This article assesses the efficacy of several statistical methods used for detecting male sexism in ratings of social power. Sexist ratings would be indicated through the observation of an ordinal interaction where female powerholders suffer when being rated by male targets. Such an interaction is typically assessed through an ANOVA. However, Bobko (1986) and Strube & Bobko (1989) recommend alternate methods that may be more appropriate. These methods are reviewed and evaluated using data obtained from 381 undergraduate students. Participants provided ratings of their male and female supervisors' power usage. Analysis indicates that the methods for detecting ordinal interactions recommended by Bobko & Strube are more appropriate than the traditional ANOVA. Furthermore, sexist ratings were observed for 3 of 6 power bases.