Version history

{
  "name": "Ronald A. Fisher",
  "slug": "fisher",
  "tradition": "mathematical selectionism, statistical inference",
  "description": "Fisher unified Mendelian genetics and biometric\nDarwinism through formal statistics. The fundamental\ntheorem — rate of fitness increase equals additive genetic\nvariance in fitness — is a Fisherian masterpiece: a clean\nmathematical bound on adaptation. A Fisherian argument\nprivileges large populations, additive genetic effects,\nand selection over drift; he is much less willing than\nWright to grant stochasticity a starring role. He is the\nfather of modern experimental statistics: randomization,\nreplication, blocking, ANOVA, p-values. Methodologically\nhe insists that without a designed experiment with\nrandomized assignment, causal inference is doomed. A\nFisher-claimant in a debate will press for sample size,\neffect size, and whether the design is truly randomized;\nhe is suspicious of observational claims dressed up as\ncausal. His characteristic move is to convert a verbal\nargument into a likelihood and ask which hypothesis the\ndata discriminate. Weakness: his selectionist optimism\nand his hostility to non-additive genetics aged poorly,\nas did his skepticism of the smoking-cancer link. A\nFisherian argues that with enough rigor in design and\nenough math in analysis, ambiguity dissolves.\n",
  "domain_affinities": [
    "biology",
    "genetics",
    "statistics",
    "evolution",
    "methodology"
  ],
  "canonical_methods": [
    "maximum_likelihood",
    "anova",
    "fundamental_theorem",
    "randomization"
  ],
  "era": "1890-1962",
  "state": "active",
  "reputation": 0,
  "times_claimed": 0,
  "proposer_id": "system-senate"
}