Communication Stereotypes and Perceptions of Managers

Previous research has documented that gender and racial stereotypes affect beliefs about communication style. This study sought to investigate whether these stereotypes would be replicated in a sample of White working adults and whether participants thought that a social skills training program that is usually targeted at women would have an impact on managerial targets’ speech. Results indicated that racial stereotypes were more salient than gender stereotypes, with participants viewing White managers’ speech as more socially appropriate and less emotional, but also as more dominant and articulate than Black managers’ speech. Participants also perceived female managers’ speech as more emotional than male managers’. After training, participants thought that men’s and White managers’ speech would become more emotional, despite the fact that this training has been targeted specifically at female managers. Overall, our findings highlight the importance of examining race and gender in evaluating managerial communication.

Stereotypes of speech are important to study because they not only describe how things are but also prescribe how things should be in social interactions (Popp, Donovan, Crawford, Marsh, & Peele, 2003). Stereotypes may actually prevent effective communication between people (Biernat & Vescio, 1983;Hughes & Baldwin, 2002). Importantly, research has indicated that when White women deviate from stereotypes by speaking assertively, they may be judged negatively as a result (Carli, 2001;Crawford, 1995). This phenomenon may be particularly true for women in leadership positions because they already face a trade-off between being perceived as likeable or competent (Eagly & Karau, 2002). It is also likely that this double bind exists for other stigmatized groups in the workplace, such as ethnic minorities.
Previous studies have documented that gender and racial stereotypes (specifically, stereotypes regarding African Americans) include beliefs about communication style (Edelsky, 1976;Hughes & Baldwin, 2002;Kramarae, 1981;Lakoff, 1973;Leonard & Locke, 1993;Ogawa, 1971;Popp et al., 2003;Rich, 1974;Siegler & Siegler, 1976). With only one exception (Popp et al., 2003), however, studies have failed to examine race and gender stereotypes about speech style in the same study. This is problematic because research has found that separate examination of race and gender categories can lead to normative assumptions (Fiske, 1998;Hamilton, 1991). More specifically, this research has found that when participants are not made explicitly aware of a target's gender or race, they default to the categories of White and male.
In Popp et al.'s (2003) study, gender and race stereotypes were jointly investigated by having White college students categorize the speech style of a fictional Black male, Black female, White male, or White female college student. Participants rated targets' typical talk style using 31 bipolar words and phrases (e.g., emotional-unemotional, articulate-inarticulate, soft-loud, active-passive). The results indicated that Black people's speech, in general, was viewed as more direct, less socially appropriate, more emotional, less playful, and tending to be more dominant than White people's speech. Women's speech was believed to be more emotional and slightly less direct than men's speech. Thus, gender and race affected perceptions of speech style. The effects for race were stronger than those for gender. However, the two variables did not interact, suggesting that a target's communication style might be judged according to whether race or gender is more salient in a particular situation. These findings are important not only in the context of Popp et al.'s study but also in other areas where people interact and communicate, such as the workplace.
Researchers have dedicated an increasing amount of study to stigmatized groups, and how stigma is related to status and hierarchy. Meta-analyses conducted by Davison and Burke (2000) and Swim, Borgida, Maruyama, and Myers (1989) have shown that women are stigmatized in the workplace 464978S GOXXX10.1177/2158244012 464978SAGE OpenCarlson andCrawford 2012 1 when they take on stereotypically male-dominated roles such as being a leader. Manifestations of gender-and racial discrimination in the workplace can be subtle but are important to examine because they can have significant consequences. Although women account for about 50% of managerial and professional positions in organizations today, the percentage of women in top managerial positions remains in the single digits (Helfat, Harris, & Wolfson, 2006;Tyson, 2003). Women hold only 15.2% of corporate officer jobs in Fortune 500 companies (Catalyst, 2008) and comprise only 1.8% of CEOs in the Fortune 500 (Helfat et al., 2006). A similar situation exists for Black managers. Current statistics indicate that Black men account for 2.9% of managers and officials, whereas Black women account for 2% of the total number (Nkomo & Cox, 1990). Moreover, women of color account for only 3.2% of Fortune 500 corporate officers (Catalyst, 2008). Scholars have hypothesized that a "double stigma" exists for women of color making it even more difficult for them to obtain leadership positions in the workplace (Botsford Morgan, Gilrane, McCausland, & King, 2011). The goals of this study were twofold. First, we aimed to replicate and extend Popp et al.'s (2003) study of communication stereotypes by using the same design but using a sample of working adults rather than the college students used in the original study, and by changing the targets to White male, White female, Black male, and Black female managers. Our second goal was to examine whether participants drew upon communication stereotypes in assessing the effects of a social skills training program on the communication style and managerial effectiveness of targets.

