Cultural influence on gender-bias in the allocation of tasks

1

Cultural influence on gender-bias in the allocation of tasks

Master Thesis

Student: Randi Ong-A-Swie Student Number: 10695842 Topic: Gender inequality

Track: Msc. in Business Administration - Leadership & Management Institution: Amsterdam Business School, University of Amsterdam Supervisor: Tanja Hentschel

2 Statement of originality

This document is written by Student Randi Ong-A-Swie who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its

references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

3 Abstract

Inequalities in the workplace between women and men are still present in today’s society. Unequal treatment in the past can have a negative impact on women’s future careers. Therefore, it is of importance to know the source of this inequality. The current study proposed that inequality is caused by disparity in the tasks to which men and women are allocated. Hence, it is of importance to know how these task allocation decisions are made. It is stated that these decisions are influenced by stereotypes. Therefore, it is hypothesized that men are more likely allocated to agentic tasks and women to communal tasks. Besides, as countries are becoming more multicultural and stereotypes differ in their held degree across cultures, its role in these decisions is of rising importance. This study states that these stereotyped allocations are more likely to be made in masculine cultures than feministic cultures, which can be explained by managers having higher evaluations of competence, well-being and comfortableness for men and women on the stereotypical tasks. To test the hypotheses, an online experiment was conducted, with a total sample of 52 Dutch and 101 German participants. Several analyses have been conducted which did not result in any significant findings. This indicates that men and women are not more likely to be allocated to their stereotyped tasks and that managerial task allocation decisions and evaluations of masculine and feminine countries are not distinct. However, it has been found that both cultures hold stereotypical views regarding female and male characteristics.

4 Table of contents

1. Introduction………....6

2. Literature review……….……...9

2.1 Gender in the allocation of tasks………...9

2.2 Evaluations………..…13 2.2.1 Competence………...………..13 2.2.2 Well-being………...…14 2.2.3 Comfortableness………..16 2.3 Culture………17 3. Conceptual model……….…19 4. Methodology………...…..20 4.1 Research design………...20 4.2 The sample………...21 4.3 Procedure……….23

4.4 Stimuli and manipulation………...…..24

4.5 Measures………..……25

4.5.1 The tasks………..25

4.5.2 Evaluations………...25

4.5.3 Communality and agency……….27

5. Analysis and results………..29

5.1 Correlations………..…29 5.2 Hypotheses testing………...…32 5.2.1 Testing Hypothesis 1………....32 5.2.2 Testing Hypothesis 2………...….33 5.2.3 Testing Hypothesis 3………34 5.3 Exploratory analysis………36 6. Discussions………38 6.1 Summary……….38 6.2 Theoretical implications……….….39

6.3 Strengths, limitations and directions for future research……….41

6.4 Practical implications………..44 6.5 Conclusion……….….46 7. Acknowledgements………...…46 8. References……….47 9. Appendix………...55 8.1 Appendix 1 – Distribution………...55

8.2 Appendix 2 – Dutch Experiment……….58

5 Index of tables and figures

Figure 1. Conceptual model………..20

Figure 2. Stereotypical allocation decisions………...35

Figure 3. Average rates communality and agency………....38

Table 1A. Means and standard deviations sample………22

Table 1B. Gender………..…23

Table 1C. Education………..23

Table 1D. Occupation………...23

Table 2. Cronbach’s alpha, means and standard deviations competence……….…26

Table 3. Cronbach’s alpha, means and standard deviations physical well-being………….…27

Table 4. Cronbach’s alpha, means and standard deviations communality and agency……....28

Table 5. Means, standard deviations and correlations Germany………..29

Table 6. Means, standard deviations and correlations The Netherlands………...31

Table 7. Means and standard deviations evaluations………....33

Table 8. Results T-test hypothesis 2……….33

Table 9. Allocation decisions in percentages and results Chi-square test ………...34

Table 10. Means and standard deviations evaluations………..35

Table 11. Results T-test hypotheses 3B and 3C………35

Table 12. Means and Standard deviations………37

6 1..Introduction

For decades women have been considered inferior to men. The discrepancy between equal treatment for women and men goes all the way back to the time before Christ (Allen, 1975). Plato for example stated that women are the weaker character. Men who were not able to control their emotions were considered as weak and were therefore punished by being reincarnated as a woman (Goldberg, 1968; Allen, 1975). This perception of inferiority is not only held by men, but by women themselves as well (Goldberg, 1968). This has caused women to be put at a disadvantage throughout history: they lacked legal rights such as voting (Sklar, 2000), were highly dependent on their husband and were not allowed to own property (Ikechukwu & Kanu, 2012), have been kept from education (Vasquez, 1982), and were thought to be less intelligent than men (Powell, 2003). Although some of these social

inequalities have vanished over the years, they have cleared the path and made way for other inequalities to emerge, namely at the workplace.

