ABSTRACT
The first chapter has shown that sets of data are usually composed of individual values which vary from one another to a greater or lesser extent. When, however, it is necessary to describe the quantitative characteristics of such a population or sample data set briefly and succinctly, i.e. to employ descriptive statistics, a lengthy recital of all the individual values is not of much use. Even the graphical representation of these, as illustrated in Chapter 1, is not a great help, for it neither allows of a speedy and easy comparison between different sets of data nor of a ready expression of these characteristics in words or numbers. It is therefore often very useful to be able to summarize these varying values within the one set of data by one value alone. This one value is chosen so as to give as reasonable an approximation as possible to what is ‘normal’ or, in other words, to summarize the data by some ‘measure of central tendency’. It is immediately apparent that however this measure is chosen it must involve certain generalizations and must also obscure many characteristics of the set of data that the distribution curve shows.