Theoretical Underpinnings for the Stigma of Female and Minority Leaders
The social stigma of female leaders can be explained by role congruity theory, which is based on social role theory, and the stereotype fit hypothesis. The former posits that the social role of being "feminine" is incongruent with the social role of being a "leader," which occurs because people observe men in management roles and then correlate the characteristics of men with the definition of an effective leader (Eagly & Karau, 2002). In support of this assertion, Eagly and Karau (2002) also found that female managers are seen as effective when characteristics such as cooperation or affiliation are required, a situation in which there is perceived congruence between the two roles. However, Eagly, Makhijani, and Klonsky (1992) found that women who lead in an autocratic, stereotypically masculine way are evaluated negatively. This role incongruity results in three main consequences for women in management: They are evaluated less favorably than comparable men, they have more difficulty becoming leaders than men, and they are not recognized as being as effective as men (Eagly & Karau, 2002). Heilman's (2001) concept of stereotype fit leads to a comparable hypothesis, which can help to explain the manageras-White stereotype. The thinking here is that stereotypes of managers and ethnic minorities might not coincide, or "fit" with one another. In support of this hypothesis, Tomkiwicz, Brenner, and Adeyemi-Bello (1998) found that participants' ratings of Whites and managers were significantly correlated, whereas the relationship between perceptions of Blacks and managers were not. In other words, they found that managers were judged to have attributes more typically associated with Whites than Blacks. Moreover, there is a relatively large body of literature documenting Black-White differentials in performance appraisals. These findings were confirmed in Roth, Huffcut, and Bobko's (2003) meta-analysis, which found that Whites were rated more highly than Blacks on objective performance measures. Similarly, Cox and Nkomo (1986) used the performance appraisal ratings of 125 first-level managers to investigate the extent to which criteria for Black and White managers differed. They found that social-behavioral factors were more strongly correlated with the job performance of Black managers. Finally, Greenhaus, Parasuraman, and Wormley (1990) found that Black managers felt less accepted in their organizations, received lower performance and promotion ratings from supervisors, and had a lower amount of career satisfaction, as compared with White managers.

Gender and Racial Stereotypes of Speech
There is a generally held belief that women and men speak differently (Crawford, 1995(Crawford, , 2001, and these communication styles are very often portrayed as dichotomous. Lakoff (1975) identified a cluster of speech features that she believed constituted a distinctive "women's language" style conveying hesitancy and powerlessness. Tannen (1990) theorized that girls and boys learn different communication styles from early childhood, and characterized their speech as "report talk" (for men) and "rapport talk" (for women). Mulac and his colleagues have conducted an extensive research program aimed at uncovering gendered speech features (Mulac, 2006). Researchers have looked for sex differences in emotional communication (Guerrero, Jones, & Boburka, 2006), instigating and detecting deception (Burgoon, Blair, Buller, & Tilley, 2006), supportive talk (Burleson & Kunkel, 2006), and many other features and types of speech style. It is only recently that the search for sex differences has been put in a larger context of differences and similarities (Dindia & Canary, 2006) and considered from a social constructionist perspective (Crawford & Kaufman, 2006).
Similarly, research indicates that Black people's speech is stereotyped as loud and aggressive (Hughes & Baldwin, 2002;Leonard & Locke, 1993;Ogawa, 1971;Popp et al., 2003). The little empirical research that has been done in this area indicates that White participants typically rate Black people's speech as being more emotional, aggressive, talkative, straightforward, and argumentative-in sum, less socially appropriate (Leonard & Locke, 1993;Niemann, Jennings, Rozelle, Baxter, & Sullivan, 1994;Ogawa, 1971). Leonard and Locke's (1993) research indicates that racial stereotypes of speech may also be dichotomous in nature. When asked to describe the speech of White people and Black people on a 21-item adjective checklist, their participants differentiated between Whites and Blacks on 18 of the 21 adjectives.

