While not telling pollsters the truth may certainly be part of the problem, I think past assumptions about who participates, and in what percentages, may also be to blame. There also may be an ideological/bias component which is contributing to these failures, either blatant or hidden/concealed.
There's no other way to explain the outcomes almost always going in one direction. Simple sampling errors or other procedural problems would be far more likely to err in either direction; when the errors nearly ALWAYS go one way, the problem is not random in nature.
While not telling pollsters the truth may certainly be part of the problem, I think past assumptions about who participates and in what percentages may also be to blame. There also may be an ideological/bias component which is contributing to these failures, either blatant or hidden/concealed.
There's no other way to explain the outcomes almost always going in one direction. Simple sampling errors or other procedural problems would be far more likely to err in either direction; when the errors nearly ALWAYS go one way, this points to the problem not being random in nature.
When one political party deploys literal mobs outside the homes and businesses of WrongThinkers, and people realize there is NO SUCH THING as an anonymous poll (they have to keep a record of the numbers they dialled to prevent re-dialling, if nothing else), people who support the other party will lie or refuse even to pick up the phone.
Whenever I see a poll with no internals provided I ignore it. Over polling of certain groups to get the results the polling organization seemly want is not unusual.
These pathetic insulated leftist clown Want to deluded themselves as long as they are able and convince as many of the great unwashed that because the rigged poles were wrong the election was stolen. All part of their self loathing perversion.
One issue is "who are they polling?" Usually you will hear of a "nationwide poll of registered voters", but registered voters aren't likely voters and most elections (Brexit is an exception here) are not national plebiscites. A poll of rural and suburban counties in Michigan, Wisconsin and western Pennsylvania would have offered insight into Trumps 2016 strength. But it's more fun to poll Brentwood and the Upper East Side, apparently.
My suspicion is not that people are lying to pollsters, but that people who do not participate in polls, are not a normal cross section of the population or point of view. Since in all the examples cited, the undercount was the conservative vote, my thought is that conservatives refuse to participate in polls more often than those of a left wing persuasion. Further, if those of a left wing persuasion think history is on their side, they will be inspired to respond to polls.
Four samples is nowhere near a large enough sample to say "there's not other way to explain", as another commenter has said. It could easily just be chance.
That being said, there is possibly other factors at play. My theory is that most pollsters are based in large cities and their employees skew left. Those who are leftist wish to believe they are in the lead, especially in a close race. So they are more likely to make choices that are biased in a certain same direction (confirmation bias), such as making wrong judgements about the composition of the electorate (age, party, etc), about turnout, about trends, about the relevance of certain issues at play in the election, and so on. Given these various inputs that must go into polls, they consistently err in the same direction, either consciously or unconsciously, because of their leftist skew.
The other factor is poll herding, which is a generally understood phenomenon in which some pollsters issue results with ooor methodology that are in line with other pollsters, so polling errors become magnified and solidified.
Another factor is who answers their phone. If they don't answer, they aren't polled. Given all the spam calling, at least in the U.S., quite a few people now only take calls from people they know. There could be a political slant to that.
A recent book examining why polls are doing so badly is "Data in Decline" by Steve Wood, and available on Amazon. It does a pretty good job of explaining the reasons behind the upsurge in polling failures and how changes in technology have affected the accuracy of polling firms, most of whom are still living in the 20th century.
"In 2017 and 2018, typical telephone survey response rates fell to 7% and 6%, respectively, according to the Center’s latest data. Response rates had previously held steady around 9% for several years."
Years ago Neal Stevenson wrote a novel under a pseudonym and one of its plots was that a polling firm, to get even better data, very carefully selected about a thousand people as a representative sample of the US population. Basically, the broke us down into a thousand different types, using all sorts of metrics like past votes, ethnicity, regional, religion, age, interests, political affiliations, etc. Then they counted up how many Americans fit into each type so they could properly weight the opinions of their thousand "Neilsen" people. The select few were given little devices that today would be an Apple smart watch, and the pollsters monitored their real-time reactions to anything going on in political campaigns or on the news.
It was an interesting idea because it basically got rid of some of the potential sampling errors, such as biased samples or low response rate.
William Safire noticed a couple decades ago the low response rate of polling. A huge cross-section of the public doesn't answer polling calls, and the numbers are staggering - so much so that pollsters don't mention their contact rate.
There are a few who do note their response rate. For one congressional seat in Oregon the pollster reported contacting over seventeen thousand people in the district, with 534 willing to talk to the callers. It has to be problematic to build a representative sample with so few responses.
Having been self-employed for many years, I am one of the few conservatives I know who is not very worried about what would happen if the universally radical leftist HR departments at the places they work found out that they are conservative. They have circles of friends at work who all know: "Oh don't talk to her [meaning just be nice], she is friends with HR."
90% of pollsters will be "woke" leftists, just the kind of people conservatives DON'T reveal themselves to.
Don't make the same mistake I did. In hindsight I wish I had checked the reviews before investing. Once your money goes in, it's not coming out. That simple. I called and messaged
their finance department over 20 times, trying to withdraw but not once did it go through. Not easy finding reliable help too with so many wolves posing as recovery agents out there. I did manage to however, even if it was quite unorthodox. In any case, if you have any questions about my experience with BB you can reach me on +15623847738 we can do this together
Polled people lie to pollsters. (We just don't like them.)
ReplyDeleteWhile not telling pollsters the truth may certainly be part of the problem, I think past assumptions about who participates, and in what percentages, may also be to blame. There also may be an ideological/bias component which is contributing to these failures, either blatant or hidden/concealed.
