STATISTICS is the science that deals with the collection, tabulation, analysis and presentation of numerical data. Any phenom ena that may be classified and counted may be studied by the statistical method, in fact, no other method will yield a comprehensive, quantitative view of many kinds of phenomena. Statistics thus becomes the adjunct of all the sciences and the indispensable agency of gov ernments in acquiring information concerning production, trade, distribution of wealth, unem ployment, population, births, marriages, deaths, divorces, diseases, accidents and the many other matters on which the welfare of the state depends.
The principal functions of statistics are: I. To furnish definite bases for comparison of related ficts dealing with phenomena of the same or of different periods.
This function is well illustrated by the decennial census taken by the Federal gov ernment. Such census gives a comprehensive view of the population in all the various civil divisions of the country and enables comp ari sons to be made of the composition of the population in the several divisions with respect to age, sex, marital condition, nativity, citizen ship, etc. The comparison of the data of the latest census with that of preceding censuses adds greatly to its significance. The facts relative to the increase or decrease of popu lation and the changes in composition that have taken place during the 10-year period are fully as interesting and valuable as the data contained in the census itself.
2. To reveal laws underlying phenomena that could not be determined by individual ob servers.
In dealing with large numbers, by the sta tistical method, the constancy and uniformity of natural and social phenomena become ap parent. In other words the laws underlying the phenomena are revealed. The mortality tables used by insurance companies constitute a practical illustration.
3. To show trends in phenomena that other wise would be imperfectly known.
Changes in the trend of rates of birth, death, marriage, divorce, etc., are shown ade quately only by statistics.
4. To reveal relations of cause and effect that otherwise would remain hidden.
The discovery by statistics of the relative longevity of total abstainers and moderate drinkers is a noteworthy example.
5. To serve as a guide for legislative, so cial and commercial action.
Statistics of dependents, delinquents and defectives are a prerequisite to the formulation of adequate statutes relating to these classes. The financial statistics compiled by the govern ment and by commercial agencies are indis pensable guides to investors, banking and other business houses and to the government itself.
Methods.— In the compilation of statistics three distinct processes are necessary, namely: 1. Collection• of data; 2. Tabulation and chart ing of data; 3. Analysis and summarization of data. The collection of data comprises two separate steps: the preparation of blank sched ules and the filling out of the blanks. The mak ing of the schedule is a very important piece of work as on it depends the success or failure of the statistical undertaking. The principal points to be observed in preparing inquiry blanks are: (a) The schedule must be made to fit in with the general tabulation scheme. The whole study must be very carefully thought' out and a plan of tabulation made that will clearly show the answers to the general ques tions it is desired to ask. The schedule should contain the items needed for tabulation and nothing more. It is a serious mistake to load up an inquiry blank with superfluous or use less questions. (b) The schedule must give due regard to precedents. It is of the highest im portance that the data obtained by the partic ular study should be comparable with those obtained in similar previous studies. If the study is one of a series taken at regular inter vals, changes in standardized schedule items should not be made unless results had shown the necessity therefor. The study may be en larged by adding new items but the contin uity of the series should be preserved. (c) The schedules should lie as clear and simple as possible. The captions or questions should be short and leave no doubt as to their meaning. Where necessary, brief explanations should ac company the items. (d) Except in institutions, questions should not relate to the habits, abil ity or private life of those expected to answer. It is obvious that the one replying will not place himself in an unfavorable light. (e) Sched ules should be printed on sheets or cards of standard size so that no difficulty will be found in obtaining filing cabinets in which they will fit. (f) If the schedules are elaborate and are to be filled out by many different persons, they should be accompanied by a booklet of detailed instructions covering all the questionable points that might arise. If enumerators are sent out from a central office they, should be given full instructions before beginning the work.