Gender, Race, and Managerial Effectiveness Interventions
Solutions to the lack of parity in management would require change at many levels; however, most interventions, have focused on the individual level, and most have focused on women, implicitly on White women. These change efforts assume that women are in need of compensatory training. Examples include assertiveness training programs, special courses designed for women in management, and self-help books that provide tips on getting ahead in corporate America (e.g., Frankel, 2004;Heim & Golant, 1992;Hollands, 2002;White, 1995). A large amount of self-help literature in the 1970s and 1980s was directed at women in an effort to increase their assertiveness-the underlying message being that women should stop being doormats and stand up for themselves (e.g., Bloom, Coburn, & Perlman, 1975).
In recent years, however, there has been a growing backlash which recommends training in the opposite direction, suggesting that women are too assertive or even aggressive. For example, Hollands's (2002) Same Game Different Rules: How to Get Ahead Without Being a Bully Broad, Ice Queen, or "Ms. Understood" offers management strategies to female managers and executives. Her self-improvement strategies include using a softer voice, adopting a submissive body language, smiling when speaking, dressing less "militantly," learning to apologize more, even disparaging oneself and affecting a stammer to appear less threatening to others. These strategies prescribe many of the same behaviors that the assertiveness training of the 1970s and 1980s sought to eliminate. Indeed, the executive Bully Boss training program (GLC website, n.d.-a) can be thought of as "anti-assertiveness" training.
As Ashcraft and Allen (2003) pointed out, organizational communication scholars have focused increasingly on how organizations are gendered, yet ignored the ways in which they are raced. Allen (2007) argued that research has drawn from and circulated "White universalistic paradigms without problematizing or even acknowledging Eurocentric limits . . . [and that] . . . this bias impedes our discipline from effecting social change related to race" (p. 259). Although Hollands's (2002) intervention is particular to women, it can be argued that the same type of self-help strategies might be perceived as improving all nonnormative groups as stereotypes of these groups' communication styles do not coincide with White men's.
There are numerous problems with individualistic approaches to gender and racial inequalities in the workplace. First, they ignore important interactional and social structural level changes that need to be made for women and ethnic minorities to achieve equality in management. Second, they are implicitly White and do not encompass diversity. Finally, they create double bind situations for women and ethnic minorities in management. More specifically, they create a trade-off between perceived competence and perceived likeability. Individual-level interventions are particularly troublesome given that reviews of gender and assertiveness show few gender differences and that these differences are even smaller in African American populations than White (Carli, 2001;Crawford, 1995).
In this study, Hollands's (2002) self-help "Bully Broad" intervention was made gender neutral ("Bully Boss") and used as stimulus material that participants evaluated. The theoretical rationale for using this particular intervention was that it contained self-help advice for female managers in particular. Furthermore, after reviewing the most recent self-help literature geared toward managers, this intervention was one of the more publicized, as evidenced by frequent appearances in the media, and was even the basis for the plot in one episode of the television show Ally McBeal (http://www.thefreelibrary.com/Silicon+Valley+ Author+Jean+Hollands+Appears+on+Oxygen+Network's . . . -a093259822). Finally, this intervention uniquely lent itself to our purposes because it was not only a self-help text but also a program to which companies could pay to send their employees, thus creating a "training" intervention that could be evaluated by participants.
In sum, we know that there are stereotypes about women and men's communication styles; more specifically, that men tend to be dominant and forceful, whereas women tend to be soft and polite. These stereotypes may be weaker than in the past, but they are still there (Popp et al., 2003). We also know that there are stereotypes about African Americans' communication style, for example, being loud and inappropriate, and these stereotypes may be quite strong (Popp et al., 2003). However, we do not know much about how these gender and racial stereotypes interact, how general they are, how they extend to targets other than college students, or how they affect perceptions of competence and effectiveness. This study replicates and extends previous research by (a) examining gender and racial speech stereotypes of managers from the perspective of working adults, (b) jointly studying gender and race, and (c) investigating the perceived effect of a female-targeted social skills training on White male, White female, Black male, and Black female managers' communication style and managerial effectiveness.

