@FitzgeraldJack_ @phl43 So how do you judge power transformations in the grand scheme of things as it pertains to your argument?
English
Luke Macon
3.5K posts
Luke Macon
@LukeMacon
Just a guy. All views something I heard my friend say.
Katılım Mayıs 2012
446 Takip Edilen91 Takipçiler
@LukeMacon @phl43 Log-like transformations approach ln(Z) for large Z but are defined when Z = 0. Power transformations are different, though we have a section in the paper showing that they exhibit similar undesirable properties (for transform Y^p, you get sweet spots in p).
English
New preprint! We re-analyze 46 papers that use log-like specifications (ln(Z+1), inverse hyperbolic sine, etc). We find widespread non-robustness, and we show through theory + simulation how these models drive spurious significance. 1/
doi.org/10.31222/osf.i…
English
@MishaTeplitskiy Also, your cite count should be divided by a co-author count. The fact that this isn’t already factored into the system slightly baffles me
English
Interesting argument for why researchers are so reluctant to do away with journal impact factors: the majority of us look "better" when evaluated by impact factor instead of article-level citations
English
For decades, peer review has been treated as the gold standard of scientific validation.
Yet many scientists know the reality: the system is far from perfect. Peer review is broken and sometimes even corrupted.
The process can be slow, inconsistent, and vulnerable to bias. Reviewers are sometimes asked to judge work outside their true expertise. In other cases, they may be evaluating ideas that challenge the very paradigm in which they were trained. And occasionally, reviewers are simply competitors.
Ironically, the most prestigious journals can also be the most conservative. Truly new ideas are often met with skepticism, while safer work that fits the current narrative moves more easily through the system.
Increasingly, papers are judged less by the originality of the idea and more by the volume of data, the sophistication of statistics, and the beauty of the figures. Science risks becoming data-rich but idea-poor.
But there is an important reality to remember: journals do not ultimately decide the impact of scientific work. Impact is decided later, by the community. By the scientists who read it, test it, debate it, and cite it.
In the end, citations and ideas determine the legacy of a paper, not the impact factor of the journal that first published it.
Science has always advanced by questioning assumptions. Perhaps it is time we also question the system that filters scientific ideas.
English
Another lower bound on likely fraud in biomed literature -- using suspicious copy/pasting in Excel files -- comes in at 3%.
It feels very uncomfortable but I think we all have to update our priors to fraud being quite common
Misha Teplitskiy | Science of Science@MishaTeplitskiy
How much misconduct/fraud is there in the academic literature? About 0.2% of papers get retracted, but that's obviously a severe underestimate. Probably the best estimate comes from a manual (!!!) inspection of 20K (!!!) Western blot images. Estimate is 3.8% (1/2)
English
tenured/permanent academics don't generally deserve higher pay--and some should probably be paid less than they are.
The PhD Place@ThePhDPlace
What unpopular academic opinion would get you in this situation?
English
@paulnovosad Yes and for every one tweet like this there’s two anti-tweets where the reviewers talk about reviewing your paper with AI. Anti because that’s what they do and could tweet but would never dare!
English
I wrote a decent paper with AI.
It took me about 3 hours from start to finish, including an interactive choose-your-own-border-RD-adventure, it’s a what are we even doing here kind of day.
Paul Novosad@paulnovosad
I was writing about land reform in West Bengal last night and was curious if it had persistent effects on the ownership distribution. So I did what anyone would do, I* wrote an academic paper on it Turns out — yes! 1/
English
@alz_zyd_ Yes, it’s somewhat like offering President Trump the opportunity for introspection 😁
English
@maxgoedl No one is really meeting his comments where they’re at. Start with oridinal utilty. Why is he wrong about that point?
English
That is certainly not what I have been taught and it's not what I teach students.
The fact is that you DO NOT NEED to make interpersonal utility comparisons (cardinal utility) for welfare economics. But that does not mean you CANNOT make interpersonal utility comparisons.
UChicago | Stone Center on Inequality & Mobility@UCStoneCenter
We rarely think of economics as scandalous, but maybe we should. Sam Bowles, in conversation with @sndurlauf & @ethanbdm, argues that a core assumption in the field impedes moral reasoning about wealth redistribution. Watch the full panel → bit.ly/3Yj4F3B
English
An example of "lagging prestige" is that econ PhDs are still, for some reason, viewed as much higher prestige than other business school PhDs (finance, mkting, accounting, etc.). However, the econ PhD job market is collapsing, and PhD admissions are getting much harder
English
The reality is, modern economic models are rich enough that you can justify almost any policy with them:
- Tariffs vs free trade
- Wide ranges of progressive income tax
- Price controls
- Free financial markets vs no financial markets
- Monetary policy autonomy vs not
English
@alz_zyd_ Is the project quality cardinal or ordinal? I’m curious if you have “hope” for the potential of several of those 126, or if those projects are truly dead like you won’t come back to them. Do you have another different folder of “alive but dormant” projects?
English
Luke Macon retweetledi
An America that behaves like an utterly self-interested predator on the world stage will not grow stronger. It will grow lonelier.
My take:
English
Video message from Federal Reserve Chair Jerome H. Powell: federalreserve.gov/newsevents/spe…
English
@davidbessis @Rohanburdened I’m sorry, but “no” would have sufficed 🤣
English
Harvard guy never took a class on non-linear dynamical systems.
Steven Pinker@sapinker
They got the same DNA. Genes are the same. Same mother, same father, same older siblings, same younger siblings, same number of books in the house, same number of TVs in the house, same teachers, same everything. So they should be absolutely indistinguishable. They should be, you know, clones. Well, if you know a pair of identical twins, you know they’re not indistinguishable. They have separate personalities. I know this intimately, having two identical twin uncles, Barry and Mark, who are more similar than people picked off the street at random, but no one in the family had any doubt which was which. They each had their own personality, and that’s true of all individuals. Well, that leads to a real puzzle that I think psychologists have not really thought through enough, which is: how can they be different? It’s not their genes. It’s not their environment—at least not anything stable about the environment, like who your mother is or what neighborhood you live in. @thinkableevents
English
@soumitrashukla9 There’s always the argument that the low paying fields are low paying due to positive externalities. Aaaaand that’s probably correct, at least in part.
English
Totally agreed. Majoring in low-paying fields isn’t a human right and I don’t want to subsidize your UBI. Thank you and voting lines are open.
Alex Imas@alexolegimas
I don’t understand this proposal. We already struggle graduating high-skilled workers in STEM and related fields. Taxing successful fields while making college free for less successful fields seems like a recipe for getting less of the former and more of the latter (& further loss of public support for higher ed). We have evidence of income-contingent re-payment from Australia. There, people do reduce labor supply to reduce their repayment, though this response seems to be outweighed by insurance benefits. But this does not take into account the much larger scope of major choice we have in US context.
English
The idea of Natural Capital is simple — when you extract and sell finite natural resources, you get +GDP but –Savings, because the resources are gone.
It's obviously correct, but not incorporated into core models, I guess because we like simple models.
Maybe its time is here.
Steven N. Durlauf@sndurlauf
3. Partha Dasgupta, On Natural Capital: The Value of the World Around Us A career summarizing work by one of the great minds in the economics profession, addressing deeply and comprehensively on how nature and the environments must be integrated into economic models and measurement. (This book is, I believe available in the US in early 2026; I read the final book draft this year.)
English
