Aces High Bulletin Board
General Forums => The O' Club => Topic started by: blur on March 22, 2003, 08:30:43 AM
-
A recent New York Times headline declares:
“Support for Bush Surges at Home, but Split Remains, Poll Shows”
They then go on to say that 70% of Americans support the President.
Farther down they say that 463 people were questioned and that there’s an error of +/- 5 percent. This error rate is compared to what? Wouldn’t you have to ask all 282 million Americans then compare the results to the sample?
I did a recent poll of my own and discovered that 100 percent of polls were horse****. The error rate is +/- 0 percent.
-
Yeah, especially when they don't agree with your position, eh?
:D
-
I did a recent poll too and found that 100% agree. Ain't that right Rude, Toad, Milo?
-
actually, blur, you don't have to ask everyone to get a representitive sample of everyone's opionion
you can ask about 1200 people, (truly random) and get a projection of all 280 million
any statistical analysis majors know the exact number?
-
With the help of media, you can easily get people to support wars.
Goverment just spouts BS to the media some BS and they send it to the people.
Basics of the propaganda at works... great work US gov.
Of course people are too blind to notice how easily the goverment, with the help of media, can affect their personal opinion.
No conspiracy things needed, thats the fact how goverments works in democratic countries today.
They wanted to go to Iraq and had a longish campaign to persuade people to support it.. the goverment will always win the opposing side in this matter.
However persuading other countries isn't just as easy, since the local medias out there doesn't care to report all the US goverments doings, like it would matter to the people of their countries.
-
Hawk220 is right. If you keep the sampling completely random, a small sampling will be representative of a VERY large group. You just couldn't have all 500 people be from NYC or the Midwest or whatever. The numbers are actually the easy part. Keeping everything truly random is the hard part.
MiniD
-
Originally posted by Fishu
Of course people are too blind to notice how easily the goverment, with the help of media, can affect their personal opinion.
Unless, of course, you agree with that opinion, at which point you arrived at that conclusion through intellegent consideration of the data. Manipulation only works when you dissagree lol.
-Sik
-
Originally posted by hawk220
you can ask about 1200 people, (truly random) and get a projection of all 280 million
any statistical analysis majors know the exact number?
You should be able to achieve a generalizable sample with as few as 40 respondents, but you'll have a large margin of error. Basically, the more people you interview, the smaller the margin of error with diminishing returns. For most polls, this is a cost consideration; while it's possible to achieve a +/- 1% margin of error (or less), you'll need to interview thousands of people to do so. If you're willing to deal with +/- 3% error, you'll only need around 1000 people. *
As it requires magnitudes more respondents to achieve the +/- 1% margin of error, from a cost perspective it is much more attractive to go with fewer respondents and a slightly higher margin of error.
All of this assumes truly random sampling.
* These numbers are just off of the top of my head. I can provide the actual sampling numbers required if you're really interested in that.
-- Todd/Leviathn
-
Originally posted by Fishu
However persuading other countries isn't just as easy, since the local medias out there doesn't care to report all the US goverments doings, like it would matter to the people of their countries.
And of course, the people in those other countries are also subject to propaganda from their own governments, governments that are no less controlled by business interests than the US government, governments whose business masters may or not have a vested interest in keeping Saddam in power.
-
First rule of statistics... you can't use convenience samples and opinion polls and then apply them to the population.
-
Originally posted by funkedup
And of course, the people in those other countries are also subject to propaganda from their own governments.
More people in UK are against the war than for it ;)
-
Originally posted by Fishu
More people in UK are against the war than for it ;)
It's only because they are so close to France. :)
-
Originally posted by Fishu
More people in UK are against the war than for it ;)
Yea? How do you know? Did they count everybody? Or did they take one of these unbelivale polls... I cant believe they reversed position right in the thread when they found a poll they agreed with...