Sunday, October 10, 2004, 10:57:07 PM, JimKandJulieB@xxxxxxx wrote: Jac> _http://www.cnn.com/2004/ALLPOLITICS/10/08/electoral.map/index.html_ Jac> (http://www.cnn.com/2004/ALLPOLITICS/10/08/electoral.map/index.html) Jac> _http://race2004.net/_ (http://race2004.net/) Jac> _http://www.rasmussenreports.com/Presidential_Tracking_Poll.htm_ Jac> (http://www.rasmussenreports.com/Presidential_Tracking_Poll.htm) Jac> Would someone who understands polling please explain to me how various polls Jac> such as those linked above can give such hugely disparate results? Is it Jac> the demographics of those being polled? The number of people questioned? Can Jac> any particular poll be considered more "valid" than any other? If polls are Jac> this wildly inaccurate (since they're so hugely different most must be Jac> inaccurate), what is their value? First, the CNN/Gallup report (electoral college votes) is based on a survey, not a poll (but polls are included) >If the election were held today, Bush likely would win 301 electoral >votes to Kerry's 237, according to a new CNN survey based on state >polling as well as interviews with campaign aides and independent >analysts so is not comparable to poll results (I didn't have time to look at its polling data) Second race2004.net's using other people's polls; the methodology behind their (race2004's) analysis is on their web page. Third, the Rasmussen poll, which is indeed a poll(!), is a nightly 'phone poll averaged out over 3 days; it hasn't yet fully taken into account reactions to the debate. Zogby International's polls are the ones whose method differs from others, in just one, possibly fairly crucial, respect. I can't decide whether they're more likely to be right than the others -- I'f say if the apparently very small % of undecided voters we've been cited from the beginning masks reasonably large shifts (and I'd now say it does) they could be. (They tend to show Kerry ahead in the swing states. But that's what Gallup/CNN's moving towards too.) Insofar as polls differ it won't be the demographics as the poll results will have been adjusted for those, it probably won't be numbers either as accurate polling can be done with remarkably small numbers. -- Judy Evans, Cardiff, UK mailto:judithevans001@xxxxxxxxxxxxxx ------------------------------------------------------------------ To change your Lit-Ideas settings (subscribe/unsub, vacation on/off, digest on/off), visit www.andreas.com/faq-lit-ideas.html