One of the latest polls that broadcast pundits and news anchors quoted this week is an NBC-Marist poll touting that democrat candidates are gaining ground in the ten closest senate races. The question is whether voters should be influenced by polls when deciding how they should cast their ballot in the fast approaching midterm election. For anyone who has spent time learning the ropes of conducting surveys and studies (quantitative research is a prerequisite for most graduate degrees) it’s time to address the art of statistics. The terminology “art” carries more fact than pollsters care to admit because the science behind polling is largely psychology-based despite all the mathematical theories and quotients required to formulate the survey models. It is essential to recognize how poll questions are engineered1 to reflect the data mining of potential voters, all in order to receive the answers that best suit a political agenda, which actually includes all subjects from medicine to the environment because every subject surveyed boils down to politics. Not all polling organizations set out to obtain a desired result by developing surveys with expected outcomes and, frankly, many opinion gatherers hardly realize how much they inject their own perspective into the questions they craft. It occurs all too naturally because bias induces individuals and even polling colleges to expect others to agree with their assumed “right” thinking, which is generally not at all correct, just prejudicially elitist. No matter what the subject matter to be studied, it is nearly impossible for a polling agency to create a survey with questions that will not glean results favoring their own partiality, and this is more than rampant in media-run polls. All one has to do is check the educational and professional history of those conducting the surveys and the indoctrinated bias will become evident. Most often that prejudice inclines toward the progressive mindset.
Posted on: Thu, 30 Oct 2014 10:47:52 +0000
© 2015