ABSTRACT
Nonparametric tests are used when the corresponding parametric procedure is inappropriate. Normally, this is because the dependent variable is not interval- or ratio-scaled. It can also be because the dependent variable is not normally distributed. If the data of interest are frequency counts, nonparametric statistics may also be appropriate. The chi-square goodness of fit test determines whether or not sample proportions match the theoretical values. A significant chi-square test indicates that the data vary from the expected values. A test that is not significant indicates that the data are consistent with the expected values. The chi-square test of independence tests whether or not two variables are independent of each other. The chi-square test of independence is a component of the Crosstabs command rather than the Nonparametric Statistics command.