Various studies have found that women are being treated unfairly and are discriminated with regard to recruitment and hiring decisions (Rice & Barth, 2016), performance evaluations (Friend, Kalin, & Giles, 1979), promotions (Heilman, 2001), development and supervision (Rosen & Jerdee, 1974; Acker & Van Houten, 1974). When they overcome this inequality in recruitment and hiring and eventually do get hired for a certain position, they have to deal with the gender wage gap. This gap is the unequal pay of men and women for the same job (Petersen & Morgan, 1995). Moreover, this inequality continuous in their career mobility. Women’s careers used to be considered to be limited by the so-called ‘glass ceiling’, which are the invisible barriers that completely limit their upward mobility in organizations (Hymowitz & Schellhardt, 1986). However, this ceiling states that women are limited in attaining high-level positions, but have equal access to low and middle-level positions, which is not the case (Eagly & Carli, 2007). A labyrinth is therefore

considered a concept more fit todescribe women’s career mobility. This metaphor describes the complex journeys women must complete before attaining top positions, unlike men who walk easier paths (Kark & Eagly, 2010). Although women face inequalities, barriers, and complexities in their careers, their participation in the labor force is rising. Since 2006, the participation rate of women in labor in Europe has increased by 2.40%, whilst the

participation rate of men decreased with 1.70% (based on the total participation of European labor force) (International Labor Organization, 2016).

7

Despite an increase of women in the labor force, females are still underrepresented in male stereotyped occupations and fields which are associated with power and status (Heilman, 2012). Merely a quarter of the top positions is held by women and this rate is barely growing (Catalyst, 2017). However, this increase in labor force presence did not come with a

corresponding expansion in the range of their work activities and roles (Heilman, 1983). A possible explanation might be that women experience less rapid and limited career

progression than men (Lopez-Claros & Zahidi, 2005). Although every individual differs in their career attainment, there is a notable difference between women and men (De Pater, Vianen & Bechtoldt, 2010). When reaching for the top, there are several factors which are essential (Oakley, 2000). One of the most important factors is job experience, more specific experience in challenging jobs (Berlew and Hall, 1966; Woodall, Edwards & Welchman, 1997). It has been found that women are assigned less to challenging tasks, which causes them to have gained less experience and are therefore less prepared than men for handling top jobs (Kirchmeyer, 2006; Ohlott, Rudermand & McCauley, 1994). A plausible explanation for this might be managers’ biased decisions. As found by Waldman (1984), the type of tasks managers assign employees to, depends on the employee’s perceived abilities. However, managers’ personnel decisions can be influenced by stereotypes, which therefore may influences their view of an employee’s abilities, resulting in stereotyped task allocation decisions (Rosen & Jerdee, 1974; Wellington, Kropf & Gerkovich, 2003).

Various studies have researched discrimination of women in the employment process but did not extensively research how they are allocated to tasks. As an individual’s previous experience influences his/her future career (Martell, Lane & Emrich, 1996), it is of

importance to know the underlying motives of task allocation decisions and the role of gender (Berlew and Hall, 1966; Woodall et al., 1997). Thereby, the scarce research that has been conducted on this topic mainly analyzed the difference in the level of challenge the tasks hold. Researchers found that women were assigned less to challenging tasks than men, which negatively influences their advancement pace (Van Velsor and Hughes, 1990; Ohlott et al., 1994). However, these studies did not research the reasons for managers’ allocation decisions. A study that did research this, is that of De Pater et al. (2010). They found that allocation decisions were made on basis of the managers’ perception of the employee’s ambition and job performance. The perception on these characteristics led to a difference in the tasks men and women were allocated to, resulting in women having less challenging task experience. However, by looking at these two characteristics, solely the current state of the employee is taken into account, whilst the employee’s future state is ignored. As job performance and

8

ambition are dependent on the characteristics of a task (Fried & Ferris, 1987), they are not a reliable indicator of the employee’s state in another task. Therefore, the current study analyzes how the employee’s anticipated future state influences allocation decisions.

Another critique on previous gender inequality studies is that they solely focus on material outcomes like earnings and promotions (e.g., England, Farkas, Kilbourne & Dou, 1988; Maume, 1999), while ignoring subjective outcomes such as work intensity (Kalleberg, 2011), emotional exhaustion (Maslach and Jackson, 1981), and social resources (Lazarus and Folkman, 1984). Conceptualizing inequality solely in terms of material outcomes misses the possibility of finding inequalities in the content of work (Hodson, 2001). It is of importance to know the differences in the content, such as tasks, because this is a key determinant of

inequalities. However, this content-inequality has been overlooked by previous research and has to be investigated thoroughly (Chan & Anteby, 2016).

Lastly, as the world is globalizing, culture is becoming an important factor to take into account when making organizational decisions (Molinsky, 2007). Globalization stimulates the integration of cultures, making countries more multicultural, with the result that managerial positions are being occupied by individuals with different cultural backgrounds (Cox & Blake, 1991). Since cultures differ in the degree to which they hold stereotypes (Williams and Best, 1990), managerial decisions, such as allocating choices, may be influenced in different ways. It is therefore of importance to create a better understanding of the role of culture and how it influences these decisions. Altogether, the aim of this study is to provide a better understanding of gender in the allocation of tasks and the role of culture in these allocation decisions.

It is of importance to understand the source of this inequality because it has several negative consequences for women. By being assigned to different tasks than men, women’s future careers are being put at a disadvantage (Mai-Dalton & Sullivan, 1981). Lastly, since the number of cultures has increased in the Netherlands and it has become a more multicultural society, it is important to know the role of culture in this inequality (Foukalne, 2016). Therefore, this study aims to answers the next research question:

“Is there a gender-bias in task allocation decisions of men and women and how does culture influence these decisions?”

