Also this week: the bibliometrics of ecology and evolution, Johns Hopkins vs. legacy admissions, and more.
A rare retraction in ecology and evolution, this one from American Naturalist, for “irregularities” in the raw data. Apparently, this is the first ever retraction from Am Nat, at least as far as anyone I’ve spoken to knows. Here are some details from first author Kate Laskowski, who it is important to emphasize did not collect the raw data, and who deserves great credit for reacting quickly and appropriately when the irregularities in the raw data were first noticed. There appear to be two other retractions in the pipeline, it sounds like for similar irregularities in their respective datasets. Retraction Watch story here.
Writing in Scientometrics, et al. report a bibliometric analysis of ecology and evolution from 1975-2014. Contrary to what I would’ve expected, they don’t find evidence for closer integration of evolution and ecology over time. As I would’ve expected, they find that both fields have switched from being organized around taxa to being organized around concepts. See this old post for related discussion.
On the Alberta government vs. its own universities. I mostly agree with this.
Johns Hopkins University is ending legacy admissions.
A critique of a recent high-profile PNAS paper finding no racial bias in police shootings, arguing that the original study flouts Bayes’ Rule. I link to this mostly to remind myself to look into it more closely (I haven’t yet looked at the reply to the critique), with a view to possibly adding this to my compilation of “statistical vignettes” useful for teaching intro stats courses.
Writing in TREE, Lord et al. question the evidence from the Russian “Farm-Fox” study that selection for tameness leads to rapid evolution of “domestication syndrome”–or even that there is such a thing as “domestication syndrome”. I link to this mainly to
talk about “zombie ideas” yet again give you the chance to make terrible jokes at my expense in the comments. Don’t say I never did anything for you. 🙂 There’s also an interesting Canadian connection to the Farm-Fox story that was news to me, and I’m guessing will be news to many of you as well.
I admit it; I laughed. 🙂
I liked this piece on how to stop freaking out and tackle climate change. I especially liked point 5, which I think could be useful for sharing with students in my Intro bio course. It begins:
Even though keeping global warming under 2.7 degrees Fahrenheit (1.5 degrees Celsius) would absolutely be better than 3.6 degrees Fahrenheit (2 degrees Celsius) of warming, there is no threshold that means that it is “too late” or that we are “doomed.” The lower, the better. It is always worth fighting.
As we fight, it is important for our mental health and motivation to have an image in mind of our goal: a realistically good future.
Imagine dense but livable cities veined with public transit and leafy parks, infrastructure humming away to remove carbon dioxide from the atmosphere, fake meat that tastes better than the real thing, species recovering and rewilding the world, the rivers silver with fish, the skies musical with flocking birds.
I also enjoyed this piece by Beronda Montgomery, on how she works from affirmation, not for affirmation.
Here is a compilation of podcasts about climate change that focus on innovation and empowerment, worth checking out! https://www.climatepodcasts.com/
I’m genuinely curious about this retraction thing. I feel like I’m seeing it more and more that anytime a retraction results from genuine errors, the authors are praised and (as you mentioned) deserve great credit. Don’t get me wrong, I’m not saying that’s the wrong attitude. Some of the consequences are interesting though: at least one person on the twitter thread said they now trust the author’s other published work even more.
Speaking purely personally, my default assumption is that every paper I read has made no errors, until I’m given a good reason to think otherwise. And my default assumption is that any error an author makes is a one-off, until given good reason to think otherwise. Not because I think errors never happen, but because any alternative default assumption would be highly inefficient. If I didn’t default to trusting every paper and author I read, until given a reason not to do so, I’d have to waste scads of time satisfying myself that every paper I read is error free.
So for instance, three good friends of mine have had to correct or retract papers for inadvertent errors (https://dynamicecology.wordpress.com/2016/12/02/on-finding-errors-in-ones-published-analyses/; http://retractionwatch.com/2015/04/09/stats-error-has-chilling-effect-on-global-warming-paper/). But my trust in their other work hasn’t gone up (or down) because of that, because I already default to trusting their work, and I have no reason to think that those inadvertent errors are anything other than rare one-offs.
Concerning the retraction, a few other papers by the same working group have had more potential irregularities highlighted on pubpeer:
Just in case you haven’t seen, this recent editorial (and resulting Twitter discussion) was quite shocking.
Yeah, was just sent that a few minutes ago by a colleague. It’s now an entry in the next linkfest.