Posted on 03/07/2020 9:15:13 AM PST by Hojczyk
Sharyl: When a news organization or company commissions a poll, do they get to decide what questions are asked and what the headline is? Because sometimes the headline they pick, which was usually that Trump was doing poorly in some area, if you dug in, there were really signals that showed the opposite in some instances but were not headlined.
Scott Rasmussen: Right. So obviously the the organization paying for the poll can use it however they want. They can select the questions, they can interpret it as they want. Sometimes there’s a really funny dynamic where the president’s job approval will go up, and yet on every question that the company asked, it shows that people disagree with the president. Well, what that tells you is they’re asking the wrong questions.
Sharyl: It seems like the polls are all thrown out there, and when there’s difference among polls, people decide to dismiss the polls or actually disparage polls if they don’t have the findings they like.
Scott Rasmussen: Look, people like things that they agree with, so if you find one poll out of a hundred that says your candidate is going to win, you’re going to say, “This is the greatest poll of all time.” And if you find one poll out of a hundred or if you find 99 out of a hundred that disagree, you’re going to say they’re all wrong. That is part of the human nature.
(Excerpt) Read more at sharylattkisson.com ...
Video
http://fullmeasure.news/news/politics/reality-check-polling
Sharyl: And advice going into this election, if people are asking you, “What should I believe and what should I do?”
Scott Rasmussen: The first thing is, if a poll disagrees with what you want to see or what you think is happening, don’t dismiss it as fake news. It is really important for people to look at polling data and see what they can learn from it. You take polls collectively to see where the averages are heading, and then you apply a little bit of skeptical assessment to all of it.
One major word of caution to Freepers (who frankly have been more wrong than the polls they like to disparage): the pollsters have learned from their 2016 mistakes. They have “recalibrated” their polling methodology to reflect the reality of the Trump era. We should not dismiss their 2020 polls glibly.
Scott Rasmussen is a good pollster and by nature an honest man.
Should we really trust any pollster? ;-)
Organizations and companies can now (and do!) commission polls to give them a pre-determined outcome, which they then use to push their aganda...
The emphasis is mine. That's interesting. I thought the questions in polling had to be carefully considered and constructed in order to avoid introducing questioning bias into the poll. In that regard I would expect the professionals such as Mr. Rasmussen to play a big part in selecting/wording the questions. I guess if you're a paying customer though, and you want your polls/questions phrased a certain way...
From the professional poll conductor's point of view you're the customer and the customer is always right... But if I'm Mr. Rasmussen and my name is going to be associated with the results I'd want to make sure everyone in the business understands when a poll is "mine" and when I'm just doing the legwork for the customer's poll.
Didn't he leave the company?
That’s what I wonder about. We often hear about push polls, which steer results to predetermined results.
Scott left Rasmussen Reports after the 2012 election.
I treat the best of them almost like family.
I think some pollsters learned from 2016. However, at this point we are early in the election so the polls are a bit looser than they will be in October/November. The MSM use the polls to further their message of anti-Trump whenever they can so most of the pollsters give them what they want, i.e. Biden is beating Trump!
Google search “polls show Biden beating Trump” and look at the flurry of these crap polls. Politico even reported a poll showing Biden would defeat Trump in Texas last summer. Seriously?
I do not trust the polls right now. Ask me in late Sept/Oct and if the polls start showing a shift towards Trump we will know they played the same game.... again.
I mostly agree. Let’s wait and see in a few months.
Public polls are largely paid for by the media companies. For the most part, especially early in the election cycle, polls are nothing more than a marketing tool. Real news is sparse, so polls are manufactured news, designed to tweak public interest. Secondarily, they tend to be designed to support the bias of the media company. Since early polls can never be fact checked against actual results, these results can be whatever the media wants to be. Early polls are also usually registered voters or even all adults. They are the cheapest to do and it is hard to measure likely voters early in the cycle. For all those caveats, early polls tend to be useless as a predictor, although they can spot certain trends.
As election day looms, polls switch to likely voter polls and the questions are more carefully crafted, as reputable polling organizations get measured by their accuracy against the eventual outcome. As well, the trends become more apparent and more accurate. The bias never goes away, so be careful. Campaigns, corporations, and other well heeled organizations order up private polls and these are almost never seen by the public and, of course, they are more accurate. You get your money’s worth.
Your choice, I have made mine based upon a simple premise
Polls are ostensibly used to gauge public opinion of 325 Million Americans by asking questions of typically at best a couple thousand people.
Per FTC regulations any agency or entity making phone calls to someone MUST obtain permission to do so [prior to making the call or face a $1500 fine per instance for doing so. Telemarketers get around this by moving offshore and using a bridge model to place call internationally that “seem” to be in the locale of the phone being called.
So in my view they have a list of “friendlies” that they know they can call and who knows the mix or demographics of those who answer the phone and participate.
Let us look back at all of the Polling in the last Presidential Election where we were being “assured” that Hillary would win in a blowout and Trump should just capitulate now and save face.
Gee, I guess all of the “pollsters” were flat wrong and what has really changed since then...
I make my own decisions no matter what a poll may say, others who follow them may be looking for what way the wind is blowing and jumping onboard for whatever reason they have.
ME?
I ignore them and vote my beliefs and conscience
So as to your admonition of “caution” I say choose for yourself and stop trying to tell me to listen to what others have chosen, or a trend or any other damned thing
Choose
But do so wisely
This has been my concern. I would have thought the polling companies would have figured out why they were so wrong in 16 and fixed it because after all their business depends on being accurate. So if all of them show Trump losing it should not be ignored. We better pray the virus fades when it warms up so the MSM effort to crash the economy fails.
Proper pollsters publish data about the parameters, demographics, party weighting, land lines versus cell phones, registered versus likely voters, and all the actual questions in the order asked. Crooked writers can cherry pick and distort the results. I try to have the discipline to assess all this at the source for polls that seem useful (or that need debunking).
“They have “recalibrated” their polling methodology to reflect the reality of the Trump era.”
Yes, now many pollsters will only use data samples with +9 Democrats.
Before they would use +14-16.
They narrow it down some closer to Election Day.
Figure it out.
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.