Posted on 10/03/2016 5:18:10 PM PDT by SpeedyInTexas
Slight clarification, just to be accurate.
Rasmussen used automated phone polls in 2012. The issue back in 2012 was which was more accurate “live phone calls” vs “automated phone calls”.
Just looked at rasmussen’s website. They apparently use automated and online polling now in 2016.
So, there are really 3 types of polling: online, live phone and automated phone.
The issue doesn’t appear to be between online polls vs phone. It’s continuous random population sampling vs static population sampling. The La times/PPD use a static population that was representative and statistically significant and then call 1/7 of that same population each day, hitting all by the seventh day. This avoids significant statistal errors and sampling bias. Here is a recent study that discusses this issue and craps all over pew’s methodology and praises the LA Times approach, looking at 2012 and 2014 results vs polling predictions.
Here is e link- http://www.stat.columbia.edu/~gelman/research/unpublished/swing_voters.pdf
Note figure 5 on page 13 of the study at this link - http://www.stat.columbia.edu/~gelman/research/unpublished/swing_voters.pdf
Rand is the LA times polling people, pew are the asshats referenced in the article from this original thread. Who looked better in 2012? I’ll take la times/rand any day.
Interesting. I think this time we have corruption in polling.
I don’t answer calls from area codes that I don’t recognize, especially during election seasons. I despise the current practice of almost-daily polls. They seem to be more about steering results than reporting them, and I resent it.
If, by chance, a pollster gets through, I refuse to respond.
I don’t have a problem with mashing a button on a web page, however ‘unscientific’ it may be.
Politicians, lawyers, most present-day journalists, pollsters -— don’t have much use for them, to put it politely...
Corruption no . Manipulation yes. It’s all how you fudge the numbers to skew the results.
we have to be realistic -— the polls that we believed in 2012 turned out to be wrong.
There are a bunch of mushy middle voters who are impacted from network news show to reality stars. This is what sucks
I never believed in the 2012 polls. Romney was such a bad candidate I always thought that polls showing him winning were questionable. In 2016 the whole campaign has been wild from the start. I don’t think the people taking these polls are really able to make sense of what they are seeing. I know the race is close and with 35 days to go anything can happen. It all depends on how much the American people want change.
I suggest that if the opportunity presents itself, consent to an exit poll. Then you need to lie. Tell them you were embarrassed about Trump and voted for the entire slate of democrats. The reason for this is that the operatives get feedback from polling stations continuously. Your participation at an exit poll is worth tons. The left will think they have your locality in the bag and they won’t have time to mobilize the fraud machine, before reality smacks their faces. And think how great you will feel watching the smug press wipe their “called” races back into “too close to call” status.
don’t republicans have the same response teams and tracking?
Sure, but with them thinking Hillary is ahead, they’ll work harder to turnout more voters.
Trump needs to stay the hell off twitter and grow up a little and win this damn thing.
***
I get you don’t like it.
But it works. Pretty much only way to get message out over the networks.
You have a point.
Do a search for “auto response for poll script” to see why internet polls are garbage too.
Polls are by and from the same media that lies about everything else....
And what is to keep the pollster from putting their answer on the survey rather than yours? Who would know? If a telephone pollster wants Clinton and you tell them you are supporting Trump, what is to keep them from changing your answer?
I never answer the phone when it’s a poll so I don’t really know how they work but anytime humans are involved there is a chance of cheating.
People's Pundit Daily (an online poll) shifted toward Clinton for three days after the debate.
But, then the trend reversed and Trump is now very slightly ahead.
The changes were all within the "area of uncertainty" -- not the same as "margin of error", but close.
PPD doesn't adjust for party identification. They do adjust according to demographics and geography.
The cross-tabs on demographics and "enthusiasm" are very interesting:
The reputable online polls don't use cookies to prevent multiple votes.
People's Pundit Daily requires registration, and uses that to track voting.
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.