Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 2, 2014 8:59:48 GMT -5
A survey to sate my curiosity and get a read on the board. Some likely FAQs and answers: - Q: What does "within stated margins of accuracy" mean?
A: All polls are published with statistically-determined margins of error that describe the (theoretical) reliability of polling results. For example, a Yes/No poll reporting a percentage might indicate the figure is "accurate to within +/-3% 19 times out of 20".
Newspapers, blogs, etc. frequently omit these margins of error, but "accurate to within +/-2% 19 times out of 20" and "accurate to within +/-3% 19 times out of 20" are by far the most common confidence intervals. They can safely be assumed in most cases.
- Q: What if I don't trust all of the polls in a particular organization?
A: If it makes any difference, I'm specifically looking for how much you trust polling on political and/or public policy issues. For example, polls on who supports tighter guns laws, who strongly approves of Congress' performance, or who believes the government should be doing more to support small businesses.
Also, if it makes any difference, I'm interested in national polling rather than state or local polls.
Given these restrictions, please answer "Yes" to a particular organization only if your trust 75% (3 out of 4) or more of their polls.
- Q: What if I don't trust all universities equally?
A: Note that for the "University Polls (Non- Ivy League)", I'm considering only accredited degree-granting US institutions.
Given this restriction, please answer "Yes" to this option only if you trust polls administered by 75% (3 out of 4) or more of these institutions.
- Q: What if I don't trust all cable companies equally?
A: Note that for the two "Cable News Polls" options, I'm considering the four major US cable news outlets: ABC, CNN, Fox, and NBC.
Given this restriction, please answer "Yes" to this option only if you trust polls administered by 75% (3 out of 4) or more of these cable news outlets.
- Q: What is a "professionally- or university-backed" cable news poll?
A: A "professionally- or university-backed" cable news poll is a poll conducted by a professional polling organization or university on behalf of a client cable news company. The poll carries the professional weight of the supporting organization, but might be deemed "suspect" for other reasons. For example, cable news companies might choose polling organizations with particular biases, or might choose not to publish unfavourable results.
- Q: What if I generally trust polls and I don't particularly care who administers them?
A: For organizations that you've never heard of or have no specific opinion on, please answer "Yes" to all of these organization only if you generally trust the results of 75% (3 out of 4) or more of polls in general.
- Q: What if I don't trust any of the organizations?
A: Please check the "None of the Above" button if and only if you leave all other buttons unchecked.
Please feel free to ask any additional questions in the thread.
|
|
Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 2, 2014 12:47:58 GMT -5
Hello...?
(Hello...?)
(Hello...?)
Is anybody there...?
(...body there...?)
(...body there...?)
|
|
Deleted
Joined: May 18, 2024 3:38:16 GMT -5
Posts: 0
|
Post by Deleted on Jul 2, 2014 12:52:31 GMT -5
I trust Nate Silver.
|
|
Opti
Community Leader
Joined: Dec 18, 2010 10:45:38 GMT -5
Posts: 39,779
Location: New Jersey
Mini-Profile Name Color: c28523
Mini-Profile Text Color: 990033
|
Post by Opti on Jul 2, 2014 12:59:09 GMT -5
I want DJ to post his reasoning. I don't feel I can accurately take your poll as I judge more on a poll by poll basis, often looking at the questions if it seems a bit off to me.
For example, I think both the Pew Research center and Gallup do good polls. But I've also found ones they've done, I wasn't that keen on the slant or methodology.
|
|
Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 2, 2014 13:08:33 GMT -5
He apparently isn't in the business anymore. For Gallup and Pew Research, would you estimate that 75% or more of their polls you've seen are good polls? That would be the cutoff line. If you don't trust at least that many of their polls, then you're being selective (either rightly or wrongly) and you don't "trust" the organization(s).
|
|
Deleted
Joined: May 18, 2024 3:38:16 GMT -5
Posts: 0
|
Post by Deleted on Jul 2, 2014 13:09:00 GMT -5
Then I trust no one.
|
|
Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 2, 2014 13:09:51 GMT -5
"None of the Above" is your vote then.
|
|
Opti
Community Leader
Joined: Dec 18, 2010 10:45:38 GMT -5
Posts: 39,779
Location: New Jersey
Mini-Profile Name Color: c28523
Mini-Profile Text Color: 990033
|
Post by Opti on Jul 2, 2014 13:19:48 GMT -5
Change we add to or change our answers until you lock your poll?
