Also this week: advice on social media for scientists, beating your fear of networking, a clever new study of the “researcher degrees of freedom” problem, the faculty’s bark vs. the provost’s bite, and more…
This piece gives three perspectives on how to use social media for science, based on an event at the recent meeting of the American Association for the Advancement of Science (AAAS). Maggie Koerth-Baker says that you need a goal (e.g., developing more professional connections, engaging with the public), but also recommends that you not force yourself to use social media if you’re not interested. Kim Cobb has a good list of why scientists can be uneasy about using social media. She also gives her personal reasons for using social media, which includes wanting “to help change the culture of science regarding public engagement with climate change, and women in science”. And, finally, Danielle Lee, talks about how social media can be used to promote diversity in science and to help bring science (and accurate portrayals of science) to a broader audience.
I always tell students that they need to plot data before the analyze it, and this is a great illustration of why: Anscombe’s quartet is made up of four datasets that have nearly identical basic statistical properties (mean, variance, correlation), even though the underlying data are clearly very different. (ht: Owen Jones)
Here’s a great old essay from economist Paul Krugman, on “accidental theorists”. It’s a playful–and very effective–argument for the value of highly-simplified models as an aid to thinking about complex real-world situations. Also notes that people who think they’re sticking close to the data and so avoiding abstract theorizing are always operating on the basis of some implicit (and usually seriously problematic) theory. It’s about economics, but it’s totally accessible (it’s from Slate), and the relevance to ecology will be obvious. Go on, click through–it’s the best thing you’ll read today! Relates to many of our old posts, like this one on the value of whimsical thought experiments.
Here’s a list of “lab lit”: novels featuring academic scientists. The only one on the list that I’ve read is Thinks…, which I highly recommend (most any David Lodge is worth your time). I’m glad to have found this, since I like reading campus novels and want to try some with scientists as main characters. Except for Thinks…, the various campus comedies I’ve read (David Lodge’s campus trilogy, Moo, Lucky Jim) mostly focus on the humanities, social sciences, or general campus life.
Andrew Hendry with a report from Charles Darwin’s local pub.
Congratulations to Simon Levin for winning the 2014 Tyler Prize for Environmental Achievement! (ht Chris Klausmeier)
The Molecular Ecologist has a wide-ranging interview with evolutionary biologist Charles Goodnight. Talks a lot about how the advent of DNA sequencing changed (and didn’t change) evolutionary biology. Also includes his experience starting his research blog, which he’s using to try to rethink evolution as phenotype-centered rather than gene-centered. And there are some choice quotes, such as this:
In college I learned a lot of facts and a lot of concepts about biology. Most of the concepts, in some form or another have served me well. Most of the facts have changed entirely…it is clear that teaching and learning facts is mostly a waste of time. Teaching and learning concepts are very valuable.
On avoiding statistical machismo when teaching statistics to undergraduates. By an economist, but it’s perfectly accessible to (and will resonate with) ecologists.
Retraction Watch reports on an interesting-sounding study in which 142 graduate theses in management and applied psychology were compared with the resulting publications. Compared to the theses, the resulting publications report a much higher proportion of results supporting the authors’ hypotheses. That’s down to numerous changes to the analyses reported in the theses. Some hypotheses were dropped, new ones were added, some were reversed in direction (!), data were dropped, variables were added to or removed from statistical analyses…Talk about researcher degrees of freedom! It would be interesting to see similar studies from other fields. As noted in previous Friday linkfests, other lines of evidence indicate that the importance of the “researcher degrees of freedom” problem varies a lot among fields. (Aside: I did find it bizarre that the authors of this study complain that less than half of the statistical tests reported in the theses came out significant. Why the heck should that be considered a bad thing?)
Faculty bark, the provost bites. This is true. (ht Terry McGlynn, via Twitter)