Contents

[edit]

 

[edit] Jerzy Neyman

[edit]
16 Apr 1894 - 5 Aug 1981

[edit] Russian

Image:Neyman.gif

Originally named Splawa-Neyman, he dropped the first part of his name at age 30. He studied at Kharkov University and wrote on Lebesgue integration. Sergi Bernstein influenced him, encouraging him to read Pearson's Grammar of Science. In Warsaw he lectured in mathematics and statistics and received a doctorate in 1924. Receiving a fel lowship to work with Pearson in London, he was disappointed to discover that Pearson was ignorant of modern mathematics. In Paris he attended lectures by Lebesgue and Hadamard but his interest in statistics was stimulated again by Pearson's son who sought a general principle from which Gosset's tests could be derived. Neyman went on to produce fundamental results on hypothesis testing. He worked in England from 1934 to 1938 when he emigrated to the USA working in Berkeley for the rest of his life. His wor k on hypothesis testing, confidence intervals and survey sampling revolutionised statistics.

Neyman's thoughts on model building and assessment:

``Whenever we use mathematics in order to study some observational phenomena we must essentially begin by building a mathematical model (deterministic or probabilistic) for these phenomena. Of necessity, the model must simplify matters and certain details must be ignored. The success of the model depending on whether or not the details ignored are really unimportant in the development of the phenomena studied. The solution of the mathematical problem may be correct and yet be in considerable disagreement with the observed data simply because the underlying assumptions made are not warranted. It is usually quite difficult to state with certainty, whether or not a given mathematical model is adequate some observational data are obtained. In order to check the validity of a model, we must deduce a number of consequences of our model and then compare these results with observations."


[edit] Adapted from the MacTutor History of Mathematics archive.