Having never taken a statistics or opinion research course in college, I have a tendency to look askance at pre-election polls because I don't understand them (or at exit polls because of the 2000 fiasco). How is it possible that pollsters can extrapolate from the voting intentions of a few to make predictions about the many? The leap from part to whole feels like a leap of faith, not science.
Tonight, in search of information that might demystify the process, I turned up Public Agenda's "About Polling," a layperson's guide to sampling statistics, margins of error, confidence intervals, the seven stages of public opinion, and other polling arcana.
Posted by karik at October 31, 2004 8:35 PM | TrackBack