Jerzy Neyman

"Suppose it is desired to test the efficiency of several treatments intended to destroy certain larvae on a field. The experiments are arranged in the usual way. The treatments compared are applied to particular plots with several replications and then the plots (or smaller parts of them) are inspected and all the surviving larvae are counted. Thus the observations represent the numbers of surviving larvae in several equal areas. It happens frequently that, while there is room for doubt as to whether there is any significant difference between the average number of survivors corresponding to particular treatments, there is no doubt whatever that the variablitity of the observations differs from treatment to treatment."

"The words "routine analyses" are used to denote the analyses performed by laboratories, frequently attached to industrial plants, and distinguished by the following characteristics: (1) All the analyses or measurements of the same kind, for example, are designed to measure the sugar content in beets or to determine the coordinates of a star. (2) The analyses are carried out day after day using the same methods and the same instruments. (3) While all the analyses are of the same kind, the quantity n varies from time to time, where n represents some small number, 2, 3, 4, 5."

"The future mathematical statistician needs early contacts with experimental sciences. He needs them because, at this stage of the development of statistics, the expeimental sciences are sources of theoretical problems. Also, he needs them because in almost any imaginable job which he may get after graduation, he will be called upon to apply his theory to experimental or observational problems."

"The development of modern science is marked by a pronounced tendency toward indeterminism. A somewhat brutal description of this tendency may be states as follows. In relation to some phenomena, instead of trying to establish a (deterministic) functional relationship between a variable y, and some other variables x1, x2, ... , xn, we try to build a (stochastic or probabilistic) model of these phenomena, predicting frequencies with which, in specified conditions, the same variable y will assume all of its possible values."