Wildroot

Queries may use an external AI service. Details

Wildroot Browser

TodayScienceSimilar posts

Similar posts

@mraginsky.bsky.social

The latest from John Ganz hits all the right notes (the vulgar positivism of much of quantitative social science, probabilism as religion, nonfalsifiability of probability statements, Husserl’s diagnosis of the crisis of the sciences brought about by mindless mathematization of the world):

3 posts close in meaning, closest first

  1. @zwlevonian.bsky.social

    I appreciate Julia’s charity here!

    If it’s done right though, I am quite partial to the “causal inference light” approach. It does try to do something coherent; it doesn’t lie about what’s actually of interest. And, at the same time, for those who have been raised on the mantra that you are never allowed to draw causal conclusions unless you randomized something, it is at least not the type of offense that warrants sending you right to jail, right away. If such a study even explicitly discusses the assumptions under which the resulting association can be interpreted causally, and points out specific confounders that may violate such assumptions—maybe future studies should additionally adjust for C?—the result can be a fairly neat package. All things considered, somehow we have collectively fooled ourselves into believing that the association is a genuinely meaningful and interesting inferential target when it doesn’t even exist. But frankly, life is too short to spend your working hours investigating quantities that ar
    Same collectionOpinionScience
  2. @dingdingpeng.the100.ci

    “If evidence from the experiment is free from confounds but statistically weak, and the evidence from the archival study is statistically strong and big but full of confounds, does the set of studies really show both strong and unconfounded evidence?”

    Similar topic and wordingScience