The wording of poll questions
Wednesday, March 14, 2012
Polling has become big business in the political arena. And perhaps, never as scrutinized as this election year.
Political polls are a combination of art and science. There's always a mathematic component that gauges the depth of the outcome. But there's an equal issue with the wording of the poll questions.
Within the past two weeks, two highly respected national polling firms have come to totally opposite outcomes for this year's presidential contest.
So what are we to believe? And should we even give passing notice to polls almost a year in advance of election day?
I well remember a college course back in the '60s that explored the question of whether opinions drive poll results or whether poll results drive opinion.
Almost 50 years later, I remain uncertain.
I do know this: Poll questions can be framed to arrive at a preconceived outcome.
And being of the conservative persuasion, I view with great suspicion all polls attached to The New York Times or The Washington Post. There's no need to even discuss any poll findings from MSNBC.
I tell myself every four years that I'm going to keep poll results to match them against outcomes. That way, next election cycle, it's easier to track the most effective polls.
But I never do.
Political experts tell us that darned near 92 percent of the electorate have already decided their candidate of choice this year.
So polling should gauge those undecided. Or so you would think.
But snap responses to phone surveys are different from decisions made in the voting booth. And thus, the varying accuracy of the polling process.
Poll results are useless unless you know the exact questions and the makeup of those surveyed.
But most voters don't take the time to dig quite so deep, and instead view poll results as a certain indicator -- which they are not.
Human nature is such that we root for the underdog but want to back a winner. When a certain outcome becomes apparent -- according to polls -- the bandwagon effect kicks in.
The truth is that all polls should be taken with a substantial grain of salt. And that often polls don't measure trends as much as they drive trends.
Here are two hypothetical poll questions on which you can virtually guarantee the outcome.
Do you support four more years of runaway spending and higher deficits or is it time to balance our national budget? Those poll results would likely favor the position held by the GOP.
Do you support reducing benefits for your children and grandchildren to help balance our national budget? Those poll results would more likely favor the position held by the Democrats.
All too often, it's not the poll results but the questions themselves that drive the outcome.