a procedure used to test the hypothesis of association or relationship between two variables. The observed frequencies of a variable are compared with the frequencies that would be expected if the null hypothesis of no association (i.e., statistical independence) were true. The chi-square test is often used for this purpose.