On the Criterion that a Given System of Deviations from the Probable in the Case of a Correlated System of Variables is such that it Can Reasonably Be Supposed to have Arisen from Random Sampling." In Philosophical Magazine, Series 5, Volume 50, No. 302 (
Pearson, Karl.

A very good and attractive copy, with no library markings. The complete Volume 50, in half morocco with gilt-decorated spine. Octavo. (Item ID: 10929)


This paper introduces the Ki-Square test of goodness of fit, "one of the most useful of all statistical tests" and "one of Pearson's greatest single contributions to statistical methodology" (D.S.B.). "In 1900 Pearson attacked the problem of curve fitting. Having fitted the best available curve to a series of data he asked what the probability that a sample from a population truly represented by his curve should fit it as badly as, or worse than, the sample in question But the question arises `what is a bad fit?' Pearson solved this problem by the invention of the function of observations called Ki-Square, which increases as the fit becomes worse. This has turned out to be an immensely powerful tool, and is used on a huge scale it is used as a test of agreement with hypothesis wherever the hypothesis is tested by counting individuals He obtained a solution of a problem which was of such generality that it had entirely unexpected applications" (J.B.S. Haldane, in Pearson and Kendall, Studies in the History of Statistics and Probability, Volume I, pp. 433).

Site by Bibliopolis