9

This paper will first provide the theoretical backbone concerning gender inequality, based on existing literature and arguments for the expected relationships. Thereafter an overview of how the research has been done will be provided in the methodology. Followed by the results of the research. Finally, the discussion which contains a short

summary, theoretical implications, strengths and limitations of this research, advice for future research, practical implications and the conclusion. The aim of this thesis is to analyze how the allocation of tasks is done and the role of gender and culture in the allocation.

2. Literature review

2.1 Gender in the allocation of tasks

Gender is defined as the social process of creating distinguishable social statuses for the assignment of rights and responsibilities to women and men (Lorber & Farrell, 1991). Making these distinctions is embedded in everyday activities and is done unnoted, it happens just as naturally as breathing (Lorber, 1994). These associations are the source of stereotypes, whereby certain attributes of a group are allocated to the individual members of that group, just because they belong to that group (Taylor, Fiske, Etcoff, & Ruderman, 1978; Heilman, 1983). Stereotyping based on gender is when generalizations are made about the attributes of men and women (Heilman, 2001). Mental associations and expectations are made about what men and women are like (Cuddy, Wolf, Glick, Crotty, Chong & Norton, 2015). Stereotyping serves as a cognitive shortcut to simplify and organize all that is observed and to make predictions about others. Since these stereotypes are automatically activated, widely held and have a great impact, they greatly influence impressions formed about women and men (Heilman, 2012).

Stereotypes can be descriptive as well as prescriptive. Descriptive stereotypes indicate what men and women are like and are an expectation of the actual characteristics they

possess. An example of a descriptive believe is that women are nurturing.

Prescriptive stereotypes, on the other hand, do not indicate what women and men are like, but they create expectations about how they should be (Burgess & Borgida, 1999). An example of a prescriptive believe is that women should be nurturing. These stereotypes set the norm for appropriate and inappropriate behavior for women and men (Heilman, 2012). When people deviate from this norm, it creates social disapproval and negativity, called backlash (Rudman & Glick, 2001). Both descriptive and prescriptive stereotypes have proven to be consistent

10

over time, culture and context (Williams and Best, 1990; Schein, 2001). Earlier and several other studies found certain stereotypical characteristics for both men and women. These characteristics entail that women are more likely to possess communal traits (per example showing concern for others) and men are more likely to possess agentic traits (per example being achievement-oriented) (Heilman, Block, Martell & Simon, 1989). Despite several social changes that took place over the years, twenty-three years later Hentschel, Heilman, and Peus (2013) found similarities in the characteristics for men and women. Moreover, Williams and Best (1990) did research on the gender stereotypes held in 25 different countries and

discovered that these countries hold the same stereotypical view of men being agentic and women being communal. Furthermore, stereotypes can be found in various contexts. They are present in several employment settings, as well as in social domestic settings (Brenner,

Tomkiewicz, & Schein, 1989).

Amidst these two type of stereotypes, the most well-known way to differentiate men and women is by means of communality and agency. Communality entails showing concern for others (being kind, caring, considerate), affiliative tendencies (such as warm, friendly, collaborative), deference (such as obedience, respectful, self-effacing) and emotional sensitivity (such as perceptive, intuitive, understanding). These qualities are assumed to be most likely found in women. On the contrary, men are assumed to most likely to possess qualities concerning agency. Agency entails achievement orientation (such as being competent, ambitious, task-focused), tendency to take charge (such as being assertive, dominant, forceful), autonomy (such as being independent, self-reliant, decisive) and rationality (such as being analytical, logical, objective) (Eagly & Sczesny, 2009; Heilman, 2012, p. 115).

Two theories which describe stereotypes are the social role theory (Eagly, Wood & Diekman, 2000) and the lack of fit theory (Heilman, 1983). The social role theory explains that stereotypes originate from the contrasting social roles men and women are distributed to, which explains the differences between these two sexes (Eagly, Wood & Diekman, 2000). This theory states that gender stereotypes arise due to the characteristics that are required for activities typically carried out by each sex and the occupational and family roles they fulfill. These gender roles create inferences about the characteristics individuals of each sex should have, which equip them for tasks that their gender is typical ought to perform (Wood & Eagly, 2002). The general view that is held nowadays is based on the traditional perception of labor: ‘men as breadwinners outside the home and women as caretakers inside the home’ (Eagly & Sczesny, 2009; Lachance-Grzela & Bouchard, 2010, p. 1). Thus, women are thought

11

more often to fulfill unpaid roles, whilst men are thought to fulfill more paid roles (Koenig & Eagly, 2014). This difference not only exists in the type of payment, but also in the

requirements of the job. Research has found that women are more likely (than men) to fulfill roles that require communal, domestic and subordinate behavior. Men on the other hand are more likely to occupy roles that require agentic, resource acquisition and dominant behavior (Wood & Eagly, 2002). This idea is retrieved in real life. Although the number of working women has increased over the years, they continue to work mainly in occupations which require communal attributes (Scheibmayr & Reichel, 2017). It has been found that the top three jobs that women occupy are assistant, nurse and teacher (Department of Labor, 2015), which are jobs that require social sensitive, service and nursing characteristics (Heilman, 1983). Men, on the other hand, mainly occupy agentic positions such as senior manager, construction worker or engineer, which require characteristics of being independent, analytical and dominant (Bureau of Labor Statistics, 2016). Although they mainly work in stereotypical jobs, there is an increase in women in agentic occupations and vice versa (Williams, 2017).

