May 12 Editorial
In light of several botched polls that have garnered widespread notoriety I decided to conduct one of my own, querying 10 people as to whether they believe polls are most often accurate. Four said “yes,” three said “no,” two were undecided and the other refused to answer, muttering something about a hatred for these types of dopey surveys.
By the way, I’m claiming my highly scientific poll is accurate to within three percentage points 19 times out of 20. Why not? The big guys do it and their accuracy rating often seems to range somewhere between moderately correct and downright abysmal.
Never have I been able to comprehend how a small sampling – often with no more than 1,000 participants – can be a true indicator of people’s thoughts, actions and beliefs on a much wider scale, and furthermore serve as an accurate barometer as to the thoughts, actions and beliefs of thousands or even millions not polled on the very same topic. Complicating matters even more: who’s to say with any degree of certainty that every person who participates in a poll is really telling the truth? Let’s just say I’m skeptical of the entire skewed process due to its many inherent flaws.
Exhibit A: the Alberta election, the results of which turned out to be the exact polar opposite of what the so-called expert pollsters were saying. “A majority government for Danielle Smith’s Wildrose Party” were the unwavering pronouncements from a number of them right up until voting day.
Maybe the pollsters got their right-wing Alberta parties mixed up and really meant to say “Conservative majority” – again. Yeah, that must be it. Or, it could have been the dreaded, seldom discussed 20th poll that reared its ugly head when accuracy ratings aren’t guaranteed to be within the promised three per cent window. Instead, the Alberta prediction went right out that same window and into the nearest dumpster. We’ll call it the “One in 20 Throwaway Poll,” sans attached accountability.
“Dewey Beats Truman” screamed the infamous front page headline of the Chicago Tribune on November 3, 1948, when in fact incumbent Harry S. Truman beat the former New York Governor in the U.S. presidential race. Admittedly, that wasn’t inaccurate polling, but rather poor editorial judgment. But it’s all a form of jumping the gun without having concrete facts. As Mark Twain famously stated: “Never let the facts get in the way of a good story.” Just something to have in mind the next time you see, hear or read about the latest poll numbers. Keep it all in perspective.
No matter how simple or elaborate the setup, the hypothesis is not necessarily the conclusion.