lukestein’s Twitter Archive—№ 15,494

When only the “wrong” regression coefficients are significant but alas, @ben_golub/1636186677562707969