so the thing is, despite the fact that anyone who actually answers the phone these days is a weirdo, you can still get high quality polling result this way
what this shows is that you can’t get them cheaply, which is why the zone is flooded with RW bullshit polls who don’t care about accuracy
In the past couple years I have participated in several phone polls, because I am A Dude With Opinions, and because my phone was also my business phone, so I answered any random number that wasn't flagged as a scam. But if they got me multiple times, the pool is *small*...
pretty sure most pollsters are moving to a combination text/phone methodology where they initially contact you and do the demographic screener via text and only call you after that for the rest
"so the thing is, despite the fact that anyone who actually answers the phone these days is a weirdo, you can still get high quality polling result this way"
Could you explain that a bit more? I didn't know that.
I think the question is screening methodology.
If you just take everyone who answers, the results will be skewed; but if you carefully screen/assess the respondents a representative sample can be collected.
Yes, but that was a problem of who had phones. There was no good way to winnow a representative sample.
Is it easy? No. Is it cheap? No.
Can it be done. Yes, but you have to care about getting good results; and spend the time/money to craft good questions, and find valid subjects.
But the method can only be representative of [“people who pick up the phone” within demographic] which may not be representative of *who votes* within that demographic. Which you can only model based on the prior election.
It’s a really hard problem that’s nowhere close to being solved for 2024.
That's where, "did you vote in the x election?", "Did you vote in the y election" comes into it.
Because that was a problem 30 years ago (and it wasn't that easy then to get people to take a telephone poll, my mother did polling).
As I said, it's not easy, it's not cheap, but it is doable.
“All of our respondents, answering their land line telephone with regularity, said their man concerns this election were prescription drug prices, all this woke DEI at The Target, and violence in video games”
Nope. It shows if you throw enough money at the problem, you can build a polling sample that’s demographically representative of who voted in 2020* among people who picked up the phone. 1 of 2 problems is who answers an unknown number anymore.
* Best case, used demographic shifts to model, still…
