Tagged replication · show all articles
Peer review, the research laboratory, and open-access publishing were not designed in one stroke — each was assembled piecewise, under specific pressures, and still shows the seams.
Kahneman and Tversky catalogued the ways judgment departs from a normative calculus. Gigerenzer's program asks what calculus the mind actually evolved to meet. Both are live research programs with their own replication scorecards, and neither wins by default.
Evolutionary mismatch explains myopia, obesity and doomscrolling with one sentence, which is why it deserves less trust than it gets. One claim here is close to airtight, one is a live dispute, and one borrows a mechanism's vocabulary without its evidence.
Archival history, bibliometric metascience, and reform trials each study how science's institutions changed—but they answer different questions, at different resolutions, and none can substitute for the others.
A circuit that looks clean, a feature that reconstructs well, an intervention that moves the metric — each is compatible with a story nobody has actually tried to break.
A feature-visualization narrative that looks right and a circuit that has been checked against a known answer are different kinds of evidence. Most published claims sit closer to the first than the coverage of them usually admits.
Every neuroscience measurement is a trace left by an instrument in contact with tissue. The claim about the brain that follows is an inference, and the inference chain is where most of the uncertainty lives.