Download Gender and Leadership Expectations: A Quantitative Study on Femininity and Masculinity and more Study notes Communication in PDF only on Docsity!
Running head: GENDER DIFFERENCES IN LEADERSHIP EXPECTATIONS
Does gender matter?
A quantitative study on women’s and men’s expectations of femininity and masculinity for female leaders, male leaders, and leaders in general. Sifra Mol ANR 879527-SNR 2005162 Master’s Thesis Department Communication and Information Sciences Specialization Business Communication and Digital Media Faculty of Humanities Tilburg University, Tilburg Supervisor: A.P.C.I. Hong, MA Second reader: Dr. P.J. van der Wijst July 2018
Table of Contents
- Abstract
- Does gender matter?
- Gender and Leadership: social roles, role congruity and lack of fit
- Social Role Theory
- Role Congruity Theory...........................................................................................................
- Lack of fit model
- Gender and Leadership: three paradigms
- Think-Manager, Think-Male Paradigm
- Agency-Communion paradigm
- Masculinity-Femininity Paradigm........................................................................................
- Gender and Leadership: influence of perceivers’ gender.........................................................
- Method
- Study Design
- Participants
- Materials
- Procedure
- Measures...............................................................................................................................
- Results
- Manipulation check
- Congruencies in expectations for leaders
- Expectations for leaders in general
- Expectations for female and male leaders
- The effect of perceivers’ gender on leadership expectations
- Discussion
- Findings
- Theoretical implications
- Practical implications
- Limitations and future research
- References
- Appendices
- Appendix A – Consent form and introduction
- Appendix B – Cases
- Appendix C – BSRI scale, manipulation check
- Appendix D – General questions and debriefing
- Appendix E – Assumption of normality
Does gender matter? As reported by the European Union in the Report of Equality Between Men and Women, the proportion of women on boards of large companies in 2017 in the Netherlands was 29.5% (European Union, 2018). Comparing this to the proportion in 2010, which was 14.9%, an increase in the number of female leaders is evident. Nevertheless, the proportion of male leaders (70.5%) is still significantly larger than that of female leaders. This indicates that there is still an unequal division of women and men in leadership positions. A phenomenon that explains the unequal division is the glass ceiling effect (Federal Glass Ceiling Commission, 1995). The glass ceiling effect can be explained as an invisible barrier that keeps women from rising beyond a certain level in hierarchy (i.e., to leadership positions), as the barrier causes negative prejudices that exist regardless of women’s qualities and past achievements (Federal Glass Ceiling Commission, 1995; Kay & Hagan, 1995; Eagly & Carl, 2007 ). These prejudices depend on expectations of certain characteristics for women, men and leaders. Two theories that elaborate on the expectations for women, men, and leaders are social role theory (Eagly, 1987) and role congruity theory (Eagly & Karau, 2002), which serve as the base for the present study. Social role theory is based on the concept that people’s behavior is determined by their social roles. People can have multiple social roles that hold different expectations and prescriptions (Eagly, 1987). For example, a woman can be a teacher, a mom, and a friend at the same time; but her behavior will be different in each of those social roles. In addition to social roles, the theory also considers gender roles, which concern the descriptive and injunctive norms that are associated with women and men specifically. These norms elicit the expectation that certain social roles must be occupied by a specific sex (Eagly, 1987). A traditional example would be that women are expected to occupy the role of housewives (i.e., raising children and cleaning the house), while men are
expected to occupy the role of providers for the family (i.e., being breadwinners) (Eagly, Wood & Diekman, 2000; Cotter, Hermsen, Oviada & Vanneman). Eagly and Karau’s (2002) role congruity theory of prejudice toward female leaders extends the social role theory by considering the congruity between gender roles and other roles, such as leadership roles. The theory states that role congruity occurs when a woman or man occupies an expected social role, and that role in congruity occurs when one occupies a role she or he is not expected to occupy. Considering leadership, research showed that characteristics expected from men (e.g., dominant, assertive, confident) are similar to those expected from leaders. In contrast, characteristics expected from women (e.g., sensitive, gentle, warm) are not associated with those of leaders (Schein, 1973; Eagly & Karau, 2002; Koenig, Eagly, Mitchell & Ristikari, 2011). This suggests that leaders are, to a greater extent, expected to be male rather than female. However, abovementioned