thing more than self-perception. Do these differences remain significant when key demographic variables are included? (Note: The next chapter introduces contextual variables into the analysis.)
One way to test the results is to determine whether the traits, behaviors, and motivations found to be significantly associated with male and female committee chairs remain important when controlling for other differences such as age, mentorship, or legislative experience. The statistical technique known as logistic regression is a method of analysis that uses other variables to predict a dichotomous variable such as sex. In other words, if we know a chair's score or response on several key variables, does that information improve our chances of guessing a respondent's sex more than would be the case by chance? 96
Knowing a chair's score with regard to integrative motivations, traits and behaviors and controlling for female mentorship, age when first elected and years of legislative service, it is possible to correctly predict a chair's sex almost 79 percent of the time. Given that the makeup of the sample is 56 percent male and 44 percent female, these traits provide substantial predictive power and link integrative leadership with women committee chairs.
Men and women who chair state legislative committees are alike in many respects. The participants in the focus groups emphasized that skills such as managing time, planning, and running effective meetings are not sex specific but rather the mark of a good chair. They saw these traits as a reflection of personal competence more than sex. Nonetheless on matters such as consensus and decision-making approach, differences do emerge.
This chapter provides evidence that the theoretical distinctions between aggregative and integrative leadership have a significant link to gender. This evidence supports the hypothesis of a more integrative style of leadership practiced by women. Thus, the "feminization of leadership" proclaimed by Vera Katz seems to be affirmed by a subtle but important shift in leadership traits, motivation, and behaviors.
Several major findings emerge from this chapter. First, women chairs are far more motivated by people and policy goals than are male committee chairs. Moreover, women talk about their policy motivations in terms of personal commitments and connectedness. Citing her own experiences, a California chairwoman noted, "They [women] care about the bills that they are willing to sponsor, and they think in terms of how [a bill] affects their loved ones."97
Second, as they gain more experience as committee chairs, women also seem to be less motivated by the classic and most aggregative goal of serving constituent interests in policymaking. Women committee chairs also articulated a view of policy detached from particular interests. For example, a Missouri representative distinguished between a woman's perspective on issues and a more traditional conception of women's interests: