Urban75 Home About Offline BrixtonBuzz Contact

Jeremy Corbyn's time is up

If that's heavy weighting what would you expect 'non-heavy' weighting to look like in this scenario, and why? Could you also explain what you mean by weighting proportionately or selectively?

Well Ashcroft's weighting in the ref exit poll amounted to half a dozen or so being moved geographically to get the weighting right (in a much larger sample). Not 25% of respondents from London being counted as though they represented opinion in the Midlands. Or sampling which included 910 of 1019 respondents admitting to voting Remain. It's way out of line with reality.

I was getting at whether their views are moved from London to the Midlands based on proportional demographic information only, without looking at what they said, or chosen more selectively for their views.
 
Well Ashcroft's weighting in the ref exit poll amounted to half a dozen or so being moved geographically to get the weighting right (in a much larger sample).

That can't be right can it? Do you have a link to the poll?

Not 25% of respondents from London being counted as though they represented opinion in the Midlands. Or sampling which included 910 of 1019 respondents admitting to voting Remain. It's way out of line with reality.

What are the real Remain figures then? IIRC previous polls have found it to be >80%.

I was getting at whether their views are moved from London to the Midlands based on proportional demographic information only, without looking at what they said, or chosen more selectively for their views.

I was under the impression that all weighting was done 'proportionally' i.e. according to some number of variables which are known for the population (which looks like the case here). Not that I've ever conducted a scientific poll. How does 'selective' weighting work then?
 

The nation mourns

procession-142188087035102601
 

Thing is, are us proles supposed to think Aaargh Eagle was a half way decent choice to begin with, though tainted with the blood of the innocents of Iraq, and then grateful for a less gaffe-prone, more people friendly NHS privatiser?

Maybe the people promoting Aaargh's bid had thought Iraq was as far from the membership's minds as it is from theirs and then needed just anybody at all that cannot be branded a warmongering, election-rigging bastard.
 
Last edited:
That can't be right can it? Do you have a link to the poll?
well, eg Ashcroft sampled 1292 voters in London and allocated 8 to elsewhere.
http://lordashcroftpolls.com/wp-content/uploads/2016/06/How-the-UK-voted-Full-tables-1.pdf
What are the real Remain figures then? IIRC previous polls have found it to be >80%.
48% nationally. 60% in London. Full results are in wiki. Ashcroft estimated 63% of Labour voters voted Remain. A poll based on 90% Remain voters being representative lacks credibility.

I was under the impression that all weighting was done 'proportionally' i.e. according to some number of variables which are known for the population (which looks like the case here). Not that I've ever conducted a scientific poll.

I posted a bit about Ashcrofts take on variables and compared it to the census on one or other of these threads.

Nor me.
How does 'selective' weighting work then?
Unethically.
 
That isn't how weighting works newbie. The respondents aren't allocated elsewhere, the figures are inflated (or reduced) for each demographic sampled so they reflect the actual % demographic split is all.
 
Hello


Thanks for your email.

The deadline to vote in the Labour leadership elections has now passed. If you have voted by 12pm noon Thursday 10th September your vote will be counted. If you did not vote by then it is too late to change this. The results will be announced at a special conference on Saturday 12th September.

As I’m sure you can understand it is a very busy period for us right now so please be aware it could take us up to a week to get back to you.

Best wishes,


The Labour Party

Sent by email from the Labour Party, promoted by Iain McNicol on behalf of The Labour Party, both at Southside, 105 Victoria Street, London, SW1E 6QT Website: www.labour.org.uk to join or renew call 0345 092 2299.

Some people are getting this automated email, its from the last leadership election!
 
Smith says he's just as radical as Corbyn. He also said on Sunday he's a supporter of austerity. They really, really, really do not get it do they? Interestingly I spoke to a woman who comes into my shop, early 20s, politics student and a big supporter of Corbyn. She said she's a member and will be voting for him but she's so disgusted with the way the party has tried to rig this with its voting barriers and exclusionary tactics that she'll leave if Corbyn doesn't win. I actually think she was a member before Corbyn. If that's repeated across the board, which I believe it will be then that's cause for some optimism regardless of the result.
 
48% nationally. 60% in London. Full results are in wiki. Ashcroft estimated 63% of Labour voters voted Remain. A poll based on 90% Remain voters being representative lacks credibility.
The population in this case is Labour members though, not the general population.
 
That isn't how weighting works newbie. The respondents aren't allocated elsewhere, the figures are inflated (or reduced) for each demographic sampled so they reflect the actual % demographic split is all.
If I've misunderstood I'd appreciate a decent explanation based on the numbers. In the Yougov poll 1019 people responded, 1019 responses counted. 216 people were in Midlands/Wales but 339 responses were allocated there.
 
Smith says he's just as radical as Corbyn. He also said on Sunday he's a supporter of austerity. They really, really, really do not get it do they?

Plenty of reasons to be suspicious of Smith and attack him, but I think the sunday austerity comment was just him mangling his words and Eagle & Marr doing a somewhat confusing job of correcting the error.

I'm pretty sure he misspoke because elsewhere the non-mangled version of his oh so cleverly planned sound-bite was reported. e.g.:

'The biggest idea I will be offering is yes we need to be anti-austerity but we need to be pro prosperity'.
 
If I've misunderstood I'd appreciate a decent explanation based on the numbers. In the Yougov poll 1019 people responded, 1019 responses counted. 216 people were in Midlands/Wales but 339 responses were allocated there.
each Midlands/Wales response was weighed at x1.56944444444 its original value - so, 21 polled would become 33 responses in the final figures
 
If I've misunderstood I'd appreciate a decent explanation based on the numbers. In the Yougov poll 1019 people responded, 1019 responses counted. 216 people were in Midlands/Wales but 339 responses were allocated there.
You're probably best just reading about survey methodology rather than have me attempt a further explanation tbh. This is as good as anything I can find after a cursory google.
 
Back
Top Bottom