Did you know Gove said this?

[ToxicLadybird]

Apparently during the referendum campaign Michael Gove claimed that people who supported the EU were the same as nazi sympathizers during the war.

The first I knew about this was watching Channel4 news last night.

yes. And it may have been analysed comprehensively but it is NOT the same level of certainty as the location breakdown. THAT is the only one that is undeniable.

polls are small samples.

not saying it is the same level of not-true as the £350m a week figure, but it is still heavily caveated.

That's fine, special subject, but the location breakdown matches the polls. So what's the problem?

Polls have been shown to be really unreliable in the last 2 or 3 elections. There was a lot of commentary on it before, during and after the referendum. Apparently it is a new UK trend, people are not telling the truth in polls, exit or otherwise!

Which makes quoting them a bit nonsensical!

Are you seriously suggesting that all the polls, which are corroborated by analysis of the results by location demographics, are wrong in finding that older people predominantly voted to leave and younger people predominantly voted to stay?

www.telegraph.co.uk/news/2016/06/24/eu-referendum-how-the-results-compare-to-the-uks-educated-old-an/

Fucking hell! Way to overreact and misunderstand a thing. Well done!

Here, try these, you might understadn the difference bewteen the information, special and I were offering and what you thought we meant!

www.independent.co.uk/news/uk/politics/eu-referendum-exit-poll-who-has-won-remain-leave-brexit-live-updates-a7094886.html

www2.warwick.ac.uk/fac/sci/statistics/staff/academic-research/firth/exit-poll-explainer/

I can't see the relevance of this at all to specialsubject's disputing of the facts in the LSE article about voting demographics.

?? That explains the cross posting then!

Special mentioned the apparent crytal ball inherent on those demographics, you mentioned polls. I pointed out that polls have been decreasing in accuracy for a number of years now...

Put the 2 together and you have information based on a blindingly obvious bit of double guesswork.. that's all!

Let me spell it out: 37% turn out... so you how do you know which part of any geographic locale voted? You can try to correlate against the local stats, but that would be a weak case to make. Stick it into a Venn diagram and see how much could remain 'outside' you could be making a 100% error in your assumption (which is apt as he makes a case for wanting more economic education in schools).

Add all of that to the apparent tendency to lie at exit polls, and other types, and you have further misinformation!

Basically, you can make the stats, numbes, assumptions say almost anything you want! So it is safest to just accept that we really don't know and nor do the headline writers!

So you think it's just a guess that older people voted leave and younger people voted remain?

Yes! An educated one and one that seems to be born out as more information is gathered. Though there is as much anecdotal evidence to say the opposite! And that is the only level of accuracy we can safely assume!

But it was being said as the votes came in, repeated as though it were fact... and the data it was based on then, as now, was fragile, at best.

The biggest problem is that if you accept that the age demographic data is accurate many assume that the rest is also accurate. And it isn't!

As special said, there can be no tying up of vote with individual... so, given the trend noted by the people who collate the data - see the Warwick link - it isn't safe to wholly accept such stats!

That's nonsense.

If you say so!

It really is

I'll bite.

Which bit do you think is nonsense? Specifically...

Not that it really matters , only the total counts. I know old people who voted remain and young ones who voted leave. Anecdata.

Thanks to ourblanche who gets it.

Otherwise, clearly there is as much point in trying an open discussion here as in talking about atheism on a religion thread, so i will stop wasting electrons.

Well for starters, where is the equivalent amount of anecdotal data suggesting the opposite?

Erm... anecdotal... I have heard. In my life that is true!

It is heard where it is said, and not heard where it isn't said.

That is the issue with anecdotal evidence - which was the point I was making!

That's Research 101!

Special I have had similar conversations about research data this summer. I was trying hard not to laugh and to stop myself banging my head on the desk.

I felt so sorry for the PostGrad taking the session. Nothing he did seemed to make sense to a small group of students. They just kept on and on, not really being able to interrogate the data samples he gave us. One just took his examples at face value and then shouted at him for lying and making it hard to trust him again!

I don't think I am cut out for sharing my learning space with 20 somethings

You seem to be acknowledging that anecdotal evidence carries no weight while simultaneously using it to suggest that the data alluded to in the article is not solid. Otherwise, why mention it at all? You are not making sense.

Surveys/polling research is also a function of the questions that are asked, which can themselves be inherently biased.

I was asked at the weekend, as part of a YouGov survey, something like (I really should have taken a screen shot of it)....

"The NHS currently relies on many EU workers. Should they need to take the same language and residency tests as non-EU workers?

• No, they should have to pass a lesser test that wasn't the exact wording but it still involved passing some sort of requirement
• Yes they should sit the same language tests and fulfil the same visa requirements as non-EU applicants"

There was no option to say "There should still be freedom of movement". The choices given assumed that there would be constraints.

Now when that is reported, it may well be along the lines of "say 60% of respondents said that EU workers should pass a language test."

Or are you suggesting that the combined poll results and voting demographics by location are of no more significance than anecdote?

In which case I think I'm getting the picture. You think the moon landings were faked, right?

You seem to be acknowledging that anecdotal evidence carries no weight while simultaneously using it to suggest that the data alluded to in the article is not solid. No, really I am not. I was trying to show that the data alluded to was no better than anecdotal evidence.

Otherwise, why mention it at all? Because that is the issue with the data alluded to.

You are not making sense. I am, you just don't understand.

You are sounded daft, now, Kaija.

You have made some weird assumptions and lept to a strange conclusion. For clarity, fo all the reasons given above, that you have iether ignored or misunderstood

Or are you suggesting that the combined poll results and voting demographics by location are of no more significance than anecdote? Yes! That is what I am suggesting. That is what Warwick University suggest - read the link!!!

Polling data is less reliable than it has been for a while.

Demographic data is correlation only. Have you tried the Venn diagram I suggested? If you do you will see what I mean. There is insufficient data to say who voted how within any geographical area. The data simply assumes that the voters are an even representation of the area as a whole.

Example What happens in areas that voted 80+% for Leave or Remain? Did only one age group turn out, perhaps? Show your working out, one way would be to use a Venn diagram!

Ourblanche, the Warwick Uni link refers to completely different polls. It bears absolutely no relation to the demographic breakdown of the Eu vote.

Can you not follow a discussion, Kaija?

I offered that as information on the changes in exit poll data!

It followed an clear explanation of why exit polls were not commissioned for the EU Referendum.

Together the 2 links offered more information about polls in general
That linked to the now truly belaboured point about demographics should give you an idea why any conclusions drawn form demographic and other polling data is not of good quality!

That's how discussions progress!