Alejandro Schuler
banner
aschuler.bsky.social
Alejandro Schuler
@aschuler.bsky.social
Assistant Professor of Biostatistics UC Berkeley

semiparametric statistics, machine learning, causal inference, stats/ML pedagogy, social justice

Modern Causal Inference Book: alejandroschuler.github.io/mci/
Or does it have as much to do with the shared social context of these objects? If we learned letters in, idk, religious school to read the sacred texts, and numbers in secular school, would we still "get" the question?
December 5, 2025 at 8:08 PM
love this because it illustrates how we often conflate the signifier and signified. Letters=numbers because both are written sigils, but of course that's not what they "are". We all do it: every person immediately "understands" or sees the intuition of this question.
December 4, 2025 at 11:16 PM
but everything was better when I was 12 years old
December 4, 2025 at 6:00 PM
Reposted by Alejandro Schuler
Lesson 4: If you are peddling the "trust in govt" has dropped because of proceduralism, you either don't understand the study of trust or don't care. It has dropped in all countries and all institutions over the same time period. It is mostly driven by perception of politics not govt services.
December 3, 2025 at 1:59 PM
Good people know about both good and evil: bad people do not know about either."

- CS Lewis
December 2, 2025 at 9:54 PM
You understand sleep when you are awake, not when you are sleeping. You can see mistakes in arithmetic when your mind is working properly: while you are making them you cannot see them. You can understand the nature of drunkenness when you are sober, not when you are drunk.
December 2, 2025 at 9:53 PM
"When a man is getting better he understands more and more clearly the evil that is still left in him. When a man is getting worse, he understands his own badness less and less. A moderately bad man knows he is not very good: a thoroughly bad man thinks he is all right. This is common sense, really.
December 2, 2025 at 9:53 PM
Reposted by Alejandro Schuler
If the field did an order of magnitude fewer studies total, and instead pooled resources to do more of these, knowledge production would dramatically accelerate.

We learn more from this one paper than from 50 one-offs examining the predictive validity of implicit measures.
December 2, 2025 at 2:46 PM
Reposted by Alejandro Schuler
We first note that hand-wringing about the decline in US college enrollments has mistakenly linked such declines to the price of four-year colleges.

But the decline is entirely driven by two-year community colleges (and by for-profit colleges). The four-year sector is the dog that didn't bark.
December 1, 2025 at 1:34 PM
I'm dead serious
November 29, 2025 at 7:03 AM
Control is great
November 28, 2025 at 11:55 PM
Reposted by Alejandro Schuler
Don't fine ICE, put ICE agents who did crimes in jail.

Don't fine companies, put CEOs who gave the president obvious bribes in jail.

And so on.
November 28, 2025 at 6:15 PM