studies and theories focus on expectations for leaders in general. This implies that leadership expectations depend on perceivers’ interpretation on whether the gender of the leader is female or male. As characteristics attributed to women are different from those of men, the question arises whether female leaders and male leaders are also expected to have different characteristics. Consequently, it would be interesting to investigate whether expectations of femininity and masculinity for leaders in general are different from those for female and male leaders in specific. Furthermore, since women and men are expected to have different sets of characteristics, it is arguable that these characteristics lead to different views towards leadership. Since women are generally more kind, caring, and gentle, they may have different expectations of femininity and masculinity than men, who are generally more independent, assertive, and confident (Schein, 1973; Eagly & Karau, 2002; Koenig, Eagly, Mitchell & Ristikari, 2011). Research already suggests that men generally have a more masculine
according to society (Eagly, 1987; Heilman, 200 1). Descriptive norms are linked to stereotypes. A stereotype is a vision (positive or negative) of how someone is expected to behave (Cambridge University Press, nd) and make the behavior of people that occupy a certain social role predictable. For example, a nurse is expected to be caring and a military officer is expected to be willing to take risks (Eagly, 1987). In addition to social roles, social role theory also includes gender roles, which consist of beliefs and expectations for how women and men should behave. These gender-specific expectations can be defined as communal and agentic (Eagly, 1987). Communal characteristics (e.g., gentle, helpful, kind) are ascribed to women and are considered to be feminine. Agentic characteristics (e.g., independent, dominant, assertive) are ascribed to men and are considered to be masculine (Eagly, 1987; Eagly & Karau, 2002). Role Congruity Theory Role congruity theory extends social role theory by considering the extent to which there is congruity between one’s social role and her or his gender role (Eagly & Karau, 2002). Considering gender and leadership, one might state that there is role congruity between leader roles and the male gender role, as characteristics of both leaders and men are expected to be agentic rather than communal (i.e., feminine) (Schein, 1973; 1975; Eagly & Karau, 2002; Koenig et al., 2011). Thus, there is role congruity when men occupy leadership positions, but role incongruity when women occupy leadership positions. Role congruity theory focuses specifically on the prejudices towards female leaders. Eagly & Karau (2002) argue that the incongruity between the female gender role and typical leader role causes prejudices against female leaders in two ways. The first prejudice arises from descriptive norms and holds that women are expected to be less effective leaders than men, which results from the incongruity between characteristics of women and those of leaders (Franke, Crown & Spake, 1997; Eagly & Karau, 2002; Koenig et al., 2011). The
second prejudice stems from injunctive norms and contains that women are evaluated less favorably than men when occupying a leadership position (Franke, Crown & Spake, 1997; Eagly & Karau, 2002). The latter prejudice may results from women’s adoption of agentic characteristics when occupying leadership positions. More specifically, when female leaders are likely to adopt agentic characteristics to measure up to the expectations people generally have for leaders. For example, women may be less sensitive and more dominant when occupying leadership positions. This contradicts expectations belonging to their typical female gender role. Consequently, women are not behaving as expected, which may lead to less positive evaluations (Eagly & Karau, 2002). This notion is supported by Rudman et al. (2012). Their study investigated prejudices towards female leaders and found that female leaders are evaluated less positive than male leaders. Lack of fit model While social role theory and role congruity theory focus on social roles and gender roles in general (Eagly, 1983; Eagly & Karau, 2002), lack of fit model by Heilman (1993; 1995; 2001) focuses specifically on gender roles in workplaces. Lack of fit model is based on expectations people have about how successful an individual will be in a certain job or position. These expectations are decisive in employment situations such as whether a woman or a man will be employed. A perceived lack of fit is described as the mismatch between the perceived characteristics of an individual and the required skills for a specific position (Heilman, 2001). A perceived lack of fit leads to expectations of failure, while a perceived fit leads to expectations of success. As research has provided strong evidence that characteristics expected for women are not similar to those expected for men and leaders (Eagly & Karau, 2002; Koenig et al., 2011), a lack of fit will be perceived when women occupy leadership positions (Heilman, 1993; 1995; 2001).
