One of the reasons I try not to fill this column with giblets from polls is that most polling concerns are Cheney-esqe about disclosing their methodology. Absent real knowledge of the voter screen, the winnowed universes, the exact wording of the questions and demographics, I'd be better off with a divining rod.
"Proprietary information" is occasionally cited as the reason for nondisclosure. But that's a sham. It's like a case officer refusing to tell an analyst anything about the source of an intelligence tip, leaving the analyst with no background information to judge the validity of the information.
Good polling is background dependent; you cannot analyze a poll's results without understanding the built-in prejudices of the survey itself.
Mark Blumenthal of Pollster.com has a solution. He wants the community of active Internet political enthusiasts to pressure polling firms who don't reveal their methodology.
Starting today we will begin to formally request answers to a limited but fundamental set of methodological questions for every public poll asking about the primary election released in, for now, a limited set of states: Iowa, New Hampshire, South Carolina or for the nation as a whole. We are starting today with requests emailed to the Iowa pollsters and will work our way through the other early states and national polls over the next few weeks, expanding to other states as our time and resources allow.