Among the more distracting givens of any political season are the interminable number of polls that are sure to ensue. Each one with a context, including inquiry language and sampling specifics, that falls shy of scientific certitude despite that presumptuous “margin-of-error” claim. At best they are snapshots of something, often too far out to be meaningful. At worst they are self serving. Ever wish Quinnipiac University was better known for its football team or its chemistry department than its polls?
Well, last week Quinnipiac rolled out a poll showing Mitt Romney ahead of President Barack Obama in Florida, 47-41. At the same time, however, an NBC/Marist poll had Obama ahead 48-44. The devil, of course, was in the sampling details. Among other variables, Quinnipiac’s included more Republicans than Democrats. The Marist version had more Democrats than Republicans. Then there were the quibbles over the difference between party “identification” and party “registration.”
In the end, all sides got what they wanted. Campaigns could claim their candidate was “ahead” in the polls and that could keep morale up and fund-raising on track. And the polls could claim that their products were well referenced once again.