Friday links: I want to work but I can’t, fame vs. h-index, and more

Also this week: online events for early career researchers in evolution, the coming college apocalypse, and more.

From Jeremy:

I want to work but I can’t. This resonated with me.

Cornell University plans to hold in-person classes in the fall, arguing that it will actually result in fewer coronavirus infections among students and staff than an online semester would. That argument of course depends on various assumptions, such as that all undergrads would return to Ithaca even if classes were online-only. But the crucial assumption is that, by holding classes in person, the university can force even asymptomatic students to take frequent coronavirus tests (e.g., by denying untested students access to email accounts and residence halls). Meanwhile, at USC…

Following on from the previous links, here’s professor of US higher education finance Robert Kelchen on why the next month is going to be awful for US colleges and universities.

Upcoming online events for early career researchers in evolution, from the ASN, SSE, and SSB.

Ken Hughes on the half-life of citations, and why he cites so much old stuff.

Writing in Ecology and Evolution, Roxanne Beltran et al. report data on UC Santa Cruz undergraduate student demography (race/ethnicity, socioeconomic status, gender, first-generation), major choice, GPA, and graduation rate in relation to whether or not the students took field courses. They also report survey data on self-efficacy gains. They find that students from historically underrepresented groups are just as likely as others to be interested in EEB majors when they start university, but are less likely to graduate as EEB majors, both because they’re less likely to graduate and because they’re more likely to switch to other majors. That may be in part because students from historically underrepresented groups are less likely to take field courses.

Here’s a forthcoming paper on what a pioneering advocate of experimental approaches in economics, Charles Plott, had to say to get reviewers and editors to accept his papers. I’d be curious to see studies like this of other pioneers of new approaches in other fields. Particularly curious whether private debates about the merits of new approach X (in the form of reviews and responses to reviews) mirror public debates.

Examples of famous scholars with low h-indices. Examples include Nobel Prize winner Peter Higgs (of Higgs boson fame) and Andrew Wiles (prover of Fermat’s Last Theorem). HT a correspondent, who noted that ecologists who could be added to that list include Robert MacArthur (WoS h-index of 21), Stephen Fretwell (11), and Ray Lindeman (2). From evolutionary biology, George Price is the first person who comes to my mind. Clearly, the best ways to become a famous scholar with a low h-index are (i) make a major discovery or solve a really important problem, (ii) have the foresight to become famous many decades ago, and (iii) die young.

15 thoughts on “Friday links: I want to work but I can’t, fame vs. h-index, and more

  1. A list of FIFTY things to stop you procrastinating?! Seriously? I appreciate the sentiment and especially agree with 1., about the relationship between mental health and procrastination, but this seems over the top. To whit:

    51. Stop reading long lists of ways in which to stop procrastinating….

  2. In my opinion, the three leading reasons why MacArthur, Lindeman, and Fretwell have such a low h-index are (1) much smaller pool of academic peers who would cite papers when those authors were active, (2) an ever-increasing tendency for present-day authors to cite very recent papers and not the original contributions that opened a field and made the primary intellectual contributions, and (3) short to very short lifetimes for MacArthur and Lindeman, respectively, limiting their total output.

    • Yup, (1) and (3) are surely important here. IIRC, the evidence on (2) is actually mixed, but I may not recall correctly; I haven’t looked at the data in years.

      Which raises the question: if you restricted attention to, say, post-tenure ecologists who were active for many years after, say, 1985, who are the ones who are most famous, *relative to their h-indices*?

  3. maybe the question could be rephrased as ‘how much of one’s reputation rests on the 10-20 most cited publications over one’s career? Even if the career is/has been long and resulted in , say, 100-200 publications, with an h-index of 60+. I suspect for many of our top ecologists/evolutionists the top 10 papers/books are the reason they are tops. So perhaps the h-index is a poor way to measure impact, certainly not an original suggestion in the h-lndex literature. Personally I think it does not capture much of interest in scientific excellence: who wants /expects to be remembered for the papers you wrote that were cited > 60 times each! 500 times in web-of-science for ecology/evolution seems a better cut-off for REAL impact [ maybe 700 times in Google Scholar].

