When interpreting these data and drawing conclusions, care
should be taken to consider the limitations of the study. Although
our use of archival organizational data is a strength of the research,
an accompanying limitation is that we cannot rule out the possibility
that the differences in performance ratings were based on
real performance differences, not on biased assessments. The idea
that women in line positions are in fact lower performers seems
unlikely, however, given that our data show that in order to
advance, women managers generally had to perform better than
their male counterparts. Moreover, although our data presented a
unique opportunity to learn about ratings and promotion decisions
concerning upper-level men and women managers, they are from
a single organization, and that may limit generalizability of our
results....
Another limitation was the somewhat
short 2-year time period for measuring promotions, suggesting
that our study provided a relatively conservative test of predictions
about promotions.

Finally, more complete information
about promotion candidacies and the types of jobs involved would
have allowed us to better explore effects of gender by job type
interactions in predicting promotion decisions.

There are limitations of the present research that should be noted. First, the use of paper-and-pencil stimulus materials in these three studies, although it allowed us to test the role of prescriptive stereotypes in a controlled manner, limits the degree to which conclusions can be drawn about how people react to successful women in actual work situations. There is no question that participation in the research session lacked the richness and intricacies of involvement in an actual organizational setting and in true-life relationships and that the questions we address in this study need to be explored further in a field setting where work relationships are of consequence and more textured information about coworkers is readily available. Moreover, because our participants were put into the role of potential subordinates, it is important to determine whether there are similar reactions on the part of those who are organizationally senior to the successful female manager. In addition, although many of the undergraduate participants were soon to be entry-level employees themselves and although assessment of the participant pool indicated that 91% of them had worked for more than a 1-year period (with an average of 3.4 years), they still might not have had the organizational experience that would ensure that their responses were representative of people in work settings. Therefore, it is also important to test out these ideas using a sample of working people. Thus, although we have demonstrated that the penalties women incur for being successful in male-dominated areas can, under some circumstances, be mitigated by communality information, specification of when, for whom, and under what conditions this occurs remains to be determined.