Are we sure of the survey methodology being used? It says the results are weighted so the sample demographics match Census Current Population Survey parameters for gender, age, education, race, region, and population density.

Doesn't this mean if their sample includes 422 Democrats and 324 Republicans (in a sample of 1000, 42.2% and 32.4%), and the population of the US is 37% Democrat and 35% Republican, that they multiple, or weight, the results to bring it inline with the actual representative percentage. Meaning they would multiply the results of the Democratic respondents by something less than 1 like .87 and multiple the Republican respondents by something more than 1 like 1.08

42.2 x...

Are we sure of the survey methodology being used? It says the results are weighted so the sample demographics match Census Current Population Survey parameters for gender, age, education, race, region, and population density.

Doesn't this mean if their sample includes 422 Democrats and 324 Republicans (in a sample of 1000, 42.2% and 32.4%), and the population of the US is 37% Democrat and 35% Republican, that they multiple, or weight, the results to bring it inline with the actual representative percentage. Meaning they would multiply the results of the Democratic respondents by something less than 1 like .87 and multiple the Republican respondents by something more than 1 like 1.08

42.2 x .87 = 37% and 32.4 x 1.08 = 35%

Isn't this what is happening?

Those aren't the numbers they are using to weight the results.

The weights are what would throw the results, if the weighting was innacurate.

The numbers you are showing are the post weighting results of the poll.

What I am trying to show is that most people look at the numbers of respondents and when it shows 400 Democrats polled 300 Republicans polled immediately jump to "THE POLL WAS RIGGED". I was trying to show that pollsters are fully aware of what the political association of the overall populace is and correct the numbers of those polled by weighting the results. 400 democrats times 85% and 300 Republicans times 110%.

All the guy who wrote the article that was clipped is saying ...

Those aren't the numbers they are using to weight the results.

The weights are what would throw the results, if the weighting was innacurate.

The numbers you are showing are the post weighting results of the poll.

What I am trying to show is that most people look at the numbers of respondents and when it shows 400 Democrats polled 300 Republicans polled immediately jump to "THE POLL WAS RIGGED". I was trying to show that pollsters are fully aware of what the political association of the overall populace is and correct the numbers of those polled by weighting the results. 400 democrats times 85% and 300 Republicans times 110%.

All the guy who wrote the article that was clipped is saying is these two polls seem to not agree with each other. He makes no comment about the science behind polling and statistics.

Did you read what they wrote on the poll?

Results are weighted so that the sample demographics match Census Current Population Survey parameters for gender, age, education, race, region, and population density.

Does it say political party? No, it doesn't. And I have seen polls where they do have political party and they change it to actually increase democrat representation. But since this one gives no weighted numbers based on party, and doesn't say it even takes that into account, what evidence do you have other than your wish for this poll to be true that they made any adjustments based on party affiliation? It isn't up to us to show that the poll was actually figured dif...

Did you read what they wrote on the poll?

Results are weighted so that the sample demographics match Census Current Population Survey parameters for gender, age, education, race, region, and population density.

Does it say political party? No, it doesn't. And I have seen polls where they do have political party and they change it to actually increase democrat representation. But since this one gives no weighted numbers based on party, and doesn't say it even takes that into account, what evidence do you have other than your wish for this poll to be true that they made any adjustments based on party affiliation? It isn't up to us to show that the poll was actually figured differently than it says it was, it is up you you to prove otherwise.

Well I am glad you have finally found where the numbers come from that you are talking about. Finally progress can be made.

None of us know what the weighting numbers that were used are, you or me. You just want to presume it is wrong.

My position is that this is the latest poll and if something significant has changed that has altered public opinion it will be reflected in subsequent polls. The one big thing that has happened is the veto of the funding of the troops. We will see what the subsequent polls show.

One thing that has been proven this evening is that you don't understand what weighted polls are because you had continually tried to point out that less Republicans are contacted....

Well I am glad you have finally found where the numbers come from that you are talking about. Finally progress can be made.

None of us know what the weighting numbers that were used are, you or me. You just want to presume it is wrong.

My position is that this is the latest poll and if something significant has changed that has altered public opinion it will be reflected in subsequent polls. The one big thing that has happened is the veto of the funding of the troops. We will see what the subsequent polls show.

One thing that has been proven this evening is that you don't understand what weighted polls are because you had continually tried to point out that less Republicans are contacted. Which is corrected for by weighting.

End of story.

None of us know what the weighting numbers that were used are

Yes we do. They were given to us right there in the answer to question number 26.

My position is that this is the latest poll and if something significant has changed that has altered public opinion it will be reflected in subsequent polls.

So, it takes years for a shift of 5% and that is news worthy, but you believe a shift of 8% can happen in a month?

One thing that has been proven this evening is that you don't understand what weighted polls are because you had continually tried to point out that less Republicans are contacted. Which is corrected for by weighting.

You obviously ...

None of us know what the weighting numbers that were used are

Yes we do. They were given to us right there in the answer to question number 26.

My position is that this is the latest poll and if something significant has changed that has altered public opinion it will be reflected in subsequent polls.

So, it takes years for a shift of 5% and that is news worthy, but you believe a shift of 8% can happen in a month?

One thing that has been proven this evening is that you don't understand what weighted polls are because you had continually tried to point out that less Republicans are contacted. Which is corrected for by weighting.

You obviously don't understand what it is because question number 26 shows that it obviously wasn't corrected, and the pollsters themselves NEVER CLAIMED to correct for political party. They said they adjusted for gender, age, education, race, region, and population density. Unless you can prove they adjusted for political party then you are wrong.

Well it is interesting to read what Rasmussen says about its own polling:

At Rasmussen Reports, we ask if people Strongly Approve, Somewhat Approve, Somewhat Disapprove, or Strongly Disapprove of the way the President is performing his job. This approach, in the current political environment, yields results about 3-4 points higher than if we simply ask if people if they approve or disapprove (we have tested this by asking the question both ways on the same night). Presumably, this is because some people who are a bit uncomfortable saying they “Approve” are willing to say they “Somewhat Approve.”

So if Rasmussen asks the poll the same way Newsweek does there result would be on...

Well it is interesting to read what Rasmussen says about its own polling:

At Rasmussen Reports, we ask if people Strongly Approve, Somewhat Approve, Somewhat Disapprove, or Strongly Disapprove of the way the President is performing his job. This approach, in the current political environment, yields results about 3-4 points higher than if we simply ask if people if they approve or disapprove (we have tested this by asking the question both ways on the same night). Presumably, this is because some people who are a bit uncomfortable saying they “Approve” are willing to say they “Somewhat Approve.”

So if Rasmussen asks the poll the same way Newsweek does there result would be only 34% approval. So what we really seem to be arguing here is that although Bush's approval rating is dismally low it just isn't the lowest ever. That must provide a lot of comfort.