indicates that expectations of masculinity for leaders have become less pronounced in comparison to 1973 (Koenig et al., 2011). Agency-Communion paradigm The agency-communion paradigm examined feminine and masculine characteristics of leadership roles with information retrieved from 22 articles (Koenig et al., 2011). Participants rated leaders on the Bem Sex-Role Inventory Scale (BSRI) (Bem, 1974). The BSRI-scale consists of 30 communal (i.e., feminine) characteristics (e.g., gentle, warm, tender), 30 agentic (i.e., masculine) characteristics (e.g., independent, assertive, aggressive) and 30 neutral characteristics (e.g., sincere, conventional, tactful). This scale will also be used for the present study in short form, consisting of 10 feminine, 10 masculine, and 10 neutral characteristics. Results of the agency-communion paradigm showed that leaders are expected to behave in an agentic way rather than in a communal way. This result confirms the think- manager, think-male paradigm, as well as other studies that claim that leaders are expected to act in a more agentic than communal way (Eagly & Karau, 2002; Heilman, 2001; Berkery, 2013 ). Masculinity-Femininity Paradigm In contrast to the other two paradigms, the current paradigm tested expectations of femininity and masculinity for multiple specific leadership roles (e.g., school principal, PR director, politician) instead of one general leadership role. In the 7 used articles, participants rated the leadership roles as either feminine or masculine on a bipolar scale. Mean ratings showed that most types of leaders were rated as masculine, which is in line with the think- manager, think-male and the agency-communion paradigm (Koenig et al., 2011). To conclude, the study by Koenig et al. (2011) demonstrates that most articles in the period from 1973 to 2007 confirm the notion that characteristics of leaders are seen as similar to the characteristics of men. Furthermore, the study showed that the masculine construal of
leadership has decreased over time. This may indicate that the overall leadership expectation has become less masculine and more feminine. However, the data for the meta-analysis was gathered in the period from 1973 to 2007, which makes the information slightly outdated. It may well be that the masculine construal of leadership has increased even more since 2007. This provides opportunities for new studies. The present study will contribute to existing literature by providing a quantitative analysis of the differences between women’s and men’s expectations for female leaders, male leaders, and leaders in general. As role congruity theory (Eagly & Karau, 2002), lack of fit model (Heilman, 1983; 2001) and the study by Koenig et al. (2011) demonstrate that people have a masculine construal of leadership; the first step of the present study is to test whether these findings are still accurate the present-day. Thus, the first hypothesis of the present study predicts that leaders are expected to have more masculine than feminine characteristics. H1: Leaders are expected to have more masculine than feminine characteristics Research showed that the characteristics that are expected for men are similar to those that are attributed to leaders (Schein, 1973; Eagly & Karau, 2002; Koenig et al., 2011, Berkery et al., 2013). However, it remains unclear whether expectations for women (e.g., being kind, gentle and warm) are also applicable for female leaders, or whether female leaders are expected to have the same characteristics as leaders in general (e.g., being dominant and independent). As there appears to be role incongruity between the female gender role and the leader role, Hypothesis 2a predicts that female leaders, similar to the female gender role, are expected to have more feminine characteristics than masculine characteristics. Furthermore, since research has shown that the male gender role is similar to the leader role (Schein, 1973; Eagly & Karau, 2002; Koenig et al., 2011, Berkery et al., 2013), Hypothesis 2b predicts that male leaders are expected to have more masculine characteristics than feminine
Johnson et al. (2008) found in their study ‘ The strong, sensitive type’ that feminine individuals expect an effective leader to be sensitive (i.e., feminine), while masculine individuals expect leaders to be strong (i.e., masculine) (Johnson, Murphy, Zewdie, & Reichard, 2008). Furthermore, feminine individuals perceived leaders as more successful when this leader behaved in a more communal and sensitive manner, while masculine individuals perceived leaders as successful when being agentic and masculine. The study also showed that feminine individuals expected female leaders to be strong and sensitive and male leaders to only be strong. Masculine individuals, on the other hand, expected all types of leaders to be strong, since sensitivity was not related to leadership (Johnson et al., 2008). From the study by Johnson et al. (2008) it may be indicated that women expect female leaders to adopt both feminine and masculine characteristics, while male leaders are expected to only