Polsters call 1005 people and from this they conclude that Bush's ratings are at 32%?

Question

I took a statistics class in college and didn't understand it, but I read on Yahoo today that the latest polls show that Bush's rating on his job performance is 32% while Congress' is 22%. At the very end of the article it said this poll was conducted by contacting 1005 people by phone.

How in the heck can pollsters gauge the mood of the country by only contacting 1005 people>

rhino-----I didn't say I flunked, I said I didn't understand it. Maybe you should take some remedial reading classes so you can understand what people write.

Answer 1

There are polls and there are polls. You can ask anybody or only certain types of people. You can conduct polls at times when more people are likely to participate or not likely to participate. For instance, phoning homes during normal working hours will provide a much higher number of unemployed or retired people. Calling on the weekend will find fewer single people.

There are few polls that provide very real, scientifically accurate results. Believe me, the politicians have these results. They are rarely publically released.

Instead, you and I will see poll results designed to recieve a desired result. They are then presented influence our thinking.

.

Answer 2

I will agree that some pollsters delibertly skew there polls but actually very few, and the ones that do, skew the results by the wording of the questions they ask.

Yes they poll a small portion of the population, but they make that portion of the population representative of the entire population of what they are polling. But there are a few things you need to realize about polls...

First of all the 1005 respondents in no way represents the number of calls they make. They probably make 4 to 5 thousand calls to get 1000 responses. Secondly there are a number of polls conducted with similar results. With 5 of the major polls conduting sepearte independant polls with similar results you can bet there is a good representation of the poulation as a whole.

BUSH"S APPROVAL RATING
ABC/Washington Post - 33%
FOX/Opinion Dynamics - 34%
CBS - 29%
Gallup - 36%
Pew - 31%
AP - 33%
NBC/Wall Street Journal - 33%
CNN - 36%
USA Today - 33%

If you look at the above polls, this would represent 9,000+ respondents and the average comes to the same figure 33% approval rating.
------------------------------------------------------------------------------------
btw, I have been polled a number of times.
Typically those being polled are one's who vote in the regulary, and have updated information in regards to address and phoen number. I know and understand this as I have done volunteer work for various candidates, conducting polls, knocking on doors requesting signatures to get a candidate on the ballott.

Answer 3

There are many polls and all come to the same numbers (+/-5)... While one poll asks 1000 people, there are countless number of polls asking 1000 people. This has happened in the history of modern politics and is very accurate. The polls do not ask the same 1000 people... if the polls were not somewhat accurate then a lot of political consultants would not have a job... (for that matter, the same could be said for market research).

This is a science but not a exact science... that is why there is a margin of +/-5 polling. But it does not have to be a exact science either... this is not a moon landing experiment, there is room for error but it is somewhat accurate.

Answer 4

Glad that you admit that you did not understand statistics in college. Surveying 1000-1500 people, if selected from a random sample of the target population, generally gives an accurate impression of the answers the population would give. There is a margin of error of plus or minus 3 points. You can see the accuracy of these polls in election after election. Have they been wrong, ever? Of course, but they have been correct many more times.

Answer 5

If the pollsters want to get accurate results then they have to make sure of the following:

Time of call (aft 7PM more conseratives)
Day of call (Weekends less single people)
Sex of resident (more women, more democrats)
Race of resident (blacks tend to vote for democrats)
Age of resident (older folks tend to vote conserative)

IF they follow proceedures to ensure an accurate sample then the error rate should be between 3 to 5%.

Answer 6

It seems we know why you flunked. Dictator Dumbya is ALSO a miserable abysmal failure. The exit polls were right in 2000 AND 2004.

Answer 7

A poll shows more about The person giving the poll than the mood of the people being asked.

Most polls have a set out come and you set up the poll to give you what you want to see.

Answer 8

It depends on how they target their demographics. Without having that information, I can't tell you how accurate their results are. At any rate: it is a poll, even though they usually say +/- 3% off, it could be 99% off what with all the remaining people in the US.

Answer 9

Well you have to read the statistical control methods used by each poll you read. This tells you how the controls make predictions.

Actually I'm surprised Bush scored that high. Often this is determined by how the questions are framed.

Answer 10

Pollsters call 1005 people before 7 PM deliberately excluding people with long-hour, white-collar jobs, deliberately skewing the results.