Thought Id start this one going because its rather dead on this board at the moment and I suppose all my usual muckers are either at the Stella tennis event watching Dim Tim (lose again) or at Henly Regatta eating cucumber sandwiches (they wish,...NOT).

Anyway please feel free to just talk to yourself blast away and let it go on any company or subject you wish. Just wish Id thought of this one before.

cheers GF.

Dil
Get some stastical education. You can't measure a % change from something that is unreliable to start with. The error factor for good polls is quite high. For unreliable polls is would be unknown and probably higher than any perceived swing. If I came up with a guess of a % of something as 20% and later my guess was 25% then all you know is they were two guesses and nothing else.

Hays, disagree completly. The 2 polls were only 3 days apart are you saying those people polled were changed significanly ........... I dont agree.

Relatively speaking we saw a 37.5% swing to labour.

The queens speech was a let down even a lot of torries say the same.

lol thats crap Haystack.

If the basis of the two polls is the same then any movement is valid.

Queen needs a new speech writer imo.

fred your post is bollocks i know quite a few who have very good degrees
and cant even get interviews ,the managers wo are immigrants only interview
fellow immigrants and run it for their friends .
the same has on friut farms do not employ brittish people it is all under the counter
and time for investigations to happen and check this out which is FACT

They don't use the same set of people to see of their views have changes. That means they have to select a new population. That means a new unreliable poll from scratch. I am signed up to YouGov and I don't get repeat polls asking about change of views on polls that I have previously filled in. The polls were unreliable partly because they only use people who sign up to be polled. You can't get reliable results from self selected populations.

Not disputing the reliabilty of the selection process or the reliability of the result but this wouldn't invalidate a swing.

Of course it would. It wouldn't even be a swing. It would just be two different sets of data with no connection between them. To calculate a swing, you have to be sure that the two populations are statistically representative of the general population regarding their demographics.

It would be like comparing the results of two biased dice. The results are only internally consistent within the results for each of the dice. the results would show you how the dice differed but nothing about the real odds of good dice. If I get a 6 half the time with one die and then get a 6 three quartets of the time with the other die, is that a swing?

If you reinterviewed the same set of people the second time them you might measure a swing. However, the result might have little actual meaning if the selected group were not representative of the general population. For instance if I interviewed, for a poll, only members of the Socialist Workers Party, I could measure a swing in their views if I then reinterviewed the same people. But their views are not representative of the general population.

Of course the data is connected. They use the same selection criteria for the second poll as the first just the same as everyone else.

No one reinterviews the same people they just use the same selection criteria based on age , race , income , location blah blah blah.

merely a one-liner (poetic licence!) as i now pretty much avoid this thread - it's too repetitive

the incumbents have certainly cocked-up quite badly in a number of areas - though done very well in others
what is most noticeable however, is that labour, who should be romping ahead, are doing no such thing
further, i'm sure i saw a survey just the other day that showed EM's public rating was even lower than DC's

No, of course they don't. That's really the point. The main pollsters can choose a representative sample just as good as the last one to test swing because they are going out into the general population and using random methods.

YouGov can't do that because they don't have representative populations to start with. Any selection they make of their members is just as unreliable as another selection. The people who have joined YouGov are not a randon sample. They have self selected. It is a lot like when Sky TV have online polls. Their results have little meaning because firstly they are just Sky viewers and secondly they are the ones who voted in the poll. All that can be said is that they are a group that voted that way.

This, from a few days ago

Asked about his awkward performance during yesterday's BBC Radio 4 World at One – when he was asked 13 times how he would fund policies intended to boost economic growth – Mr Miliband admitted the interview could have gone better.

Prime Minister David Cameron plunged to a rating of -27, his lowest rating in a poll since the awkard alliance of opposites called ConDem snuck into government without a true mandate from the people.

He’s faring best of the bunch just because he is adroit at managing to shift the blame to coalition partners, the Liberal Democrats.

Ed Miliband fares even worse, the Labour leader drops to an approval rating of -41.

With all the charisma of Mr Bean on valium and a collection of equally useless policies to match, poor Ed can’t even manage to make ground in the face of a coalition government which is almost universally despised by the electorate.

As for Deputy PM Nick Clegg, he’s languishing on a rating of -53 and is about as popular as a dose of flu – at least the flu is gone in a week or two.

A woman has been pulled out alive from the collapsed building in Bangladesh. I am sure that whatever religion she belongs to will claim it as a miracle for the home side.

Spire installed atop World Trade Center 'Freedom Tower'

The spire has been installed atop the skyscraper of New York's World Trade Center, making the Freedom Tower the tallest building in the Western hemisphere.

Seven members of a child prostitution ring, guilty of exploiting girls as young as 13, have been jailed.


The convictions include brothers Ahdel and Mubarek Ali, who ‘systematically groomed’ several teenage victims in Telford, Shropshire, alongside five others.

Mubarek, 29, and younger brother Ahdel, 24, have been jailed for 32 years for grooming teenage girls, before ‘pimping them out to workers at a curry house for £150 sex sessions’.


they should be returned to Pakistan for their crimes and let them decide their punishment not in our prisons put the on a plane and tell the Pakistan gov to do the right think with them .stone them to death

They were jailed for 32 years in total. One got 18 years and one got 14 years. That's not enough. There were several others in the gang.

It would be nice to deport them if they are foreign. however I suspect that some of the gang may be British born so we are stuck with them.

And if they aint British born and bred Haystack ?

At least Cam in his own way is trying to put issues like this back in our own hands only thanks to his kicking by UKIP ..

and that was a real poll voted on by real people as predicted by yougov et al !!!