Analysis
In this study, the sample consists of the wait staff at randomly selected restaurants. At each restaurant either the man or the woman assumed the in-charge role. Thus, experimental procedures resulted in two separated samples of subjects -- wait staff serving an "in-charge" man and wait staff serving an "in-charge" woman. Given that a couple was dining together at each restaurant, these are independent groups; one subject could not be assigned to both the "in-charge man" and the "in-charge woman" experimental groups. Thus, the statistic selected must be one for independent groups.

The scale of measurement of the dependent variable in this study is nominal (check presented to man or woman), thus the appropriate statistic would be the chi-square test of independence. This statistic allows us to determine whether the frequency of check presentation to the male or female diner varied by whether the male or female was assuming the in-charge role.

Our last step in selecting the appropriate statistic for analysis is to determine whether the assumptions of the statistical procedure have been met.