Also this week: crowd-sourcing natural history, Pimm vs. Possingham, the economics of essay writing services, more on the 80 hour work week myth, a great in-class exercise for teaching students about sampling error and bias, and more…
From Brian (!):
In a nice intersection of two recent popular posts on DE (80 hour work weeks and lognormal productivity in intellectual fields), computer programmers wonder if their profession requires working most waking hours to be at the top of their profession. Especially check out the top ranked answer (the first paragraph is the question).
A nice tool for easily accessing species occurrence data in R across many different compendiums.
EEB and Flow’s take on prediction in ecology (Jeremy adds: see below for my comments on this one).
Statistician and blogger Cosma Shalizi just got tenure. He’s pleased and relieved, but not triumphant.
Speaking of Cosma Shalizi, in this recent post he highlights the latest work on causal inference and structural equation modeling. Apparently someone’s come up with a way to unify graphical models of causality with counterfactual approaches. I’m guessing this will be of interest to some of you (hi Jarrett!) After you absorb it, write a post explaining it to me, because it was too technical for me to even get the gist…
On the economics of essay-writing services. See the comments for some insight into who the customers are for these services, and what they’re buying.
In the late 1960s, only a minority of those working in American four-year higher educational colleges tended to publish regularly; today over sixty per cent do … In 1969 only half of American academics in universities had published during the previous two years; by the late 1990s, the figure had risen to two-thirds, with even higher proportions in the research universities. The number of prolific publishers is increasing. In American universities the proportion of faculty, who had published five or more publications in the previous two years, exploded from a quarter in 1987 to nearly two-thirds by 1998, with the rise in the natural and social sciences particularly noticeable
A great in-class exercise for introducing statistics students to the concepts of sampling error and bias, and to the importance of random sampling. And not just because it involves candy.🙂
A cynical theory of what determines the popularity of “obvious” vs. “non-obvious” methods and approaches, namely “do they give the desired answer?” A bit too cynical (and simple) for me, actually. But there are ecological examples where I wonder if something like this is going on (e.g., see this comment and the subsequent thread).
As discussed here and here, at the recent American Society of Naturalist’s meeting in Asilomar there was a formal debate on ecological vs. evolutionary determinants of species richness. I just found out that a video of the debate is online. Cool! Skip to 24:17 and 47:40 for Luke Harmon’s remarks on “zombie ideas”.🙂
CBC Radio 1 has a program called Tooth and Claw that hosts conservation-related debates. They just posted the full, unedited debate between ecologists Stuart Pimm and Hugh Possingham on conservation triage. Other recent debates concern the wisdom of de-extinction, and human-animal relationships (with Frans de Waal).
I’m a bit late to this but wanted to pass it on anyway. Writing in the New York Times, ecologist Arthur Middleton asks whether ecologists are mythologizing Yellowstone’s wolves. Somewhat in the spirit of our recent discussion of the value of biodiversity. The EEB and Flow has related commentary, asking whether the problem is that ecologists set themselves too high a bar when it comes to “predicting” the effects of the wolf reintroduction.
Speaking of prediction, a while back I reviewed journalist Nate Silver’s very good book on prediction. Silver’s new “data journalism” website, FiveThirtyEight, has now launched. Tyler Cowen is underwhelmed, and so is Paul Krugman. I agree with both of them, while also sympathizing with Silver and his staff. I also sometimes struggle with what to write and how to write it, even though I’m aiming for a much smaller and narrower audience. It’s early days, of course, so maybe FiveThirtyEight will improve. Or perhaps my disappointment just indicates that I’m not part of the intended audience? Maybe Silver is aiming for readers who have much less quantitative training? This is something I also struggle with from time to time, too. Probably every writer does. The audience we’re writing for, or wish we were writing for, isn’t always the same as the audience we actually have.
I just found ecologist Petr Keil’s quite good blog. I liked this post on different motivations for doing science, and this one on how data mining is a failure of imagination. And Brian will probably appreciate Petr’s bemusement that those calling for a new discipline of “macrosystems ecology” apparently don’t see much connection to the already-existing discipline of macroecology.
I just did this to a student of mine. Sorry Stephen.🙂