The panic in sociology, psychology, nutrition science, and pharmacology has been growing as >70% papers with “p-values” smaller than 0.05 are discovered to be unrepeatable.
One outcome of Simpson’s Paradox is that machines cannot replace statisticians in analysing results. A great deal depends on interpretation, as Marks shows. “Clustering remains largely an art.”
In his fascinating new book The AI Delusion, economics professor Gary Smith reminds us that computers don’t have common sense. He also notes that, as data gets larger and larger, nonsensical coincidences become more probable, not less.
Should we believe them when they tell us that the drinking four cups of coffee daily lowers our risk of death people with spouses live longer? A statistician and a physician team up to explain why not: A subtler manifestation of dishonesty in research is what amounts to statistical cheating. Here is how it works… […]
There may be an additional, more sinister explanation for the ongoing reproducibility crisis, he suggests: A stunning report published in the Annals of Internal Medicine concludes that researchers often make “inappropriate requests” to statisticians. And by “inappropriate,” the authors aren’t referring to accidental requests for incorrect statistical analyses; instead, they’re referring to requests for unscrupulous […]