ABSTRACT
Getting meaningful survey reports to a reporter and ultimately to the public is something akin to the childhood game of telephone. Survey researchers have tried promoting disclosure of methods, self help courses for journalists, newsroom guides about surveys for editors and reporters, discussions with all who will listen about what is wrong with particular surveys or pseudo-survey approaches. Researchers usually refer to the sample design by the single word "probability" or some variant such as a "modified probability" selection, whatever that means. A researcher might know from previous work the likelihood of these people going to the polls and casting a vote, and something about their voting behavior. News organizations can do several things. They can train a reporter or a newspaper editor or a television producer to understand the rudiments of survey research, at least enough so that they can ask researchers insightful questions.
