INTEGRATING SEVERAL STREAMS OF LEADERSHIP RESEARCH
The emergence of leadership research in the early 1900s was based on the idea that certain traits predisposed an individual I to emerge as a leader (Bass, 1990). Up until published reviews by Mann (1959) and Stogdill (1948), trait theories of leadership dominated the literature. However, conclusions drawn from early reviews set a new direction for leadership research focusing on leadership style or behaviors, which lasted the next 40 years. Indeed, the shift away from personality research and leadership was unfortunate, given that Kenny and Zaccaro (1983) later reported that 48% to 82% of the variance in leadership emergence was accounted for by the leader's attributes. Ironically, Stogdill (1974) reported that personality attributes such as surgency, emotional stability, conscientiousness, and agreeableness were all positively related to leadership effectiveness (Hogan, Curphy, & Hogan, 1994). Yet it was Stogdill's review that led to a near abandonment of research on traits.
Lord, DeVader, and Alliger (1986) argued that many previous authors misinterpreted Mann and Stogdill's reviews, and raised several concerns. First, both Mann and Stogdill examined the relationship between leadership emergence and personality, limiting the scope of research reviewed. Second, although there were a number of consistent significant relationships between leadership traits and emergence, Mann commented on the lack of relationships. Lord et al. conducted a meta-analysis of the literature reviewed by Mann, reporting that many of the relationships between personality and leadership emergence had been underestimated. They concluded that traits (e.g., intelligence, masculinity-femininity, and dominance) were associated with leadership perceptions to a much greater extent than had been reported.
Howard (2001) discussed the importance of examining not only which traits predicted leadership success and effectiveness but also those traits that predicted failures. Hogan, Raskin, and Fazzini (1990) focused on what caused leadership failures after finding that 60% to 70% of employees reported that their worst or most stressful aspect of their job
was their supervisor. Earlier work by McCall and Lombardo (1983) and by Hellervik, Hazucha, and Schneider (1992) reported that managers who failed exhibited a number of personality flaws including being overly controlling, irritable and exploitative. These results parallel findings reported by Kaplan, Drath, and Kofodimos (1991). Many bright, hard-working managers fail because they are arrogant, abrasive, selfish, and lacking what Goleman (1998) called emotional intelligence.
Judge, Bono, Hies, and Werner (2000) completed a metaanalysis of 94 studies examining the relationship among the Big Five personality traits, leadership emergence, effectiveness, and transformational leadership. Judge et al. reported a multiple R of .47 with the Big Five traits predicting leadership effectiveness. Extraversion was most consistently and positively correlated with leadership emergence and effectiveness. Extraversion, conscientiousness, and openness to experience were consistently correlated with leadership effectiveness. Judge and Bono (2000) examined the relationship between transformational leadership and the Big Five personality factors. Results based on over 200 organizations indicated that agreeableness was the strongest predictor of transformational leadership, followed by extraversion. Openness to experience was also positively related to transformational leadership; however, this relationship disappeared when effects of other personality traits were statistically controlled. Overall, the multiple-/? value between the Big Five personality traits and transformational leadership was .40.
The Big Five personality traits have also been linked as antecedents to leadership emergence in autonomous teams (Taggar, Hackett, & Saha, 1999). Earlier research on antecedents of leadership found that the leader's interests, energy, verbal fluency, confidence, and independence were each predictors of leadership success (reported in Bass, 1990).
In sum, one of the main conclusions from this literature is that personality does indeed matter and, like attributes of the context, must be taken into consideration when predicting leadership emergence and effectiveness. The accumulated research in this area indicates that there are certain attributes one might want to take into consideration when making selection decisions. Some of these attributes may also prove to be quite effective in predicting whether a more or less successful candidate will succeed at current leadership in an organization (McCauley, 2001).