Albeit, stereotypes still exist and are widely used, they are usually overgeneralizations, in many cases incorrect and do not apply to the individual group member (Heilman, 1983). They may create false judgments which lead to biased feelings and actions against others, just because individuals belong to a certain group (Heilman, 1983). These false judgments have several consequences for both men and women and their careers, especially for women. When there is a misalignment between the attributes an individual actually possesses, and the

attributes needed to succeed at a certain job, it is called ‘lack of fit’ (Heilman, 1983). This second theory states that due to an individual’s assumed lack of correspondence with the job requirements, the expectation of failure arises because the individual is assumed to lack of essential skills and abilities (Heilman, 1983).

This disparity mainly exists between women and agentic positions, such as top management and executive positions. It is believed that top positions mainly require the stereotypical characteristics of a man, namely agentic characteristics, but that of a woman to a lesser extent (Schein, 2001). Due to the stereotypical view that women lack these agentic characteristics, they are considered less fit for top positions and expected to be unlikely to succeed in their job (Heilman, 2012). Besides, when asked to describe the characteristics of a manager, research found that good and successful managers were described mainly in

masculine terms (Dennis & Kunkl, 2004; Schein, 2001). So, the mentality of the so-called ‘think manager, think male’, lives on and seems reluctant to change (Schein 1973; 1975;

12

Heilman, 2012). Thus, a lack of fit can form a barrier for women who compete equally with men for employment, specifically for top positions, which in turn impedes their career advancement.

From the previously explained theories can be concluded that men and women are more likely to occupy work-related positions that fit with their stereotype. However, progress has been made in achieving equality in these positions. Despite the growing number of men and women in non-stereotyped jobs (Williams, 2017), differences can still occur in the tasks they perform (De Pater et al., 2010). As aforementioned, the tasks individuals are allocated to has consequences for their future career and their pace of advancement (Berlew & Hall, 1966; Woodall et al., 1997). Whilst previous research has widely researched gender inequality in the workplace and the discrimination against women in several stages of the employment process (Terborg, 1977), they have neglected the importance of the allocation of tasks.

Taking the lack of fit model and social role theory in account, it is expected that men and women get assigned to tasks that fit with their stereotypes. Regarding the social role theory, men and women are assigned to tasks that are in line with their stereotypes. As women are assumed to possess more communal characteristics and men more agentic characteristics, it is likely that they will be allocated to tasks that contain these characteristics (Heilman, 1983). When deciding which task characteristics are in line with a person’s characteristics, the lack of fit theory is taken into account. It is assumed that the requirements for communal tasks are more in line with a woman’s skills and abilities. Agentic tasks, on the other hand, are assumed to require the skills and abilities of a man. Due to a compliance with the

requirements with their skills and abilities, it is expected that they will be fit for these tasks (Heilman, 1983).

Supported by previous theories, I propose that due to gender stereotypes, women are more likely to be allocated to communal tasks and men to agentic tasks. Therefore, the first hypotheses.are:

Hypothesis 1A: Men are more likely to be allocated to tasks which require agency (than women).

Hypothesis 1B: Woman are more likely to be allocated to tasks which require communality (than men).

13

2.2 Evaluations

When making employee decisions, managers rely on employee characteristics to support their decisions. Lepak and Snell (1999) stated that an employee’s skills and abilities and its

uniqueness, determine the outcome of managerial choices. Bills (1990) states that previous experience is an important factor. Tews, Stafford and Tracey (2011), on the other hand, lay focus on the criteria which can best predict future performance, namely personality and intelligence. The current study however, focuses on performance and the anticipated state of the employee in a certain task as criteria on which managers base their decisions. More precisely, managers evaluate the employee’s anticipated competence, well-being, and comfort when allocating tasks.

2.2.1 Competence

To begin with competence. Competence is related to an individual’s intelligence, power, efficacy, and skills (Cuddy, Glick & Beninger, 2011). When making hiring choices,

competence is one of the criteria that managers take into account. Individuals are perceived as competent when they meet the job demands and thus show the desired characteristics for a certain job (Burgoyne, 1988). Unfortunately, when a job is stereotyped to a certain sex, there exists a hidden bias in making these hiring decisions, resulting in benefitting the gender typical-applicant (Phelan, Moss-Raucsin & Rudman, 2008). Per example, a manager has to decide on who to allocate to a technical task, which is characterized as agentic. He can choose between a female and male employee. Due to the stereotypical view that men are technicians and more often fulfill technical roles, the manager might expect that he is more competent and have a subtle preference for the male employee. As these biases exist in the hiring phase, they might also arise in the process in which tasks allocation decisions are made.

Referring this back to the social role theory, men and women are expected to work on tasks, for which they possess the needed qualities (Eagly & Sczesny, 2009; Heilman, 2012). Women are seen as more competent for communal tasks, because they possess these

characteristics. Equivalently, males are perceived as competent for tasks which require agentic skills (Eagly & Sczesny, 2009; Heilman, 2012). On the contrary, women are seen as less competent for stereotypically male tasks and men are seen as less competent for

stereotypically female tasks. This view of being less competent is caused by the fact that they are perceived to lack skills that fit with the task, which can be explained by the lack of fit theory (Eagly & Sczesny, 2009; Heilman, 1983; 2012). Moreover, managers may anticipate

14

that employees who are able to exert their competencies, perform better and increase

organizational success. Therefore, managers might more likely allocate employees to tasks in which they are able to use their competencies in a way that is beneficial for the organization (Karnouskos, 2017).