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 2, 2014 13:54:30 GMT -5
I want DJ to post his reasoning. I don't feel I can accurately take your poll as I judge more on a poll by poll basis, often looking at the questions if it seems a bit off to me.
For example, I think both the Pew Research center and Gallup do good polls. But I've also found ones they've done, I wasn't that keen on the slant or methodology.
i think all of the polls are good save one: Rasmussen. if you want to know why, i can post a link. since Rasmussen became a subsidiary of FOX, they have shown systematic bias and inaccuracy that is laughable. Gallup made some huge mistakes in the last presidential election- changing their methodology just before November. i give them demerits for that, but hope they learned their lesson. i don't really trust any poll more than another, with exceptions noted. i look for polls that are statistical outliers, and i tend to dismiss them.
|
|
sesfw
Junior Associate
Today is the first day of the rest of my life
Joined: Dec 21, 2010 15:45:17 GMT -5
Posts: 6,268
|
Post by sesfw on Jul 2, 2014 14:10:40 GMT -5
None of the above.
I've had the same poll call me several times within a week and each time I give a different answer.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 2, 2014 14:14:04 GMT -5
None of the above.
I've had the same poll call me several times within a week and each time I give a different answer. why? so you can post about how much you distrust polls?
|
|
Opti
Community Leader
Joined: Dec 18, 2010 10:45:38 GMT -5
Posts: 39,779
Location: New Jersey
Mini-Profile Name Color: c28523
Mini-Profile Text Color: 990033
|
Post by Opti on Jul 2, 2014 14:16:37 GMT -5
None of the above.
I've had the same poll call me several times within a week and each time I give a different answer. What is the poll about?
|
|
billisonboard
Community Leader
Joined: Dec 20, 2010 22:45:44 GMT -5
Posts: 37,512
|
Post by billisonboard on Jul 2, 2014 14:21:42 GMT -5
None of the above.
I've had the same poll call me several times within a week and each time I give a different answer. I find it hard to believe that any polling organization would call the same person\number multiple times in a week. Anyone have any knowledge of why they would do that?
|
|
billisonboard
Community Leader
Joined: Dec 20, 2010 22:45:44 GMT -5
Posts: 37,512
|
Post by billisonboard on Jul 2, 2014 14:29:58 GMT -5
"Trust but verify" should be an option in the poll.
|
|
Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 2, 2014 17:33:23 GMT -5
I should offer free candy for voting or something. We had hundreds of votes in the mod election. Do people care so little about polling--the eyes, ears, and very lifeblood of US politics? I'm not sure how you would "verify" poll results aside from consulting multiple polls and using a consensus algorithm such as adjusted median, etc. But if this is the case, you generally trust/don't trust polls and the answer to FAQ #6 applies. If you selectively reject poll results because they seem incredible, you don't trust the organization. Why would you do that? If you don't want them calling you, tell them. Polling outfits aren't like telemarketers. They'll take "no thanks" for an answer. Yes, and the poll won't be locked.
|
|
billisonboard
Community Leader
Joined: Dec 20, 2010 22:45:44 GMT -5
Posts: 37,512
|
Post by billisonboard on Jul 2, 2014 17:52:41 GMT -5
... I'm not sure how you would "verify" poll results aside from consulting multiple polls and using a consensus algorithm such as adjusted median, etc. But if this is the case, you generally trust/don't trust polls and the answer to FAQ #6 applies. If you selectively reject poll results because they seem incredible, you don't trust the organization. ... It is me so of course the comment was only half serious The half that was serious was that I like to go look at the poll itself and not rely on some news story or posting that tells me what the poll said.
|
|
sesfw
Junior Associate
Today is the first day of the rest of my life
Joined: Dec 21, 2010 15:45:17 GMT -5
Posts: 6,268
|
Post by sesfw on Jul 3, 2014 12:25:28 GMT -5
What is the poll about?