ReplyDeleteThere's no other way to explain the outcomes almost always going in one direction. Simple sampling errors or other procedural problems would be far more likely to err in either direction; when the errors nearly ALWAYS go one way, the problem is not random in nature.
While not telling pollsters the truth may certainly be part of the problem, I think past assumptions about who participates and in what percentages may also be to blame. There also may be an ideological/bias component which is contributing to these failures, either blatant or hidden/concealed.
ReplyDeleteThere's no other way to explain the outcomes almost always going in one direction. Simple sampling errors or other procedural problems would be far more likely to err in either direction; when the errors nearly ALWAYS go one way, this points to the problem not being random in nature.
When one political party deploys literal mobs outside the homes and businesses of WrongThinkers, and people realize there is NO SUCH THING as an anonymous poll (they have to keep a record of the numbers they dialled to prevent re-dialling, if nothing else), people who support the other party will lie or refuse even to pick up the phone.
ReplyDeleteWhenever I see a poll with no internals provided I ignore it. Over polling of certain groups to get the results the polling organization seemly want is not unusual.
ReplyDeleteThese pathetic insulated leftist clown
ReplyDeleteWant to deluded themselves as long as they are able and convince as many of the great unwashed that because the rigged poles were wrong the election was stolen. All part of their self loathing perversion.
One issue is "who are they polling?" Usually you will hear of a "nationwide poll of registered voters", but registered voters aren't likely voters and most elections (Brexit is an exception here) are not national plebiscites. A poll of rural and suburban counties in Michigan, Wisconsin and western Pennsylvania would have offered insight into Trumps 2016 strength. But it's more fun to poll Brentwood and the Upper East Side, apparently.
ReplyDeleteMy suspicion is not that people are lying to pollsters, but that people who do not participate in polls, are not a normal cross section of the population or point of view. Since in all the examples cited, the undercount was the conservative vote, my thought is that conservatives refuse to participate in polls more often than those of a left wing persuasion. Further, if those of a left wing persuasion think history is on their side, they will be inspired to respond to polls.
ReplyDeleteFour samples is nowhere near a large enough sample to say "there's not other way to explain", as another commenter has said. It could easily just be chance.
ReplyDeleteThat being said, there is possibly other factors at play. My theory is that most pollsters are based in large cities and their employees skew left. Those who are leftist wish to believe they are in the lead, especially in a close race. So they are more likely to make choices that are biased in a certain same direction (confirmation bias), such as making wrong judgements about the composition of the electorate (age, party, etc), about turnout, about trends, about the relevance of certain issues at play in the election, and so on. Given these various inputs that must go into polls, they consistently err in the same direction, either consciously or unconsciously, because of their leftist skew.
The other factor is poll herding, which is a generally understood phenomenon in which some pollsters issue results with ooor methodology that are in line with other pollsters, so polling errors become magnified and solidified.
The pollsters are now “Woke”, they’re not pushing wrongthink that might be embarrassing to leftists.
ReplyDeleteAnother factor is who answers their phone. If they don't answer, they aren't polled. Given all the spam calling, at least in the U.S., quite a few people now only take calls from people they know. There could be a political slant to that.
ReplyDeleteA recent book examining why polls are doing so badly is "Data in Decline" by Steve Wood, and available on Amazon. It does a pretty good job of explaining the reasons behind the upsurge in polling failures and how changes in technology have affected the accuracy of polling firms, most of whom are still living in the 20th century.
ReplyDelete"In 2017 and 2018, typical telephone survey response rates fell to 7% and 6%, respectively, according to the Center’s latest data. Response rates had previously held steady around 9% for several years."
ReplyDeletehttps://www.pewresearch.org/fact-tank/2019/02/27/response-rates-in-telephone-surveys-have-resumed-their-decline/
Fewer than 1 in 10 survey attempts are completed. Why? Nobody knows. They ain't talking.
Years ago Neal Stevenson wrote a novel under a pseudonym and one of its plots was that a polling firm, to get even better data, very carefully selected about a thousand people as a representative sample of the US population. Basically, the broke us down into a thousand different types, using all sorts of metrics like past votes, ethnicity, regional, religion, age, interests, political affiliations, etc. Then they counted up how many Americans fit into each type so they could properly weight the opinions of their thousand "Neilsen" people. The select few were given little devices that today would be an Apple smart watch, and the pollsters monitored their real-time reactions to anything going on in political campaigns or on the news.
ReplyDeleteIt was an interesting idea because it basically got rid of some of the potential sampling errors, such as biased samples or low response rate.
William Safire noticed a couple decades ago the low response rate of polling. A huge cross-section of the public doesn't answer polling calls, and the numbers are staggering - so much so that pollsters don't mention their contact rate.
ReplyDeleteThere are a few who do note their response rate. For one congressional seat in Oregon the pollster reported contacting over seventeen thousand people in the district, with 534 willing to talk to the callers. It has to be problematic to build a representative sample with so few responses.
Having been self-employed for many years, I am one of the few conservatives I know who is not very worried about what would happen if the universally radical leftist HR departments at the places they work found out that they are conservative. They have circles of friends at work who all know: "Oh don't talk to her [meaning just be nice], she is friends with HR."
ReplyDelete90% of pollsters will be "woke" leftists, just the kind of people conservatives DON'T reveal themselves to.
Don't make the same mistake I did. In hindsight I wish I had checked the reviews before investing. Once your money goes in, it's not coming out. That simple. I called and messaged
ReplyDeletetheir finance department over 20 times, trying to withdraw but not once did it go through. Not easy finding reliable help too with so many wolves posing as recovery agents out there. I did manage to however, even if it was quite unorthodox. In any case, if you have any questions about my experience with BB you can reach me on +15623847738 we can do this together