Also this week: why girls shy away from math-heavy STEM fields, do preprints make authors patient with reviewers, statistical vignettes as teaching tools, and more.
Interviews with several experienced ecologists, reflecting on changes in the discipline in the last few decades.
A big Gates Foundation-funded initiative to identify college students at risk of failing and provide them enhanced advising had absolutely no effect on student outcomes in a massive randomized controlled trial across three US colleges and universities. Indeed, at one of the three institutions, there were hints that the program worsened student outcomes. Perhaps because students were discouraged by being flagged as at risk, or dropped out rather than take extra meetings with advisers. Kudos to the Gates Foundation for funding this study, designing it well, and publishing the results–but there’s no getting around that the results are discouraging.
Why are women massively underrepresented in heavily-mathematical fields like physics, engineering, and, um, mathematics? Various reasons, no doubt. But one big one, that affects educational and career choices starting many years before college, seems to be that girls have a comparative advantage in reading, boys have a comparative advantage in math. That is, girls tend to be better at reading than they are at math, whereas the reverse tends to be true for boys. So if everybody plays to their own strengths, you end up with girls avoiding math-heavy fields and boys going into math-heavy fields. The link goes to a new PNAS paper by Breda and Napp, based on PISA test data, survey responses, and subsequent educational choices for 300,000 15-year old students from 64 countries. I skimmed it, it seemed pretty convincing to me, but I freely admit I’m no expert here so would welcome comments from those who know more than me. I liked that the paper at least addresses alternative hypotheses like girls misperceiving their own mathematical abilities, and boys and girls choosing the fields where their absolute (not comparative) advantages lie. Though it’s worth noting that some of those alternatives aren’t mutually exclusive with the comparative advantage hypothesis. One take-away for me was that, rather than asking “why are girls choosing not to go into some STEM fields?”, we might be better off asking “why do girls, and boys, make the educational and career choices they do?” After all, every girl who doesn’t go into a math-heavy field is a girl who goes on to something else instead, and every boy who goes into a math-heavy field is a boy who doesn’t go on to something else instead. But what do you think? (ht Marginal Revolution)
Chill out about self-citations. The vast majority of among-researcher variation in self-citation rate merely reflects which researchers have run cohesive research programs in one study system for a long time, rather than bouncing among topics and study systems. And scroll down to the comments, where Steve Cooke and Andrew Hendry catch a potentially-serious technical flaw in that Plos Biology paper on self-citations that everyone is talking about. And that’s before we even start talking about how blanket criticism of any and all self-citation is likely to differentially impact the wrong people.
I love this idea for teaching graduate students how to write papers!:
I think an equally good version would be to assign students a published paper by a good writer, that they haven’t yet read. Don’t show them the abstract and intro. Just give them the rest of the paper, have them write an abstract and intro, then compare to the actual abstract and intro. (ht @noahpinion)
Interesting hypothesis: it’s now less important for reviewers to complete peer reviews quickly, because authors can just post preprints to make their work available to anyone who would like to see it immediately:
Curious to hear what folks think of this. As an author, are you now happy to wait however long it takes to receive good careful reviews, knowing that your preprint has already been posted?
Hoisted from the archives: statistical vignette of the day as a teaching tool. Feel free to steal these for your own undergrad stats courses.
The obituary for Cordelia Scaife May, who passed away recently, makes for sobering reading for anyone who supports environmental causes (which I do), or thinks the Earth is or soon will be “overpopulated” (which I don’t). May was a wealthy philanthropist who started out supporting bird conservation. She moved into supporting family planning via Planned Parenthood due to concern about overpopulation. But she eventually dedicated her life and funds to virulently anti-immigrant movements, feeling the US was being “invaded on all fronts” by foreigners who “breed like hamsters”. She’s far from the only prominent white supporter of environmental causes to find it a very short step from caring about nature to Donald Trump’s position on immigration–think for instance of Madison Grant or John Tanton. The latter was inspired in part by ecologist Paul Ehrlich’s arguments on overpopulation and the anti-immigration implications he drew from them. I definitely wouldn’t suggest that environmentalism is always or necessarily anti-immigrant or racist. Like any view it can (and for many environmentalists, has) evolved away from its racist roots. But still, this is history ecologists should know and reflect on.