Statistical concepts such as the t-test can be explained so intuitively that a first grader can understand. In the app's "Experiment" mode you play with things. In "Guess" mode, you only see the balls, and guess the location of the carts. homepageambuehl.s3.us-east-2.amazonaws.com/webapps/tTes...

1/ New review paper out today! "Mental Models of Causal Structure in Economics and Psychology" with MIT cognitive scientist @RahulBhui and econ theorist @HeidiThysen. How do people come to believe that one thing causes another—and what happens when those beliefs are wrong? 🧵

Delighted to announce that Ingvild Almås will join the Department of Economics at the University of Zurich in June 2025 as the 𝐋𝐚𝐫𝐬𝐬𝐨𝐧-𝐑𝐨𝐬𝐞𝐧𝐪𝐮𝐢𝐬𝐭 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐏𝐫𝐨𝐟𝐞𝐬𝐬𝐨𝐫 𝐨𝐟 𝐂𝐡𝐢𝐥𝐝 𝐚𝐧𝐝 𝐘𝐨𝐮𝐭𝐡. Welcome, Ingvild! More information and official press release: www.econ.uzh.ch/en/news/insi...

I am just a simple man, begging you not to include constants in your tables when you include a bunch of fixed effects.

Unsolicited sci writing advice: Writing that something always or never happens (e.g., Ɐ(x)) makes your claims vulnerable because it only takes one counterexample to break them. The most dangerous version of this? "No prior research has addressed this question."

economists' creed

DeepResearch is helpful but you really can't trust it. A review I just had it do claims that "promise tactics" are a well-used influence strategy, but the paper it cites does not contain the word "promise" at all. For a different paper, DeepRes claims the opposite of what's in the paper's abstract

Reviews of Economic Literature has been launched True open access, managed by academics for academics, it follows our mass resignation from JES where the commercial publisher imposed very problematic conditions. Submit your literature reviews in economics! rel.journals.sup.org/index.php/re...

Here's a nice "proof without words": The sum of the squares of several positive values can never be bigger than the square of their sum. This picture helps make sense of how ℓ₁ and ℓ₂ norms regularize and sparsify solutions (resp.). [1/n]

What do people think is the probability the SAVE act will get passed? Can't find anything on prediction markets, but the act seems to be an important (and scary) step to entrench the current admin (by making voting harder disproportionately for more liberal-leaning groups such as married women)

The currently most famous swiss person is probably Roger Federer. But the most important Swiss person to ever have lived is probably the guy below (Euler), besides the Bernoulli dynasty, perhaps. Remarkably, all of them came from the city of Basel.

We know a lot about how fake news spread, what makes propaganda effective, etc. But what to do about it? This is one of the few pieces I've come across that offer solutions, based on what the British tried in Germany before and during WWII. www.youtube.com/watch?v=sgHH...

Harvard completely caved... they're even regurgiating the propaganda

Current US federal admin clearly respects the Pareto ordering: if a policy harms somebody without benefitting anyone, they’ll do it

I think the main point about ridiculous plagiarism debates such as the one currently affecting Mark Carney is to trivialize academia. What non-academics will perceive is that acad. work = copying + rephrasing others and that the "value added" is in the rephrasing. Which nobody would want to pay for.

TIL Velcro is a Swiss invention modeled after those pesky little burdock flower seeds that are so hard to get off your clothing once they cling to it: en.m.wikipedia.org/wiki/Velcro_...

We had an wonderful couple of days hosting a strong group of PhD applicants. I am looking forward to an exciting incoming cohort at the Zurich Graduate School of Economics this fall!

Here's another instance of these "if it sounds too good to be true, it probably is" cases. The guy has 17 papers in 2024 alone. Not so hard if you don't care about whether the results are actually true... (the development econ scandal described here: bsky.app/profile/i4re...)

The longer I'm in this line of work, the more often I see instances of "if it sounds too good to be true, it probably is"

Experimental validation, which links economic preferences to choices in incentivized elicitations, may fail to produce reliable new measures, from Jonathan Chapman, Pietro Ortoleva, Erik Snowberg, Leeat Yariv, and Colin Camerer https://www.nber.org/papers/w33520

Sounds about right…

PhD admits visiting day at @econ.uzh.ch Drinks in the Dozentenfoyer #NoFilter

Trump's 1987 (!) NYT ad. Stunningly similar to what he's doing now (except for the "let's help .. our sick, our homeless" part). [Of course, this is not an endorsement]

LLM agents behave unlike humans in standard experimental finance (bubble) experiments.

The main purpose of the "comment" function in LaTeX is making it psychologically easier to delete things that should be deleted.

It’s so important for people to realize that authoritarianism in America will look like this, not like a police state with martial law and people disappearing into prisons. It will be a dispirited opposition that people are afraid of supporting, bc otherwise they’ll get in trouble.

When you board a Swiss ski lift, a ski instructor might hand you a random child. It's now your responsibility to make sure the child stays safe and doesn't fall to death. You hand the child back at the top. I think this is awesome-but it has to feel pretty weird to non-Swiss, no?

I did a NYT video interview—along with voices from Hungary, Nicaragua, and Russia— about how repression works in authoritarian states. www.nytimes.com/video/opinio...

Thank you ABA. "We have seen attempts at wholesale dismantling.. This is chaotic. It may appeal to a few. But it is wrong. And most Americans recognize it is wrong. It is also contrary to the rule of law." www.americanbar.org/news/abanews...

22 states suing over NIH overhead cut www.chronicle.com/article/22-s...

Trump Claims He Can Overrule Constitution With Executive Order Because Of Little-Known ‘No One Will Stop Me’ Loophole

The implications of this for academia are likely to be catastrophic. If you care about science, pay attention 👇🏻👀

Here's another cool one, from D-TEA Paris. Ann looks at Bob. Bob looks at Mary. Ann is married. Mary is unmarried. Does at least one married person look at an unmarried person? Most people say "can't say." Until you point out that either...

From Fiery Cushman's talk: One of the shapes below is the magic symbol. It's the middle one. Select exactly one of the three words below such that somebody who does not know which is the magic symbol will know it for sure. [People do it well, though it needs quite a high level k]

Here's a trick if matlab is very slow to start, or slow to start a parallel pool or other package: switch off the internet and try again. It might be slow because it's trying to check licenses but fails. When you shut off the internet, it no longer tries that.

I really like the way this paper (out this morning via NBER) conceptualizes the idea of corporate social impact - the social welfare loss from the firm's exit. By that definition, most existing 'social impact' metrics are way off the mark (according to the authors) www.nber.org/papers/w31803

I had zombie observations in Stata--more data points than I should have. Here's why: max(1, missing) does NOT evaluate to missing, but to 1. (I don't even want to know how many wrong results there are out there due to such Stata quirks...)

3yo: "I want to hear this book" Me: "No, it will scare you" 3yo: "No, I won't be scared because of the wolf"

"We are proud to announce that the 2023 Nobel Prize in Physiology or Medicine will be awarded to insurance broker and Findlay, OH resident Darren Weston for his astounding discoveries that vaccines don’t work and will kill you."