On the Methods of Measuring Association Between Two Attributes

I. Introduction.1. IF in any case of statistical observation we classify the objects or individuals observed into two classes only-e.g., peas into yellowseeded and green-seeded, or the members of any group of mankind into male anid female-the resulting data are of the simplest possible statistical form.If, for each object or individual, we note two characters instead of one, dividing again into two classes only, the data become slightly more complex.We have four classes resulting from the two successive divisions-the class all the members of which possess both characters, the class all the members of which possess the first character but not the second, the class all the members of which possess the second character but not the first, and finally the class all the members of which possess neither of the two characters noted.The data resulting from any such count may, if space is no great consideration, be conveniently represented in the form of a small table such as the following (Macdonell, 10, Table II), which shows the recoveries and deaths amongst vaccinated and unvaccinated patients during the small-pox epidemic at Sheffield in 1887-88.There were 4,703 cases, of which 4,I51 were vaccinated TABLE I.-Sheseffld smalt-pox epidemic, 1887-88: cited fronm lacdonell (10).