Reply to: MP Attempts to subvert the will of parliamnet, colludes with foreign power.

[Yorkie62]

I understand fully how surveys are constructed; I have a statistics degree and used to be a statistican, back in the day. So I also understand the guff you have posted and that you don't understand how surveys and sample sizes are designed.

You're wrong. Get over it. You lost

Ooooh get you with your statistics degree.
The guff I posted was taken directly from the yougov link you so kindly supplied.

Btw
The 2nd page of the tabulated results makes even more interesting reading where they split the sample by class, income, and political attention. Wrt to income 40% didn't declare an income. They must be the 40% tax dodgers then[emoji23]

Sent from my SM-G955F using Contractor UK Forum mobile app

[Whorty]

You completely fail to see the point I am trying to make. The point was all these polls are nothing but tulip.
Your stats are based on a poll, a small (compared to the size of the electorate) sample of people and then the result is extrapolated. If the sample of labour votes was taken in say London which voted predominately for remain you get a completely different result than if your sample of labour voters was from say Grimsby, which voted 70% to leave, and perform the same extrapolation.
The problem with polls is you can make them say pretty much what ever you want. If you can't see that then you have a very closed mind to data, data extrapolation, and holistic thinking.
As I said previously it all....

Lies damn lies and statistics.
Xxx

Sent from my SM-G955F using Contractor UK Forum mobile app

I understand fully how surveys are constructed; I have a statistics degree and used to be a statistican, back in the day. So I also understand the guff you have posted and that you don't understand how surveys and sample sizes are designed.

You're wrong. Get over it. You lost

[GreenMirror]

I did when it was originally posted but that was an age ago and I've slept since then.

Sent from my SM-G955F using Contractor UK Forum mobile app

OG's posts send most of us to sleep.

[Yorkie62]

So the data was a very small sample 5455 individuals. But the unweighted sample size is claimed to be 5242. Where did 213 people go?
Of which there were only 1355 labour voters...even smaller sample size.
The total sample was made up of slightly more female than male voters, no idea why this makes a difference other than to show that the total sample size was split roughly 50/50 amongst only 2 genders! What was the actual split with respect to labour only voters?
There were slightly more older voters ( older than 50) on the total sample. Again what was the split for labour only voters?
The percentages quoted do not include those that didn't vote. Therefore the percentages quoted are warped.
When you add up the male only voters split by age you get 3130 yet they claim there were only 2463 male voters. Where did the extra 667 males come from? Were these transgender females switching at the last minute. Seriously mistakes like this throw any credibility out of the window.
The total number of males and females by age is given as 5242. Therefore given the male sample size has increased by 667 then the female sample size must have decreased by 667. These must be from the LBGT community. For the avoidance of doubt I am taking the p*ss.
I still don't know where in the country the sample was taken and how many voters from each location or how many labour voters from each location.

Lies damn lies and statistics



Sent from my SM-G955F using Contractor UK Forum mobile app

[Yorkie62]

Did you check out the PDF?

https://d25d2506sfb94s.cloudfront.ne...Reweighted.pdf

I did when it was originally posted but that was an age ago and I've slept since then.

Sent from my SM-G955F using Contractor UK Forum mobile app

[Old Greg]

Somewhere between 0% and 100%.

Unless you can give me precise information regarding the demographic of the sample used.

Sent from my SM-G955F using Contractor UK Forum mobile app

Did you check out the PDF?

https://d25d2506sfb94s.cloudfront.ne...Reweighted.pdf

[Yorkie62]

What do you estimate the margin of error to be in the poll I cited?

Somewhere between 0% and 100%.

Unless you can give me precise information regarding the demographic of the sample used.

Sent from my SM-G955F using Contractor UK Forum mobile app

[Old Greg]

You completely fail to see the point I am trying to make. The point was all these polls are nothing but tulip.
Your stats are based on a poll, a small (compared to the size of the electorate) sample of people and then the result is extrapolated. If the sample of labour votes was taken in say London which voted predominately for remain you get a completely different result than if your sample of labour voters was from say Grimsby, which voted 70% to leave, and perform the same extrapolation.
The problem with polls is you can make them say pretty much what ever you want. If you can't see that then you have a very closed mind to data, data extrapolation, and holistic thinking.
As I said previously it all....

