to waste money right now.
‘You ruined my life’, Brown yelled in showdown before Blair departed.
ha ha ha
SNP lead Labour at Westminster in Scotland?
HAHAHHAHAHA. Get a grip Anonymous.
Those 20 seats are a real prospect now - balance of power my arse.
That was a sub sample in yesterdays poll.
The sample size per Political Betting was a massive 128.
Another poll sub-sample has:
Lab 37% (-2) : SNP 25% (+7) : Con 18% (+2) : LD 13% (-10) (changes since 2005 in brackets) The Angus Reid ample size was 296.
Do a little research before you panic.
Hopefully this is just a pre-election blip and a result of all the negative nat-bashing in the press over the past couple of weeks.
After Nicola Sturgeon's speech yesterday over the Abdul Rauf affair, after the complete vindication of the SNP leadership over the Lunchgate saga, after Jim Murphy's embarrasing comments over religion, after the publication of the proposed independence referendum questions, and going forward as the campaign moves on to the real issues, hopefully SNP support will pick up again.
Politico
It wasn't too many months ago that the SNP were persisitently leading Labour in Westminster voting intentions in full polls rather than sub samples.
So not that funny after all.
However, it appears that the SNP are getting squeezed as Scots plump for the lesser of two evils with Labour in the face of the real prospect of a Tory government.
The sample size was only 128 - and not political weighted - totally meaningless. Just pro-Labour propaganda by the Scottish Sun. Murdoch may not be pleased.
So if only 128 in the sample, margin of error exponential, and all taken in the pub at Anderston Quay, at lunchtime, or the offie next door to pick up the "buckfast" voters.
How many of those 128 came from Banff and Buchan, or did they do 2 per constituency. Pathetic.
As I told Roe re the Sun's headline, another useless poll, just as the COMRES poll was for the SNP in reverse.
Contact Nicoll at Sun HQ and find out if the 2 people were even from each consituency.
To not mention the sample size in the rant, (and how could you weigh that sub sample accurately), shows them to be conniving devious and malicious, perhaps even liars. David Roe to be fair correctly excluded as he was also concerned the sample was not provided.
But a good headline based on deception.
The media may try to downplay UKIP, the BNP and the SNP as irrelevant as they tried to do at the Euro elctions, but people do not like any of the main parties and come the day I think the others vote will be substantial.
People voted UKIP in England once, and I think they might surprise people again, which will cost the Tories as without them existing they would surely have another 3 to 4%.
If the sample size was, as one commentor suggests, 128, then you are looking at an error of around 10%.
Until the Sun and YouGov start publishing these polls in full or have them verified by another organisation, I'm going to call bullshit on them. It seems far too much like it's manipulating data for an editorial line.
Which, of course, is pretty common for the gutter press.
This is not a sub-sample - poll is based on 667 responses, detail here:
http://www.yougov.co.uk/extranets/ygarchives/content/pdf/ScottishSun02.24.pdf
Fieldwork 17 - 24th Feb - does it really take them a week to collate this stuff, given it's all online?
Get a grip! Next you will be believing pigs will fly! Or that UK PLC is no longer in recession.
I treat anything from the scum press with the contempt it deserves.
I think a few people need to pull their heads out of the clouds. There was a lot of similar howling at a similar poll that was dismissed as having no detail. It had a sample size of 562.
If this one was a sample size of 128 then yes, it's nonsense.
667 sounds a bit meatier to me though and looks like it's been backed up.
People can entertain and delude themselves thinking the SNP is still up in the mid 30s if they wish.
I won't be one of them.
I think there are issues with these polls. I simply don't believe that the YouGov/Scottish Sun one posted by Too Old for the YSI with the sample of 667 is in any way representative.
The reason for that is not because of the voting intentions themselves; it is because so many people appear to have voting intentions.
You have 667 people of whom only 44 say either that they will not vote or that they are undecided. That is 6.6 per cent of the sample.
If that is supposed to be representative of public opinion at present it is simply rubbish.
Everyone who is out chapping doors at the moment, of whatever party, knows that the undecideds and won't votes massively outstrip any other single voting intention.
The undecideds are running at between 30-40% in most campaigns. When I see a poll which reflects that, I'll believe it.
I know some polls are taken among people certain to vote - but if the YouGov poll comes into that category then why does it include don't know/won't votes.
The other issue I noted with it was the weighting data on political party identification. The weighted sample included 240 people who identified with the Labour Party - 36% - as compared with 93 people who identified with the SNP, which is just 14%. That strikes me as a bit strange.
Jeff,
You really must stop this.
YouGov are doing daily UK wide polls for the Sun. These are published every day.
It seems that the Sun are collating the Scottish sub-samples from these polls to try and manufacture a 'meaningful' poll for Scotland.
As I have said before, if the weighting is not applied for a specifically Scottish context (which it has not), the sample size is absolutely irrelevant.
On these figures, taking the UK weighting YouGov apply out of the equation (where Labour support is multiplied by about 1.4 and 'Others' support is multiplied by about 0.6) it would seem that the SNP and Labour are roughly level around the 30 - 32% mark.
Adding 5 or 6 inappropriate polls together does not make a good or meaningful poll.
And by the way, if anyone disagrees with this then they might want to explain why anybody ever bothers to do Scotland only polls given the number of UK wide polls that are done on a monthly and currently an even more frequent basis.