2.2.2 Well-being

The second factor that managers take into account is employee being. Employee well-being is defined as “the overall quality of an employee’s experience and functioning at work” (Warr, 1987, p. 2). Well-being is a trending topic in organizations and managers are devoting considerable resources to fostering it. This is reasoned by the belief that it increases

employees’ effort, contributions, and productivity (Fisher, 2003). Grant, Christianson and Price (2007) found that the type of the task has an influence on employee well-being. It is therefore expected that when making allocation choices, managers take the employee’s expected well-being in that task into account.

This concept can be divided into three dimensions: happiness, health, and relationship (Finn, 1992). Happiness, the psychological part of well-being, is aroused by subjective work experiences. It is concerned with the balance of positive and negative thought and feelings in employee’s judgments (Grant, Christianson & Price, 2007). The second dimension is health, also known as physical well-being, which represents the stressors and strain at work. Stressors arise when individuals are exposed to stressful working conditions, such as a high workload or work intensification and can lead to strains such as stress and burnouts (Spector & Jex, 1998; Van De Voorde, Paauwe & Van Veldhoven, 2012). The third dimension is

relationships, also called social well-being. This dimension is concerned with the relationship between an employee and its leader. Whilst the other two dimensions are focused on the individual, social well-being is more concerned with interactions. As this study will only focus on the individual and lays no emphasis on the relationship between manager and employee, this dimension will be excluded (Van De Voorde et al., 2012).

According to Dunn, Mount, Barrick, and Ones (1995) managers emphasize the importance of emotional stability in their hiring decisions. Happiness is created when there is job satisfaction, which has several causes. The most notable cause of job satisfaction is the nature of the work itself, called intrinsic job characteristics. When these characteristics are in line with an employee’s abilities, goals, psychological needs, and interests there is a person-job fit (Sekiguchi, 2004). This match between what an employee wants, needs and receives

15

from a job, creates a high level of fit, which arouses feelings of happiness (Edwards, 1991). The same has been found by Herzberg (1966) in the motivation-hygiene theory. According to this theory employee satisfaction is also created by the characteristics of a job. He found that to create satisfaction, employees need to be assigned to tasks which fit within their skills and abilities. Moreover, Frijda (1988) found that people are happier when they perceive a certain situation as beneficial to their interests or to their progress in goal attainment. Due to the perception that women should fulfill communal roles, it might be perceived that they will be more interested in these tasks. Furthermore, these role expectations might arouse the thought that women will be happier with these type of tasks, because they will be more likely to achieve their goals, since it fits within their stereotyped capabilities. Aside from their ability to achieve their goals, research found that women favor occupational roles that are in line with their personal goals. As women are communal, they value this greatly and would prefer their job to fulfill their communal goal (Diekman, Brown, Johnston & Clark, 2010). Due to this greater value of women for communal jobs, managers may perceive them to prefer working on these tasks.

Physical well-being, on the other hand, is caused by the absence of stressful working conditions (Van Voorde et al., 2012). Smith, Kaminstein, and Makadok (1995) have found that stress in the workplace is caused by role conflict. Role conflict occurs when the roles a person is expected to fulfill contradict each other, which makes it difficult to fulfill the roles and causes stress (Katz & Kahn, 1978; Rizzo, House & Lirtzman, 1970). The traditionally held view that work is for men and housekeeping is for women, still exists in some degree and may cause role conflict for women in the workplace (Eagly & Sczesny, 2009). For example, at home women are expected to be the communal type: being concerned about their family members and being a warm and caring mother. However, this same communal mother fulfills a managerial role at work, which demands high agentic characteristics such as being

dominant. The roles she is expected to fulfill at home and at work demand different characteristics. Making a clear distinction between what characteristics to show, at what moment might be difficult and may cause stress. Moreover, it has been found that women spend more time in combined work and family activities than men, meaning they have to switch more often between roles, which increases the likelihood that they will perceive role conflict (Pleck, 1985).

Furthermore, stress can be caused by a lack of person-job fit, because the requirements lay outside the employee’s competencies (Sekiguchi, 2004). When men or women are asked to complete tasks that lay outside their competencies, managers might expect them to be

16

stressed out. Pagana (1988) found that when people feel like being incompetent, they feel scared, are afraid of making mistakes and face the threat of failing. These factors increase the level of stress in employee experiences. In addition, employees who possess different abilities than needed, have to put in more effort and work harder than a person who possesses those skills and abilities. This increased the workload might cause stress (Van Voorde et al., 2012). As stress has negative consequences which can negatively impact the organization (Spector & Jex, 1998; Van De Voorde et al., 2012), managers might choose to allocate employees to tasks in which they are physically well. These are tasks in which their roles are in harmony; tasks that fit with their goals and tasks in which the workload does not form a burden.

