Do you think polls are and effective way to tell what Americans believe? (1 Viewer)

[Navy Pride]

Your comments please:

 

[Liberal Because I Care]

I voted "no", as polls don't tell people what to believe but rather they measure what people already believe.

 

[Kandahar]

Generally yes. As long as the poll is conducted scientifically and the questions are as straightforward and unbiased as possible, what's the problem?

 

[Hoot]

Depending on whether the polling company is reputable, I voted yes...

...plus or minus 3%! LOL

 

[Navy Pride]

I voted no........My reasons are:

1. It depends where you poll, for example if you had a poll as to the Presidents approval rating and you polled in Seattle Wash. his approval rating would be about 10%.........If you had the same poll outside the gate at the Naval Submarine base bangor his approval rating would be about 80%....

2. It is impossible to poll 1000 people when we have 300,000,000 people in this country and get and accurate approval rating.......

Nothing proved how inaccurate polling is more then when the exit polls in the 2004 presidential elections had Kerry winning in a landslide.........

 

[Kandahar]

I voted no........My reasons are:

1. It depends where you poll, for example if you had a poll as to the Presidents approval rating and you polled in Seattle Wash. his approval rating would be about 10%.........If you had the same poll outside the gate at the Naval Submarine base bangor his approval rating would be about 80%....

Generally polls specify the group being polled. If the polling company polled exclusively in Seattle and claimed that the poll represented the entire nation, they'd be wrong. If they polled exclusively in Seattle and claimed that the poll represented Seattle, they'd most likely be correct.

That doesn't mean that polls don't work, it just means that people have different opinions in different parts of the country.

2. It is impossible to poll 1000 people when we have 300,000,000 people in this country and get and accurate approval rating.......

Some basic statistics/probability would indicate that you are wrong about that.

Nothing proved how inaccurate polling is more then when the exit polls in the 2004 presidential elections had Kerry winning in a landslide.........

Umm...Please cite an exit poll that showed that. The exit polls matched the actual election results very well in 2004, both state-by-state and nationwide.

 

[Liberal Because I Care]

<deleted post>

 

[Navy Pride]

Generally polls specify the group being polled. If the polling company polled exclusively in Seattle and claimed that the poll represented the entire nation, they'd be wrong. If they polled exclusively in Seattle and claimed that the poll represented Seattle, they'd most likely be correct.

That doesn't mean that polls don't work, it just means that people have different opinions in different parts of the country.



Some basic statistics/probability would indicate that you are wrong about that.



Umm...Please cite an exit poll that showed that. The exit polls matched the actual election results very well in 2004, both state-by-state and nationwide.

Generally polls specify the group being polled. If the polling company polled exclusively in Seattle and claimed that the poll represented the entire nation, they'd be wrong. If they polled exclusively in Seattle and claimed that the poll represented Seattle, they'd most likely be correct.

Agreed that is eactly why you can't take a poll of 1000 people that says President Bush has a low approval rating and claim it reflects the feelings of 300,000,000 Americans....

 

[tecoyah]

Nothing proved how inaccurate polling is more then when the exit polls in the 2004 presidential elections had Kerry winning in a landslide.........

I also voted no....Polls are far too dependent on the way a question is worded. But the above statement is inaccurate....unless this is a landslide :


Florida
Kerry 51
Bush 49

Ohio
Kerry 51
Bush 49

Michigan
Kerry 52
Bush 46
Nader 1

Pennsylvania
Kerry 53
Bush 46

Iowa
Kerry 50
Bush 49

Wisconsin
Kerry 51
Bush 48
Nader 1

Minnesota
Kerry 52
Bush 46
Nader 2

New Hampshire
Kerry 54
Bush 44
Nader 1

New Mexico
Kerry 50
Bush 48
Nader 1

Colorado
Kerry 49
Bush 50
Nader 1

Arkansas
Kerry 45
Bush 54
Nader 1

Missouri
Kerry 47
Bush 52

New York
Kerry 62
Bush 36
Nader 2

Nevada
Kerry 49
Bush 48
Nader 1

New Jersey
Kerry 54
Bush 44
Nader 1

West Virginia
Kerry 45
Bush 54
Nader 1

 

[Kandahar]

Agreed that is eactly why you can't take a poll of 1000 people that says President Bush has a low approval rating and claim it reflects the feelings of 300,000,000 Americans....

Taking 1,000 people from Seattle and claiming they represent all of America, is not the same thing as taking 1,000 random people across the nation and claiming they represent all of America. The first would be misleading, the latter is statistically accurate.

 

[Navy Pride]

Taking 1,000 people from Seattle and claiming they represent all of America, is not the same thing as taking 1,000 random people across the nation and claiming they represent all of America. The first would be misleading, the latter is statistically accurate.

I don't care if you divide the 1000 by 50 and take 20 from each state there is no way in hell you can know how the rest of the people in that state would vote........

Its just a wild guess at best..........

 

[Kandahar]

I don't care if you divide the 1000 by 50 and take 20 from each state there is no way in hell you can know how the rest of the people in that state would vote........

Its just a wild guess at best..........

No it isn't. A basic understanding of statistics will show exactly how pollsters come up with the margin of error and what it means. They don't just pull that number out of a hat...

