Polling samples

[depotoo]

LOL!!!!!!!!!!!!

President Obama 332 Governor Romney 206

RealClearPolitics - 2012 Election Maps - Electoral Map No Toss Ups

And if Romney keeps shooting his mouth off, the President may get over 400 electorial votes.

and those maps are nothing more than a repeat of what the oversampled polls show as realclearpolitics does not decipher their sampling of the polls.

Look at this new one from the Washington Post regarding his lead in Virginia, it says he is ahead, yet when you click on show details this is what you find -

Q: (Among likely voters) If the presidential election were being held today and the candidates were (Barack Obama and Joe Biden, the Democrats) and (Mitt Romney and Paul Ryan, the Republicans), for whom would you vote?
Published: September 18

*Small sample size (99) for results among respondents with "No religion."
The Washington Post poll


If you didn't click on show details, you would have completely missed it.

(Among likely voters) If the presidential election were being held today and the candidates were (Barack Obama and Joe Biden, the Democrats) and (Mitt Romney and Paul Ryan, the Republicans), for whom would you vote? - The Washington Post

 

[Saigon]

*Small sample size (99) for results among respondents with "No religion."
The Washington Post poll


If you didn't click on show details, you would have completely missed it.

Well, I call that a good catch.

That is very odd.

Well posted, Depoto!

 

[Liability]

LOL!!!!!!!!!!!!

President Obama 332 Governor Romney 206

RealClearPolitics - 2012 Election Maps - Electoral Map No Toss Ups

And if Romney keeps shooting his mouth off, the President may get over 400 electorial votes.

BS meter ^ is sparking!

The ACTUAL state of affairs, as things presently stand WITH toss ups (to the extent the polling is not reflective of the very bullshit noted in the OP):

RealClearPolitics - 2012 Election Maps - Obama vs. Romney Create Your Own Electoral Map

 

[rightwinger]

Yes, I agree more dems are being polled, however how do you know what the ratio of dem/pub/independent should be?

You don't

Any poll that sets out to poll an equal number of Democratic and Republican voters is invalid

What a poll does is sample 1000 voters. It does not try to sample 500 Republicans and 500 Democrats....just 1000 random voters

If it turns out that 550 of those voters are Democrat it does not invalidate the poll

You are an idiot.

If those being polled overwhelmingly support one party over the other then the data will not be an accurate representation.

A more accurate poll would poll 33.333 percent of each democrat, republican and independent.

You are defining your results then aren't you?

A statistical sample has to be representative of its objective population. If the sample is indeed random then you get what you get. If ALL polls, conducted by different Polling agencies, conducted week after week are showing a majority of Democrats then that is your population

 

[rightwinger]

LOL!!!!!!!!!!!!

President Obama 332 Governor Romney 206

RealClearPolitics - 2012 Election Maps - Electoral Map No Toss Ups

And if Romney keeps shooting his mouth off, the President may get over 400 electorial votes.

BS meter ^ is sparking!

The ACTUAL state of affairs, as things presently stand WITH toss ups (to the extent the polling is not reflective of the very bullshit noted in the OP):

RealClearPolitics - 2012 Election Maps - Obama vs. Romney Create Your Own Electoral Map

Still doesn't look good for Romney and has shifted more in Obamas favor in recent weeks

 

[AmyNation]

Yes, I agree more dems are being polled, however how do you know what the ratio of dem/pub/independent should be?

You don't

Any poll that sets out to poll an equal number of Democratic and Republican voters is invalid

What a poll does is sample 1000 voters. It does not try to sample 500 Republicans and 500 Democrats....just 1000 random voters

If it turns out that 550 of those voters are Democrat it does not invalidate the poll

You are an idiot.

If those being polled overwhelmingly support one party over the other then the data will not be an accurate representation.

A more accurate poll would poll 33.333 percent of each democrat, republican and independent.