Hypotheses
Based on our review of the relevant literature, the following hypotheses were generated: Hypothesis 1: It was expected that pre-training communication style stereotypes of race would replicate those found in Popp et al.'s (2003) study, specifically that Black targets would be rated as more dominant and emotional, and less socially appropriate than White targets. Hypothesis 2: We hypothesized that female managers may be stereotyped differently than in Popp et al.'s (2003) study due to their higher status position, such that a "Bully Broad" stereotype might emerge which would counter traditional stereotypes of women (i.e., female managers may be perceived as more dominant, and less socially appropriate than their male counterparts). Hypothesis 3: Regarding communication effectiveness, it was expected that nonnormative groups' (e.g., Black and female) speech would be perceived as improving more than the normative groups (e.g., White and male) as a result of "Bully Boss" training because they are perceived to need it more. Hypothesis 4: In line with expectations about changes in communication style, it was also predicted that nonnormative groups would be perceived as improving in terms of their managerial effectiveness as a result of the training, whereas normative groups would be viewed as unchanged or negatively affected by the training. Hypothesis 5: Finally, it was expected that, when rating nonnormative groups, participants would be more in favor of company policies that give access to or require "Bully Boss" training.

Method Participants
Participants were working adults who were recruited by students in undergraduate psychology courses as part of an extra credit assignment. Students were instructed to recruit participants who were currently employed full-time, to target as many different fields of work as possible, and to preserve anonymity by having participants seal their completed surveys in unmarked envelopes and send them directly to the first author. Surveys were collected from 157 job incumbents. In all, 67 job titles were represented in the sample. The most frequently occurring job titles were manager (n = 21), administrative assistant (n = 12), teacher (n = 11), accountant (n = 5), and engineer (n = 5). Examples of other titles included consultant, interior designer, social worker, postal employee, firefighter, project analyst, nurse practitioner, financial advisor, lawyer, and sales associate. Participant ages ranged from 20 years to 73 years (M = 40.62, SD = 12.07). Education level ranged from 1 (some high school) to 6 (graduate degree), M = 3.98, SD = 1.24. Workforce tenure ranged from 1 to 55 years (M = 17.88, SD = 12.12). Fifty-five percent of the respondents were women, 84% were White or nonminority, 5% were Black/African American, 3% were Hispanic/Latino/Latina, 7% were Asian/Asian American, and 1% was Native American. As in Popp et al. (2003), participants who did not self-identify as White were removed for methodological and theoretical reasons, yielding a final sample of 133 White participants, including 71 women and 62 men. 1

