I'm excited to announce that my book of selected writings, Coexistence and Other Fighting Words, has been published, and is now available on Amazon https://t.co/EibIzyOhVH or my publisher's website, Political Animal Press. For a taste of it's content, see https://t.co/NpQ7dq5Sfi but please start with "Why I Wrote This Book" https://t.co/XufJpQeh3s.
This is a good slide, nicely summarizing what causal models enable us to do, and how little we need to assume to get them to work. Thanks for posting. @ConjectureInst
The Math of Why
Conjecture Institute Advisor @yudapearlโs structural causal models turned causation into mathematics: draw the causes, write the mechanisms, intervene with do(X), and ask what would have happened.
The math of why, in four steps ๐
@shibl@LeMondeMent If Jewish self-determination requires Arabs' approval, not the other way around, then Arabs are "more equal," and El-Sayed is deceitful, if not racist.
It is far easier to convince a Muslim that El-Sayed is a dangerous racist than it is to convince them that Muhammad was a false prophet. The latter is impossible, while the former is feasible and badly needed. The idea of "properly interpreting the Quran" helps Muslims achieve the former without denouncing their deepest faith.
@GadSaad@AsraNomani@AkaLazarus
@yudapearl@GadSaad@AsraNomani@AskAkbar@DutchRenitent Insisting on "properly interpreted" is the foundation of "Islamism". One can personally believe in Allah, reject jihadism, and subscribe to Western enlightenment values. That is not a "proper interpretation" of Islam but a reinterpretation, a selective rejection, of it.
An honest Senator would have answered: No, Jews have no right to selfdetermination. But El-Sayed represents a new breed of Senators who speak deceit to appear neutral. He says: Yes, if Arabs approve. How Noble of him. Ukraine has a right to exist if Putin approves. How thoughtful!
@f2harrell@yudapearl@dylanarmbruste3@Foreman1David@PatientStormDoc Do you not think a logical inconsistency in your position is a compelling argument to reevaluate it? Or an example where you were unable to explain to a hypothetical trainee what theyโd need to assume for a certain analysis to be correct? Are you debating earnestly?
Today it's Somerville MA -- tomorrow?
Somerville city council votes to divest from Israel
https://t.co/QY32XCdZXG
Two reactions:
1. City council member speaks against ordinance: https://t.co/37ST5PbsDC
2. Somerville - Triumph of the Will: https://t.co/HMBkpyKf4q
I am glad to share @MellingerBarry's substack on "The Genocide Lobby" with which I fully agree, especially the section on Zionophobic Jews -- Jews with no heart and no soul: https://t.co/c9kMWxbpni
@shibl@LeMondeMent If someone declares: (1) "Both side have equal right to dignity, I am neutral" and then adds: (2)
"But one of the sides is more equal than the other", then declaration (1) is "muddying" and (2) is clarifying.
The causal question most frequently asked is whether a treatment inproves or not the outcome Y for unit i (a patient or an agricultural lot), namely whether (Y_i(1) - Y_i(0)) is positive or negative.
However, this notation was proposed by Rubin in 1974 (Neyman only dealt with Randomized trials). So, for Tukey in 1962, the question was "vague" -- he could not even articulate it, let alone answer it.
@amiralex has advised me to clarify my take on John Tukey's quote: "Far better an approximate answer to the right question, which is often vague, than the exact answer to the wrong question, which can always be made precise." I agree with Tukey that an approximate answer to the right question is preferable to an exact answer to the wrong one. What I take issue with is how some people invoke this quote when discussing causal questions. In 1962, when Tukey wrote this, causal questions were indeed "vague"โwhich is no longer the case today, now that we have a mathematical language to articulate such questions and a formal calculus to answer them (or, at least, to determine whether they are answerable given the information at hand).
Tukey's piece of ancient wisdom can be expected from a generation that could not articulate the "right question" and had to resort to "exact answer to the wrong question". Future historians of science should collect these gems of wisdom to demonstrate the post-Kuhn's "Structure of Scientific Revolutions". @DavidDeutschOxf@ConjectureInst@eliasbareinboim@soboleffspaces@amiralex@horganism @mpigliucci @njstang
Thanks for the virtual nomination, but Navi Pillay is right, Hamas may have intended Oct. 7 to be just a barbaric pogrom, not a genocide. The ones who meant genocide were the participants of the 2001 Durban Conference, including Navi Pillay, where NGO forums declared Zionism "a form of racism". That racist declaration had no other intention but genocide.
@dylanarmbruste3@f2harrell@Foreman1David@PatientStormDoc How could a statistician like Tukey emphasize "the need to know what question you wanted or needed to answer" when he had no language to express that question. For him, the question "causal effect" was a metaphysical object, not a target of scientific inquiry.
What causal-phobic statisticians miss is that the solutions to problems like "truncation by death" need not beg for smart tricks; they follow by a universal calculus, from the mathematical representation of the problems themselves.
The usual sleight of hand from the esteemed causality-basher.
I cite him โA popular repair assumes that treatment changes no patientโs survival, and offers equal survival in the two treatment groups as a check on that assumption.โ
What does this claim have to do with causal inference by @yudapearl ?
Projecting statistics own sins? three ga-ha
@shibl@LeMondeMent It's true that I have arrived at my conclusion by many other observations, not merely this one answer he gave. For a summary, see @MissDiagnosis' piece: https://t.co/uYgFrlGDwk
MY LATEST via @fox_opinion America's first Muslim senator exposes a political blind spot - great reporting by @asranomani informs this peice examining El Sayed and his stance on Israel, Islamism, antisemitism and Zionism https://t.co/7O2ZnJv7A1 #FoxNews cc @IWF
MY LATEST via @fox_opinion America's first Muslim senator exposes a political blind spot - great reporting by @asranomani informs this peice examining El Sayed and his stance on Israel, Islamism, antisemitism and Zionism https://t.co/7O2ZnJv7A1 #FoxNews cc @IWF
Dawid's article "Causal Inference without Counterfactuals should never be read without my discussion paper "The logic of counterfactuals in causal inference": https://t.co/mXEveTIQjD
I doubt Dawid would write his paper today, in light of the misinterpretations shown by some experimentalists. @soboleffspaces@f2harrell
@f2harrell's view of scientific inquiry is worse; it's like saying temperature is what we read on an instrument that members of our cult call "thermometer".