RadioFreeLiberal.com

Smart Voices, Be Heard
It is currently Sun Nov 18, 2018 8:29 am

All times are UTC - 5 hours [ DST ]


Forum rules


Please click here to view the forum rules



Post new topic Reply to topic  [ 31 posts ]  Go to page Previous  1, 2
Author Message
PostPosted: Wed Nov 23, 2016 3:21 pm 
Offline
Board Emeritus
User avatar

Joined: Thu Jun 09, 2011 1:10 pm
Posts: 17725
Location: The blue parts of the map
The inconvenient truth here is that, were exit polls completely accurate before 2004, Tom Bradley would have been governor of California.

The Wiki entry at Carmen's link is quite accurate, and it describes something fairly close to a latter day Dewey Defeats Truman.

Quote:
Based on exit polls, a number of media outlets projected Bradley as the winner and early editions of the next day's San Francisco Chronicle featured a headline proclaiming "Bradley Win Projected." However, despite winning a majority of the votes cast on election day, Bradley narrowly lost the overall race once absentee ballots were included.[9] Post-election research indicated that a smaller percentage of white voters actually voted for Bradley than polls had predicted, and that previously undecided voters had voted for Deukmejian in statistically anomalous numbers.


The last exit poll I was in consisted of someone in an official looking business type of getup with a bunch of papers and stuff identifying as L.A. Times, and asking people who'dja vote for. A non-trivial number of people said things along the lines of, "none of your fracking business." Right away, you have an imperfect sample, and you haven't even taken time of day and such into account.

My real problem with polling, though, is that it looks too much like physical science. It isn't. It does return reproducible results and all that. Looks all impressive and sexy, like verifying the Higgs boson. All it means, though, is that you can do a faulty experiment over and over, and get the same faulty results every time.

_________________
We used to hang our traitors. Now we elect them to lead us.


Top
 Profile  
 
PostPosted: Wed Nov 23, 2016 3:27 pm 
Offline
Board Emeritus
User avatar

Joined: Wed Jun 08, 2011 11:53 am
Posts: 14382
The inconvenient truth here is that, were exit polls completely accurate before 2004, Tom Bradley would have been governor of California.

The Wiki entry at Carmen's link is quite accurate, and it describes something fairly close to a latter day Dewey Defeats Truman.



The last exit poll I was in consisted of someone in an official looking business type of getup with a bunch of papers and stuff identifying as L.A. Times, and asking people who'dja vote for. A non-trivial number of people said things along the lines of, "none of your fracking business." Right away, you have an imperfect sample, and you haven't even taken time of day and such into account.

My real problem with polling, though, is that it looks too much like physical science. It isn't. It does return reproducible results and all that. Looks all impressive and sexy, like verifying the Higgs boson. All it means, though, is that you can do a faulty experiment over and over, and get the same faulty results every time.



NOPE

back up



Quote:
argues that the exit polls were wrong because Bradley actually won on election day turnout, but lost the absentee vote


Exit polls are accurate unless Karl Rove and friends are involved. Or whoever the new version of that is.

_________________
"Corporate Democrat" phrase created at the same place "Angry Mob" was...People keep falling for rightwing talking points. How sad.


Top
 Profile  
 
PostPosted: Wed Nov 23, 2016 4:10 pm 
Online
Board Emeritus
User avatar

Joined: Thu Jul 02, 2015 6:57 pm
Posts: 12316
Location: Sunny South Florida
The inconvenient truth here is that, were exit polls completely accurate before 2004, Tom Bradley would have been governor of California.


Well, interestingly enough, exit polls are used as a tool for checking the integrity of foreign elections. In fact, several times, in recent history, when the U.S. has alleged election fraud in other countries, it is sometimes based on separation between final results and .... the exit polls.

Now, that said, yes, one can read Edison Research's web site, and they admit an honest distinction. Edison is a for-profit private corporation. Unlike NGOs that do exit polling in other countries to check the accuracy/integrity of the vote, Edison is only interesting in gathering demographic voter data to provide to the media and other corporations for sale. They say themselves they do not have any intention or purpose to check the integrity of the election; that's not their function; they therefore warn other people not to use their un-adjusted data that way.

The NGOs in other countries usually do very quick and easy spot checks for exit polling - often asking voters only 1 question about 1 race. More participate, since it doesn't take much time. Edison hands you a long questionnaire. And let's face it, regardless of who you voted for, after already filling out 1 long ballot, who wants to sit down and fill out 1 long exit questionnaire, especially since they don't pay you for your time. Yeah, non participation will be high, and they say as much.