Political
Polling outfits aren't like telemarketers.
These were computer generated questions and you pushed a number for an answer. First time I said 'undecided' and the computer gave me a statement on how great this guy was and asked the question again. Same answer ........... therefore a different comment from computer.
Then I started playing games.
|
|
billisonboard
Community Leader
Joined: Dec 20, 2010 22:45:44 GMT -5
Posts: 37,512
|
Post by billisonboard on Jul 3, 2014 12:36:09 GMT -5
What is the poll about?
Political
Polling outfits aren't like telemarketers.
These were computer generated questions and you pushed a number for an answer. First time I said 'undecided' and the computer gave me a statement on how great this guy was and asked the question again. Same answer ........... therefore a different comment from computer.
Then I started playing games. So it was a push poll for a particular candidate. That is significantly different thing than an opinion poll. That is why I look at the poll itself instead of what someone says it says.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 13:35:29 GMT -5
i find this survey profoundly disturbing. but before i explain why: does anyone know the history of polling, by any chance?
ie: why was polling developed?
|
|
The Captain
Junior Associate
Hugs are good...
Joined: Jan 4, 2011 16:21:23 GMT -5
Posts: 8,717
Location: State of confusion
Favorite Drink: Whinnnne
|
Post by The Captain on Jul 3, 2014 13:40:00 GMT -5
I could be wrong, but wasn't polling first developed as a means to test or predict election outcomes?
I think it's since devolved (yes that is the word I choose) to find out popular opinion and tailor the message in a way designed to gather the most votes.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 13:56:54 GMT -5
I could be wrong, but wasn't polling first developed as a means to test or predict election outcomes? nope. it was way more benign and hopeful than that. I think it's since devolved (yes that is the word I choose) to find out popular opinion and tailor the message in a way designed to gather the most votes. i think that is a very negative view of polling, but concede that it is widely believed.
|
|
Lizard King
Senior Member
It's an anagram, you know.
Joined: Nov 6, 2013 16:22:24 GMT -5
Posts: 2,589
Favorite Drink: La Fee Verte
|
Post by Lizard King on Jul 3, 2014 14:25:25 GMT -5
What if you trust aggregates of polls to iron out the individual biases inherent in each polling organization's methodology?
Nate Silver's rep is based on the fact that he looks at all the polls. Granted, he trusts some more than others because he's making considered judgements of the validity of their methodology - he's not a fan of Pew or Rasmussen, so them saying a candidate is six points up doesn't cut as much ice with him as a more-favored pollster, but it does go in the mix. If eleven polling organizations are calling a race the same way, and one's not, that one had better have a pretty peerless record to justify following their lead rather than the consensus.
I tend to use RCP's poll aggregates as a very rough and ready go-to, personally. They don't weight contributions of individual polls, so they tend to be disproportionately driven by Gallup and Rasmussen on a lot of things just because Gallup and Rasmussen contribute more data points to their aggregate. I guess that means that indirectly, to some extent, I favor Gallup and Rasmussen, but the two of them are often so wildly divergent from one another and from the current consensus as reflected in the RCP average that I wouldn't be prepared to say so in the OP poll.
|
|
The Captain
Junior Associate
Hugs are good...
Joined: Jan 4, 2011 16:21:23 GMT -5
Posts: 8,717
Location: State of confusion
Favorite Drink: Whinnnne
|
Post by The Captain on Jul 3, 2014 15:38:07 GMT -5
I could be wrong, but wasn't polling first developed as a means to test or predict election outcomes? nope. it was way more benign and hopeful than that. I think it's since devolved (yes that is the word I choose) to find out popular opinion and tailor the message in a way designed to gather the most votes. i think that is a very negative view of polling, but concede that it is widely believed. Well, are you going to share how polling really began and how it was more hopeful? As far as my second point, yes I agree it's very negative but based on the polling questions I've been asked (would you vote for a democratic candidate if they were opposed to abortion) I don't think it's too far out of line. As far as if our government functions to serve itself, or it's constituency, remember I'm from Illinois, so please forgive me if I'm a bit jaded.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 15:45:29 GMT -5
i think that is a very negative view of polling, but concede that it is widely believed. Well, are you going to share how polling really began and how it was more hopeful? sure, but that will kill the discussion as well as any spoiler. i would rather see if anyone knows the history of polling before i post it.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 15:53:08 GMT -5
What if you trust aggregates of polls to iron out the individual biases inherent in each polling organization's methodology?
