Urban75 Home About Offline BrixtonBuzz Contact

Political polling

brogdale -- do you reckon there are there good reasons to distrust phone polls then? Recent results from a few of them seem out of line with other ones.
No. It's not a matter of trust, but there is certainly good empirical evidence to suggest that different polling methodologies are (unsurprisingly) yielding different outcomes. Recently Smithson graphed up some polling to illustrate one aspect of this relating to the notion of the 'shy 'kipper'...

b3949d80-6832-4593-b002-e40b740a27e9_zps4tlmiisq.png


It's clear that the actual human interaction of the phone call appears to depress support expressed for UKIP, hence disproportionately raising tory numbers as a consequence. That's all.
 
danny la rouge brogdale Ta. That shy voter factor is interesting ...

UKIP aside, the Tory figure was a bit higher in that most recent one, Labour's a bit lower ... probably best to be cautious and not jump to conclusions though (yet). Just interested is all.

Hmmm - Labout have now been behind in four of the last five polls and UK Polling report average has the Tories in a one point lead. The nearer to the election the greater reluctance for change in swing voters. It's not looking good this week for Labour.
 
Hmmm - Labout have now been behind in four of the last five polls and UK Polling report average has the Tories in a one point lead. The nearer to the election the greater reluctance for change in swing voters. It's not looking good this week for Labour.
 
Hmmm - Labout have now been behind in four of the last five polls and UK Polling report average has the Tories in a one point lead. The nearer to the election the greater reluctance for change in swing voters. It's not looking good this week for Labour.



on its own, means little. Far more depends on the marginals, especially Con - Lab ones; posts have been on this thread recently explaining. A Tory boost by a couple of points nationally means little if they're taking them off Lib Dems in the south west. Labour losing share to SNP similarly means nothing.

Also, most (all?) polls have a built in Tory swing back factor ..many are level until they build that in and 'how did you vote in 2010?'..that means less this time with many Labour staying at home then, many UKIP not voting, and obviously overstates Lib Dem and understates those who have now switched to Labour.

Plus many weight by intention to vote.. The Ipsos Mori I think only had those definitely voting, discounting others totally...advantaging the Tories.

Cameron has a very difficult path to No 10, even with a late swing ..and his best chance includes Clegg holding his seat and being able to carry the members again this time.
 
None of it good news for Labour as they need to buck trends to have a majority. It's not good to lose even if the Tories don't end up with a majority coalition to form.
 
None of it good news for Labour as they need to buck trends to have a majority. It's not good to lose even if the Tories don't end up with a majority coalition to form.






have you been looking at the numbers recently?! Its well known Labour can't get a majority..

virtually certain they will with SNP (more Left than Labour)..Seems as if you've just parachuted in after a few weeks away.
 
so 3 of last 4 polls have Labour ahead yet Guardian comes out with another 'clear lead for Tories' 'polls have definitely shifted' today .. without any details of Ashcroft sampling and Ipsos weighting , whether phone or on-line etc.
 
Ashcroft had his final marginal out today. (does he only publish those that he wants?..he's a Tory after all, and may earlier have published polls as he could perhaps change their policy..too late to do so now..) His sample of marginal is very selective. he says bottom end of Labour targeted for a reason - Peterborough Lab +2, high UKIP, also high migrant labour force.
Anyway, Pudsey seems tied, Norwich North Lab +2, McVey looks in trouble in Wirral, but has Battersea 12 pt Tory lead?! also ahead in Croydon South despite London average swing going the other way.

He also found some shift in Murphy's seat from SNP to Lab..Tory tactical maybe?

http://lordashcroftpolls.com/2015/05/mixed-fortunes-in-my-final-round-of-marginals/



Ipsos the other day had crossbreaks of Tories winning (big sometimes) amongst DE's, 18-24s, London...some posts saying internally they were nearly not going to publish due to the incredibility of this.

Finally on Methodology, Populus the other day have just changed their allocation of DKs to 2010 vote , putting Labour back and increasing Tory, Lib Dem giving a tie.. Labour ahead before that.
 
Im not sure i can take another week of obsessive poll watching and the stress it produces. You Gov and panelbase polls are consistantly showing lab and tory in a dead heat - with labour maybe shading it by ball hair. Other pollsters are less consistent but MORI, Ashcroft - keep recently given the tories leads of 4-6% - enough for them to scrape some sort of majority together.

I think that the you gov polling is more reliable - but I am definitely not certain. The spectre of 1992 still lingers. There may be a number of 'shy' tories out there.

Thursday is going to be a long, anxious, night.
 
Last edited:
I'd imagine there's proportionally a lot less people who voted Labour last time among the 'don't knows' - so it'll be much more heavily weighted toward the vermin parties.
In all honesty I'm quite suspicious of the notion of DKs altogether. I think that a majority of interviewees claiming DK status are merely those people unwilling to reveal (shy/embarrassed/guilty) or lacking the confidence/self-esteem to express a preference. I think most people do know tbh.
 
I'd think the lib-dems would be the biggest beneficiaries of that. There was a lot of them - "don't know, but not them" is probably a common answer among the 50ish % of their voters they've dropped, yet would be reallocated to their 2010 vote.
 
Back
Top Bottom