shift the irregular curve along and draw only short portions of the curve line at any one stroke of the pencil or pen. Care in shifting the irregular curve will permit drawing a clean, smooth curve line, even though the irregular curve used is quite different in shape from the curve line which is drawn.
An image should appear at this position in the text. To use the entire page scan as a placeholder, edit this page and replace "{{missing image}}" with "{{raw image|Graphic methods for presenting facts (1914).djvu/222}}". Otherwise, if you are able to provide the image then please do so. For guidance, see Wikisource:Image guidelines and Help:Adding images. |
Karl Pearson in "Biometrika"
Fig. 164. Correlation Curve Showing Probable Span of a Daughter from Finger Tip to Finger Tip, for any Given Span of the Mother. 1,370 Cases were Studied in Making up this Chart
Here the dots for the 1,370 observations are not all shown. The dots seen represent the averages of all
cases studied in each 1-inch range of mother's span according to the horizontal scale. The curve line
is then drawn through these dots. Instead of using the curve line, the mathematical relation may be
expressed by the formula D = 34.18 + .473 M
This chart would have been easier to interpret clearly if both the vertical and the horizontal scales began with the same figure, say 51 inches. The reader is not apt to notice that the vertical scale here begins at 58 while the horizontal scale begins at 51
Curves like those seen in Fig. 163 are properly drawn as smooth curves because they are based upon some definite laws of nature. It is only because of the crudity of the observations of mankind that the dots are so widely scattered. If human knowledge were sufficient to obtain measurements with exactness, curves for data relating to the laws of nature would fall exactly on points plotted according to the observations and there would be no difficulty in getting smooth curves. The curves in this book are nearly all plotted from statistical data for which there are ordinarily some hundreds of variables, of which many do not follow any definite laws of nature. For most statistical work it is much better to join the points showing the observations by straight lines without any attempt to draw smooth