Andrew Gelman's @StatModeling take on the Brian Wansink debacle is the best I've seen: bad science often comes from wishful thinking, not ill intent.
Wansink told me his goal was libertarianism, to limit the need for govt by helping people self-regulate. Wishful, not wicked.
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I was quoted in that NPR piece, but I agree, I wish there had been less focus on the *p-hacking* (and a better explanation of p-values...) and more on the overall chicanery apparent in Wansink's work. For example, the three studies that all ended up with *770* questionnaires...
-
...one study allegedly mailed out 2000 questionnaires, a second mailed out 1600, a third mailed out 1002, and they all got *exactly* 770 questionnaires returned? Some of those issues looked like stronger evidence of tomfoolery than the p-hacking.
End of conversation
New conversation -
-
-
I think a relatively simple way to deal with this is for the PI to be asked to sign a certification that the hypothesis was specified prior to analysis. A COI - “certification of intent.” This is by far the most common form of scientific malpractice.
-
That’s not practical. As
@StatModeling wrote, “Nononononono. Sometimes it’s fine to make a single hypothesis and test it. But work in a field such as eating behavior is inherently speculative.” - 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.