Posted on 10/21/2012 11:37:40 AM PDT by dila813
He said they weight using the US Census Data, basically, if the Census give statistics by a category, they weight by it in the polls.
Polls weight by standard demographics -- race, age, gender, income, education, etc.
And, in doing so, they are obviously over-weighting their polls toward Democrats. And it won't be over-weighted by black Democrats...or Hispanic Democrats. It is, instead, by elimination, white Democrats -- at the expense of white Republicans.
The cause of this phenomenon is equally easy to determine: Only 9% of all random phone calls result in an interview. Obviously, white Democrats are significantly more likely to pick up the phone and submit to a polling interview than are white Republicans.
Net:net, the polls are right, as far as they go -- but they are not necessarily representative of the final turnout figure. But you can adjust them to whatever turnout model you want.
[ We only ask the person after we have polled them who they currently self-identify with at the end of the poll for informational purposes. ]
WEll DUUUUUGH... talk about a “TELL”... i.e. poker
True. But no poll will ever be 100% representative, no matter how it’s weighted or massaged, the target is always moving. In close elections it’s an advantage to be under represented in a poll, especially if you’re essentially tied within the MoE. Democrats compensate by cheating, of course.
If we had an honest media, there would be fewer Democrats and conservatives would win every election. That’s because 40 percent is conservative, while 20 percent is liberal. The remainder are swing voters who more likely to vote conservative than not.
But we have a corrupt media that ads 10 to 20 points to the Democrat side of the ledge, IMHO. Thank God, then, that not everyone votes because the vast majority have had nothing but leftist lies and propaganda stuffed into their heads by a corrupt media.
Yes — after pointing out that they ask the affiliation question second. That would still enable them to use the second method I mentioned (weighting during analysis). Gallup chose not to do this. If you have reason to believe that Republicans are less inclined than Democrats to talk to pollsters, then Gallup has made a serious error.
If the proportion of registered Republicans, Democrats and Independents does not closely match the known proportion of each in the population; then you have a selection bias problem. Plain and simple. And, BTW, this has nothing to do with the way each respondent actually intends to vote. It has everything to do with matching the sample to the population.
Yup.
"Weighting" [which means "tampering"] destroys the entire science of statistics. A truly random sample should tell us what we're looking for.
Shouldn't it?
I mean, look at what "weighting" has done to the climate record.
Party self-identification and actual registration are very different things. Someone could be registered unaffiliated (in some states they don't register by party) and tell a pollster they're a "Democrat" one week and "Republican" the next.
O.K. that’s an important point. I was labouring under the illusion that the registration numbers were known everywhere. Obviously, party registration wouldn’t serve as a control, where there are no reliable registration records.
I never suggested that party self-identification would suffice as a check (and I know you weren’t suggesting I was). Clearly, self-identification would tell you nothing about the quality of the sample.
So Gallup not weighting by party affiliation doesn't concern me. They've been doing this since the 1930s and have an almost perfect record with respect to presidential elections so when they say Romney is up by 7 just two weeks away from Election Day, well, I'm feeling pretty good about that.
One thing I will throw in there is that millions of people have eliminated landlines in favor of cellphones. Typically these are people "on the go" who spent most of their time outside the home working, dining out or running errands, ergo, conservatives. Many Democrats tend to sit around the house by a landline, watching Oprah or something.
So there is a chance that Gallup might still be undersampling the conservative vote.
Bflo wrote: “”Weighting” [which means “tampering”] destroys the entire science of statistics. A truly random sample should tell us what we’re looking for.”
i agree but the problem is that if Republicans have a high rate of nonresponse, then the “sample” the pollster ends up with is not a random sample of the people who voted... and reweighting by party might be able to deal with the “shy republican” issue, but it is still a guess...
I agree. You are correct except that most of us have not voted. Everyone is trying to guess just who will vote. That's what causes the abuse of statistical methods.
The pollster has to "adjust" his results to come up with something plausible [to whom?] Distorting statistics to predict elections is a fool's game. The guy wins who distorts best.
Full disclosure: I love following the polls. But they are, IMO, using the patina of science to make their guessing respectable. Much like the global warmingists.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.