- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
How can Fox have a national poll that only samples 1000 people?
Posted on 8/14/20 at 10:04 am
Posted on 8/14/20 at 10:04 am
They open every segment this morning touting their new national poll.
Posted on 8/14/20 at 10:05 am to loogaroo
They call till they can get data from 1000 people.
This post was edited on 8/14/20 at 10:05 am
Posted on 8/14/20 at 10:06 am to loogaroo
It's "registered voters".
Posted on 8/14/20 at 10:06 am to loogaroo
1000 people is enough. It’s that they over sample Dems and undersample Republicans
Posted on 8/14/20 at 10:06 am to redneck hippie
Lies and damn lies...
Posted on 8/14/20 at 10:07 am to loogaroo
1000 is not the problem. The problem is that it uses registered voters rather than the more accurate likely voters. Merely using registered voters usually favors democrat candidates.
Posted on 8/14/20 at 10:09 am to loogaroo
Are you not familiar with statistics and sample size?
Posted on 8/14/20 at 10:18 am to oogabooga68
quote:
Lies and damn lies...
Nice reference.
Posted on 8/14/20 at 10:19 am to MikeBRLA
Hard for me to believe they reach an accurate representation with only 20 people per state.
Posted on 8/14/20 at 10:22 am to MikeBRLA
quote:Statistics really should be a mandatory course for every major, even at the 101 level like Econ or Eng. Even a Statistics Terminology class at the least so people can understand what is being talked about.
Are you not familiar with statistics and sample size?
Posted on 8/14/20 at 10:22 am to loogaroo
The posters always give themselves plausible deniability by stating, "as the race stands now" but the people reporting on the poll promote the illusion that this poll forecasts the result of election day.
In other words, the poll's only purpose is to provide subject matter for pundits and media personalities to talk about.
In other words, the poll's only purpose is to provide subject matter for pundits and media personalities to talk about.
Posted on 8/14/20 at 10:24 am to loogaroo
They never reveal the methodology, and they never mention Rasmussen anymore, because it doesn't fit their newfound dogma.
Posted on 8/14/20 at 10:26 am to SirWinston
quote:
1000 people is enough. It’s that they over sample Dems and undersample Republicans
This, and it was a national poll of registered, not likely voters.
Posted on 8/14/20 at 10:27 am to loogaroo
quote:
How can Fox have a national poll that only samples 1000 people?
Gaussian distribution, yo.
Posted on 8/14/20 at 10:28 am to loogaroo
Kushner was on Fox saying almost all these polls are bullshite because they greatly undersample Republicans. Some as low as 24%
This post was edited on 8/14/20 at 10:28 am
Posted on 8/14/20 at 10:29 am to loogaroo
Might have interviewed 3000 and picked the 1000 they wanted to use.
Posted on 8/14/20 at 10:31 am to loogaroo
Has to do with sample size and margin of error.
There comes a point where you reach a saturation level, and more poll respondents don’t change the outcome.
There comes a point where you reach a saturation level, and more poll respondents don’t change the outcome.
Posted on 8/14/20 at 10:34 am to loogaroo
quote:
How can Fox have a national poll that only samples 1000 people?
That's a lot more than what's considered to be a sufficient sample size.
Posted on 8/14/20 at 10:41 am to TheBoo
You don't need a whole lot of precision at this time and point of the election, as long as your sample size is accurate and representative (which it may not be in this case.)
1,000 voters yields an error of plus-or-minus 3 percent, which is reasonable for August. It can't tell you if one candidate is ahead by a razor margin, but it can tell you if one candidate is comfortably ahead.
1,000 voters yields an error of plus-or-minus 3 percent, which is reasonable for August. It can't tell you if one candidate is ahead by a razor margin, but it can tell you if one candidate is comfortably ahead.
Back to top
Follow TigerDroppings for LSU Football News