Researched
Statistics & Least Squares
Least squares, the normal curve and population statistics (1805-1835) let scientists extract reliable signals from noisy data.
Open in the interactive tree →Legendre published the method of least squares in 1805, and Gauss, who had used it to recover the lost asteroid Ceres in 1801, gave its theory and the normal error curve in 1809. Laplace proved the central limit theorem, and Adolphe Quetelet (1835) applied averages and distributions to people and societies. William Farr's use of registered deaths from 1839 made public health a numbers science, and Fisher's significance tests in the 1920s completed the toolkit.
Prerequisites
- Probability Theory1654
- Calculus (Analysis)1684
- Bayesian Inference1763Laplace and Gauss grounded error theory and least squares in Bayesian inverse probability
Unlocks
- Epidemiology1854Snow's cholera analysis used the registered death statistics built up by William Farr
- Forensics & Fingerprints1892
- Markov Chains1906
- National Accounts & GDP1934
- Smoking causes lung cancer1950
- Genome-wide association studies2005
- Replication Crisisopen