Chapter 9 introduced a test for the difference between two proportions. The test had the advantage that it was a straightforward extension of the one-sample test of a proportion, and used the by-now-familiar normal distribution. However, the test has limitations. Most importantly, it cannot be extended to more than two samples. It is ne if we have two assembly lines and want to know if they differ in the proportion of defectives they turn out. But what if we have three assembly lines? It is ne if we want to know whether the proportion of Natural Science majors differs by sex. But what if we want to know whether the whole distribution of majors differs by sex?