Home / Brinton, Willard C. Graphic Methods for Presenting Facts. New York: The Engineering Magazine Company, 1914. Internet Archive identifier: cu31924032626792 (Cornell University Library copy). The first American textbook on what we now call data visualization. / Passage

Graphic Methods for Presenting Facts

Brinton, Willard C. Graphic Methods for Presenting Facts. New York: The Engineering Magazine Company, 1914. Internet Archive identifier: cu31924032626792 (Cornell University Library copy). The first American textbook on what we now call data visualization. 310 words

Total Loans Made to Employees by a Large Industrial Corporation and Total Amount Paid Back, Shown Monthly Since the Beginning of Loans

The two upper curves are plotted on a cumulative basis. The bottom curve shows the amount outstanding and is equivalent to the vertical distance between points on the two upper curves. The capital devoted to loans is indicated, and the chart permits easy reading of the amount of capital not on loan at the end of any month

''Paid-back," in Fig. 136. Because of the great fluctuation in the amount loaned and paid back each month, the operation of the department as a whole can be seen much more easily from the cumulative curves of Fig. 136 than from the actual monthly-loan curves of Fig. 137. In general, the cumulative curA^e is of very great assistance in showing phenomena in which there are violent fluctuations such as are seen in the loan curve in Fig. 137.

It is interesting to note in Fig. 137 that the peaks and valleys in the ciu've showing the amount paid back follow ordinarily one or two months behind the peaks and valleys in the curve showing the

CUMULATIVE CURVES

amount loaned. This is entirely natural, for these loans were made only in eases of extreme emergency when employees were in temporary need. This relation of two curves to each other, with the peaks of one curve following at some definite distance behind the peaks of another curve, is generally referred to as "lag." Students of economics are continually finding curves which are seen to have a very great dependence upon each other when "lag" is taken into account. If "lag" is not considered, a great difference in the shape of the curves might be taken to show that there was absolutely no relation between the facts from which the curves were plotted.