As someone who is very well-versed in survey research, I'll say political polls should always be viewed with the lens of: which polling co. and what is their methodology. Most of them are liberal-leaning companies and their methodology (intentionally or not) skews their results to the liberal side. It's all about the underlying methodolo…
As someone who is very well-versed in survey research, I'll say political polls should always be viewed with the lens of: which polling co. and what is their methodology. Most of them are liberal-leaning companies and their methodology (intentionally or not) skews their results to the liberal side. It's all about the underlying methodology. Who takes their survey? How do they weight the results? How do they define a "likely voter?"
The problem with most polls is that they get massive underrepresentation of Trump supporters taking them b/c Trump supporters generally do not trust the pollsters. This is true for corporate Marketing surveys as well. You have to REALLY work to find Trump conservatives and get them to take a survey.
So, yeah, if the poll is showing Trump tied with "Biden," it's likely Trump is well ahead of "Biden" (whoever or whatever "Biden" is at this point).
The only "poll" I believed in the 2016 and 2020 elections were turnouts at Trump rallies vs. Biden "rallies". I stood in nearly 1/2 mile long lines at one in 2020, and got turned away from Trump's final 2016 rally in GR MI, while watching 50 people sitting in circles at the very few Biden 2020 "rallies". 2020 was the swing state fix the demoncrats didn't think they needed in 2016. Don't tell Grandma Garland I said this, I'd rather not be persecuted by the Department of "Justice".
The turnouts for Trump rallies were astounding,I've never in 73 years seen anything like it, and it happened every single time, over and over. The few pathetic public events for Biden were ridiculously small, embarrassingly so. There were so few "supporters" there that the MSM cameras had to use very tight angles so you wouldn't see how sparsely attended Biden's events were. I simply do not believe based on what I observed with my own eyes that Biden got more votes than Trump. I don't think it was even close.
The media likes to pin it on Idiots getting riled up like at a tractor pull with a KKK speaker--after all those church attending Bible believers need some entertainment off of the gun range. Maybe the Koch brothers are paying for people to attend? Studious, educated professionals and poor oppressed folks of color have NO NEED to go to a Biden event in mass, that would be undignified and his quaint stories don't add anything to the mix. They already all know how good they have things as long as Ol'Joe runs the show. See, even when no one shows up you're still the winner in MSM spin.
Yes, just posted about this on another thread as I do some occasional survey work for small businesses, although informal, such as surveying their current customer list, so I am always interested in methodology, how they word it, and who is paying for it. I record them and send them to whoever I think would be interested.
"likely voter' is suspect to me, as people may lie about how often they vote.
I've compiled door knocking lists for conservative school board candidates the past 3 years based on frequent voters, which I define as those who have voted in either 1, 2 or 3 of the last municipal elections (which have about a 12-14% turnout). Get the list from the county election board. Amazing when you cross reference that with the people who are the most vocal on social media, look them up to see if they have voted in past elections, and oftentimes, they have not. Actually 2 of the conservative candidates who have won in the last 2 SB elections had not voted in past SB elections - but they do now! Heaviest turnout in local elections is in blue precincts, R's need to step it up!
Well, different polling houses use different likely voter models. Some use self-report as in someone saying "yes, I plan to vote." Others use voting history (which in a survey is also self-reported). Others use age, gender, race, and strength of opinions to try to predict likely voters. So, for example...younger people are less likely. Whites and Asians are more likely. If they have strong opinions to their answers (rating feelings at the high or low end of scales), they would be more likely b/c supposedly more passionate.
It's really a guessing game, and when polls go horribly wrong (when polls attack, lol) it's usually b/c the likely voter models were wrong.
It's also something that left-leaning pollsters frequently get wrong, rating Dem voters as more likely to vote b/c from their biased viewpoint there is always something to be fired up about. Dems and leftists especially do not have good historical awareness. So, when the Supreme Court makes a ruling, they automatically assume that means every Dem is going to be hopping mad and ready to run to the polls. In fact, Republican voters tend to be older, whiter, more steadily employed (aka paying more taxes), and therefore more likely to vote, regardless of what outrage du jour the left whips itself into a frenzy over.
Yes. In my 'frequent voter' list for school board, (frequent voters in low turnout municipal elections) I sorted by date of birth. ~60% were over age 55, and only 6% were under age 35. On a national scale, this is why we will never get a good 'fix' for social security. Seniors vote, and woe to any politician who talks about reform.
So the polling companies don't have access to state voter databases showing how often particular voters have voted? I know political parties do have such access.
Probably depends on how hard they want to work at it. If they are polling just to get the answer they want, rigorous methodology is just an unnecessary expense.
Absolutely! You can get any result you want if you phrase the question correctly. I used to do census work and we were trained to use the questions as written to prevent skewing the results.