How would that be more accurate? Is the country made up of 33% dems, 33% pubs and 33% independents?

 

[healthmyths]

..Polling samples are skewing results to Obama's favor.
Some polls are as much as +15% democrat responders.
Also many polls are "registered voters" rather than "likely voters". The latter category tend to vote GOP.
Obama leads in skewed CBS poll weighted 13 percent in favor of Democrats - Spokane Conservative | Examiner.com
How Biased Are Polls? - Katrina Trinko - National Review Online..
The liberal media and of course the pollsters know they must change the narrative. They know the President cannot run on his record. Polls are being used as a weapon to discourage likely GOP voters to give up and stay away from the voting booth.
It will not work.

That's it....it's the poor polling data. I am sure it has NOTHING to do with Romney's constant gaffes, nor his out of control campaign workers....

Let's see ... If I asked 100 Democrats who is better Obama or Romney I would stake my life that at least 70% would say Obama! NO brainer right.
And If I asked 50 GOP who is better at least 70% would say Romney.. again no brainer!

So if I add 70 Democrats for Obama and 15 GOP that adds up to 85 for Obama and 65 for Romney!

SEE that's the point.. I know this has to be simplified for MATH challenged like you ..
WHEN YOU ask MORE Democrats then GOP YOU WILL get a "SAMPLING" error!

AND this has been going on for several decades... i.e. Polls are slanted with MORE democrats!

AND IN spite of that... Obama is doing worse!
YOU can't conclude Obama will WIN because MOrE democrats are asked then GOP.. IT'S that simple!

 

[Liability]

LOL!!!!!!!!!!!!

President Obama 332 Governor Romney 206

RealClearPolitics - 2012 Election Maps - Electoral Map No Toss Ups

And if Romney keeps shooting his mouth off, the President may get over 400 electorial votes.

BS meter ^ is sparking!

The ACTUAL state of affairs, as things presently stand WITH toss ups (to the extent the polling is not reflective of the very bullshit noted in the OP):

RealClearPolitics - 2012 Election Maps - Obama vs. Romney Create Your Own Electoral Map

Still doesn't look good for Romney and has shifted more in Obamas favor in recent weeks

LOL.

It looks pretty damn good for Romney and hideous for the incumbent precisely because the toss up states are going to go to Romney.

And it is silly and false to claim that things have recently gotten BETTER for The ONE. The stupid cow hasn't yet felt the impact of the voters' distaste for how horribly he has handled the Libyan and Egyptian attacks.

I realize it boggles your minds, libs, but I kind of enjoy rubbing your noses in your own poo:

President Obama is going to lose his bid for re-election.

 

[Lonestar_logic]

Lonstar -

Rightwinger is correct, IMHO.

If one party has 70% real world support, then 70% of the people I call should be his supporters, correct?

But no party has 70 percent real world support.

 

[Saigon]

SEE that's the point.. I know this has to be simplified for MATH challenged like you ..
WHEN YOU ask MORE Democrats then GOP YOU WILL get a "SAMPLING" error!

What if there ARE more Democrats than GOP?


I do love it when posters go on about people being maths-challenged...when they are wrong.

 

[rightwinger]

..Polling samples are skewing results to Obama's favor.
Some polls are as much as +15% democrat responders.
Also many polls are "registered voters" rather than "likely voters". The latter category tend to vote GOP.
Obama leads in skewed CBS poll weighted 13 percent in favor of Democrats - Spokane Conservative | Examiner.com
How Biased Are Polls? - Katrina Trinko - National Review Online..
The liberal media and of course the pollsters know they must change the narrative. They know the President cannot run on his record. Polls are being used as a weapon to discourage likely GOP voters to give up and stay away from the voting booth.
It will not work.

That's it....it's the poor polling data. I am sure it has NOTHING to do with Romney's constant gaffes, nor his out of control campaign workers....

