Letters to the Editor
-
Nature of Polls
I am currently in the process of watching a remarkable series of videos over the last few days at YouTube presented by a science teacher named Eric Sodomka, username “Wonderingmind42” (just google the name and select the link with his username at YouTube or go to www.wonderingmind42.com), that discusses global warming in a more logical and at least as compelling manner as “An Inconvenient Truth.” For those who don’t have much time, the following YouTube video (titled “How It All Ends” - ten minutes) summarizes his overall risk-assessment approach to the issue at
http://www.wonderingmind42.com/.
The summary is supported by a voluminous “expansion pack” videos that are generally just under ten minutes in length and total about six hours. The “Nature of Science” section, which consists of three such videos (approximately thirty minutes total) familiarizing us with the open-ended, non-biased, skeptical approach to research, interpretation of evidence gathered, and peer review and criticism of conclusions reached to which the professional scientific community aspires, is what really grabbed me; because it relates not only to matters scientific but also to much of what we discuss here.
It directly relates to what Jay Rosen has described as the “Escape from Empiricsm” we have seen in Bush/Cheney’s faith-based, ideological approach to governing for the last seven years as well as the mindset of the authoritarian leaders and followers discussed here at length by Glenn and others, that have facilitated this disastrous approach, which can certainly claim global warming action, even its very recognition, among the many victims of this administration’s neglect and malfeasance.
But the currently-relevant application has to be polling, what it does and what it does not do - how it’s being used and how it’s being misused. Polling techniques have been developed and refined over the years by using these scientific methods, particularly in the areas of statistics, human behavior and language. A major point stressed in the videos is that science never reaches absolute certainty on anything; it can only try to make the uncertainty as small as possible and draw a conclusion as to how close the science comes while, as Sodomka puts it, “Acknowledging that you’ll never get there.”
The fact that science never “gets there” should temper our expectations of polling. If science is doing what it is supposed to do, polling should tell us how likely a particular candidate is to win or lose, not whether they will win or lose. Sodomka summarizes how the scientific profession is constantly seeking to identify and minimize “confirmational bias,” or pre-conceived notions and beliefs in absolutes that can taint overall results, courses of action and valuation and interpretation of underlying evidence. It goes without saying that conformational bias has been a major problem for Bush/Cheney.
While much of the mainstream media can be faulted for the same thing in assuming Clinton had lost, we must also look at such bias within ourselves. I know I didn’t need the media to interpret the polls for me - they were well outside the margin of error and it seemed that the only question was how handily Obama would win. In fact, I was going to write a comment regarding the effect that Hillary’s campaign was humming along pretty nicely until Bill became more visible and vocal, which may have garnered the added pleasure of possibly running into Glenn’s buzz saw. (This may be more correlation than cause-and-effect - see the video - but I still think there is a lot of concern out there in the electorate regarding dynasty issues) I was able to avoid this predicament thanks to the actual vote not cooperating with the polls.
I don’t fault the establishment media for misinterpreting the polls as much as I do for them misinterpreting the results. Chris Matthew’s confirmational biases are legend, but I don’t think Rachel Maddow or others have escaped this trap. It seems that all the explanations - "Bradley effect,” sympathy for Hillary for various reasons, no Obama coronation, etc. - meticulously avoid one thing: giving Hillary credit for a good campaign and organization. The only thing the exit polls tell us is that it appears a lot of women voted for Clinton, but do we even know that with 100% certainty, and especially, do we really know why?
A lot more to say, but gotta go …