With every new day during this election year, data is released from yet another poll; and regardless of whether you pay attention to their results, one cannot help but be filled with a combination of respect and loathing for organizations producing them. Loathing because you know that while some will make an honest effort at accurate statistical inquiry, others will in turn this delicate political tool into a mallet with which to beat public opinion into the shape they desire. Respect comes because the mind-numbing labor involved in reaching out to an unwilling public for answers to telephone surveys is incredible difficult; and because if you’re going to do the job properly, the other challenges involved can seem even more insurmountable.

One must carefully craft questions that will winnow out the information you desire without skewing the process with prejudice for a desired result. One misplaced comma or an ill-chosen word can completely ruin the data collection being attempted. Even a misplaced emphasis on a word while questioning can negate the accuracy of the data gathered.

Pollsters must likewise be concerned with those questioned. An accurate sample must be as inclusive as possible, attempting to cover parameters of education, income, political affiliation, and likelihood of voting, while fulfilling that most difficult of assignments, getting someone to answer the phone and take the poll in the first place.

The pollster must likewise recognize that, as in any other type of scientific study, the act of observation itself has an effect on the data that must be factored into the tight rope they walk on behalf of a client.

Not all, of course, make much effort to live up to these high standards when performing their duties. Some are far more concerned with confirming previously held (or paid for) assumptions, demeaning both themselves and the process. Let’s face it folks, two of the population’s least favorite groups are politicians and telemarketers. Pollsters, by the very nature of their labors, fall into both categories, and thereby suffer regard normally reserved for used car salesmen and people who pump out septic tanks (they’re potentially disreputable and likely to stink).

Even when accurate data has been gathered and disseminated, however, it often gets misinterpreted or misused by a politician looking to flip-flop and get back in front of the parade. Far too often candidates use such data not so much to “modify” their opinions and principles, as much as to “restate’ them in a way that’s completely contrary to any previously made statements. The mainstream media, meanwhile, hoping to remain at least relevant, if not as influential as they once were, attempt to go “inside baseball” with the data gathered to move opinion based on these surveys. Bringing forward their experts, and completely forgetting that their audience is one with math skills taught them by the public education system in this country (or perhaps counting on it), they insist on “breaking down the numbers for us.”

Of course few if any care about this kind of esoteric statistical interpretation, if they can understand it at all. Their eyes in fact glaze over when told that the latest polls show that 56 percent of the 82 percent who care about A, also care about B. (The answer, for those of you reaching for your calculator, or trying to take your shoes and socks off, is 45.92 percent.) They are likewise unmoved to discover that 64.3 percent of people who have orange shirts in their closets are likely to vote for one candidate over another.

The apparently capricious nature of the American public only adds to the ludicrous nature of the process. Evidently we are so impulsive, mercurial and unstable that polls must be taken daily to determine how a subtle shift in the political winds of change may have affected what should be rather strongly held opinions. Every candidate becomes so dependent on constant polling for guidance on how to reshape their image and every media source to tweak their coverage that no one is content unless the sample is less than hours old.

I say we fix them all! There’s been little enough fun this election season and maybe it’s time we took part of this “Groundhog Day” process and turned it on its ear. Not only is abusing telemarketers of any kind a right in this country, allowing politicians to govern (let alone run) on a never-ending stream of polling data has certainly done more harm than good. So instead of carefully considered answers , we should give completely random and nonsensical ones instead. Not only will it be interesting to see if they can tell the difference, but it also might be interesting to see whether they can stand on their own, without the constantly updated stream of opinion from political telemarketers.

Tags: Just Blowing Smoke, political telemarketing, polling, statistics, Tim Higgins

This entry was posted on Tuesday, February 28th, 2012 at 12:30 pm and is filed under Just Blowing Smoke, Opinion.
You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.