Let's see ... If I asked 100 Democrats who is better Obama or Romney I would stake my life that at least 70% would say Obama! NO brainer right.
And If I asked 50 GOP who is better at least 70% would say Romney.. again no brainer!

So if I add 70 Democrats for Obama and 15 GOP that adds up to 85 for Obama and 65 for Romney!

SEE that's the point.. I know this has to be simplified for MATH challenged like you ..
WHEN YOU ask MORE Democrats then GOP YOU WILL get a "SAMPLING" error!

AND this has been going on for several decades... i.e. Polls are slanted with MORE democrats!

AND IN spite of that... Obama is doing worse!
YOU can't conclude Obama will WIN because MOrE democrats are asked then GOP.. IT'S that simple!

Your problem is that before you poll people you do not ask "Are you a Democrat or Republican?"

You ask people at random and as part of the process you ask political affiliation. If it is a truly random poll it will reflect the population as a whole. If week after week, poll after poll shows there are more Democrats answering....then it probably reflects a population of more Democrats

 

[Saigon]

Lonstar -

Rightwinger is correct, IMHO.

If one party has 70% real world support, then 70% of the people I call should be his supporters, correct?

But no party has 70 percent real world support.

My example was hypoethical. Make it 51% and the same applies.

 

[Remodeling Maidiac]

Liberal responses in this thread defy both reality and credibility. Especially the first page. The tc makes a CLEAR observation about pollsters and you morons still make it about Romney.

Unreal. You can't remove your dunce caps even for a minute. And we ever wondered why we are where we are now.....

 

[Lonestar_logic]

You don't

Any poll that sets out to poll an equal number of Democratic and Republican voters is invalid

What a poll does is sample 1000 voters. It does not try to sample 500 Republicans and 500 Democrats....just 1000 random voters

If it turns out that 550 of those voters are Democrat it does not invalidate the poll

You are an idiot.

If those being polled overwhelmingly support one party over the other then the data will not be an accurate representation.

A more accurate poll would poll 33.333 percent of each democrat, republican and independent.

How would that be more accurate? Is the country made up of 33% dems, 33% pubs and 33% independents?

It would be more accurate since 40 percent identify themselves as independents, 31 percent as democrats and 27 percent as republicans than a poll favoring one group over the other. IMO

 

[Lonestar_logic]

Lonstar -

Rightwinger is correct, IMHO.

If one party has 70% real world support, then 70% of the people I call should be his supporters, correct?

But no party has 70 percent real world support.

My example was hypoethical. Make it 51% and the same applies.

No party has 51 percent of real world support either.

 

[Saigon]

Come on Mr Lonestar Liberia - TRY and understand this.

If a party has 46% support in the real world, and as a pollster I cam going to call 100 people, then 46 of the people I call should support that party, right?

That way my research will have achieved an accurate portrayal of the real world support.

 

[Liability]

Come on Mr Lonestar Liberia - TRY and understand this.

If a party has 46% support in the real world, and as a pollster I cam going to call 100 people, then 46 of the people I call should support that party, right?

That way my research will have achieved an accurate portrayal of the real world support.

In a random sample, it is doubtful that you would hit that 46% mark. Out of your 100 cold calls, somewhere AROUND 46% would (in terms of probability) be from that Party. But the margin of error (call it 5%) would mean that you might end up with anywhere between 41 or 51.

On the other hand, if we have reason to believe that slightly more folks are actually Republicans than Democratics with somewhere less than a third being Indys,* then it is rather odd that so many of the polls reveal a disproportionately higher number of Democratics getting polled.

And that DOES skew the results.

Bottom line? The left wing main stream media outright lies and distorts and does so deliberately.

__________________
* http://www.rasmussenreports.com/public_content/politics/mood_of_america/partisan_trends

 

[AmyNation]

Come on Mr Lonestar Liberia - TRY and understand this.

If a party has 46% support in the real world, and as a pollster I cam going to call 100 people, then 46 of the people I call should support that party, right?