2.2.3 Comfortableness

Thirdly, managers could make their allocation choices based on the prevention of backlash. When women show agentic qualities such as being competent, independent and assertive, they might impede gender inequality. However, deviating from the stereotypical norm has negative consequences (Rudman, Moss-Racusin, Phelan & Nauts, 2012). One of these consequences is the backlash effect. Rudman (2008, p. 64; 1998) defined backlash as social and economic repercussions for disconfirming prescriptive stereotypes. As previously stated, it is expected of women that they should be communal and men agentic (Eagly & Sczesny, 2009; Heilman, 2012). When women deviate from this stereotypically held view and excel in agentic positions, whereby they show agentic qualities, they might be viewed as social deficient and unlikable, compared to men in agentic positions (Heilman, Wallen, Fuchs, & Tamkins, 2004). On the contrary, when they show gender-descriptive behavior in these agentic positions, they run the risk of being considered incapable (Fiske & Stevens, 1993). Either way, not behaving like a woman or behaving too much like a woman leads to negative consequences (Rudman, Moss-Racusin, Glick & Phelan, 2012; Lyness & Thompson, 1997), which may cause women to be dissatisfied at work.

In addition, backlash has been found to exist in every stage of employment and therefore might also occur in the allocation of tasks (Phelan, Moss-Racusin & Rudman, 2008). Therefore, managers might perceive that women are more comfortable when they perform tasks which fit with their stereotypical gender, so negativity and social disapproval from others are prevented. The same goes for men, to conform with the norm and prevent negative consequences of deviating, managers might perceive them to be more comfortable in performing male stereotyped tasks (Paddock & Kray, 2011; Bowles & Babcock, 2013).

17

Managers, therefore, might be more likely to allocate women and men to tasks which fit with their stereotype and prevent backlash.

From the previous paragraphs can be concluded that there are various reasons why managers might allocate employees to certain tasks. However, managers’ view is influenced by stereotypes. Their allocation decisions are dependent on their view on the employee being competent, psychological well, physically well, and comfortable in that position. Therefore, the next hypotheses are proposed:

H2A: Men’s anticipated competence, well-being and comfortableness will be evaluated higher, when working on agentic tasks, compared to women working on agentic tasks.

H2B: Women’s anticipated competence, well-being and comfortableness will be evaluated higher, when working on communal tasks, compared to men working on communal tasks.

2.3 Culture

Culture comes from the Latin word cultura, meaning the cultivation of soil. Over the years it has changed from an activity into a condition, namely a state of being cultivated (Freilich, 1989). Culture is a broad concept, which has been widely researched and has been given several definitions. Unfortunately, on the day of today, no consensus has been created on one definition (Baldwin, Faulkner, Hecht & Lindsley, 2006). The general idea that is held about culture is that it is the deeply shared values and beliefs held by a particular group of people (Baldwin, Faulkner, Hecht & Lindsley, 2006). Plog and Bates (1990, in Hofstede, 1994) extended this general vision and define culture as a system of shared beliefs, values, customs, behaviors and artifacts used by societal members to cope with the world and one another, which is passed on from generation to generation. D’Andrade (1984) on the other hand defined culture as a learned system of meaning, which is capable of creating cultural entities and particular senses of relating, through which people survive and adapt to their

environment. From these various definitions, we can conclude that culture is a shared learning system held among a certain group of people, which is used to survive and adapt and is passed on from generation to generation.

One of the most well-known authors studying culture is Geert Hofstede. He states culture to be important, due to its psychological nature. Meaning that the way people think is affected by national cultural factors, which is a result of early life experiences and later

18

experiences in organizations. These national cultural factors differ across national borders and are difficult to change (Hofstede, 1983). To describe culture, he created a framework

consisting of several dimensions (Hofstede Insights, 2018), of which the dimension of masculinity versus femininity is relevant for the current study. This dimension explains the division of roles between sexes in a society. The division of these social roles is arbitrary and differs for each society. Some societies have a clear division between male and female roles, in which the male occupies assertive and dominant roles, whilst women occupy caring and service-oriented roles. As there is a strict division, exchanging roles between men and women is often considered unacceptable. The institutions of these type of societies are characterized by an assertive mentality. This strong division between roles is called masculinity (Hofstede, 1984). Femininity, on the other hand, represents societies in which the roles of men and women are nearly the same. In these societies, it is accepted that women and men fulfill both gender roles. Moreover, the society is characterized by a caring, quality of life-oriented mentality (Hofstede, 1984).

Since culture affects the way people think, it can be expected that it influences managers’ choices when allocating tasks (Hofstede, 1983). Individuals with a masculine cultural background are used to the great division between male and female roles (Hofstede, 1983). Their values and beliefs have been imprinted (Kitayama & Park), which may influence how they consider men and women. Due to their masculine background, it could be

anticipated that they consider women as communal and men as agentic. Therefore, the expectation arises that managers with one of these cultural backgrounds will more likely allocate men to agentic tasks and women to communal tasks. This can be explained by their increased judgment about employees’ competence, well-being, and comfortableness in these stereotypical positions. To prove this, the current study will make a comparison between a culture which scores above average on masculinity and a culture which scores above average on femininity. These scores are based on a normal distribution.

The culture representing masculinity is Germany, which has a high score namely 66 out of 100. In general, Germans are driven by competition, achievement, and success. Many value performance highly, which might be recognized in the school system in which children of the age of ten are already separated into different types of schools. At this young age they have to make the first choice of what their future will look like and need to decide which of the three schools they will attend. These schools are 1) Gymnasium for students with the highest grades, who want to go to college, 2) Realschule, mainly for average or white-collar jobs and 3) Hauptschule, overall for the future blue-collar workers (The German way, 2018).