    • “maybe the question could be rephrased as ‘how much of one’s reputation rests on the 10-20 most cited publications over one’s career? ”

      Yes, agreed. And perhaps also “how much does your reputation rest on non-research contributions, such as writing a textbook, or leadership of a large organization?”

      • Yes, I agree that reputation rests on more than just impact of original science. And text books and high level service are part of the more. so is graduate training.
        In applied ecology [ fisheries, etc] so is development and implementation of management plans. Or development of techniques that may be widely used in practice if not widely cited .
        But some of us value research contributions over all else.

      • “But some of us value research contributions over all else.”

        Yes, “my fame/reputation is higher than you’d predict based on my h-index, because of my non-research contributions” is different than “my fame/reputation is higher than you’d predict based on my h-index, because a few of my research contributions were enormously important”.

    • I think the problem is we have all let ourselves off the hook for assessing quality of other people’s contributions (‘too far from my field’ but probably we really mean is I don’t have time to read their top 3 papers and think) so we now just go for quantity. # of papers and # of citations being easiest, and h-index being a very fancy combination of the two that obfuscates how simplistic we’re being.

      But as you point out, even if we want to be lazy and quantitative there are other ways of assessing quality. Of course the challenge is the 500 citation threshold will vary with field and career stage. But citations are awfully linear with age – it wouldn’t be hard to modify slightly to a rate – say 20 citations/year – how many papers reach that thresh hold?

      • The h-index seems to have been proposed to erase the possibility that one’s reputation could/should rest on a very few high impact publications. and of course there is no reason that a single measure ever captures all we want; even the original h-index paper proposed several measures in addition to the h. or to be used along with h.
        very few papers reach a life-time total of 500 WebofSci cits [ I dont know the percentile for ecology/evolution], and about 25-30 yrs ago Eugene Garfield [ inventor of WOS, Science Citation classics measure, etc] pegged the Citation Classic cut-off at 400, but noted that it varied by field; a few recent studies in Economics and fisheries used 500 ( ref?). Jane Lubchenco, while at NOAA, claimed to have authored 8 science citation classics, and must have meant she had 8 papers above 400 citations.
        Papers cited highly/yr are certainly candidates for for high impact [ say 25-30/yr], but to be a classic there must be ‘staying power’ too. The science citation people have long used a wide range of measures and time scales too.

  4. Comments from Paul Romer on how Cornell’s reopening plan depends on pooled testing. Argues that that’s the way forward for both colleges and universities, and K-12 schools.

    • A little light on details to my mind. Specifically relies on scaling something that worked or is hoped to work at two of the premier and richest institutions in USA – Cornell and Stanford with their own in-built testing capability. Follow on estimates I see estimate the cost of replicating this in all K-12 schools to be a $1 trillion at a time when most districts are expecting to see the largest budget cuts in a generation (10-20%) due to falling state revenues.

      And assumptions about college-faculty & college-community interactions are naive at best (both assume that college students return to dorms, faculty & university operate at full speed and community interacts with campus as its null model to measure against).

      Could this all be still achievable if US federal government steps up big time? At the edge. But really who is betting on the US government right now? Seriously even if they ponied up the money, we haven’t even gotten testing up to adequate levels for testing suspected COVID cases. Where is the technological capacity for testing every school child every week (even with pooling) going to come from?

      • In Romer’s defense, I think he does say elsewhere that for K-12 schools to do this, the feds need to pay for it. Though I don’t know if he’s ever costed it out anywhere.

        And yes, one of the most dismaying/infuriating/horrifying aspects to all these policy debates in the US right now is debating “what’s optimal?” vs. “what’s optimal, given that large important chunks of the US government are some combination of incompetent, hamstrung by partisanship, and corrupt?”

  5. Possibly of interest, especially to Meghan due to the inclusion of Daphnia: A friend of a friend has developed a very simple cellular-automata-based game of trying to keep a simple aquatic ecosystem alive, focusing on balancing nutrients, Oxygen, and C02. Amusing, and a good way to visualize cycles in population and environmental variables (plus simple aspects of spatial structure).

    Perhaps these very much toy models could be (or are already) used in teaching?

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.