Given FindOutNow polls keep making news it's worth saying a bit more about their methodology.Unlike other online pollsters it doesn't use a panel of people who've signed up to do polls. Instead they ask a small number of Qs to people playing the Pick my Postcode lottery. (Thread)
This means their sample is far less politically engaged than for other pollsters. They would argue this is a better reflection of reality. And it's true other pollsters to have an issue with signing up too many politically engaged people.But…
It also means they get a lot more people saying "don't know" than other pollsters. At the moment this really hurts Labour (and the Tories to a lesser degree) and really helps Reform/Greens because their voters tend to be more enthusiastic. (This is generally true but exaggerated in FON polls).
Again they would argue this simply reflects current reality. And maybe they're right. Outliers can be correct if they've picked up on an issue others haven't.But there is a clear methodological reason why their numbers are so different. And I'm not convinced…
Towards the end of the week we get the regular Find Out Now poll which has a tendency to get Reform supporters excited but as Sam Freedman notes this is down to their methodology and sampling and like him I am not convinced.
If you have a wildly unrepresentative sample then the chances are the end product will be utter bobbins.
Here’s the question I would ask anybody reading this article. Do you really think the people who pick my postcode lottery are representative of the wider Great British public?