YouGov has UKIP at 6%
August 27th, 2004Today’s YouGov poll has CON 34%(+1), LAB 34%(nc), LD 21%(-2), Oth 11%(nc).
UKIP is at 6% compared with 4% with MORI and 1% with ICM.
They cannot all be right but a sixfold difference must be a record.
.
MessageSpace Advertising
YouGov = \”What ever you want Gov\”
They gave IDS a five point boost in the polls after his last conference speech, kind of like Zogby in the States, their spot on some of the time while the rest of the time their usually off by a far bit… that said I think Mori and ICM look about right, the reason I dislike YouGov is that their internet based and its gotta be hard as hell to rely on data gathered over the internet IMHO
Internet-based polling seems to be at least as reliable as
telephone or face-to-face interviews. Rasmussen came
closer than MORI or ICM to predicting the last election
result.
Yougov\’s final poll for the European elections was pretty
accurate, as were its polls for the London Assembly and
last year\’s Scottish elections.
Rasmussen, wasn\’t an internet pollsters, the method it employed was using pre-recorded telephone calls… YouGov however has failed to publish its methodology and has come under a great deal of criticism from MPs and previous clients.
An America Firm I would love to see do some polling in the UK would be Mason-Dixon who have a strong record in the states as do Teeter-Hart, it would also be very interesting to hear what Philip Gould’s research is showing him at this time… sadly I doubt we’ll get either
I found out about this rather interesting program (PAAN or Political Advanced Analysis Network) on the observer website, interesting and rather suspiring…
http://www.paancommunications.co.uk/home.php : P.A.A.N Website
http://www.paancommunications.co.uk/press.php?pagenum=05 : A copy of the Observer article.
YouGov\’s methodology is up on their website here (it\’s a pdf articel by Peter Kellner, which includes details of their methodology).
A few weeks back Andrew Cooper said on here that Philip Gould\’s current prediction is for another 3-figure Labour majority.
I wonder what makes Gould think this…hmm… sorry I was rather behind the curve on YouGov, that said my misgiving about internet polling remains, shame we don\’t have groups like Teeter-Hart and Mason Dixon over here… ah well
..The often repeated claim that YouGov is biased to the Tories doesn’t stand up to scrutiny. The pollster underestimated the party at the ‘01 GE, the ‘03 Scottish Parliament elections and June’s Euro Elections.
Then YouGov overestimated Labour by 1.4% and underestimated the Tories by 0.7%. Leaving their UKIP debacle aside YouGov have the most proven record of any pollster measuring the Lab-Con split.
I dislike YouGov aswell, not because they use the internet - if it works then fine.
The problem with YouGov doesn\’t actually appear to be their accuracy but rather their honesty. Somehow they miraculously seem to come up with their clients looking consistently stronger in their polls than in oyther peoples.That having been said - when they are being paid by an independent they seem quite close to the mark.
I can\’t help but feel that Kellner is tarnishing his rep being associated with an outfit that appear willing to tailor their results to suit the client.
I think that\’s unfair Bullseye. No pollster has a vested interest in providing his clients with anything other than the most accurate information available.
One can only judge by results. Yougov have had (in this Parliament) the best record of measuring support for the Conservative and Labour parties in actual elections of any pollster.
Can we all be careful about what we say about YouGov and other pollsters? I have never had to moderate comments on the site but we could be getting on to dangerous ground if we cast doubt on peoples\’ integrity.
It is fine to criticise methodologies or raise a pollster\’s record but not to accuse them of bad faith.
Any comments which do this will be deleted.
fair enough - ill bite tongue
As has probably been noted in earlier threads, something which complicates translating the figures into seats (a la Martin Baxter) is variable swing. In 2001, for example, there was a 2% swing to Labour in Wirral West, but a 1% swing to the Conservatives in Wirral South. 1992 saw the breakdown of the old nationwide swing (something which I remember noting not just to myself on the night, but also hearing Patrick Seyd note in a lecture some months after the said election at a sixth form conference), and the breakdown continued in 97 and 01. Even in the Labour surge of 97, the swing was variable, not just regionally, but also often in adjacent seats (a 7% swing in Bristol North East, compared to a 12% swing in Bristols East and South, 14% in Brighton Kemptown, 16% in Brighton Pavilion).
That said, even if the national swing has disappeared, there were notable regional swings at those elections, whether those enjoyed by Labour in Greater London in 92 and 97, or the Tories in south-east Essex in 01 (providing three of their five gains from Labour, their only two in Greater London, and their solitary gain in the South East).
I agree that Yougov seems to lack the integrity of other pollsters. On a recent poll for the Mail on Sunday they asked \”Which wife of the three party leaders was most likely to make you not vote for their husband\”. I cannot see Mori, ICM or Populus sinking to that level whoever their client!
Re. the question alluded to above, I wonder if that was YouGov\’s idea, or that of the Mail on Sunday, with its pronounced dislike (to put it mildly) of Cherie Blair/Booth?
solitare Heh. How it goes? Buy it all. ASAP. Last discount in your live (AAAAA!!!!!). Take a rest.