Read this and marvel, once again, at Mother Nature.
. . . but Stata 15's ad does. (They're up to 15!--it seems like just yesterday when I was happily running regressions in release 8.) Part of the ad says, "Nonparametric regression--When you know something matters. But have no idea how."
- If you have "no idea how" something matters, maybe--just maybe--you should revisit your presumption that you know that it does.
- And if you think you've found out how solely using nonparametric regression--or any other statistical technique--you run a grave risk of committing a problem receiving a large, and increasing, amount of criticism these days: data mining, p-hacking, "torturing the data long enough until it confesses" are three of the names it's known by.
But that risk doesn't tend to worry the folks who sell statistics programs.
"NASA picked the 100 greatest images of Saturn from its Cassini mission — here are some of the best"
Pretty amazing that we can now take these pictures.
B-school professor gives a nice answer: he imagines what Big Al's faculty annual report for 1906 would have looked like.
Of course, he seems to have been wrong about the whole God and dice thing. (But the jury may still be out.)
"The Week My Husband Left And My House Was Burgled I Secured A Grant To Begin The Project That Became BRCA1"
For everyone, a great story . But if you were a fan of Joe D. it is a must read.
Link courtesy of Michael Greenspan.
In light of the recent award against Johnson & Johnson I found this interesting. It's seemingly well-balanced.
Good news and bad news. Good news: somebody is giving a little thought to this. Bad news: they work for the federal government.
Prof. Andrew Gelman being both true and funny.
And this brings up the question I want to address today: What sort of errors can we expect peer review to catch? . . .
To jump to the punch line: the problem with peer review is with the peers.
"Hexagonal clouds"? Really? Well, at least the explanation is not supernatural.