19

Moreover, Germans live to work. It is expected from managers that they are decisive and assertive (Hofstede Insight, 2018). Contradictory, The Netherlands scored 14 on masculinity, meaning that it is a feministic culture (scoring 86 out of 100 on the femininity scale). The Dutch overall, are prone to work-life balance. Managers are seen as effective when they support and involve their subordinates. Moreover, managers strive for consensus and equality, whereas solidarity and quality at work are highly valued (Hofstede Insights, 2018).

As German managers are used to the great division between women and men, it is expected that they will be more likely than Dutch managers to have high ratings on the evaluations and make stereotyped decisions. This means that the Germans will be more likely to allocate men to agentic tasks and women to communal tasks. The Dutch managers, on the other hand, are used to the equality between men and women and are therefore more likely to make decisions and have evaluations based on equality. Therefore, the following hypotheses can be proposed:

H3A: Managers from high masculine countries (Germany) are

more likely to make stereotyped task allocation decisions than managers from high feminine countries (The Netherlands).

H3B: Managers from high masculine countries (Germany) have higher ratings on

competence, well-being and comfortableness for men working on agentic tasks than high feminine countries (The Netherlands).

H3C: Managers from high masculine countries (Germany) have higher ratings on

competence, well-being and comfortableness for women working on communal tasks than high feminine countries (The Netherlands).

3. Conceptual model

The next figure displays the conceptual model of the previously explained hypotheses. This model assumes that the independent variable gender (employee) influences the independent variable task allocation decisions (manager). These choices are made on basis of judgments called evaluations, which are competence, well-being and comfortableness. These evaluations mediate the relationship between the independent and the dependent variable. Moreover, the reason why these evaluations are made is influenced by the cultural background of the

20

H3BC

manager. Therefore, culture represents a moderator variable between the independent variable and the mediator. To conclude, this study will look at the role of gender in the allocation of tasks, why these allocations are made and the role of culture in these choices.

Figure 1. Conceptual Model

4. Methodology

4.1 Research Design

To research whether there is a gender-bias in the allocation of tasks and how culture

influences this, quantitative research was done. The relationship between these variables was tested by means of an online experiment. The design of the experiment is a 2 (gender

employee: male, female) x 2 (task: communal, agentic) x 2 (culture: Dutch, German) mixed design, whereby the first two factors are within subjects and the third factor between subjects.

Gender employee Culture rater

Independent variable _{Dependent variable}

21

4.2 The sample

The sample of this study consisted in total of 293 participants, with a Dutch or German nationality. The sample was collected in The Netherlands and Germany and was collected differently for each country. In The Netherlands, the sample was collected through social

media [e.g. Facebook; Linkedin; Whatsapp] and personal contacts, whilst the sample in

Germany was collected through personal contacts and the career-oriented social networking site called XING (see appendix 1 for details on the distribution). The Dutch experiment was launched on April 16th and closed on May 18th, whilst the German experiment launched on May 3rd and closed on May 19th.

The Dutch sample consisted in total of 118 respondents. To ensure the quality of the data, only the respondents which met certain criteria were part of the final sample. Firstly, the participants were selected on basis of their nationality, meaning that every participant who did not state to have a Dutch nationality was excluded from the sample. Of the total sample, 44 participants stated to have another nationality. Thereafter participants were selected on their answers on the two control questions that have been added. The first question requested the following: “Could you please respond to this question by clicking very unlikely?”, which led to the exclusion of another 19 participants. The second control question was stated as follows: “Did you give your honest opinion and honest answers when filling out this survey?”, which did not lead to the exclusion of any participants. Finally, participants were excluded when they did not have distinct answers in their allocation decisions and thus chose the same employee for both tasks. Three employees allocated the same employee to both tasks and were therefore excluded from the dataset. Participants who met all four criteria formed the final sample, which were in total 52 participants.

The age of these 52 Dutch participants ranged from 19 to 60 years, with an average age of 31.58 years (SD= 12.37). Of this sample, 65.38% was female and 34.62% identified as male. The majority of the participants completed higher education (Bachelor, Master, Post-graduate) (44.23%), followed by 36.54% with an HBO degree, 15.38% with an MBO degree and 3.85% without an educational degree. Of this sample 63.46% was employed, either self-employed or working for an employer, of which 1.92% was a working student. The other 34.62% were full-time students and 1.92% was unemployed. The participants’ work

experience was on average 12.17 years (SD= 12.28). The occupations held by the participants varied greatly, from bartender, to mechanic, to cardiothoracic surgeon. However, the majority worked in the service industry. Only 28.85% had experience in a leadership role,

22

of which 19.23% is currently holding a leadership position. These leadership positions were on average held for 5.70 years (SD=7.55).

One hundred and seventy-five Germans participated in the experiment. Similar to the Dutch experiment, solely the participants who met the four criteria were included in the dataset. Checking for the first criteria, which is nationality, 51 participants stated to have another nationality than the German and were therefore excluded. Secondly, 21 participants answered anything other than unlikely on the first control question, which led to their

exclusion. Furthermore, employees who stated to not have filled in the questionnaire honestly and thereby did not meet the criteria of the second control question were excluded. This was not distinct from the participants who answered honestly, leading to no further exclusion. Lastly, participants were checked on their allocation choices. Here too, participants who allocated the same employee to both tasks were no part of the dataset. Two participants failed to meet the allocation criterion, leading to the final sample of 101 participants.

