I don't think it's opt-in in that way though. They have a pre-existing list of millions of pre-screened people and they're selecting a representative sample from that list. Fivethirtyeight ranks them fairly highly among other polls -- certainly high enough for this opinion poll to be considered accurate.
Fivethirtyeight only grades them on their ability to predict american election results. I don't think that's the same as advocating for their efficacy in producing leading public opinion polls.
You’re right, kinda. Issue polling is generally better than horse race polling and YouGov is no exception.
It's a serious issue when there's a clear political bias in the founders. They put more effort into steering the narrative than objectively reporting it.
Sure, and were there a clear bias your comment would have value.
If you haven't seen any clear bias then you probably have a close enough outlook to them to not notice. And that's fine, I don't require everyone to have the same opinion as me for their comments to have value.
My impression of bias is probably born out of the leading polls that rightwing media and thinktanks in the UK commission them to do. You can fairly argue that these polls are externally commissioned so their tenor is a product of their issuer not yougov. But the overall impression I got was that they could be readily depended on to produce misleading propaganda against labour when it wasn't being run by corporate technocrats.
It's still selecting from a list of people who have something to say, though.
As far as how accurately they represent broad swaths of america... well, that's a different matter. I would expect your average american to be far more luke warm to any given subject than respondents to a poll.
That's probably just a problem with polls though -- people who won't answer aren't included. But they're saying that 26-32% of Americans are "unsure," and that sounds pretty lukewarm. Their methodology does sound odd to me too but if it was flawed it would show in the election data, right? Elections are a brutal testing ground. Hundreds of surveys have been predictive and high quality on average.
Agreed on all of that.
I would have guessed 1/3 are “wtf! stop it”, a 1/3 are “bomb them harder!” And then there’s everyone else just doing their thing, going to work. Going to school.
I don't think it's opt-in in that way though. They have a pre-existing list of millions of pre-screened people and they're selecting a representative sample from that list. Fivethirtyeight ranks them fairly highly among other polls -- certainly high enough for this opinion poll to be considered accurate.
Fivethirtyeight only grades them on their ability to predict american election results. I don't think that's the same as advocating for their efficacy in producing leading public opinion polls.
You’re right, kinda. Issue polling is generally better than horse race polling and YouGov is no exception.
It's a serious issue when there's a clear political bias in the founders. They put more effort into steering the narrative than objectively reporting it.
Sure, and were there a clear bias your comment would have value.
If you haven't seen any clear bias then you probably have a close enough outlook to them to not notice. And that's fine, I don't require everyone to have the same opinion as me for their comments to have value.
My impression of bias is probably born out of the leading polls that rightwing media and thinktanks in the UK commission them to do. You can fairly argue that these polls are externally commissioned so their tenor is a product of their issuer not yougov. But the overall impression I got was that they could be readily depended on to produce misleading propaganda against labour when it wasn't being run by corporate technocrats.
It's still selecting from a list of people who have something to say, though.
As far as how accurately they represent broad swaths of america... well, that's a different matter. I would expect your average american to be far more luke warm to any given subject than respondents to a poll.
That's probably just a problem with polls though -- people who won't answer aren't included. But they're saying that 26-32% of Americans are "unsure," and that sounds pretty lukewarm. Their methodology does sound odd to me too but if it was flawed it would show in the election data, right? Elections are a brutal testing ground. Hundreds of surveys have been predictive and high quality on average.
Agreed on all of that.
I would have guessed 1/3 are “wtf! stop it”, a 1/3 are “bomb them harder!” And then there’s everyone else just doing their thing, going to work. Going to school.