Many will have already seen my analysis of Matt Braynard’s poll asking people in several states whether they requested an absentee ballot, and whether or not they returned that ballot. Gist: maybe 150 thousand votes, plus or minus, went missing.
I was not the first one to peer into this poll. The first was Stephen Miller, a math prof at Williams. His analysis was different than mine in small details, but his conclusion and mine were the same.
His analysis somehow found its way into the hands of his professorial colleagues. Being sober, serious people—they are academics!—you’d guess they examined Miller’s report carefully and offered collegial, behind-the-scene criticisms in their earnest and honest attempt to improve the analysis and grasp the truth.
What if I tell that instead they reacted to the news of potential cheating like teenage girls being deprived of their cellphones?
See which version more accords with the facts.
The Williams Record has the story.
Among the statement’s critics is Associate Chair of Statistics Richard De Veaux, who is also the vice president of the American Statistical Association. De Veaux described Miller’s document as “completely without merit” and “both irresponsible and unethical” as part of a longer rebuttal that also accused Miller of violating at least seven out of the 10 guidelines for ethical statisticians laid out by the American Statistical Association.
This De Veaux said Miller got the poll data from an “obviously biased source”. It is unethical and immoral to falsely accuse a colleague of unethical behavior based on an unproven assumptions, as De Veaux did—repeatedly. Loathsome slimy behavior.
De Veaux has no proof that the poll Braynard commissioned was biased. To him the mere logical possibility that the data could be biased is sufficient proof it is biased. This is like saying that because it is logical possibility De Veaux misconducts himself with horses, he therefore has, and that we should now wear gloves if we’re forced to shake hands with him.
There was and is no reason to suppose the polling firm Braynard used was any better or any worse than any other. Or does De Veaux condemn all polls on principle? Let him say so, if he does. The Braynard survey is simplicity itself, using official sources. The calls the firm made were recorded—not so usual, that!
Did any of the attacks against Miller have any merit?
De Veaux, [George] Marcus and [Charles] Stewart all pointed out that Braynard’s data is likely flawed in a number of ways, including that response rates were relatively low, that Braynard only reached out to registered Republicans and that voters often tell pollsters untrue statements on their voting history.
Sure, the response rate was “relatively low”, which does not imply “too low”, which is obvious. It is more than large enough to make the reasonable extrapolations Miller made. De Veaux should know this. Perhaps he was only pretending not to.
These self-congratulating smug-bunnies make the same consorting-with-horses error here as De Veaux made above. Yes, it’s true voters sometimes lie about voting history—though this wasn’t about voting history per se, a distinction they all missed—but it does not follow that therefore this poll’s respondents did lie, or did in such a way to bias the results in only one direction.
Their conclusion is reached by soy-drenched anemic noodle-brained reasoning. Biased reasoning.
Also: in Pennsylvania the respondents were Republicans; the other four states had voters from all parties. But that it was only Republicans in PA in no way invalidates the results. Good grief, how could it! The data still applies to Republicans who, it may well turn out, got shafted.
This isn’t the end of the story. Miller is a nice guy, new to politics. He honestly believes his colleagues only have in mind what’s best for him. He doesn’t understand that his “friends” were happy to sacrifice him to make themselves look better, and to show Miller his place.
We know this thanks to the Berkshire Eagle, which ran the misleading article “Williams prof disavows own finding of mishandled GOP ballots”.
Part of the story is a collection of quotes from other academics who might (it is logically possible) have misconducted themselves with various farm animals. Example: “Lior Pachter, a computational biologist at the California Institute of Technology, said that simple issues, such as incorrect phone numbers, could have accounted for some of the concerning patterns that Miller saw.”
Yes, they could, Pachter. Just like your silly comments could be evidence of a disturbing desire to hang out down at the morgue.
All this would, could, might, possible nonsense is not evidence. These luminaries should know better, and probably do, which makes the whole thing sad.
Because they beat Miller into submission. He apologized for a “lack of clarity and due diligence”, an apology that wasn’t necessary, because it isn’t true. Well, poor Miller will come to understand just how much charity is in the hearts of his “friends”.
To support this site and its wholly independent host using credit card or PayPal (in any amount) click here