An introduction to statistics and biometrics

Abstract

In their present meaning the words "statistics, "statistician" and "statistical" are barely a century old. They have, however, been in use longer than that and it is interesting to trace briefly the origin of their present meaning. They are all derived more or less indirectly from the Latin status, in the sense acquired in mediaeval Latin, of a political state. During the 18th century, the word "statistics" meant simply the exposition of the noteworthy characteristics of a state, the mode of expression being almost preponderantly verbal; however with the commencement of the 19th century the growth of official data became more and more numerical, and the word thus insensibly acquired a narrower significance, viz the exposition of the characteristics of a State by numerical methods.

Once the above change of meaning was accomplished, further changes followed; from the name of a science (sensu lato), the word was transferred to those series of figures on which it operated, so that one now spoke of vital statistics, shipping statistics and so on. It was then applied to similar numerical data which occurred in other sciences, such as anthropology and meteorology.

The word statistics has thus come to have two meanings: when used as a plural noun it has the meaning given in the last paragraph, namely a collection of numerical data, whilst in the singular it refers to the methods of the science of statistics - statistical method.

The theory of statistics is of comparative recent growth. Its roots may be traced to the work of Laplace, and Gauss on the theory of errors of observation, but the study itself did not really begin to flourish until the last quarter of the 19th century under the influence of Galton and Karl Pearson (1857-1936) who developed the ideas of regression and correlation using problems from genetics. (Pearson also studied extensively the effects of errors of sampling.) In the early part of this century Gosset developed methods for interpreting data obtained by sampling, and these were later popularised in England by R. A. Fisher (1890-1962) and his colleagues who greatly extended the theory and fields of application, especially in agriculture, biology, and genetics.

Biometrics is the field resulting from the application of the concepts and methods of mathematical statistics to biological problems. Not only must the biometrician be able to interpret the results he obtains, but he must also understand the assumptions that lie behind the mathematics employed, for these determine the applicability of the theory to the situation. Where no theory fits your particular case, a mathematical model must be specially constructed, and this is a task you should hand over to a professional biometrician.