adopt masculine characteristics. According to men, however, all leaders are expected to adopt masculine characteristics instead of both masculine and feminine ones. Given the finding by Koenig et al. (2011) that men’s masculine construal of leadership has decreased over time, the present study investigates whether men still expect leaders to have more masculine characteristics than women expect leaders to have. H4a: Men have a more masculine construal of leadership than women Although the study by Johnson et al. (2008) indicates that feminine and masculine individuals have different leadership expectations for female and male leaders, it remains whether women’s and men’s expectations of feminine and masculine characteristics are significantly different for female and male leaders. Therefore, the present study investigates whether perceivers’ gender influences expectations of femininity and masculinity for leaders from both sexes. Since men have a more masculine construal of leadership than women, it is expected that men expect both female and male leaders to be more masculine than women
expect them to be. Thus, Hypothesis 4b predicts that men expect female leaders to be more masculine than women expect them to be, and Hypothesis 4c predicts that men expect male leaders to be more masculine than women expect them to be. H4b: Men expect female leaders to be more masculine than women expect them to be H4c: Men expect male leaders to be more masculine than women expect them to be Figure 1. Conceptual model Method Study Design In order to test the hypotheses of the present study, an online experiment was conducted. The used design was a 2 (gender participant: female vs. male) x 2 (gender leaders: female vs. male) mixed-subjects design. The two independent variables were gender of participants and gender of leaders, both between-subjects variables. The gender of leaders was manipulated, which resulted in two conditions (female leader, male leader) and a control condition (leaders with no sex apparent). The dependent variables were leadership expectations of femininity and masculinity, measured as within-subject variables. For this study, a quantitative, self-administered online questionnaire was conducted which was comprised of an existing validating scale (i.e., BSRI short form scale, see Appendix C). Methods of convenience and network sampling were used to gather participants, whereby Gender Leader Female Male Leadership Expectation Femininity Masculinity Gender Perceiver Female Male
Procedure To investigate the leadership expectations of the participants, an online experiment in the form of a questionnaire was conducted using Qualtrics. A link to the experiment was distributed to participants via Facebook, Whatsapp, LinkedIn and social networking sites of Tilburg University. Upon opening the link, participants were shown a short and general introduction, which explained the aim of the study and provided the contact information of the author. Furthermore, a consent form was provided that explained that all data was anonymous and that participation in the study was voluntary. When not willing to sign the form, participants were thanked for their time and the experiment was concluded. When participants agreed to participate in the study, they were directed to the next page that briefly explained the procedure of the experiment (see Appendix A). Thereafter, participants were presented with the manipulation of the study in the form of the fictive case. After reading the case, participants rated 27 characteristics that measured their expectations of femininity and masculinity for the leader they were exposed to. The items were randomly shown to participants. The Dutch translations of the items that were used in the experiment can be found in Appendix C. After measuring participants’ expectations of femininity and masculinity, a manipulation check was conducted by asking participants what gender the director in their specific case had (Appendix B). As gender of the leader was the manipulated variable, given answers might not be reliable when the gender was identified incorrectly. After the manipulation check, several questions were asked to explore the demographics of the sample (e.g., gender, age, leadership experience, and level of education). Lastly, participants were thanked for their participation. Participants were able to enter their e-mail addresses if they wanted to receive a summary of the study once it was finished.