Nate Silver's rep is based on the fact that he looks at all the polls. Granted, he trusts some more than others because he's making considered judgements of the validity of their methodology - he's not a fan of Pew or Rasmussen, so them saying a candidate is six points up doesn't cut as much ice with him as a more-favored pollster, but it does go in the mix. If eleven polling organizations are calling a race the same way, and one's not, that one had better have a pretty peerless record to justify following their lead rather than the consensus.
I tend to use RCP's poll aggregates as a very rough and ready go-to, personally. They don't weight contributions of individual polls, so they tend to be disproportionately driven by Gallup and Rasmussen on a lot of things just because Gallup and Rasmussen contribute more data points to their aggregate. I guess that means that indirectly, to some extent, I favor Gallup and Rasmussen, but the two of them are often so wildly divergent from one another and from the current consensus as reflected in the RCP average that I wouldn't be prepared to say so in the OP poll. i know it is a left wing site, but Huffington weights polls in a way more scientific way than RCP does.
|
|
Lizard King
Senior Member
It's an anagram, you know.
Joined: Nov 6, 2013 16:22:24 GMT -5
Posts: 2,589
Favorite Drink: La Fee Verte
|
Post by Lizard King on Jul 3, 2014 15:57:31 GMT -5
Well, are you going to share how polling really began and how it was more hopeful? sure, but that will kill the discussion as well as any spoiler. i would rather see if anyone knows the history of polling before i post it. Please identify an arbitrary time limit after which you'll be satisfied nobody else is going to do it, so I can return then to see what you eventually have to actually say on the subject.
|
|
Lizard King
Senior Member
It's an anagram, you know.
Joined: Nov 6, 2013 16:22:24 GMT -5
Posts: 2,589
Favorite Drink: La Fee Verte
|
Post by Lizard King on Jul 3, 2014 15:59:15 GMT -5
I don't think RCP weights the polls at all.
I'm not really interested in how accurately a poll reflects what respondents told the pollsters.
I'm more interested in what impact the reportage of the poll has on the public.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 16:02:25 GMT -5
I don't think RCP weights the polls at all.
they do. they weight as if all polls are created equal. which is false.
I'm not really interested in how accurately a poll reflects what respondents told the pollsters.
I'm more interested in what impact the reportage of the poll has on the public.
i am not at all interested in the latter. but to each their own.
|
|
djAdvocate
Member Emeritus
only posting when the mood strikes me.
Joined: Jun 21, 2011 12:33:54 GMT -5
Posts: 75,233
Mini-Profile Background: {"image":"","color":"000307"}
|
Post by djAdvocate on Jul 3, 2014 16:04:25 GMT -5
sure, but that will kill the discussion as well as any spoiler. i would rather see if anyone knows the history of polling before i post it. Please identify an arbitrary time limit after which you'll be satisfied nobody else is going to do it, so I can return then to see what you eventually have to actually say on the subject. psssssssst.....there is this thing called the internet. if you go to this thing called a search engine, and look up "history of polling" you will only have to wait MILLISECONDS to find out.....if you HAVE to know. but to answer your question, i figure a day is about right for most people. most people browse the threads at least once a day. maybe tonight? but i have a gig until after midnight EST.
|
|
Virgil Showlion
Distinguished Associate
Moderator
[b]leones potest resistere[/b]
Joined: Dec 20, 2010 15:19:33 GMT -5
Posts: 27,448
|
Post by Virgil Showlion on Jul 3, 2014 19:46:40 GMT -5
According to Wiki, polling was originally used in the US for the purpose of determining election outcomes. If that's not really the case, give us the beef already, Socrates.
|
|