A Formula for Drawing two Correlated Curves so as to Make the Resemblance as Close as Possible

Abstract

One of the problems with which statisticians are confronted almost daily is that of drawing two curves in such a way that the correspond? ence, so far as the eye can judge, shall be as close as possible. The statistician, as a rule simply experiments with his scales, varying the two units until the curves fall into some sort of correspondence, aban? doning his graphical illustration if the appearance of the two curves is not such as to warrant a positive conclusion. As is shown below, it is possible to derive a formula which will, without imposing too much labor upon the statistician, make this resemblance as close as could be desired. If, for instance, we are comparing yearly marriage rates with yearly prices of wheat, the two groups will be brought into closest cor? respondence if, after plotting the marriage rate, we multiply each