|
Jeremy Koster@jeremykoster.bsky.social |
86 followers 58 following 77 posts
Aligning incentives for better science - quality over status
Signer of DORA: sfdora.org/
Co-director of ENDOW project: endowproject.github.io/
Interdisciplinary socio-ecological scientist advocating for congruence of theory, data, and stats
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
"A survival analysis reveals that the median age of first marriage in this community was 17 for women and 19 for men." "Neither the husband’s age at marriage (β = −0.01) nor the wife’s age at marriage (β = −0.02) are predictive of the risk of divorce." - Sounds like cultural differences are relevant
1 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
For additional contrasts to the Malawi data, the paper by Kristin Snopkowsi about divorce in Bolivia is a good one:
link.springer.com/article/10.1...
0 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Interestingly, having children together reduces the probability of divorce among married Mayangna/Miskito couples:
www.mdpi.com/2076-0760/10...
1 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
One of the suggested references in this compelling piece by Gonzalez and Rice is a new textbook, Human Behavioral Ecology. Instructors can request an exam copy from the Cambridge website:
www.cambridge.org/us/universit...
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Anyone who follows these debates probably guessed correctly that the abstract is referring to MDPI and Frontiers. But can we also agree that calculating journal impact factors over short time frames (the previous two years) is a terrible metric in any case?
0 replies 1 reposts 3 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
For mentors of smart and ambitious undergraduate students who are looking to publish their research, the Cambridge Journal of Human Behaviour could be worth a look.
cjhumanbehaviour.com/about/
0 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Kermyt's book with Peter Gray, Fatherhood, is also a great addition to the reading list:
www.hup.harvard.edu/books/978067...
0 replies 1 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Blue Sky should have an extra big heart tapback option for the skeets that I really love. The counting impulse is strong, but it would be ideal to incentivize lasting science.
www.cell.com/trends/cogni...
1 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To what extent does this resonate?
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Practically all of the most useful statistical knowledge I have gleaned over the years was imparted by full-fledged statisticians. Rather than hiring methodologists in Psychology (and other disciplinary) departments, what about fostering opportunities for students to take courses with statisticians?
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Ecological Momentary Assessment scholars should hang out with animal behavior scholars because the structures of their dataset are basically the same:
link.springer.com/article/10.1...
0 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Thanks, Rebecca. Maybe our suggestions will stand the test of time, or maybe not. In either case, it seems worthwhile to reflect on the outstanding questions of our time and the ways in which the science and incentives can be realigned to answer those questions convincingly. @rmcelreath.bsky.social
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Exam copies of Cambridge's new textbook, Human Behavioral Ecology, can be requested by instructors who are considering the volume for use in their classes. Copies can be requested by following the link on this page:
www.cambridge.org/us/universit...
0 replies 0 reposts 4 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
There are not many human behavioral ecologists who contend that there is a "strict sexual labor division" presently or in the Paleolithic. A recent rebuttal summarizes some of the evidence:
www.biorxiv.org/content/10.1...
Also see this new chapter: www.cambridge.org/core/books/a...
0 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
“Everybody thinks of the immediate gains they might get from having a paper out somewhere, but we really should be thinking about the long-term sustainability of the research ecosystem. After you’ve been dead for 100 years, are people going to be able to get access to the things you’ve worked on?”
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
One of the worst things about academia circa 2024 is that researchers are incentivized to think not only about novel scientific questions but also, "Will I be cited a lot if I publish this study I'm proposing?" And those sorts of considerations then impact everything you listed, unfortunately.
0 replies 0 reposts 3 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Goodhart's law seems relevant here: "When a measure becomes a target, it ceases to be a good measure."
1 replies 0 reposts 3 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
The corresponding values from the Mayangna data are 19 minutes per day for men and 1 minute per day for women, on average. (see Table 5.3)
etda.libraries.psu.edu/files/final_...
There are lots of additional time allocation data out there.
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
For the "Hunting is not a binary phenomenon" discussion, it could be helpful to note relevant work on time allocation. For example, Hames (1989) compiled cross-cultural data, showing that men (left) hunt substantially more minutes per day than females (right) in a sample of lowland societies.
2 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Oh, no doubt there is gaming of Google Scholar in that way, including black and white cases. (My comment about gray areas was meant to add more to the overall picture.) A tenure committee (or hiring committee) that relies heavily on Google Scholar metrics is going to make some bad decisions.
0 replies 0 reposts 3 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
There are potentially gray areas to consider, too. For example, I once published a chapter in this book, edited by Matt Gompper. The chapter is in my Google Scholar profile, but the book and its citations are not. Some other folks might lobby that the book should count, too.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Here's a paper that can hopefully catalyze incentives that aren't based on problematic citation metrics
royalsocietypublishing.org/doi/full/10....
0 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To clarify, do you know if the job was filled or if it closed unsuccessfully?