Procedure
Participants were told that they were taking part in a study about management training. Modeling the procedure from Popp et al.'s (2003) study, participants were randomly assigned to one of four experimental conditions based on the target's race and sex (i.e., White male manager, White female manager, Black male manager, or Black female manager) and were then asked to rate the average or typical target group member's "usual everyday communication style" on 31 bipolar items which ranged from 1 to 7. Some example items are "sensitive-insensitive," "polite-impolite," "strong-weak," "competent-incompetent." For example, in the Black female condition, Please rate the following adjectives on the extent to which they reflect the usual, everyday communication style of the typical Black female manager, boss and executive on a scale from 1 to 7. We are asking for your opinion about this group in general. Please describe the average, or typical, Black female manager's communication style in your ratings.
Each participant was then given a vignette to read regarding "Bully Boss Syndrome" and its associated antibully training, which was explained to be particularly useful with their target's group (see the appendix for vignette; the description was generated using phrasing and wording from Hollands's [2002] book). The last sentence of the vignette varied based on target group. For example, the last line of the vignette in the Black female condition read as follows: This training has been found to be particularly useful with Black female managers, bosses and executives.
Therefore, the remainder of these materials will focus on Black women as participants.
Next, participants were asked to rate what effect they believed this training would have on the typical target group member's behaviors and outcomes on a scale ranging from 1 (very negative effect) to 5 (very positive effect). This scale consisted of 10 items that were developed for this study: "role as a leader," "role as coworker," "being liked," "ability to get ahead (move up to a higher level of management) in a company," "productivity," "effectiveness as a manager," "general job performance," "being taken seriously," "benefit from the training," and "effectiveness in change (as a result of training)." Cronbach's alpha for internal consistency was .89.
Participants were also asked to indicate the extent to which they agreed with four statements regarding organizational policies about the training, on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). Again, these statements were developed for this study and included "This training program would be a waste of a company's Human Resources budget," "There is no reason for companies to provide this training as an employee benefit," "This training would be useless in general (for any manager who took it)," and "Companies should require this training." Cronbach's alpha for internal consistency was .87.
Next, participants were presented with the same 31 bipolar adjective items from the beginning of the survey, in a different order, and were asked to rate the typical target group member's communication style after participating in "Bully Boss" training. Again, responses ranged from 1 to 7. Speech items were combined using the factor structure in Popp et al.'s (2003) study. In that study, the two strongest factors were social appropriateness, consisting of 14 items (e.g., sensitive, polite, pleasant, intelligent, warm, competent, inoffensive, soft, cooperative, does not interrupt), and dominance, consisting of 7 items (e.g., strong, dominant, active, self-assured, decisive). In this study, the alpha estimates of internal consistency for social appropriateness and dominance were .86 and .78, respectively. The other three factors in Popp et al.'s study, which were directness, emotionality, and playfulness, did not have high enough alphas to create subscales in this study; thus, the single items "articulate," "emotional," and "humorous" were used instead. Composite measures of social appropriateness and dominance were formed by taking the mean of the relevant items. Higher scores indicate more socially desirable speech characteristics (i.e., high social appropriateness and low dominance). Means were also used for the three single-item measures, with higher scores indicating more articulate speech, less emotional speech, and more humorous speech.
Finally, participants completed the Ambivalent Sexism Inventory (ASI; Glick & Fiske, 1996), the Modern Racism Scale (MRS; Swim, Aiken, Hall, & Hunter, 1995), and the Motivation to Control Prejudice Reactions Scale (MCPRS; Dunton & Fazio, 1997). The ASI is a 22-item measure that assesses two components of sexism: hostile sexism and benevolent sexism. Some example items include "Most women interpret innocent remarks or acts as being sexist," and "Women seek to gain power by getting control over men." The MRS is a 6-item measure that assesses modern, as opposed to old-fashioned, racist beliefs. An example item is "Over the past few years, the government and news media have shown more respect to Blacks than they deserve." Finally, the MCPRS is a 17-item measure of social desirability in relation to individuals' attempt to control prejudiced reactions. An example item is "I always express my thoughts and feelings, regardless of how controversial they may be." The alpha estimates of internal consistency for these three scales were .80, .73, and .73, respectively.

Managerial Speech Stereotypes
As modern racism, hostile sexism, motivation to reduce prejudice reactions, and age were correlated with dependent measures, they were used as control variables in this analysis. Composite scores were analyzed using a 2 (target sex) × 2 (target race) × 2 (participant sex) multivariate analysis of covariance (MANOVA). The MANCOVA revealed a significant main effect for target sex, F(5, 113) = 2.47, p < .05, and for target race, F(5, 113) = 4.32, p < .01. No other main effects or interactions were found. Given the overall experiment-wise error protection of this procedure, it was followed by five 2 (target sex) × 2 (target race) × 2 (participant sex) univariate analyses of covariance (ANCOVAs) -one for each of the five dependent measures.

Social Appropriateness
In support of Hypothesis 1, there was a significant main effect for target race, F(1, 122) = 5.49, p < .05, η

Emotional
This analysis revealed a significant main effect for target sex, F(1, 122) = 4.16, p < .05, η p 2 = 0.03, and for target race, F(1, 122) = 10.37, p < .01, η p 2 = 0.08, with women and Black targets perceived as more emotional (M = 3.63, SD = 1.24; M = 3.5, SD = 1.17) than men and White targets (M = 3.99, SD = 1.22; M = 4.14, SD = 1.24). Higher scores on this item indicate less emotional speech. These findings support Hypothesis 1, which predicted that Black targets would be perceived as more emotional than White targets, but counter Hypothesis 2, which predicted that women would be seen as less emotional than men.

Humorous
There were no significant main effects or interactions for this dependent measure.