That way my research will have achieved an accurate portrayal of the real world support.

In a random sample, it is doubtful that you would hit that 46% mark. Out of your 100 cold calls, somewhere AROUND 46% would (in terms of probability) be from that Party. But the margin of error (call it 5%) would mean that you might end up with anywhere between 41 or 51.

On the other hand, if we have reason to believe that slightly more folks are actually Republicans than Democratics with somewhere less than a third being Indys,* then it is rather odd that so many of the polls reveal a disproportionately higher number of Democratics getting polled.

And that DOES skew the results.

Bottom line? The left wing main stream media outright lies and distorts and does so deliberately.

__________________
* Partisan Trends - Rasmussen Reports™

And yet, even very conservative pollsters like FOX poll more dems.

Perhaps instead of a national conspiracy to defraud conservatives, its something simple like, more people call themselves dem but consider themselves conservative, or more dems are willing to be polled.

 

[Liability]

Come on Mr Lonestar Liberia - TRY and understand this.

If a party has 46% support in the real world, and as a pollster I cam going to call 100 people, then 46 of the people I call should support that party, right?

That way my research will have achieved an accurate portrayal of the real world support.

In a random sample, it is doubtful that you would hit that 46% mark. Out of your 100 cold calls, somewhere AROUND 46% would (in terms of probability) be from that Party. But the margin of error (call it 5%) would mean that you might end up with anywhere between 41 or 51.

On the other hand, if we have reason to believe that slightly more folks are actually Republicans than Democratics with somewhere less than a third being Indys,* then it is rather odd that so many of the polls reveal a disproportionately higher number of Democratics getting polled.

And that DOES skew the results.

Bottom line? The left wing main stream media outright lies and distorts and does so deliberately.

__________________
* Partisan Trends - Rasmussen Reports™

And yet, even very conservative pollsters like FOX poll more dems.

Perhaps instead of a national conspiracy to defraud conservatives, its something simple like, more people call themselves dem but consider themselves conservative, or more dems are willing to be polled.

To some extent the reason for the oversampling of Dims is too consistent to be mere coincidence.

But regardless, the point remains that the fact of such over-sampling DOES tend to skew the result.

And just as (intentionally or inadvertently) oversampling Dims skews the result, so too does interviewing REGISTERED voters rather than LIKELY voters.

It really isn't even debatable.

The main stream liberal media's hell-bent intent to breathlessly then report these erroneous figures does seem to be designed to dishearten GOP voters.

 

[rightwinger]

In a random sample, it is doubtful that you would hit that 46% mark. Out of your 100 cold calls, somewhere AROUND 46% would (in terms of probability) be from that Party. But the margin of error (call it 5%) would mean that you might end up with anywhere between 41 or 51.

On the other hand, if we have reason to believe that slightly more folks are actually Republicans than Democratics with somewhere less than a third being Indys,* then it is rather odd that so many of the polls reveal a disproportionately higher number of Democratics getting polled.

And that DOES skew the results.

Bottom line? The left wing main stream media outright lies and distorts and does so deliberately.

__________________
* Partisan Trends - Rasmussen Reports™

And yet, even very conservative pollsters like FOX poll more dems.

Perhaps instead of a national conspiracy to defraud conservatives, its something simple like, more people call themselves dem but consider themselves conservative, or more dems are willing to be polled.

To some extent the reason for the oversampling of Dims is too consistent to be mere coincidence.

But regardless, the point remains that the fact of such over-sampling DOES tend to skew the result.

And just as (intentionally or inadvertently) oversampling Dims skews the result, so too does interviewing REGISTERED voters rather than LIKELY voters.

It really isn't even debatable.

The main stream liberal media's hell-bent intent to breathlessly then report these erroneous figures does seem to be designed to dishearten GOP voters.

How do you account for oversampling vs an actual reflection of the population?

Why is EVERY poll from both the left and the right showing the same thing?