If polls were "just a wild guess at best," you'd expect to see election results differing wildly from the final polls (as you claimed earlier in this thread). But we in fact see just the opposite: election results tend to be very similar to the poll results.

 

[Navy Pride]

No it isn't. A basic understanding of statistics will show exactly how pollsters come up with the margin of error and what it means. They don't just pull that number out of a hat...

If polls were "just a wild guess at best," you'd expect to see election results differing wildly from the final polls (as you claimed earlier in this thread). But we in fact see just the opposite: election results tend to be very similar to the poll results.

Then how do you explain when they are so far off?

 

[Kelzie]

Then how do you explain when they are so far off?

They were off by 2%. All reputable polls have a margin of error. The error in the exit polls was well within that margin.

 

[tecoyah]

How do you know they "Were so far off"....did you take another poll....or just pull that out of your hat?

 

[Kandahar]

Then how do you explain when they are so far off?

It's very rare that poll results differ significantly from election results. When they do, it is generally only in statewide contests where A) there was only one company polling, and B) the methodology was flawed or unscientific.

Perhaps if you cited some examples of these many polls that are "so far off" from the election results? The 2004 presidential polls were very close to the actual result; I know of none that predicted the "Kerry landslide" you claim.

 

[Navy Pride]

It's very rare that poll results differ significantly from election results. When they do, it is generally only in statewide contests where A) there was only one company polling, and B) the methodology was flawed or unscientific.

Perhaps if you cited some examples of these many polls that are "so far off" from the election results? The 2004 presidential polls were very close to the actual result; I know of none that predicted the "Kerry landslide" you claim.

Then you must not remember the exit polling early on election day that had the Kerry people celebrating a certain victory....

I was watching the election returns and all the pundits said Kerry was on the way to a huge victory...

 

[Navy Pride]

They were off by 2%. All reputable polls have a margin of error. The error in the exit polls was well within that margin.

Is that so.......The experts might disagree with you on that one..


http://www.usatoday.com/news/politicselections/vote2004/2004-11-03-polls-burn-pundits_x.htm

Predictions burn pollsters, pundits — again
By Mark Memmott, USA TODAY
The polling industry is on the defensive for its surveys of voters on Election Day, which some TV pundits relied on Tuesday to overstate support for Sen. John Kerry.

 

[Kandahar]

Then you must not remember the exit polling early on election day that had the Kerry people celebrating a certain victory....

I was watching the election returns and all the pundits said Kerry was on the way to a huge victory...

Hmm, I'm going to call bullshit on this, as it is simply not true. Here are all of the major polls leading up to the 2004 election:

http://www.pollingreport.com/2004.htm

 

[Kandahar]

Is that so.......The experts might disagree with you on that one..


http://www.usatoday.com/news/politicselections/vote2004/2004-11-03-polls-burn-pundits_x.htm

Predictions burn pollsters, pundits — again
By Mark Memmott, USA TODAY
The polling industry is on the defensive for its surveys of voters on Election Day, which some TV pundits relied on Tuesday to overstate support for Sen. John Kerry.

Of the ten polls listed in that article, seven of them showed Bush leading just prior to the election. One showed Bush and Kerry tied, and two showed Kerry ahead by 1% or 2%. None of them show the Kerry landslide you claim.

 

[Navy Pride]

Hmm, I'm going to call bullshit on this, as it is simply not true. Here are all of the major polls leading up to the 2004 election:

http://www.pollingreport.com/2004.htm

whatever, so much for polling reports.This is what happened......

http://www.usatoday.com/news/politic...-pundits_x.htm

Predictions burn pollsters, pundits — again
By Mark Memmott, USA TODAY
The polling industry is on the defensive for its surveys of voters on Election Day, which some TV pundits relied on Tuesday to overstate support for Sen. John Kerry.

 

[Navy Pride]

Of the ten polls listed in that article, seven of them showed Bush leading just prior to the election. One showed Bush and Kerry tied, and two showed Kerry ahead by 1% or 2%. None of them show the Kerry landslide you claim.

I will say this one more time........Prior to the elections (I am not saying the day before) the liberal media was full of glee because the president's approval rating was in the mid forties and no president had ever been re elected with and approval rating of below 50%............

So much for polls on approval ratings.........

 

[Connecticutter]

I don't care if you divide the 1000 by 50 and take 20 from each state there is no way in hell you can know how the rest of the people in that state would vote........

Its just a wild guess at best..........

Actually the margin of error is based upon the number of people polled, and is not related to the total number of people in the country. Theoretically, if you have a pool of an infinite amount of people, you can still gain a great deal of information on the distribution of those people by polling 1000 randomly. There's a formula that relates the number polled to statistical error.

Of course, a poll is no good if you don't select randomly. No poll is completely random - for example a phone poll would select random phone numbers, which is a little different from random people.

 

[Kelzie]

I will say this one more time........Prior to the elections (I am not saying the day before) the liberal media was full of glee because the president's approval rating was in the mid forties and no president had ever been re elected with and approval rating of below 50%............

So much for polls on approval ratings.........

What? There just saying it had never been done before. The fact still remained that his approval rating was in the mid 40s, regardless of if he was voted in or not.

 

[Navy Pride]

I might also add that in a lot of polls it depends what questions are asked and how they are asked........