Jeff, you never seem to miss an opportunity to post on 'poor' polls for the SNP these days. Conversely, last weeks ComRes poll passed without comment.
Either you're trying to downplay expectations so that we're not too disappointed come the GE or...something else.
Lets wait for a properly weighted Scotland only poll before slitting our wrists, eh?
Valid concerns Indy but I guess I'm just wary of novating a 'biased press' grudge into a 'biased polling company' one. The air of 'they're all out to get us is an unseemly one', not that I'm suggesting you have adopted it.
There's no replacement for knocking doors for getting a sense of public opinion so good to know it's still a bit up in the air. I would suggest however that those undecideds on the street might end up just staying indoors which would mean this poll is valid after all?
In terms of the political party identification, I've always found that strange and I don't know if the SNP's relatively low showing in 05 means they are artificially low in polls now?
For me it comes back to not treating YouGov with undue suspicion and trusting the professional firm to deliver which, to be fair, they tend to do election after election.
Jeff
I despair, yet another unscientific poll with no proper weighting being promoted as gospel.
The Sun will produce these every week until the general election so we need to get used to it.
Lets wait until we get a proper weighted poll with a representative sample (above 1000) before we slit a wrist.
I will of course be quick to post and eat humble pie if a proper sample showed similar results to this. That will be just before I slit a wrist.
P.S. Do you think the Sun by promoting these results, if achieved, could deny the Torys of an overall majority at Westminster.
Anon, absolutely fair enough. If this is an amalgamation of lots of small polls then the results are garbage (and the SNP should consider a PCC complaint)
I've always blogged on polls with 500+ sample sizes, good or bad, so nothing different here and also the reason the tiny ComRes poll didn't make the cut.
If this is a slice and dice of several UK polls from a tracker then it'll be the last such post you'll see on here.
Dubbieside, good shout. Waiting for a proper poll is well worth it I reckon.
I really, really don't think people vote as a result of what polls say so really doubt the Sun could force a hung Parliament with this but hey, you never know.
Jeff
They are not doing it for "no reason" if that makes sense.
Most people would acknowledge that "propaganda works" the message do not bother to vote as you have no chance of winning will start to get through if it is allowed to go unchallenged.
Dubbieside - I wonder if the polls showing Labour so far out in the lead could be for the purposes of trying to lull Labour voters into a false sense of security, and to make Tory voters ensure they get out and vote?
Doug Daniel
I suppose that, that could be the "flip" side to what I have written about.
"We are so far ahead in the polls that I do not need to go out and vote as we will win easily"
I am not really sure about that, but what I think for sure, most people would think, "could we just get some honest polls so we can make up our own bloody minds, without all the spin and mind games"
Obviously you will realize that my mind is already made up no matter the poll.
I have taken to ignoring YouGov "scotland" polls unless they're a real poll, conducted only in Scotland, with a sample size of over 1000.
Otherwise, whilst YouGov's polling might be just, the way it's compiled, and the bias inherent therein (for UK-wise, or for the Sun etc.) tends to inflate Labour over the SNP, and the Tories over Labour, in a Scottish context.
I'd imagine, after the past few weeks of sleaze-crap in the papers, we've slipped from SNP 42, Labour 32, to SNP 30, Labour 32. Or thereabouts. Either way still within the margin of error.
We're not going to get 20 seats I think. It was always a big ask, and the party just haven't had a good run or luck, nor a loud enough voice, but I think 15 is obtainable, local factors depending…
Jeff,
it's an amalgamation of unweighted subsamples from properly conducted polls. No attempt has been made within the subsamples to reflect demographics or party ID, as that is done over each UK-wide poll as a whole. Just adding it all together to give a "sample" (which it's not, as a sample is representative, and this isn't) is statistical nonsense.
You'll have Anthony Wells of UK Polling Report banging his head off the monitor!
I might not be happy at the headline figures, but it's the GBH inflicted by The Sun on opinion polling, and your unquestioning regurgitation of their nonsense, that makes me despair.
Bugger all to do with this, but I have noticed on several Question Time programmes that there is always a member of the audience who comes out with the New Labour latest party line and applause always ripples around him/her. The significant thing is that this person is always wearing a brightly coloured shirt/top, last night from Cardiff it was red.
I now understand the overheard conversation by someone who works as a spark on these programmes. "It is red tonight David." We should be told?
Steven, your objection and despair is duly noted but it looks like some weighting has taken place going by the YouGov website but I'm not going to sit and audit a poll just to keep you happy, "unquestioning regurgitation" or not.
Yes I think that a lot of the undecideds will stay at home, without a doubt. But that is precisely why I find the poll so strange. If you took it as being representative of public opinion it would be indicating a really big turnout which we all know isn't going to happen.
I dom't really know the methodology of polling however so I don't know how they select the people they poll. How do they weed out the people who won't vote and find the people who will?
The simplest way to find confirmed voters, and some campaigns are doing this, would be to survey people who voted at the euro elections. They are your hard core voters. But I don't know how YouGov select the people they sample or what geographic balance they have.
It just looks odd - my strong belief is that the number of constituencies where support for the Tories is pretty much running neck and neck with the SNP could be counted on the fingers of one hand.
Post a Comment