As someone who is very well-versed in survey research, I'll say political polls should always be viewed with the lens of: which polling co. and what is their methodology. Most of them are liberal-leaning companies and their methodology (intentionally or not) skews their results to the liberal side. It's all about the underlying methodology. Who takes their survey? How do they weight the results? How do they define a "likely voter?"
The problem with most polls is that they get massive underrepresentation of Trump supporters taking them b/c Trump supporters generally do not trust the pollsters. This is true for corporate Marketing surveys as well. You have to REALLY work to find Trump conservatives and get them to take a survey.
So, yeah, if the poll is showing Trump tied with "Biden," it's likely Trump is well ahead of "Biden" (whoever or whatever "Biden" is at this point).
The only "poll" I believed in the 2016 and 2020 elections were turnouts at Trump rallies vs. Biden "rallies". I stood in nearly 1/2 mile long lines at one in 2020, and got turned away from Trump's final 2016 rally in GR MI, while watching 50 people sitting in circles at the very few Biden 2020 "rallies". 2020 was the swing state fix the demoncrats didn't think they needed in 2016. Don't tell Grandma Garland I said this, I'd rather not be persecuted by the Department of "Justice".
The turnouts for Trump rallies were astounding,I've never in 73 years seen anything like it, and it happened every single time, over and over. The few pathetic public events for Biden were ridiculously small, embarrassingly so. There were so few "supporters" there that the MSM cameras had to use very tight angles so you wouldn't see how sparsely attended Biden's events were. I simply do not believe based on what I observed with my own eyes that Biden got more votes than Trump. I don't think it was even close.
Agree. There was no evidence that Biden winning was credible or made any sense at all.
The media likes to pin it on Idiots getting riled up like at a tractor pull with a KKK speaker--after all those church attending Bible believers need some entertainment off of the gun range. Maybe the Koch brothers are paying for people to attend? Studious, educated professionals and poor oppressed folks of color have NO NEED to go to a Biden event in mass, that would be undignified and his quaint stories don't add anything to the mix. They already all know how good they have things as long as Ol'Joe runs the show. See, even when no one shows up you're still the winner in MSM spin.
Yes, just posted about this on another thread as I do some occasional survey work for small businesses, although informal, such as surveying their current customer list, so I am always interested in methodology, how they word it, and who is paying for it. I record them and send them to whoever I think would be interested.
"likely voter' is suspect to me, as people may lie about how often they vote.
I've compiled door knocking lists for conservative school board candidates the past 3 years based on frequent voters, which I define as those who have voted in either 1, 2 or 3 of the last municipal elections (which have about a 12-14% turnout). Get the list from the county election board. Amazing when you cross reference that with the people who are the most vocal on social media, look them up to see if they have voted in past elections, and oftentimes, they have not. Actually 2 of the conservative candidates who have won in the last 2 SB elections had not voted in past SB elections - but they do now! Heaviest turnout in local elections is in blue precincts, R's need to step it up!
Well, different polling houses use different likely voter models. Some use self-report as in someone saying "yes, I plan to vote." Others use voting history (which in a survey is also self-reported). Others use age, gender, race, and strength of opinions to try to predict likely voters. So, for example...younger people are less likely. Whites and Asians are more likely. If they have strong opinions to their answers (rating feelings at the high or low end of scales), they would be more likely b/c supposedly more passionate.
It's really a guessing game, and when polls go horribly wrong (when polls attack, lol) it's usually b/c the likely voter models were wrong.
It's also something that left-leaning pollsters frequently get wrong, rating Dem voters as more likely to vote b/c from their biased viewpoint there is always something to be fired up about. Dems and leftists especially do not have good historical awareness. So, when the Supreme Court makes a ruling, they automatically assume that means every Dem is going to be hopping mad and ready to run to the polls. In fact, Republican voters tend to be older, whiter, more steadily employed (aka paying more taxes), and therefore more likely to vote, regardless of what outrage du jour the left whips itself into a frenzy over.
Yes. In my 'frequent voter' list for school board, (frequent voters in low turnout municipal elections) I sorted by date of birth. ~60% were over age 55, and only 6% were under age 35. On a national scale, this is why we will never get a good 'fix' for social security. Seniors vote, and woe to any politician who talks about reform.
So the polling companies don't have access to state voter databases showing how often particular voters have voted? I know political parties do have such access.
Probably depends on how hard they want to work at it. If they are polling just to get the answer they want, rigorous methodology is just an unnecessary expense.
Kind of like with the Covid clinical trials 😑
And always, what questions are asked and how are they phrased? That can hugely influence the results of a poll too.
Absolutely! You can get any result you want if you phrase the question correctly. I used to do census work and we were trained to use the questions as written to prevent skewing the results.