Manoel Galdino
@mgaldino.bsky.social
Polisci professor at USP. Former executive director at Transparência Brasil. Father of Serena, Martín and Nina. Causality, Bayes, antirracism, black is beautiful, R, Corinthians. Mostly PT, sometimes EN.
I have a dataset where I used SDiD and would love to test the new estimator. If you need someone to test a beta version in an applied setting, let me know that I enlist myself to do it.
November 26, 2025 at 2:45 PM
I have a dataset where I used SDiD and would love to test the new estimator. If you need someone to test a beta version in an applied setting, let me know that I enlist myself to do it.
O ponto importante não é que o PT não de envolve em corrupção. Se envolve e muito, como os outros. Mas a diferença tem sido de que não impede a PF de trabalhar. Pelo menos em relação ao Bolsonarismo.
November 26, 2025 at 12:58 PM
O ponto importante não é que o PT não de envolve em corrupção. Se envolve e muito, como os outros. Mas a diferença tem sido de que não impede a PF de trabalhar. Pelo menos em relação ao Bolsonarismo.
I'd argue the problem is with an I'll defined estimand. As an estimate for the ATE, the effect is indeed small. If the estimand of interest is the CATE (ATE conditional on some measure of risk of heart attack), then the estimate is bigger than .03 and the effect size is big.
November 24, 2025 at 1:25 PM
I'd argue the problem is with an I'll defined estimand. As an estimate for the ATE, the effect is indeed small. If the estimand of interest is the CATE (ATE conditional on some measure of risk of heart attack), then the estimate is bigger than .03 and the effect size is big.
Alterno entre aqui e o substack.
October 29, 2025 at 3:07 PM
Alterno entre aqui e o substack.