Replies: 15
| visibility 340
|
Oculus Spirit [97719]
TigerPulse: 100%
Posts: 64859
Joined: 7/13/02
|
You know, in the past 25 years I have never once
Feb 28, 2019, 8:14 AM
|
|
received a call from a national pollster. NEVER. I've received only 2 poll calls EVER, and they were push polls for local elections. I eventually sniffed those out and confronted the pollster and asked who he was conducting the poll for and they hung up. And I had a landline up until just a few years ago.
I have never been picked to serve on a jury. EVER. Three times on jury duty, never once picked to even make the first cut at being on a jury.
Anyway, this article reminded me of the above:
https://www.politico.com/story/2019/02/27/phone-polling-crisis-1191637
And trust me I'd JUMP at the opportunity to participate in a national political poll. Mainly because I've never been contacted and I know I would blow their curve.
Polls have been way off for a long time. I think I understand why.
|
|
|
|
All-In [42167]
TigerPulse: 100%
Posts: 38250
Joined: 11/30/98
|
If I don't recognize the number, I ain't answerin'.***
Feb 28, 2019, 8:21 AM
|
|
|
|
|
|
|
All-In [26968]
TigerPulse: 96%
Posts: 44823
Joined: 7/6/10
|
Everyone is responsible for using the information
Feb 28, 2019, 8:24 AM
|
|
correctly. The number of respondents influences the margin of error of the poll. If the response is low, it should be reflected in the margin of error. Most people form opinions from polls, without considering what the poll actually means.
If a poll says 55% of people like Donald Trump, and the margin of error is plus-minus 3%, then all we can say is that we can be 95% confident that a majority of people like Trump. It does not mean that it is a fact that more people like Trump.. The only way you know that is by asking the entire population. If the same poll has plus-minus 6%, then what that poll says is that we cannot be be 95% confident that most people like Trump. It definitely does not mean that more people like Trump.
|
|
|
|
|
All-In [26968]
TigerPulse: 96%
Posts: 44823
Joined: 7/6/10
|
The toughest thing is probably to get a good random sample.
Feb 28, 2019, 8:26 AM
|
|
Online polls are always self-selecting, and are pretty much trash.
I'm not an expert, but phone polling may border on self-selecting. Are certain people more likely to answer the phone than others?
|
|
|
|
|
Oculus Spirit [97719]
TigerPulse: 100%
Posts: 64859
Joined: 7/13/02
|
Even the margin for error is wrong IMO
Feb 28, 2019, 9:16 AM
[ in reply to Everyone is responsible for using the information ] |
|
I mean if half the people hang up, you have no clue the propensity for that hangup caller to get to a voting booth, or how they may vote. Your percentages of unemployed people answering and elderly will be skewed WAY high because they have time for a poll. Most people working don't have the time to waste, and/or just hate spam calls and don't even answer an unknown number.
Leaving a message with a callback number would be helpful but they don't do that.
|
|
|
|
|
All-In [26968]
TigerPulse: 96%
Posts: 44823
Joined: 7/6/10
|
Having people call back would obviously be self-selecting.
Feb 28, 2019, 9:18 AM
|
|
Wouldn't be any good.
|
|
|
|
|
Oculus Spirit [97719]
TigerPulse: 100%
Posts: 64859
Joined: 7/13/02
|
Wouldn't be self selecting because they already attempted to
Feb 28, 2019, 9:31 AM
|
|
call you first. You would at least grab a higher percentage of the hangups. There would have to be a code or something to enter for a one-time use or something to keep it legit, but those who call back would be a more likely person to go to the polls and vote because they care enough to call back.
Either way, polls are not accurate, and really have never been. Exit polling is much more accurate. But given the logistics of polling, there's no way they're accurate, no matter how many statistical alterations you do.
|
|
|
|
|
All-In [26968]
TigerPulse: 96%
Posts: 44823
Joined: 7/6/10
|
You don't want the people who call back to be
Feb 28, 2019, 9:35 AM
|
|
"more likely" to do anything. You want a representative, random sample of the population.
Polls are "accurate" according to the constraints of the polling, which most people looking at polls don't understand. No poll is ever or will ever be 100% "accurate", unless you literally poll every person in the population. That is not a poll...that is an election.
|
|
|
|
|
All-In [26968]
TigerPulse: 96%
Posts: 44823
Joined: 7/6/10
|
Even if a poll did show with 95% confidence
Feb 28, 2019, 9:38 AM
|
|
that event "A" was true, then if "A" isn't true, that doesn't mean the poll was wrong. That's why it's 95% confidence, not 100% confidence. There's no such thing as 100% confidence in statistics.
|
|
|
|
|
Heisman Winner [137937]
TigerPulse: 100%
Posts: 63812
Joined: 10/22/00
|
It's probably because Gallup, Harris, etc., could see that
Feb 28, 2019, 8:35 AM
|
|
you're on Spectrum, and they're like "yeah, this guy doesn't vote."
|
|
|
|
|
Oculus Spirit [97719]
TigerPulse: 100%
Posts: 64859
Joined: 7/13/02
|
Voted every election and midterm since graduating college***
Feb 28, 2019, 9:17 AM
|
|
|
|
|
|
|
Heisman Winner [111581]
TigerPulse: 100%
Posts: 73753
Joined: 9/10/03
|
Re: You know, in the past 25 years I have never once
Feb 28, 2019, 9:29 AM
|
|
sounds like by some sort of weird bureaucratic error, you are not an actual american citizen. Hold still while I call immigration to carry you off to mexico.
|
|
|
|
|
CU Guru [1405]
TigerPulse: 54%
Posts: 3412
Joined: 1/10/13
|
Re: You know, in the past 25 years I have never once
Feb 28, 2019, 9:31 AM
|
|
I was called by Gallup once. It was approve/disapprove questions about W's job performance. And I've been called two other times by other national polls, but I can't remember which ones.
I got a push poll on behalf of Mark Sanford in 2002 and when I told the pollster that her facts were incorrect, she hung up.
|
|
|
|
|
Oculus Spirit [97719]
TigerPulse: 100%
Posts: 64859
Joined: 7/13/02
|
Think I got that same push poll.***
Feb 28, 2019, 9:49 AM
|
|
|
|
|
|
|
Oculus Spirit [81061]
TigerPulse: 100%
Posts: 56140
Joined: 9/13/04
|
I had jury duty last week.
Feb 28, 2019, 9:55 AM
|
|
Some folks in there had been on jury duty 2 or 3 times in their life.
|
|
|
|
|
110%er [9664]
TigerPulse: 100%
Posts: 11415
Joined: 9/10/99
|
It's funny to me that news agencies still cite polls
Feb 28, 2019, 10:17 AM
|
|
after the 2016 election.
|
|
|
|
Replies: 15
| visibility 340
|
|
|