Also this week: the
joy weird satisfaction of cleaning data, why publishing false and unjustified scientific claims might (sometimes) be good, unintentional entertainment in scientific writing, <- > =, and more.
Sad news: Philip Grime has passed away. He was a hugely influential plant ecologist. He proposed the CSR triangle model of plant life history strategies (Grime 1977), and a hump-backed model suggesting that plant diversity peaked at intermediate primary productivity. His work also was important to the development of the intermediate disturbance hypothesis. He was a prominent figure in several high-profile debates in community ecology, such as his debate with David Tilman over how to model plant competition, and the debate over whether sampling effects drive the relationship between biodiversity and ecosystem function. His ideas didn’t always pan out in the long run (see, e.g., here and here). But that doesn’t take away from their enormous influence. Among his many awards and honors, Philip Grime was a Fellow of the Royal Society of London.
Interesting and accessible discussion of the work of econometrician Isaiah Andrews. Raises issues about causal inference and robustness that are relevant to ecology.
The #pruittdata saga started with an Am Nat paper (see here for background). Now, #pruittdata is (hopefully!) over as far as Am Nat is concerned. Pruitt et al. 2012 Am Nat has received a correction and an unusually detailed Expression of Concern. Which means that all of Jonathan Pruitt’s Am Nat papers have now been retracted, corrected, and/or subjected to EoCs. As someone who’s been heavily involved with Am Nat’s investigations, boy do I wish I could say the saga is over for everyone else too. But it’s not. Not even close. There are still investigations ongoing at various other journals (see below). The McMaster University investigation is still ongoing (or it was as of a couple of months ago, anyway). Is Tennessee still investigating?
Speaking of ongoing #pruittdata investigations…I think I missed this at the time, but back in late Feb. 2021, Nature attached an Expression of Concern to Pruitt & Goodnight 2014. So, if you’re scoring at home, it took Nature ~13 months to get around to publishing an EoC that merely says (in so many words) ‘concerns have been raised; we’re investigating’. Even though they have in fact been investigating (or at least, have said that they’re investigating) that whole time. That’s Nature for you. Bob Trivers had to fight them for years and self-publish an entire book to get them to retract his own paper.
One more #pruittdata link: I second Flo Débarre’s comments. Investigating #pruittdata for Am Nat was an enormous amount of work, but it needed doing and it needed to be done right–thoroughly, promptly, and fairly. I’m proud to have so many colleagues who care so much about getting it right. I hope we’ve set an example that others will be inspired to follow if–god forbid–they’re ever faced with a similar situation in future.
Alex Usher analyzes what’s happening, and what happened, at bankrupt Laurentian University.
And here are Alex Usher’s comments on what the 2021 Canadian federal budget means for higher education in Canada. A bit disappointing that so much of the R&D funding is going to a few specific high-tech areas rather than basic science more broadly.
Matt Levine on negative logging: companies obtaining C credits by paying timberland owners not to cut down trees.
Forthcoming in Synthese, Dang & Bright argue that it can be appropriate to publish scientific claims that aren’t true, aren’t justified, and aren’t believed by their authors. I haven’t read it yet but it’s on my list. I’m curious if they think the only such claims are speculative claims (in which case I think their argument is correct), or if they think their argument applies to other sorts of claims as well (in which case I’m dubious but curious). Continuing to think about which ways of organizing science result in the most rapid collective progress.
Ecologist David Lindenmayer has won an out of court settlement in his defamation lawsuit against timber industry advocates. As part of the settlement, the timber industry advocates were required to publish apologies. (ht Retraction Watch)
Will Wilkinson on selection effects, personal identity, personal preference, self-actualization, and occupational diversity (or lack thereof). I want to think more about selection effects on different dimensions in the context of academia.
Ken Hughes on unintentional entertainment in scientific writing.
Wait, there are JSTOR hats? Or at least there used to be? Now I need to know if there are Data Dryad hats. There have been a couple of Dynamic Ecology t-shirts over the years (here’s one), but they’re not for sale.
Kieran Healy on cleaning data. Funny, to the point, and true. 🙂
And here’s Kieran Healy on why you shouldn’t use = as an assignment operator. 🙂
Why is a tree species listed as an author of the EoC in Am Nat? That gave me a chuckle.
It’s the pseudonym of a PubPeer commenter who submitted a comment to Am Nat on Pruitt et al.
As a Pubpeer denizen I understood why at once, but it still made me laugh! Good work, serviceberry tree person.
While I imagine Lindenmayer is probably right on the scientific and policy questions, I’m shocked at what counts as defamation in Australia!
If you read into the article they repeatedly called him a liar. That is an objectively true or untrue statement and it was untrue. And saying that a scientist has published intentionally untrue work is certainly damaging to their professional reputation and ultimately economic well being. The standard in the US for defamation is published, untrue, injurious and unprivileged (certain arenas like courtrooms and legislative assemblies are privileged, but debate over policy in a newspaper is not privileged). It certainly could have met the standard in the US (or at least been close enough that lawyers wouldn’t want to risk it going to court – this was settled out of court).
Pingback: Schneider Shorts 7.05.2021 – For Better Science