Changes

A '''public opinion poll''' is a survey of attitudes from a particular sample of the population at interest.  Opinion polls represent the opinions of the total population by asking a sample of size N, where N is large enough for reliable results (usually expressed as a 95% confidence of +/- Z%. For most polls, N = 400 to 1000 people.  

A '''public opinion poll''' is a survey of attitudes from a particular sample of the population at interest.  Opinion polls represent the opinions of the total population by asking a sample of size N, where N is large enough for reliable results (usually expressed as a 95% confidence of +/- Z%. For most polls, N = 400 to 1000 people.  

The sample had to be a ''valid'' representation of the population, which means that each person in the target population has a known chance of being sampled.<ref> In unweighted samples, each person has an equal chance. If people have a known but unequal chance, they are weighted to make the chances appear equal.</ref>  Unscientific public opinion polls, made by collecting information from a haphazard sample, date from the '''straw polls''' in American elections in the 1820s. As soon as modern statistical sampling theory was invented around 1930, [[George Gallup]] and other advertisers started created nationwide public opinion polls, using about 1500 cases. The American public became fascinated with polls and they spread to many democratic countries by 1940. Totalitarian countries did not allow them.

The sample had to be a ''valid'' representation of the population, which means that each person in the target population has a known chance of being sampled.<ref>In unweighted samples, each person has an equal chance. If people have a known but unequal chance, they are weighted to make the chances appear equal.</ref>  Unscientific public opinion polls, made by collecting information from a haphazard sample, date from the '''straw polls''' in American elections in the 1820s. As soon as modern statistical sampling theory was invented around 1930, [[George Gallup]] and other advertisers started created nationwide public opinion polls, using about 1500 cases. The American public became fascinated with polls and they spread to many democratic countries by 1940. Totalitarian countries did not allow them.

==History of Polling==

==History of Polling==

All polls based on  samples are subject to [[sampling error]] which reflects the effects of chance in the sampling process. Too many of this group happen to fall in the sample, and too few of that group.  This produces a bias that is inevitable when people are picked at random.  The percentage error gets smaller as the sample size increases. The uncertainty is often expressed as a [[margin of error]].  A poll with a random sample of 1,000 people has margin of sampling error of 3% for the estimated percentage of the whole population. A 3% margin of error means that 95% of the time the procedure used would give an estimate within 3% of the percentage to be estimated. The margin of error can be reduced by using a larger sample, however if a pollster wishes to reduce the margin of error to 1% they would need a sample of around 10,000 people. In practice pollsters need to balance the cost of a large sample against the reduction in sampling error and a sample size of around 500-1,000 is a typical compromise for political polls. (Note that to get 500 complete responses it may be necessary to make thousands of phone calls.)[http://www.publicagenda.org/polling/polling_error.cfm]

All polls based on  samples are subject to [[sampling error]] which reflects the effects of chance in the sampling process. Too many of this group happen to fall in the sample, and too few of that group.  This produces a bias that is inevitable when people are picked at random.  The percentage error gets smaller as the sample size increases. The uncertainty is often expressed as a [[margin of error]].  A poll with a random sample of 1,000 people has margin of sampling error of 3% for the estimated percentage of the whole population. A 3% margin of error means that 95% of the time the procedure used would give an estimate within 3% of the percentage to be estimated. The margin of error can be reduced by using a larger sample, however if a pollster wishes to reduce the margin of error to 1% they would need a sample of around 10,000 people. In practice pollsters need to balance the cost of a large sample against the reduction in sampling error and a sample size of around 500-1,000 is a typical compromise for political polls. (Note that to get 500 complete responses it may be necessary to make thousands of phone calls.)[http://www.publicagenda.org/polling/polling_error.cfm]

The margin of error does not reflect other sources of error, such as poorly defined or ambiguous questions, or overlooking groups of people who might have different opinions. The ''Literary Digest'' sampled automobile and telephone owners in 1936 and thus got too many rich Republicans and too few poor Democrats. Today nearly all interviews are done by telephone or over the internet. Researchers today are studying whether people who only have cell phones are much different from people with land lines. (In general the differences are quites small.) Researchers have found that people who volunteer for a poll--say, on the internet--tend to be different; they are younger, better educated and more sure of their opinions.

The margin of error does not reflect other sources of error, such as poorly defined or ambiguous questions, or overlooking groups of people who might have different opinions. The ''Literary Digest'' sampled automobile and telephone owners in 1936 and thus got too many rich Republicans and too few poor Democrats. Today nearly all interviews are done by telephone or over the internet. Researchers today are studying whether people who only have cell phones are much different from people with land lines. (In general the differences are quites small.) Researchers have found that people who volunteer for a poll—say, on the internet—tend to be different; they are younger, better educated and more sure of their opinions.

===Nonresponse bias===

===Nonresponse bias===

In Britain, most polls failed to predict the Conservative election victories of 1970 and 1992, and Labour's victory in 1974. However, their figures at other elections have been generally accurate.

In Britain, most polls failed to predict the Conservative election victories of 1970 and 1992, and Labour's victory in 1974. However, their figures at other elections have been generally accurate.

==Exit polls==

==Exit polls==

[[Exit poll|Exit polls]] ask voters just after they leave the booth how they voted. The results are tabulated within minutes and allow news agencies to make a prediction of who will win (or if it is "too close to call".)

[[Exit poll]]s ask voters just after they leave the booth how they voted. The results are tabulated within minutes and allow news agencies to make a prediction of who will win (or if it is "too close to call".)

==Influence ==

==Influence ==

* [http://www.ncpp.org/?q=home National Council on Public Polls - An association of polling organizations in the United States devoted to setting high professional standards for surveys]

* [http://www.ncpp.org/?q=home National Council on Public Polls - An association of polling organizations in the United States devoted to setting high professional standards for surveys]

* [http://answers.vizu.com/ Vizu Answers - Online Opinion Polling]

* [http://answers.vizu.com/ Vizu Answers - Online Opinion Polling]

==See Also==

==See also==

*[[Gallup Poll]]

*[[Gallup Poll]]

[[Category:Media]]

[[Category:Media]]

[[Category:Statistics]]

[[Category:Statistics]]

[[category:political Terms]]

[[Category:Political Terms]]