1 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Maybe someone can figure out how to redo the Google Scholar search algorithm to show us the best studies on a particular topic, not the most highly-cited ones.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Ideally, one or two judicial rulings in favor of data sleuths will set a precedent that protects subsequent generations of open science advocates who are committed to ferreting out the malfeasant work.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
It's also worthwhile to acknowledge that there is an important role for institutions to prioritize and support longitudinal work, even (especially?) in settings with challenging logistics. From anthropology, HBE inherited a culture of cross-sectional studies, but cultures and priorities can change.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Often, though, a retrospective study of fertility outcomes ends up omitting important time-varying confounders, which can then result in biased estimates. The chapter additionally suggests that longitudinal studies offer a promising path toward stronger inferences.
1 replies 1 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
First, let's acknowledge the formula that is commonly used in HBE work: 1. This study addresses a particular behavior that humans do. 2. Our findings show that people who do this thing well had relatively high fertility in the past. 3. Therefore, we need to rethink theories of human evolution.
1 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To clarify, the papers we describe are stereotyped HBE publications that don't actually exist (yet), but which appear in various other guises.
The idea for the box arose from Peter Elias's 1958 similarly tongue-in-cheek piece about papers in information theory:
www.oikosjournal.org/sites/oikosj...
2 replies 0 reposts 7 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
As a discipline, anthropology is seemingly due for reorganization and rebranding. It has much to offer, but the message has been bogged down.
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
I like this paper for many reasons, including (a) it’s a compelling alternative to the status quo and (b) organization scholars could potentially study processes of collaborative group formation and divergent interests in academic settings:
www.sciencedirect.com/science/arti...
0 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
AI provides good fodder, though. Thanks! (Original post inadvertently deleted)
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
As someone who is not familiar with the Virtual SRCD, I wonder to what extent graduate students are disproportionately represented among poster presenters (since graduate students would seemingly have less ability to sample hard-to-reach study populations).
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Here's another good paper on the topic:
journals.sagepub.com/doi/full/10....
1 replies 2 reposts 4 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
It is very rare to see a genuinely informative and helpful letter of recommendation.
0 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
In our multinomial methods paper, Richard and I use 96% intervals:
link.springer.com/article/10.1...
Assuming readers know something about probability theory (including the familiar ol' normal distribution), then virtually any interval can be useful for conveying the posterior distribution.
🤷♂️
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
I was asking in part because I was reminded that the same author (Bell) recently addressed the problem of distinguishing age, period, and cohort effects in event history analysis, a problem that may be more likely with terminal events.
www.tandfonline.com/doi/abs/10.1...
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Out of curiosity, are there potentially repeated events within the units? That is, animals die once but they can potentially have multiple reproductive events.
1 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
As an alternative to conventional fixed effect specifications, there are compelling reasons to consider the REWB alternative (a reparameterization of a Mundlak model):
link.springer.com/article/10.1...
Whether or not it remedies the IPP ... I haven't seen papers on this specific point.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To clarify, are you referring to the incidental variables problem as it arises in probit/logit models when using econometric fixed effect dummy variables?
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
The volume came together beautifully.
And in addition to the substantive contributions from the all-star lineup of authors, the bibliography stands to be a unique resource.
Perhaps we should share a version with Kermyt Anderson to mine for the online HBE bibliography:
kermytanderson.com/hbeb/
0 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Here's a paper from the orgs literature that uses multiple membership models:
onlinelibrary.wiley.com/doi/abs/10.1...
The calculation of variance partition coefficients in these models can be tricky, though. I would consult this paper for guidance:
arxiv.org/abs/1907.04148
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
At least in terms of data structure, there are clear parallels between the founding of entrepreneurial teams and the formation of foraging groups.
I'm also glad to advocate for the use of multiple membership models when outcomes are at the level of teams.
www.science.org/doi/full/10....
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Social foraging is low-hanging fruit in terms of topics for human behavioral ecologists to study.
2 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Here's the full lineup of chapters:
0 replies 1 reposts 3 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
We hope that the next textbook will help with this. (Doug Bird and I wrote the chapter on foraging theory.)
www.cambridge.org/core/books/h...
1 replies 0 reposts 2 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To the extent that names and labels matter, some rebranding on the part of human behavioral ecologists might help to effect a sensible transition.
1 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
"As a social science discipline, anthropology stands at a crossroads."
Jones, JH. 2009.
web.stanford.edu/~jhj1/papers...
1 replies 0 reposts 0 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
Understood. And the diverse sampling in cross-cultural research is a beneficial step even though, from an outside vantage, the similarities (in terms of parental education, market integration, family structure) across the sampled populations often seem high from an anthropological perspective.
0 replies 0 reposts 1 likes
|
|
Jeremy Koster
@jeremykoster.bsky.social
[ View ] |
To many anthropologists, for what it’s worth, cross-cultural psychological studies represent valuable progress even as they continue to seem relatively homogeneous, oversampling urban populations and ethnic majorities.
1 replies 0 reposts 2 likes