Stephen M. Stigler
University of Chicago
- Research field
- History of statistics, statistical theory
Stephen Stigler, the Ernest DeWitt Burton Distinguished Service Professor Emeritus of Statistics at the University of Chicago, is the leading historian of statistics. His 2007 paper traces the idea of choosing the parameter that makes the data most probable through Lambert, Lagrange, Daniel Bernoulli, and Gauss to Fisher's papers of 1912 and 1922, reconstructs from letters how Harold Hotelling's questions of 1930 pushed Fisher to change his argument for the method, and follows the proofs of its properties to the counterexample of 1951 that showed where they stop. The unit presents this history next to the theorems it concerns.
In the masterclass
Section 4, after the theorems on the properties of maximum likelihood estimates, presents his account of how the method was found several times over and how its proofs were made.
The research is presented in Unit 9: Likelihood and Maximum Likelihood, in a research spotlight box that explains the result at the level of the unit and links it to what you have just learned.