Measures Leadership expectation. Leadership expectations were measured using the BSRI short-form scale as derived from Bem (1981). The BSRI short-form scale consists of 10 feminine characteristics (e.g., warm, affectionate, gentle), 10 masculine characteristics (e.g., aggressive, dominant, assertive), and 10 neutral characteristics (e.g., reliable, sincere, adaptable). The neutral items were measured as a part of a PhD project but will not be included in the present study. To make the scales more compatible for the present study, three items were excluded. ‘Loving children’ was excluded from the feminine scale and ‘jealousy’ from the neutral scale, since those items are not directly related to leadership. Furthermore, the item ‘has leadership abilities’ was excluded from the masculinity scale, since all items aim to measure leadership expectations in the present study, which makes this specific item unnecessary. The remaining 27 items were measured on a five-point Likert scale (e.g., ‘ I expect a leader to be adaptable ’, 1 = completely agree, 5 = completely disagree). The questions belonging to the short-form BSRI-scale can be found in Appendix C. All items and illustrating sentences were translated from English to Dutch to fit the Dutch target group of the present study. Reliability tests showed that the femininity scale had a good reliability α .8 1 (M = 3.53, SD = 0.51) and that the masculinity scale had an acceptable reliability of α .6 5 (M = 3.69, SD = 0.41). The reliability scores per scale per condition can be found in Table 1. Table 1 Reliability scores for the femininity, masculinity, and neutral scale per condition Condition Scale Cronbach’s A M SD Female leader Femininity scale .79 3.56 0. Masculinity scale .62 3.76 0. Male leader Femininity scale .79 3.51 0. Masculinity scale .62 3.66 0. Neutral leader Femininity scale .83 3.53 0. Masculinity scale .68 3.66 0. 42
Table 2 Means and standard deviations per condition Overall Women Men Condition Outcome M SD M SD M SD Leader in general Femininity 3.52 0.03 3.61 0.04 3.43 0. Masculinity 3.70 0.03 3.64 0.04 3.75 0. Female leader Femininity 3.56 0.06 3.59 0.08 3.53 0. Masculinity 3.76 0.05 3.71 0.07 3.81 0. Male leader Femininity 3.50 0.05 3.64 0.06 3.35 0. Masculinity 3.66 0.05 3.61 0.06 3.72 0. Expectations for leaders in general The first hypothesis predicted that leaders in general are expected to have more masculine than feminine characteristics. Results of the ANOVA revealed that there was a significant difference between the scores for feminine characteristics ( M = 3.52, SD = 0.03) and masculine characteristics ( M = 3.70, SD = 0.03), F (1, 242) = 19.25, p < .001, 𝑛!^!^ = .07. This supports Hypothesis 1. Expectations for female and male leaders To test whether expectations of femininity and masculinity are different for female and male leaders, it was tested what effect the gender of the leader had on the dependent variables. The ANOVA revealed no significant interaction effect of the gender of the leader on expectations of femininity and masculinity, F (2, 242) = 0.0 8 , p = .924. This indicates that there is no significant difference between expectations of femininity for female ( M = 3.56, SD = 0.06) and male ( M = 3.50, SD = 0.05) leaders and expectations of masculinity for female ( M = 3.76, SD = 0.05) and male ( M = 3.66, SD = 0.05) leaders. This finding is in contrast with Hypotheses 2a, 2b, 3a and 3b. The effect of perceivers’ gender on leadership expectations The ANOVA revealed a significant main effect of gender on leadership expectations F (1, 242) = 12.43, p = .001, 𝑛!^!^ = .05. A repeated contrast was performed to compare
expectations of femininity that women have for leaders to those that men have, and to compare expectations of masculinity that women have for leaders to those that men have. Results of the contrasts showed that women expect leaders to be more feminine ( M = 3.61, SD = 0.04) than men expect leaders to be ( M = 3.43, SD = 0.05), F (1, 242) = 7.1 9 , p = .008, 𝑛!^!^ = .03. In turn, men expect leaders to be more masculine ( M = 3.75, SD = 0.04) than women expect them to be ( M = 3.64, SD = 0.04), F (1, 242) = 4.3 3 , p =. 039 , 𝑛!^!^ = .02. These results are in line with Hypothesis 4a, which stated that men have a more masculine construal of leadership than women. Furthermore, results also showed that men expect leaders to have more masculine than feminine characteristics, F (1, 242) = 28.4 1 , p < .001, 𝑛!^!^ = .11. Results were similar for women, although not significant, F (1, 242) = 0.41, p = .521. Lastly, the ANOVA revealed no significant interaction effect of gender on the expectations of femininity and masculinity for different sex leaders, F (2, 242) = 0.8 0 , p = .450. For female leaders, there is no significant difference between women’s expectations of femininity ( M = 3.59, SD = 0.08) and masculinity ( M = 3.71, SD = 0.07) and men’s expectations of femininity ( M = 3.53, SD = 0.08) and masculinity ( M = 3.81, SD = 0.06), which is not in line Hypothesis 4b. Furthermore, for male leaders there also appears to be no significant difference between women’s expectations of femininity ( M = 3.64, SD = 0.06) and masculinity ( M = 3.61, SD = 0.06) and men’s expectations of femininity ( M = 3.35, SD = 0.09) and masculinity ( M = 3.72, SD = 0.07), which does not support Hypothesis 4c. Discussion The goal of the present study was to examine the influence of perceivers’ gender on leadership expectations for female leaders, for male leaders, and for leaders in general. Based on previous findings (Schein, 1973; 1975, Johnson et al., 2008; Koenig et al., 2011) and relevant theories such as social role theory, role congruity theory, and lack of fit model
