Also this week: US Supreme Court justice vs. linear regression, more on the Notre Dame fire vs. environmentalism, and more.
Jeff Ollerton on the response of (some) ecologists and environmentalists to the Notre Dame fire–“How come people are so upset about this building, which can be rebuilt? What about irreversible human impacts on the environment?”, or words to that effect. Jeff suggests that both the public outpouring of grief over the Notre Dame fire, and the whataboutism of some ecologists and environmentalists, are emotional responses and therefore equally “reasonable” and valid. I think he’s got a good point that they’re both emotional responses, and so debating their rights and wrongs is probably pointless at some level.
ESA President Laura Huenneke summarizes and comments on ESA member feedback on Plan S, the European initiative to force all scientific journals to move quickly to an author-pays open access model. Includes a very kind link to Brian’s post on the challenges of moving to open access.
A faculty position at a teaching-focused institution isn’t any easier or less stressful than one at a research-focused institution, it’s just different. Whether your faculty job is difficult or stressful for you is going to depend on factors specific to you and your institution, not on the type of institution.
Question to which I’m honestly unsure of the answer: in the grand scheme of things, how concerned should I be that a Supreme Court justice apparently doesn’t understand linear regression? Note that this terrible interpretation of regression comes from a justice with whom I disagree about many things, but I’d think it just as terrible if it came from any other justice. And of course, the question could (and probably should) be broadened to the way courts in general handle technical evidence of all sorts. I only know what I read in the news, but I’m pretty concerned about how US courts handle forensic “science”. OTOH, I’m sure judges, like ecologists or members of any other profession, would benefit from knowing more about lots of stuff. But there are only so many hours in a day, or a law degree, and I’ve no idea what (if anything) judges should learn less of to free up time to learn more statistics and science. (ht @noahpinion)
Insert your own mtcars joke here. 🙂 (ht @kjhealy) I like to imagine that somewhere somebody’s driving around a Model T Ford with a “STATVIEW” license plate. 🙂 I wonder if this guy still has his “SAS USER” plate? 🙂 In the comments, tell us: what’s your dream nerdy license plate? Nerdiest gets +1000 Internet Points.
RVWER3 – of course
Nice try, but I think Skip (below) has the nerdiness lead so far.
Hmmm. Maybe I *should* get the
H2C=CH2 license plate after all. Seems appropriate to have a wind stress hormone on a car driving hurricane speeds down the interstate.
A strong early contender for the +1000 Internet Points!
Since most of the commentary for DE seems to have shifted to Twitter, it’s pretty easy to stand out in the small sample size here.
“Since most of the commentary for DE seems to have shifted to Twitter”
Well, we do get fewer comments here than we used to. 😦 But we don’t really get more comments on Twitter than we used to, even if you count subtweets and other tweets that discuss our posts without linking to them. So I don’t think it’s commentary about our posts moving to Twitter so much as commentary about our posts slowly dying away.
About stats and courts, this is a nice ted talk: https://www.ted.com/talks/peter_donnelly_shows_how_stats_fool_juries
I’m less concerned that the supreme court justice didn’t understand regression. What concerns me more is that he put such complete lack of understanding down on paper in an official decision. This is a decision that was presumably reviewed by multiple clerks. An occasional typo or grammar mistake is extremely rare (at least on a per page basis) in supreme court decisions. The fact that this got in shows that folks don’t care enough about maths mistakes to run it by someone and that most folks don’t understand regression.
I believe this is a transcript of an oral argument, not a written decision.
Re Notre Dame, if you’ve not seen it, take a look at Spamletblog’s long comment on my post; very beautifully written and well argued.
Jeremy’s licence plate has got to be IDH1 surely?
“Jeremy’s licence plate has got to be IDH1 surely?”
-1000 Internet Points for you.
Sorry, low punch 🙂 Would DELTAZ be better?
WDELTAZ for the win!
The fact that Gorsuch even knows there is such a thing as linear regression is pretty good. I’d love to hear what the average American judge knows about linear regression.
But here he appears to be speaking in jest, exaggerating for effect and joking about questionable statistical claims of certainty.