Also this week: 1930s Japan vs. the US right now, the #pruittdataversary, R vs. Groundhog Day, Haldane’s diary, and more.
I know that for many of you it’s a welcome escape to read an ecology blog for a few minutes rather than spend yet more time paying attention to the appalling events in Washington, D.C. earlier this week. But although I certainly don’t have anything to say about those events that’s worth reading, they don’t feel to me like events that should pass totally unmarked even on an ecology blog. So here are a few things I read this week that I thought were good. If you only want the ecology links, just scroll down a bit.
The lessons of the right-wing coup attempts in Japan in the 1930s. I learned a lot from this.
Worried that updates to R packages will break your (or other people’s) old scripts? Arguably, you should be. Now there’s a easy to use solution, in the form of an…R package (ht Stephen Heard) Related: Meghan’s old post on how R improved the reproducibility of her work. Whether you think R is good or bad for reproducibility depends on what you’re comparing it to.
Am Nat EiC Dan Bolnick on how Am Nat handles concerns and complaints about papers it publishes, including anonymous PubPeer comments and comments from others who wish to remain anonymous. Speaking as an Am Nat editor: Am Nat has had a lot of experience with this issue in the past year; other journals should learn from us.
Andrew Hendry with the first and second in a series of posts on how he taught online-only intro bio to 600 students at McGill this past fall. My immediate reaction is that Andrew took full advantage of McGill’s early decision (compared to many other institutions) to go to online-only instruction. They gave their profs months to prepare.
Epidemiologist Whitney Robinson on the controversy over prioritizing coronavirus vaccine recipients. Precise and insightful. Also a good example for intro biostats classes.
Epidemiologist Deepti Gurdasani on the contributions of children attending in-person schooling to Covid-19 transmission. Useful clarifying thread; curious to hear what others think of it.
This is charming:
It’s approximately one year since the #pruittdata saga began. Somehow it seems appropriate to mark the #pruittdataversary* like this: there’s now a correction for Pinter-Wollman et al. 2016 Am Nat, a paper on which Jonathan Pruitt is last author. Summarizing briefly: this is mostly a theoretical paper, and no concerns have been raised about the theoretical portions. But a small portion of the paper refers to data collected Pruitt, which contains anomalies that Pruitt was unable to explain satisfactorily. Kudos to Pruitt’s co-authors for doing the right thing and promptly agreeing to the correction once they were made aware of the anomalies. Depressingly but unsurprisingly, Pruitt did not agree to the correction. Oh, and after the correction was published, an alert PubPeer commenter noticed that the anomalous data seem to have been duplicated in large part from Keiser & Pruitt 2014 Behav Ecol (a paper that itself already has an Expression of Concern attached for data anomalies). Here’s the kicker: according to Keiser & Pruitt 2014, the anomalous data in question were collected on spider species from Tennessee. According to Pinter-Wollman et al. 2016, the anomalous data in question were collected on different spider species from, um, Africa. Which is not in Tennessee. I have many…thoughts on this, but I am not going to share them right now because I can’t even. Instead, maybe I’ll go see if there’s anything good on Netflix.
*What do you get someone for their one year #pruittdataversary? One year anniversary is “paper”, right? What’s an appropriate paper gift? 🙂 😦
I just discovered Desert Island Discworld–the first podcast I’ve ever listened to. If you, like me, prefer rereading Terry Pratchett to reading new books because you can’t find anything else that measures up, you’ll like it.
And finally, we now go live to my laptop as it tries to calculate 350+ cumulative random effects meta-analyses: