Also this week: what intro courses should teach, field safety, grant lotteries, what probability is, and more. Oh, and Happy New Year!
From Jeremy:
Writing in Plos Biology, Kevin Gross and Carl Bergstrom give a theoretical argument for funding scientific researchers based on either a partial lottery, or on the basis of their past scientific success. Interesting exercise, though as a Canadian I refuse to accept the assumption that only a few grants can be funded. I wonder whether it would be better on balance to reduce average award size so as to be able to fund substantially more awards. See old posts from Brian, me, and me again for discussion. And here’s a different motivation for grant lotteries. (ht Meghan)
Very interesting and accessible interview with philosopher Rebecca Kukla. Touches on everything from what makes for interesting philosophy, to when (if ever) to use the phrase “speaking as a woman”, to the unfollowable norms and bad science that make it impossible to be pregnant without feeling guilty. Best thing I read this week.
Why economists shouldn’t teach behavioral economics in intro courses, even though doing so would improve “realism”. I link to this because some of the issues raised generalize to all intro courses. I particularly like the suggestion that intro courses should spend the most time on material that is of applied relevance, but that most students find counterintuitive or otherwise difficult to learn. What would that material be for intro ecology and evolution?
Happy 4th birthday to Scientist Sees Squirrel! Stephen Heard celebrates by looking back on his most underrated posts. Lately I’ve been thinking a lot about the point Stephen makes in this one.
Advice from a prof who often appears on cable tv news for other profs who want to do the same.
Andrew Gelman on what probability is. Interesting to me because I’ve long wondered what his view was on this, and because I hadn’t encountered this view before in my admittedly-scattershot reading in philosophy of probability.
The University of Ghent is radically changing how it evaluates academic staff. Not many details at the link, but as best I can make out evaluations will be much less frequent (every five years rather than annual), and will no longer be based (primarily? at all?) on quantitative metrics such as publication counts. Instead, “All staff members will make commitments about how they can contribute to the objectives of the department, the education programmes, the faculty and the university.” (ht @dsquareddigest)
However you wear academic robes, you’re doing it wrong. This is how you do it. 🙂 (ht Jacob Levy, via Twitter)
From Meghan:
I enjoyed this post by Robert Talbert on planning as an act of hope, reflecting on life, work, productivity, and hope. Here’s one section:
There are two choices: to spend my life purposefully, or to give up on the concept of purpose. To do the former makes assumptions. To do the latter is to give up hope.
My life, your life is finite one way or the other, so the question isn’t about how to prolong that life but how to use it, whether that’s a month or 50 more years, so that’s bigger than just the sum of our moments.
And the way we do this, is to have the audacity to make the assumption that we’ll have a life to lead in the future, and then guide how that life unfolds. That’s what planning is. That’s what GTD and productivity, otherwise just silly corporate buzzwords, is all about. Being intentional about your time, tasks, projects, calendar, and all the rest is stepping out on faith, a radical act of hope that says Whatever time I have left here, matters.
Daniel Bolnick had an important post advocating for better training for field emergencies and risk management related to field research. I don’t think I’ve ever been cavalier about safety in the field (I credit being the child of a firefighter and a nurse — safety first!), but this twitter thread last fall by Matthew Venesky made me think even harder about it. I discuss field safety with my lab folks, and we discuss that it’s the most dangerous thing we do as a lab. Our field work is local, but in rural areas where there’s poor (or no) cell service. So, we should probably spend more time before next field season thinking as a lab about what to do. At a minimum, we should have a list of which hospitals are closest to our different field sites — this seems obvious now, but it’s not something I thought to do before Dan’s post prompted me to think more about the golden hour. (Jeremy adds: here’s our old “ask us anything” on field work safety; many good resources in the comments.)
I love this post by Jen Heemstra and especially this part:
careers are all about growth. And, if you are willing to push yourself and grow a little bit each day, by the time you get to each stage, you will have the skills and knowledge you need to be successful there
(Jeremy adds: I see that Meghan’s New Year’s resolution is “steal back any links that Jeremy steals from me.” 🙂 )
Grant lotteries are intriguing, but I agree that more, smaller grants is better. To add to your posts on the topic, here’s my (quite old) one: https://scientistseessquirrel.wordpress.com/2015/05/12/why-grant-funding-should-be-spread-thinly/
This was one of my favourite posts, in part because I wrote a 1300 word blog post complete with a graphical model – and had someone use a 140-character tweet to dismiss it as “lazy”. Ah, Twitter…
I got similar dismissals when I proposed smaller grant sizes. Change creates winners and losers and the losers in a move to more small grants (i.e. those with frequent large grants) are not going to let go easily.
To my mind as long as we agree on the assumption of maximizing overall aggregate output, it is a simple empirical fact with many data points that many small grants is more effective.
Thank you for the thoughtful commentary and posts. Interestingly, replacing a few big grants with many small ones might reduce waste and increase efficiency in proposal competitions just as effectively as a partial lottery, at least according to our model. The key mechanism that increases efficiency in a partial lottery is that many investigators win small prizes (lottery tickets for a grant) instead of a few investigators winning large prizes (grants). However, those small prizes could also be small grants, and the same efficiency gains would accrue, as long as the value of the science underwritten by the award scales with the size of the award. In other words, a $200K award would need to produce 40% as much scientific value as a $500K award. If that scaling holds, then a many-small-awards approach would have ample merit, as it would reduce waste without adding explicit randomness to the award process. However, if there is a monetary threshold below which an award is too small to be scientifically useful — perhaps because a small award wouldn’t support the necessary replication to make an experiment worthwhile — then a qualifying lottery reduces waste while preserving award sizes that are large enough to make the funded projects feasible.
Of course, all of this is an attempt to make the best of a funding environment in which there isn’t enough funding to support all the funding-worthy ideas. Increasing funding so that the available funds are better matched to the number of funding-worthy ideas would provide a more satisfying resolution, to be sure.
Happy Effin’ New Years! Many thanks for the many good reads that you folks have written or linked to. They provide so much provocative thought and many enjoyable hours of reading instead of doing whatever it was I was supposed to be doing. (Who says procrastination is all bad?) A request: are there efficient browse strategies for archives? No topical key wording hidden away? You have an embarrassment of riches.
Thanks for the kind words Chris! Positive feedback helps us keep going.
“are there efficient browse strategies for archives?”
The search bar. Googling the topic of interest plus the words dynamic ecology. Odds are pretty good that whatever your topic of interest is, we’ve posted on it!
Emailing me and asking me if we’ve posted on [topic] works too. I leave it to you to decide if that counts as “efficient”. 🙂
I suppose “scrolling down” would be an efficient browsing strategy if all you’re looking for is “random posts with interesting-sounding titles”. 🙂 Or picking a random month from the archives toolbar on the right hand side of the homepage.
We used to classify posts into broad categories like “new ideas” and “controversial”. But nobody seemed to use the categorizations to find posts, and nobody complained when we stopped bothering (well, I think Meghan still categorizes her posts sometimes).
You used to be able to tweet to @dynamicecology and ask me if we’d ever posted on [topic]. But that won’t work any more because the @dynamicecology account is now purely a robot that tweets out announcements of new posts. I no longer log in to Twitter and so I won’t see any tweets to @dynamicecology.
The author of the economics post admits straight up that “many people”, which I read as almost everyone except economists, disagree with foundational economic ideas. Here is a direct quote:
“The core ideas of economics are extremely counterintuitive and are not accepted by most people. Thus economists face a difficult challenge in teaching the subject to non-economists. As an analogy, quantum mechanics seems very counterintuitive to me, and thus I have great difficulty in understanding the subject. It’s hard work teaching basic economics.”
But instead of interrogating these flawed heuristics (homo economicus etc), and incorporating greater realism, the author appears to be advocating restricting realism (i.e. the findings of behavioural economics) in order to favour the heuristics. he underscores this with appeal to a set of “myths” about economics, at least some of which have some basis in reality.
Meghan, you might consider a satellite phone for your lab’s field work – very roughly $1000+ to buy or $50/week to rent. These days it should provide 100% 24/7 coverage. They are about the size of a cell phone. Many have emergency assistance and/or GPS features.
Here’s an example, I don’t necessarily endorse:
https://satellitephonestore.com/catalog/rent/details/iridium-9575-extreme-satellite-phone-kit-7
I think there are many things in ecology that are unintuitive and tricky to learn, and it is good for ecology classes to teach this material — as, I think, many do. But what is there in introductory ecology classes that a member of the public would find counterintuitive? I’m not sure many people have strong preconceptions about the bread-and-butter topics of ecology — unlike economics. It would be interesting to see data on this.
“I’m not sure many people have strong preconceptions about the bread-and-butter topics of ecology ”
The candidate that comes to mind is one Brian and Meghan have both noted: naive “good of the species” thinking, and its cousin, thinking of nature as a harmonious, Gaia-type whole.
Not quite lotteries, but I have made to point to a couple of funders that incentives do not currently work for academics and others to participate in things like long term data collection and survey that form part of our national biodiversity strategy https://www.npws.ie/legislation/national-biodiversity-plan
For me, a solution would be to pay modest amounts for quality-controlled open data (say €2500-€5000 per data provider per year). I think this would be enough to incentivize coordinated fieldwork by academics and others that currently find it difficult to justify unfunded research. For a country the size of Ireland, there may be quite a lot of data generated for <€100,000.
It may be that the particular incentives are different in different countries. The focus on bringing in project funds and publication means that many Irish academics (in my experience) focus on this to the exclusion of other data collection (even though in theory our national strategy is looking for this). State agencies (e.g., EPA, NPWS) do some ecological monitoring, but there are concerns that they are limited by under funding https://iwt.ie/irish-wildlife-trust-submission-to-budget-2018/
Would be interested in views on academic experience in other countries and if similar funds-for-data schemes are known.
Interesting idea! I don’t know of any examples though.