Effects of Communication Training on Managerial Speech Stereotypes
In examining the prediction that training would be differentially effective or beneficial, a 2 (target sex) × 2 (target race) × 2 (participant sex) × 10 (time) repeated-measures MANCOVA was conducted. Dependent measures were the two pre-training composite and three single-item scores (ratings of the target before Bully Boss training), and the two post-training composite and three single-item scores (ratings of the target after Bully Boss training). Because modern racism, benevolent and hostile sexism, motivation to reduce prejudice reactions, and age were correlated with the dependent variables, they were entered as covariates in the analysis. The analysis revealed a significant interaction for time × target sex, F(9, 107) = 2.00, p < .05, and for time × target race, F(9, 107) = 1.94, p < .05. No other main effects or interactions were found.
Given the overall experiment-wise error protection of this procedure, it was followed by five 2 (target sex) × 2 (target race) × 2 (participant sex) repeated-measures univariate ANCOVAs-one for each of the five dependent measures. The analyses revealed a significant interaction effect on the dependent measure "emotional" for time × target sex, F(1, 122) = 3.87, p < .05, η p 2 = 0.03, and time × target race, F(1, 122) = 5.71, p < .05, η p 2 = 0.05. More specifically, male targets were perceived as less emotional than women before training (M = 3.99, SD = 1.22 vs. M = 3.63, SD = 1.24), but were equal with women after training (male M = 3.60, SD = 1.14 vs. female M = 3.66, SD = 1.22), t(131) = 0.31, ns. Furthermore, White targets were perceived as less emotional than Black targets before training (M = 4.14, SD = 1.24 vs. M = 3.50, SD = 1.17) but were equal with Blacks after training (White M = 3.63, SD = 1.27 vs. Black M = 3.63, SD = 1.08), t(131) = −0.02, ns. These findings contradict the predictions made in Hypothesis 3, where our expectation was that nonnormative groups would be perceived as changing more than normative groups. No significant results were found on the other four dependent measures.

Effect of Training on Perceptions of Managerial Effectiveness and Company Policy Opinions
In examining our prediction that training would differentially affect perceptions of managerial outcomes, a univariate ANCOVA was conducted on the composite score for effectiveness. Benevolent sexism was added as a covariate because it was correlated with the dependent measure.
The ANCOVA on perceptions of managerial effectiveness revealed a significant interaction effect for target sex × participant sex, F(1, 123) = 7.48, p < .01, η p 2 = 0.06, where female participants gave higher effectiveness ratings to female targets than to male targets (M = 4.08, SD = .70 vs. M = 3.70, SD = .99), and male participants gave higher ratings to male targets than to female targets (M = 3.82, SD = .53 vs. M = 3.59, SD = .93). In general, participants thought that the training would have a slightly positive effect on managerial effectiveness for all targets, as evidenced by means that are just above the scale's midpoint. No other significant main effects or interactions were found for managerial effectiveness.
In examining our prediction that training would differentially affect participants' opinions of company policies, a univariate ANOVA was conducted on the composite score for policy opinions. No covariates were used in this analysis, as none of the control measures were correlated with the dependent variable. The analysis revealed no significant main effects or interactions. Thus, Hypotheses 4 and 5 were not supported.

Discussion
The goals of this study were twofold. First, we aimed to investigate communication stereotypes about managers, and we included both race and gender as target attributes. Previous research had demonstrated communication stereotypes about race and gender among college student samples assessing student targets (Popp et al., 2003). However, our search of the literature found no research assessing perceptions of the communication style of targets who were managers. Research on the stigma of both gender and race in the workplace has repeatedly shown that negative consequences can occur for women and ethnic minorities in leadership positions because of these stereotypes. Thus, we chose a sample of working adults as the most relevant group in which to assess such stereotypes. Assuming that we would find evidence of communication stereotypes in this population, our second goal was to examine whether participants drew upon these stereotypes in assessing the effects of a social skills training program on the communication style and managerial effectiveness of targets. We discuss these two aspects of our results in turn.

Do Working Adults Hold Race-and Gender-Based Stereotypes of Managers' Communication Style?
Overall, our results show a consistent pattern of stereotypical judgments based on race. Participants viewed the communication style of White managers, female and male, as more socially appropriate (polite, sensitive, cooperative), dominant (decisive, self-assured, strong), articulate, and less emotional, than the communication style of Black managers. There were fewer results based on gender; however, women were perceived as more emotional than men. These results can be compared with Popp et al.'s (2003) research on communication stereotypes using college student participants and targets. Popp et al. found that the communication style of Blacks was seen not only as less appropriate and more emotional but also as more direct. They also found that gender stereotypes were minimal; similarly, our results showed differences for target gender only on the dependent measure of emotionality. Overall, for college student and working adult participants, race seems to be more salient than gender in evaluating others' speech style.

