Posted on 10/03/2004 7:35:29 PM PDT by dvwjr
oops--you know what I mean.
bump!
Frank Newport 26-Sep 8:14 pm Steven writes in with a question that reflects misinformation that is boomeranging around the Net. We have posted blogs below that go into some detail outlining the whole issue of party identification in a survey context. Our Gallup samples are rigorously executed and checked and weighted against a number of know U.S. Census Bureau parameters: age, gender, region of country, race, and education. Party ID ("In politics, as of today, do you consider yourself a Republican, a Democrat, or an independent?") is not a variable that is measured by the U.S. Census Bureau, is not fixed, and in fact is to a significant degree a measure that is quite likely to change based on the environment. After 9/11, polls showed many more people identifying with the Republican Party than the Democratic Party because citizens were rallying behind the president. This winter during the primary season, polls showed more people identifying with the Democratic Party than the Republican Party because the news coverage was focused almost exclusively on the Democratic primaries. Analyses shows that polls had more identification with the Democrats than Republicans after the Democratic convention this summer, and then more identification with the Republicans than Democrats after the Republican convention. The measure of partisanship we and other pollsters use is not measuring some lifelong fixed value like gender or race. It is an attitudinal identification with one or the other party at the time of the survey. So, if there are forces at work out in the environment that are favorable to the Democratic Party, for example, they will cause more people to identify with the Democratic Party in the survey, and also cause more people to say they will vote for the Democratic candidate.
Here are links to two recent summaries by other students of polling that go over this same concept.
http://mysterypollster.typepad.com http://mysterypollster.typepad.com
This whole issue of partisan identification is one that pollsters and survey scientists have been discussing and dealing with for years in publications and scholarly conferences. It's not a new issue.
It's surprising that some people on the Net feel that they have suddenly "discovered" something about polling as if pollsters are not highly aware of the variables like party identification that we measure in each survey.
Gallup has a team of experienced editors who have been conducting polls for decades, and teams of statisticians and methodologists who work on every poll. All of this is not to say that there can't be legitimate scientific debate on this and other issues. There can be, just as heart surgeons have conferences and debate the value of different methods of conducting coronary artery bypass surgery. But I can assure all users of Gallup Poll data that the methods we use in pre-election polls are the results of about 70 years of experience in conducting them (since 1936) and intensive, ongoing study and examination of each element of the survey process.
Oops, wrong thread but I suppose it's still relevant
So we increased the number of idiots who don't know who they are voting for by now? Geez
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.