John Boehner still has the best “PF report card” average of the congressional majors and Harry Reid is still at the bottom, the rankings are the same since the first time I performed these calculations in late June. Obviously, the more rulings a person has, the less the average is affected by additional ones; Obama has dropped by a few points, as well as John Boehner, as they have been the only ones to receive additional rulings.
Bryan White objected yesterday in his blog that my calculations are not scientific….where did I ever say they were? The fact is even trying to do it scientifically would be wrought with the quandaries of the same selection bias of which he speaks. Not just selecting samples, but judging those samples. I don’t know that it’s ever been tried, because it’s most likely not possible. The only way I could see it being done is by two groups: one biased left, one biased right. Then a “random” sample of the same statements could be drawn, and both sides would judge their truthfulness. Then, a cross comparison would be made to see if there were any rulings that matched. Even that is not a good way to do it, because the two groups’ “bias” would have to be “equal” and there is no good way to measure that.
If the Truth-O-Meter is cheesy, it must be Swiss, because it has a lot of holes in it. Measuring it in “aggregate” as a point average is simply a different and interesting way to look at it. It’s easier than looking over the range of True to Pants on Fire, say, for example, when Bill Adair looked at Glenn Beck’s rulings recently on PolitiFact: “… It's fair to say that record skews toward the False end of the Truth-O-meter.” Using my percentages, it would have been more fair to say: Glenn Beck’s current average is 68.24, just below Barely True. However, including the ruling of August 25, his average is now 70.0. So Glenn Beck doesn’t skew toward false, he’s actually right spot on Barely True. He’s on average higher than Michael Savage, Bill O’Reilly and Rush Limbaugh. Those conservative pundits as a group measure 69.27, so Glenn Beck is slightly above that average. So, Bryan could say that Bill Adair, by stating the record “skews toward the False end” might have been a bit….biased. Adair should have said Beck was in the Barely True area, although toward the False side of the meter. On the other hand, the pundits (Conservative and Liberal) taken together average 71.94, so relatively speaking, ALL of them skew toward the False end of the Truth-O-Meter.
My agenda for calculating these averages was curiosity, to see if (1) there WAS a left trend, and (2) because I thought that on average, it was highly likely there’d be evidence that PolitiFact split the baby a lot (the overall average would be very close to Half True), and I proved it in my own way.
And (3) to go back to the PolitiFact comment page on Facebook, on Saturday, August 14, 2010, time stamped 11:22 a.m., Bryan posted in response to commenter Derek Grass: “There should be no serious question that PolitiFact is going to have a quota of sorts to keep the number of fact checks somewhat even. It just looks bad if one side gets checked much more than the other when you're billing yourself as "nonpartisan."
The data I compiled should provide an answer to that “quota” issue as well. I believe that they do—as also indicated by their putting out feelers to their Facebook followers for “Liberal” chain emails to offset all the Conservative chain emails they had been evaluating.
Bryan posted in the same thread in a conversation with commenter Derek Grass on August 13-14: “….I freely admit to my bias… I'm claiming that PF is biased toward the left.” Well, I freely admit to my bias as well. Whether PF is biased to the left remains to be seen.
No comments:
Post a Comment