Race as a moderator in a model of sexual harassment: An empirical test.

L. F. Fitzgerald, C. L. Hulin. and F. Drasgow (1995) proposed that victim characteristics, such as race, might moderate the relationships between sexual harassment and its job, psychological, and health status outcomes. This study describes 2 theoretical positions, tokenism and double jeopardy, that could account for this possible moderation by race, as well as the alternative view that no moderating effects exist. The effects of race are empirically examined through simultaneous path analysis. Results indicate that whereas, mean levels of harassment differ across race, the phenomenon of sexual harassment unfolds similarly across races; race is not a moderator of the relationships between sexual harassment and the variables proposed as its antecedents and outcomes. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Tokenism also works with groups as tokens: The impact of group openness and group qualification on reactions to membership in a low-status group.

Two experiments tested whether the tokenism effect would also emerge when groups as a whole rather than individual group members are denied access into a high-status group. In total, 317 pupils were led to believe that their class was not accepted into a high-status group. The openness of this high-status group toward other groups (open, minimally open, and closed) and the qualification of the own class (low, high) were manipulated. In both studies, pupils of a highly qualified class confronted with a closed high-status group differed from all other conditions by preferring nonnormative action to normative action. These results indicate that tokenism is a very robust phenomenon that occurs even when groups strongly expected to join the high-status group. (PsycINFO Database Record (c) 2010 APA, all rights reserved)