The other problem is - as I already said - if 40% of the populace has already early voted or by absentee, exit polling on election day won't sample that at all. And look, the Dem party has really aggressively been encouraging people to vote early, for good reasons, but ... then this will also affect reliability of exit polls.

Honestly, we can bitch and moan about who doesn't vote - lord nose I do sometimes :mrgreen: - but the U.S. could join the rest of the rational world, and make election day a weekend, or a holiday. And make it easier for the working folk.

_________________
-- Tis an ill wind that blows no minds.
Malaclypse the Younger


Top
 Profile  
 
PostPosted: Thu Nov 24, 2016 4:06 pm 
Offline
Board Emeritus
User avatar

Joined: Thu Jun 09, 2011 1:10 pm
Posts: 17725
Location: The blue parts of the map
The inconvenient truth here is that, were exit polls completely accurate before 2004, Tom Bradley would have been governor of California.

Now, this is interesting about the limitations of a poll conducted at the voting place. Yes, I can sort of remember when the late results in California favored Republicans, because their older demographic tended to vote absentee in greater numbers. This flip flopped somewhere along the line. Now it seems to work the other way most of the time. I would imagine that we're definitely looking at the increased inconvenience of voting in person as a cause here.

The private exit polls are a royal pain in the ass, with all their stuff on what influenced your vote and demographics and shit. You can tell that it's about marketing, like just about everything else is right now. Given the inconvenience of voting, I can definitely see some skew in the data returned. Maybe the pollsters account for this. Maybe they don't.

_________________
We used to hang our traitors. Now we elect them to lead us.


Top
 Profile  
 
PostPosted: Thu Nov 24, 2016 4:46 pm 
Online
Board Emeritus
User avatar

Joined: Thu Jul 02, 2015 6:57 pm
Posts: 12316
Location: Sunny South Florida
The inconvenient truth here is that, were exit polls completely accurate before 2004, Tom Bradley would have been governor of California.


Since you repeated that twice,

https://en.wikipedia.org/wiki/Bradley_effect

Inaccurate polling statistics attributed to the Bradley effect are not limited to pre-election polls. In the initial hours after voting concluded in the Bradley-Deukmejian race in 1982, similarly inaccurate exit polls led some news organizations to project Bradley to have won.[47] Republican pollster V. Lance Tarrance, Jr. argues that the exit polls were wrong because Bradley actually won on election day turnout, but lost the absentee vote.

[snip][end]

Just to remind, this is is a problem I also mentioned myself.

It would indicate not that the exit pollsters are creating the skew, but instead simply are not counting the not-at-the-polls early/absentee vote.

_________________
-- Tis an ill wind that blows no minds.
Malaclypse the Younger


Top
 Profile  
 
PostPosted: Thu Nov 24, 2016 5:49 pm 
Offline
Board Emeritus
User avatar

Joined: Thu Jun 09, 2011 1:10 pm
Posts: 17725
Location: The blue parts of the map
Yes, I've said it three times now. We have learned from the climate mess that inconvenient truths need a lot of repetition. However, I'll admit it's kind of obnoxious, so I'll knock it off.

I've seen various explanations of what happened with Tom Bradley. It's possible that the absentee ballots did swing it. Like I said before, I believe that they used to skew strongly Republican. In this present election, they seem to be going the other way.

There's also a theory that people just outright lied on polls, due to white guilt or whatever. One of the reasons I keep mentioning faulty experiment design is that there's no way to check the data. At least I don't think there's a way. When this many people are involved, it's hard to isolate the relevant variables. It seems as if the major check seems to be just comparing the data with that returned by other polls. This passes the test of reproducible results, but if everyone's doing the same flawed experiment, you're not getting much from that.

Nobody could reproduce the RAND/L.A. Times poll's results, and it was considered an outlier barely worth mentioning. This was true in California, but they actually had a handle on something happening in some other places, however vague. Reproducible results aren't as reliable a check as in, say, particle physics, even though both are using statistics.

I think we can all agree that polling is a lot better than in the Dewey Defeats Truman era. Even so, it's still not the holy writ of God handed down on stone tablets and read (in Latin) by a guy in a robe standing on a Doric column labeled "TRVTH." People have to get over this mystical trust in numbers.

_________________
We used to hang our traitors. Now we elect them to lead us.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 31 posts ]  Go to page Previous  1, 2

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 7 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group