Page:Graphic methods for presenting facts (1914).djvu/120

 compared with the normal are expressed in percentages, making the increases above normal in figures above 100 per cent and decreases below normal in figures below 100 per cent. Unfortunately, the reader not familiar with index numbers may not realize that a chart relating to index numbers should be read from the 100 per cent line rather than from the base of the chart. It is very common to find a chart relating to index numbers so drawn that the chart does not extend to the zero of the vertical scale. Such a chart may give a false impression of much more violent fluctuation than would be interpreted from a chart plotted on the usual co-ordinate field and showing the zero of the vertical scale.

United States Government Crop Reporter

Fig. 92. Fluctuation in the Price of Eggs in the United States as Compared with the Average of the Monthly Figures for the Preceding Four Years

An untrained reader may not realize that this chart must be read from the line representing 100 per cent. Charts for index numbers, where the fluctuation is compared with 100 per cent, should have the 100 per cent line made broad, and a wavy line should be used at the bottom of the chart unless the zero of the scale is shown. An alternative arrangement is shown in Fig. 93

Fig. 92 is taken from the United States Government Crop Reporter, a magazine widely distributed to farmers. It has been the hope of this magazine to give producers of agricultural products an opportunity to study the price records of previous years, so that they may, in so far as possible, sell at the time of the year when prices are the highest. It is much to be doubted whether the average reader of charts like that seen in Fig. 92 would realize that the 100 per cent line must be used as a basis for interpreting the chart. The 100 per cent line has not been made any heavier than the other lines on the sheet. A man who had been at all accustomed to reading charts having zero at the base of the chart would be apt to read Fig. 92 as though the bottom line were the zero from which the curve had been drawn. On such a basis, he might think that the price of eggs in January, 1912, was more than eight times the price in July, 1911. Such a conclusion would, of course, be entirely unwarranted by the actual figures.