Outliers
There seems to be something about opinion polls for the News Of The World that gives them much different results to the others. Last year, they produce the poll that gave all the major parties 31%, earlier this year they had the Lib Dems at 28%, IIRC, and now there's a new one that puts Labour third with 28%, LD 29% and Conservatives 32% (which gives others 11%, by my reckoning). However, a poll for the Independent on Sunday does make slightly better reading for Labour, giving them 32% to the Tories' 30 and Lib Dems 27 (others at 11% again). I believe NOTW polls are usually done by MORI, but I'm not sure who does them for the IoS.
A quick (and probably inaccurate, given different polling methodologies) average of the two polls gives the Conservatives 31%, Labour 30% and Liberal Democrats 28%
(I've only got an Ananova link for these two polls right now - I'll update when I get links to any stories on the NOTW and IoS sites. Ananova are also a bit dodgy about static links, but you should find the story listed on the headlines page)
I've been pondering about why the NOTW polls constantly seem to be outliers, and I can see two immediate explantions. First, they could be using a different methodology to the other polls that in some way either discovers or encourages a higher level of support for the Liberal Democrats at the expense of (predominantly) Labour. However, they could (either by accident or design) also just be taking polls at times when the Lib Dem share of the vote is higher (as it would be now, straight after the party conference) and continually picking up the high points in Lib Dem support and the low ones in Labour support.
(Poll originally found via Labour Watch)
Update: Anthony provides some information in the comments (and analysis on his site) - the NOTW poll was by Populus, taken at a different time to their usual polls and the IoS poll was by a new company called Communicate Research. There's also a new MORI poll in the Observer (Con 33%, Lab 32%, LD 25%) which changes the official back-of-an-envelope rounded-up average to Con 32%, Lab 31%, LD 27%.
A quick (and probably inaccurate, given different polling methodologies) average of the two polls gives the Conservatives 31%, Labour 30% and Liberal Democrats 28%
(I've only got an Ananova link for these two polls right now - I'll update when I get links to any stories on the NOTW and IoS sites. Ananova are also a bit dodgy about static links, but you should find the story listed on the headlines page)
I've been pondering about why the NOTW polls constantly seem to be outliers, and I can see two immediate explantions. First, they could be using a different methodology to the other polls that in some way either discovers or encourages a higher level of support for the Liberal Democrats at the expense of (predominantly) Labour. However, they could (either by accident or design) also just be taking polls at times when the Lib Dem share of the vote is higher (as it would be now, straight after the party conference) and continually picking up the high points in Lib Dem support and the low ones in Labour support.
(Poll originally found via Labour Watch)
Update: Anthony provides some information in the comments (and analysis on his site) - the NOTW poll was by Populus, taken at a different time to their usual polls and the IoS poll was by a new company called Communicate Research. There's also a new MORI poll in the Observer (Con 33%, Lab 32%, LD 25%) which changes the official back-of-an-envelope rounded-up average to Con 32%, Lab 31%, LD 27%.
Nick got away with this at 10:15 PM
0 Comments:
Post a Comment
<< Home