The New Bailout

Originally Posted by classicman

redux, Who was polled, what was their political affiliation and how were the questions phrased. What was the number of people sampled? What geographic region? Was it a nationwide poll or local from a specific district or state. There are a million questions about how and who was polled that make the info presented useless without the supporting answers.

Results are based on telephone interviews with 1,018 national adults, aged 18 and older, conducted Feb. 6-7, 2009, as part of Gallup Poll Daily tracking. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of sampling error is ±3 percentage points.

Other results are based on telephone interviews with 1,012 national adults, aged 18 and older, conducted Feb. 4, 2009. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of sampling error is ±3 percentage points.

Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only).

In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls. (this sentence is a standard disclaimer generally used by most pollsters in the event that is the poll is used in some fucked up manner beyond its intent)

http://www.gallup.com/poll/114202/Ob...lus-Fight.aspx

Gallup, like all reputable national polling organizations, uses widely accepted statistical techniques for the sample to be representative of age, income, region, political affiliation, etc. then the data is "weighted" to even more accurately represent the total population.

That is why Gallup can they say with 95% confidence that the maximum margin of sampling error is ±3 percentage points.

Standardized protocols are also used to minimize question bias.

Again, a poll is simply a snapshot of public opinion on a particular issue (s) at a particular point in (or over) time. Its not something you should stake your life on but it is also not something to dismiss as useless or not credible if you are at all interested in what the public may be thinking (in very general terms) about a particular person or issue.

added:
I can say that with a high degree of confidence and Merc can say its bullshit until he finds a poll that he likes.

[classicman]

Originally Posted by Redux

Results are based on telephone interviews with 1,018 national adults, aged 18 and older , conducted Feb. 6-7, 2009, [/b]
Other results are based on telephone interviews with 1,012 national adults, aged 18 and older , conducted Feb. 4, 2009.

Who answered? Unknown
What was their political affiliation? Unknown
How were the questions phrased? Unknown
What was the number of people sampled? Answered
From what geographic region? Unknown
Were the respondents evenly distributed? Unknown

Originally Posted by Redux

In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.

Gallup, like all reputable national polling organizations, uses widely accepted statistical techniques for the sample to be representative of age, income, region, political affiliation, etc. then the data is "weighted" to even more accurately represent the total population.
Standardized protocols are also used to minimize question bias.[/quote]

I don’t think anyone here is a poll expert, but many times these things seem to be biased. They repeatedly state how they want to or try to minimize the bias or statistical inaccuracies. They try to accurately represent a huge number of people based on percentage wise, a miniscule sampling. I’m not against polls, but there are too many variables that are never answered. Not the least of which in this case is who actually responded. Aside from their age there is no answer. It was a national poll, but was there an even sampling of the data nationally or did more people respond from one area versus another? I think the outcome would be very different if they asked those in the northeast versus the west coast, for example.

Or more importantly let’s say they attempted to reach an even number of people from as diverse a group as possible. It is not clear whether one party responded more than another did. Another example is the time of day polls are conducted. If they are calling during normal business hours, are they getting a relative sampling of those people or is the data skewed? Of course the data is skewed, it has to be. I am not implying that this poll or any other is intentionally doing this. It is just the way it is. It is nowhere near an exact science. Also, how are they weighting the data?

Originally Posted by Redux

added:
I can say that with a high degree of confidence and Merc can say its bullshit until he finds a poll that he likes.

Of course you can. But isn't that why you posted this poll, because you liked it?