Lies damn lies and statistics.
Xxx

Sent from my SM-G955F using Contractor UK Forum mobile app

What do you estimate the margin of error to be in the poll I cited?

[Yorkie62]

Been offline just logged on ..... I see others have corrected you already. Do you understand where you are wrong now?

Firstly, my data was based on how individuals voted in the 2015 GE and how these same people voted in the referendum. Your data is based on the 2017 GE voting.

Secondly, your data only talks about constituencies. My data was individuals. I don't think it's under discussion that a majority of Labour constituencies voted out, but this does not follow that Labour voters voted out. You do understand that not everyone who lives in a Labour area votes Labour? You do realise that others parties get votes too?

Old Greg even helped you with his post at 14:02 "Maybe they are confused about the difference between Labour voters and voters in Labour constituencies. An easy enough mistake to make for a cretin." Yet you still fall into the same error.

Well done though. You found a page with data which doesn't support your argument. You, OPM and Shauny are on fire this last few days

You completely fail to see the point I am trying to make. The point was all these polls are nothing but tulip.
Your stats are based on a poll, a small (compared to the size of the electorate) sample of people and then the result is extrapolated. If the sample of labour votes was taken in say London which voted predominately for remain you get a completely different result than if your sample of labour voters was from say Grimsby, which voted 70% to leave, and perform the same extrapolation.
The problem with polls is you can make them say pretty much what ever you want. If you can't see that then you have a very closed mind to data, data extrapolation, and holistic thinking.
As I said previously it all....

Lies damn lies and statistics.
Xxx

Sent from my SM-G955F using Contractor UK Forum mobile app

[Whorty]

And yet if you look at a different source of information you get a slightly different answer

Did a majority of Conservative and Labour constituencies vote to leave in the EU referendum? - Full Fact

just saying

Been offline just logged on ..... I see others have corrected you already. Do you understand where you are wrong now?

Firstly, my data was based on how individuals voted in the 2015 GE and how these same people voted in the referendum. Your data is based on the 2017 GE voting.

Secondly, your data only talks about constituencies. My data was individuals. I don't think it's under discussion that a majority of Labour constituencies voted out, but this does not follow that Labour voters voted out. You do understand that not everyone who lives in a Labour area votes Labour? You do realise that others parties get votes too?

Old Greg even helped you with his post at 14:02 "Maybe they are confused about the difference between Labour voters and voters in Labour constituencies. An easy enough mistake to make for a cretin." Yet you still fall into the same error.

Well done though. You found a page with data which doesn't support your argument. You, OPM and Shauny are on fire this last few days

[WTFH]

FTFY

The joys of lies, damn lies, and statistics.

No, the joys of your link. You chose to "fix" my quote to remove the word that points out the flaw in your argument.

[Yorkie62]

...

Then what happens is someone takes the stats and then converts them, to suit their reporting.

FTFY

The joys of lies, damn lies, and statistics.

[Old Greg]

Your report was an extrapolation of a poll conducted ???? Not the actual result of the referendum.

Sent from my SM-G955F using Contractor UK Forum mobile app

Of course it is.

[WTFH]

Your report was an extrapolation of a poll conducted ???? Not the actual result of the referendum.

Sent from my SM-G955F using Contractor UK Forum mobile app

Your report was based on the same data, but then they attempted to aggregate it by constituencies, then reported back based on the constituencies.

Let's say there are 150 voters and 75 of them voted one way. That's 75/150 = 50%
But if you then say 100 of those voters are in constituency A, and 65 of the people who voted one way were in that constituency, then 65% from constituency A voted that way
The other 50 voters are in constituency B, and so 10 of them voted the same way. That's 20% of constituency B.
So, based on constituency data, the average is 42.5% ((65+20)/2)

Based on the actual vote, the difference is considerable.

Then what happens is someone takes the constituency stats and then converts them back up, to suit their reporting.

[Yorkie62]

Try using simpler language for him.

Your report was an extrapolation of a poll conducted ???? Not the actual result of the referendum.

Sent from my SM-G955F using Contractor UK Forum mobile app