1/ Most of the time, when I try to replicate the statistical analysis in a paper, I find errors. - Most are trivial; - Some are not. No one is really immune from the former variety because, - There's too little time; and, - Publish-or-perish. It's a recipe for haste.
-
Show this thread
-
2/ Does this doom research? Of course not! Science is robust to most of these errors because paper's aren't science -- they're just loosely-coupled maps that scholars use to explore the underlying terrain.
1 reply 0 retweets 2 likesShow this thread -
3/ This analogy is useful to me when evaluating claims outside of my domain. If I read an abstract and can't decide if the uncovered effect is surprising, then I know my priors are too weak to guard against errors that wouldn't mislead domain experts. Danger: Here Be Dragons!
1 reply 0 retweets 0 likesShow this thread -
4/ As a corollary, when I see someone touting some hot paper as strong evidence supporting a position they claim is heterodox without relating it to the orthodoxy, I just assume they're full of shit. Implicitly, they claim the paper *is* the complete terrain.
1 reply 0 retweets 1 likeShow this thread -
5/ This is a really round about way of saying: I can't believe I'm still seeing posts about the Sokal Squared bullshit. The perpetrators of that hoax imagine themselves Galileo valiantly exclaiming, "and yet it moves!" But, really, they're the church.
1 reply 0 retweets 3 likesShow this thread
6/ (P.S. Publish and perish does do harm by constraining the set of administratively admissible problems -- overly small ones.)
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.