Page:Graphic methods for presenting facts (1914).djvu/341



GENERAL METHODS

There are a number of comparatively little known short cuts and convenient methods available in the collection and recording of statistical facts. If obsolete or unsuitable methods are used it may make a difference between success and failure in the work of keeping records of any complex business. When the methods of tabulation are too laborious, not only are the records so extensive as to be in disfavor, but they may occasionally include errors, in spite of the greatest care that can be taken by even the highest grade of employees. Anything which will reduce the amount of mental concentration necessary on the part of persons collecting and tabulating facts, will ordinarily assist in the production of more accurate final results. In large statistical studies, such as are made by the United States Census Office, it would be practically impossible to get all the information now obtained if tabulating machinery were not brought to the aid of the human brain and hand.

The punched-card system now widely used in statistical work has made possible an almost unlimited amount of subdivision of analysis with very little extra expense. Fig. 228 shows the card used by the United States Census Office for the 1910 census. One of these cards was punched for each inhabitant in the United States in accordance with the data obtained by the Census enumerators. It will be noticed that the card contains different columns of names or numbers and that there are twelve classifications possible in each vertical column in which a punched hole may be made. Ordinarily the different columns are used for different subjects, and the position of the punched hole in each column records the classification of the data relating to that particular subject.

The punched cards are stacked so that all are right-side up. It will be noticed from Fig. 228 that the lower right-hand corner of the card