What Are the Perceived Effects of Communication Skills Training on Managers?
Next, we investigated working adults' perceptions of the effects of a social skills training program on communication style and managerial behaviors. Despite the fact that "Bully Broad" training has been targeted specifically at female managers, our participants thought that White and male participants would be most affected by becoming more emotional. Counter to our predictions about nonnormative groups being perceived as needing to be "fixed" more than normative groups, our results indicate that participants thought the normative groups would benefit most from more sensitive communication. One possible explanation for this is that White and male managers are seen as having the most power in the workplace and occupying the highest echelons of organizations, and thus are perceived as having the most potential for "bullying" behavior. Although research has documented negative evaluations of women who deviate from gender stereotypes in their speech, female (and Black) managers were perceived to have more emotional speech than White or male managers prior to training, which is reflective of general race and gender stereotypes in society. It appears that participants viewed the training program as a way to equalize emotionality, bringing White and male managers to the same level and Black and female managers.
Another interesting finding was that participants did not think training would have much of an effect on managerial effectiveness (M = 3.82 on a 5-point scale), although female participants thought it would have more of an effect on female targets, whereas male participants thought it would have more of an effect on male targets. Participants did not have strong feelings regarding whether training should be a part of organizational policy (M = 4.42 on a 7-point scale).

Implications, Limitations, and Future Research
Since the 1970s, women have received conflicting messages about how to communicate effectively. In the 1970s and 1980s, assertiveness training for women was a therapeutic fad (Crawford, 1995). More recently, the trend has been to provide "anti-assertiveness" training, as in the Bully Broad program. The recent backlash on assertiveness may be detrimental to women attempting to climb the corporate ladder because it suggests that, to be successful, they should engage in behaviors that other working adults do not view as particularly helpful, effective, or necessary. Despite the large amount of publicity given to "Bully Broad" training (GLC website, n.d.-b), it seems that our sample of working adults did not perceive overly aggressive female managers as a major workplace problem. Neither male nor female managers were perceived as bullies; the ratings for both groups were moderate. The selling of "Bully Broad" training to corporate human resources departments may depend more on evoking outdated stereotypes of powerful women than on any real need.
Overall, these findings point to the necessity of examining communication, race, and gender in a managerial context. Importantly, our study examined race and gender together. Being a member of a minority racial category and being gendered female are sources of prejudice and discrimination in our society. Because the race and gender categories are not often studied together, little is known about whether the effects of membership in two stigmatized categories are additive, multiplicative, or whether they might even cancel each other out. In the study by Popp et al. (2003), the gender and the race of a target influenced evaluations of the target's communication style, but the effects were separate and noninteracting. In our study, race proved more important than gender with respect to initial stereotypes; however, gender played a role in the evaluation of managerial effectiveness. Although race and gender did not interact in this study, it is still important to investigate both variables simultaneously to better understand their relationship. Although scholars have hypothesized that Black women will experience a double disadvantage in leadership positions based on the dual stigma of race and gender, recent research has shown that this is not so clear. A recent study done by Livingston, Rosette, and Washington (2012) indicated that the combination of two stigmatized categories may interact in unexpected ways. In their experiment, the researchers found that using a dominant leadership style had negative consequences for White women and Black men, but not for Black women. They hypothesized that double stigmatization put Black women in a unique position, rendering them invisible and affording them the opportunity to behave in a more assertive way without penalty (even though penalties may prevent them from becoming leaders in the first place).
One major limitation of this study was the use of an all-White sample. It would be interesting to investigate the effects of this research with a more diverse population that would enable interaction effects to be examined. In addition, it would be important to explore whether the differences found in this study extend to other racial categories. Future research should further investigate beliefs about racial differences in leadership style. The strong negative stereotypes of African American communication style found in earlier research with college students were not completely replicated here. Nevertheless, Black managers were evaluated as less socially appropriate, less articulate, and more emotional than White managers. There is a need for more research on racial stereotypes as expressed in beliefs about communication style. In particular, why are Black managers perceived as communicating less appropriately? In addition, when and how does managerial status override racial stereotypes about communication?
Unlike race, gender was relatively unimportant in affecting judgments about communication effectiveness. This study underscores how much the beliefs about women's communication skills (and by implication, their managerial competence) have changed over the past decades. However, it also suggests that beliefs about African Americans' lack of social appropriateness may still undermine the effectiveness of Black managers.