Male and Female Leaders
Over the last decade, many attributes associated with effective management have been associated with women (Helgesen, 1990; Rosener, 1995). Changes in organizational structure
and a greater emphasis on inclusion have led to calls for managers to be more collaborative, cooperative, participative, em-pathetic, nurturing, and developmentally oriented. These qualities have been traditionally associated with "female advantages" (Rosener, 1995), as well as with transformational leadership (Maher, 1997). The conventional wisdom suggests that men and women differ in terms of leadership styles and behaviors. The literature on sex role types indicates that men tend to be seen as more task oriented, whereas women are viewed as more relationship oriented. Men have been shown to be less comfortable working for a female leader, while also viewing her success as being due more to luck than to capability (Forsyth & Forsyth, 1984). As noted later, however, these effects may disappear when organizational context factors are controlled (Eagly & Johannesen-Schmidt, 2000). Hollander and Neider (1978) reported that female respondents tended to cite more incidents of bad leadership being associated with male leaders, while citing an equivalent number ^)f good incidents for male and female leaders. Eagly, Karau, and Makhijani (1995) concluded from their meta-analysis of the literature that women who exhibited a more masculine style were perceived as less effective than were women who used a feminine style. Women using a feminine style were also seen as less effective than men, who exhibited a masculine style.
Men and women can lead equally effectively (Eagly et al., 1995; Powell, 1993), but may differ in terms of how they lead (Adler, 1996; Eagly & Johnson, 1990; Parker & Ogilvie, 1996). However, Kanter (1977) argued that individual differences in terms of personality were probably more important than gender in determining how male and female leaders performed in managerial roles.
Many books written for the general public contend that women's leadership styles are different than men's (Helgesen, 1990;Loden, 1985; Rosener, 1995). Rosener labeled the style of women as being interactive and that of men as being command and control oriented. Nonetheless, much of the research on this topic has not reported reliable male-female differences (Bartol & Martin, 1986; Eagly & Johannesen-Schmidt, 2000; Eagly, Karu, Miner, & Johnson, 1994). Eagly et al. (1994) reported that men generally scored higher than women on motivation to manage others in a more hierarchical manner. Eagly and Johannesen-Schmidt (2000) conducted a metaanalysis of literature comparing male to female leadership styles. Most differences were relatively small, but there was a tendency for women to be more interpersonally oriented, less autocratic, and more participative.
Comparisons of male and female leaders on transformational leadership behavior have also produced small but significant differences. Bass, Avolio, and Atwater (1996)
Integrating Several Streams of Leadership Research 283
reported that women were rated more transformational than were their male counterparts. Ross and Offermann (1997) reported that military cadets' ratings of their commanding officers' transformational leadership were positively correlated with being seen as more nurturing and feminine and were negatively correlated with attributes such as aggressiveness and masculinity. Hackman, Furniss. Hills, and Peterson (1992) reported a positive relationship between ratings of transformational leadership and communal qualities assessed by Bern's (1974) sex role inventory. Eagly and Johannesen-Schmidt (2000) examined the normative database (N = 9,000 raters) for the Multifactor Leadership Questionnaire (MLQ) Form 5X (see Avolio, Bass, & Jung, 1999), reporting that female leaders were rated higher on two aspects of transformational leadership: attributed charisma and individualized consideration. Male leaders were rated higher on all aspects of negative, or less effective, leadership.
Eagly and Johannesen-Schmidt (2000) suggested that some gender-stereotypic role differences may disappear as the specific characteristics of those organizational roles are controlled. For example, Moskowitz, Suh, and Desaulniers (1994) reported that agentic behavior traditionally associated with men was more controlled by the status of the interacting partners (e.g., any boss with a follower), whereas communal behaviors were more controlled by gender regardless of organizational roles.
In sum, literature comparing male and female leaders has generally reported relatively few differences in terms of leadership style. Future research needs to take a closer look at how the organizational and cultural contexts affect these results, especially at more senior levels. A more precise comparison of men and women leaders needs to take into account other variables that may be correlated with gender differences including level, tenure, and experience (Yukl, 1998).