The failure modes — worth more practical attention than most of the theory.
11 topics · 16 curated works
No prior grounding assumed.
The best stats you've ever seen
Hans Rosling (TED) · 2007
Shows with animated development data that the rich-world / third-world split stopped describing reality decades ago, and that aggregate categories…
+2 more at this level
Assumes you know the vocabulary.
Why Most Published Research Findings Are False
John P. A. Ioannidis · 2005
Shows that under realistic assumptions about power, bias and prior probability, most published positive findings will not replicate.
+9 more at this level
Primary sources and full treatments.
The Interpretation of Interaction in Contingency Tables
Edward H. Simpson · 1951
Shows that an association present in every subgroup of a population can reverse when the subgroups are pooled, and that no purely arithmetical rule…
+2 more at this level
12 of 16 works
Hans Rosling (TED)
Shows with animated development data that the rich-world / third-world split stopped describing reality decades ago, and that aggregate categories hide the within-country variation that matters.
link checked 17 Sept 2026Carl T. Bergstrom & Jevin D. West
A free university course teaching readers to spot the specific statistical and visual tricks - misleading axes, cherry-picked baselines, spurious correlations - that make quantitative claims look more rigorous than they are.
link checked 17 Sept 2026David Spiegelhalter (The Royal Institution)
Argues that uncertainty is a property of our knowledge rather than of the world, and that how a risk is expressed changes the decision people make from identical numbers.
link checked 17 Sept 2026John P. A. Ioannidis
Shows that under realistic assumptions about power, bias and prior probability, most published positive findings will not replicate.
6 pageslink checked 17 Sept 2026Francis Galton
Shows that tall parents' children are on average shorter than their parents and short parents' children taller, naming the statistical pull toward the population mean shown by any imperfectly correlated repeated measurement.
link checked 17 Sept 2026Abraham Wald
Argues that the returning bombers' damage pattern shows where armour is least needed, because the planes hit in the unmarked spots are the ones that never made it back to be counted.
link checked 17 Sept 2026Gerd Gigerenzer & Ulrich Hoffrage
Shows that people who badly neglect base rates when a problem is posed as single-event probabilities reason correctly once the same information is reframed as natural frequencies.
link checked 17 Sept 2026Andrew Gelman & Eric Loken
Argues that flexible, undisclosed analysis choices made after seeing the data can produce a significant result even when no single decision was made in bad faith — the 'garden of forking paths'.
link checked 17 Sept 2026Judea Pearl
Argues the paradox is not a statistical curiosity but a causal one, and that only a causal model of how the data were generated - not the data alone - tells you whether to trust the aggregated or the disaggregated trend.
link checked 17 Sept 2026Ronald L. Wasserstein & Nicole A. Lazar
The American Statistical Association's formal statement that a p-value below 0.05 is not evidence a finding is true, important or free of error, issued because the threshold had become a substitute for judgement rather than an aid to it.
link checked 17 Sept 2026Marcus R. Munafò, Brian A. Nosek, Dorothy V. M. Bishop, Katherine S. Button, Christopher D. Chambers, Nathalie Percie du Sert, Uri Simonsohn, Eric-Jan Wagenmakers, Jennifer J. Ware & John P. A. Ioannidis
Sets out concrete institutional changes — pre-registration, registered reports, incentives tied to replication — needed to fix the practices behind the replication crisis, rather than treating it as a matter of individual researcher virtue.
link checked 17 Sept 2026Anne Marthe van der Bles et al.
Proposes a framework distinguishing the different kinds of uncertainty behind a number - about the estimate, about the evidence, about the science itself - and reviews evidence that stating uncertainty honestly does not, contrary to common fear, destroy public trust.
link checked 17 Sept 2026This subject genuinely sits in more than one domain. These fields approach the same ground with different methods.