Frequencj^ curves are very often used, however, with numbers rather than percentages represented on the vertical scale, and the vertical scale then shows the actual number in each class. To assist the reader, the total number of observations made would usually be recorded, perhaps in the title of the illustration. In biological work observations are usually made in vast number, to permit making a very accurate conclusion regarding the general laws of frequency for any particular subject under consideration. For a great many problems of everyday life, however, the observations are not

30 32 34 36

Aqe at Marriage

Dala of Amy Hcwes, in Publications Am. Statistical Assn.

Fig. 143. Age at Marriage of 439 Married Graduates of Moxrnt Holyoke College who Graduated from 1890 to 1909

The vertical scale shows the percentage of the whole 439 who married at each age given on the horizontal scale. The totals of all pei'centage figures at the upper margin of the chart is 100 per cent. If a greater number of persons were included in a frequency curve of this sort the curve would be less irregular and the mode would show more distinctly

GEAPHIC METHODS

of sufficient number to permit the formation of any general laws. Thus for Fig. 143 we are not justified in saying that-all college women marry at the particular ages indicated by this chart. The number of women taken into consideration for the preparation of the chart was not sufficient to allow any final conclusion, certainly not to justify any general statement that college women are less likely to marry at the age of twenty-six than at either twenty-five or twenty-seven.