If you do a poll you need a 'representative sample', of rich, poor, young, old, Labour, Tory, Leave, Remain. This is hard and expensive work, but you can get a pretty decent snapshot from 1000-2000 voters.
-
-
Deze collectie tonen
-
A Twitter poll does not have a 'representative sample'. Twitter users are much more pro-Labour than the wider public.https://www.hansardsociety.org.uk/blog/people-who-use-social-media-for-politics-skew-labour-and-remain-but-leave …
Deze collectie tonen -
And Twitter polls take on a life of their own - if they're shared by some big partisan accounts it will totally skew the sample.
Deze collectie tonen -
I'm sympathetic to people who share stuff like this, 'bigger sample size is better' makes intuitive sense. But it's wrong. A 'random sample' matters far more than overall size. Don't make or share Twitter polls about the election, it spreads misinformation!
Deze collectie tonen -
That initial tweet has NINE THOUSAND retweets which is absolutely huge. Bleak!
Deze collectie tonen -
If UK Twitter were a constituency, it wouldn't be a swing seat, it'd be a safe Labour Remain seat in inner London, where there are loads of young graduates, but not nearly as many older people!
Deze collectie tonen -
Twitter is a bit niche, OK, but your aunt uses Facebook and this misinformation is going viral there too:pic.twitter.com/5QiNo5zIRc
Deze collectie tonen
Einde van gesprek
Nieuw gesprek -
-
-
You work for the BBC and are objecting to online polls that haven't been put through the YouGov mangle. Interesting.
-
Deze Tweet is niet beschikbaar.
- Antwoorden weergeven
Nieuw gesprek -
-
-
Deze Tweet is niet beschikbaar.
-
No we won’t.
- Antwoorden weergeven
-
Het laden lijkt wat langer te duren.
Twitter is mogelijk overbelast of ondervindt een tijdelijke onderbreking. Probeer het opnieuw of bekijk de Twitter-status voor meer informatie.