Of this total sample, 50.50% identified as female and the other 49.50% as male. Age varied between 21 and 63 years, with a mean age of 37.57 years (SD= 10.81). The majority of the participants completed higher education (Bachelor, Master, Post-graduate) (59.41%), whilst 17.82% completed college, the other 18.81% completed high school and 3.96% completed vocational education. The majority of the sample was employed (81.19%), the other 17.82% was a student and 0.99% unemployed. The service industry was the main industry they were employed in with a great variety of occupations. Work experience was on average 14.89 years (SD= 11.59). Leadership roles were currently held by 36.63% of the sample, however, in total 49.50% has experience with leadership roles, either currently or previously. These leadership positions were on average held for 7.02years (SD=8.91).

Table 1A. Means (and standards deviations)

Variables The Netherlands Germany

Mean (SD) Mean (SD)

Age 31.58 (12.37) 37.57 (10.81) Work experience 12.17 (12.28) 14.89 (11.59) Leadership experience 5.70 (7.55) 7.02 (891)

23 Table 1B. Gender (in percentages)

Table 1C. Education (in percentages)

Table 1D. Occupation (in percentages)

4.3 Procedure

Participants were recruited to take part in an experiment titled: “Evaluations and Decisions”, in which the goal was to review information and provide an impression of a person.

Participation could be done through an anonymous link via Qualtrics; therefore, the data was completely anonymous.

The experiment consisted of four parts: a general introduction explaining the objective and instructions, assigning an employee to a task, a questionnaire for the underlying reasons of the allocation decision and a questionnaire about personal information of the participants such as age and occupation. Participants were asked to assume the role of a manager and needed to assign two employees to two different tasks; one communal and one agentic (needs assessment and departmental negotiations). However, it was not made obvious for the

participants that these tasks were made distinct based on communality and agency. To support their decision, managers were given two resumes, one of a female employee and one of a male employee. Each employee needed to be allocated to solely one of the tasks, distinct from

Variables The Netherlands Germany Female 65.38 50.50 Male 34.62 49.50

Variables The Netherlands Germany Other (vocational) - 3.96

None 3.85 -

High school 36.54 18.81 College (HBO/ Fach.) 15.38 17.82 Bachelors, Master & Post-graduate 44.23 59.41

Variables The Netherlands Germany Employee 53.85 77.23 Student 34.62 17.82 Self-employed 9.61 3.96 Unemployed 1.92 0.99

24

the task they allocated the other employee to. This means that they could not allocate both participants to the same task, but when allocating one employee to one task, this automatically means you choose to allocate the other employee to the other task. After allocating the

employees, the participants were given a questionnaire with questions about their decisions. They were asked to rate how they think the employee would feel when working on one of the tasks, the likelihood of advancement of the employee, rate the employee on dimensions of communality and agency, and characterize the tasks. This questionnaire was followed by questions about the participants’ demographics.

4.4 Stimuli and manipulation

The tasks that needed to be assigned required the same amount of work and time but differed in activities. Whilst the needs assessment task contained communal characteristics (per example ascertaining needs and wishes of organizational members), the departmental negotiations task contained agentic characteristics (per example negotiating with different departments and companies). Gender was manipulated by means of name and photo. The photos were matched by means of attractiveness and competence to reduce biases. Moreover, the resumes that were shown to the managers differed in university, average grade, previous work experience, and hobbies. Of each of these characteristics there were two types, per example the university the employees could have attended, could be the University of

Amsterdam or the Vrije Universiteit. As there were two of each characteristic, there could be 2 type of resumes. Of both types, there was a female and male version, which led to 4

different resumes (2 type of female resumes and 2 type of male resumes) (See Appendix 2 & 3 for resumes). Each manager was presented to solely one of the female resumes and one of the male resumes. Additionally, the order in which the resumes were shown: female or male first, was randomized. This led to a total of four different versions, each participant was randomly assigned to. Thus, participants could be assigned to the next versions: (Female type 1, Male type 2); (Female type 2, Male type 1); (Male type 1, Female type 2); (Male type 2, Female type 1).

Furthermore, the language of the experiments differed. The experiment launched in Germany was in their native language and contained German characteristics, per example a German university or work experience in German companies. The experiment for the Dutch was in English, but contained Dutch characteristics (see appendix 2 & 3 for the complete

25

experiment). To analyze the impact of culture, a comparison was made between the data of these cultures. Both the Dutch and the Germans had an average response time of 21 minutes.

4.5 Measures 4.5.1 Tasks

Participants needed to allocate the employees to the tasks. Employees could be allocated to two different tasks, the needs assessment task and the departmental negotiations task. The needs assessment task requires the employee to be in contact with others and relationship-oriented and therefore represents the communal task. The departmental negotiations task requires strategic and negotiation skills and therefore analytical skills, representing the agentic task. After being shown the resumes, the managers were first asked to allocate one of the employees to the communal task and thereafter the agentic task. To do this, they could choose an employee from a drop-down menu. In this menu, the above option was the male employee, whilst the female was shown below. Their choices were converted into dummy variables, in which 0 = the male participant and 1= the female participant.

4.5.2 Evaluations Competence

Competence was based on 2 items with a 7-point scale (If employee x were to work on task x, to what degree do you think he/she is 1= incompetent to 7= competent and 1 = unsuccessful to 7 = successful). This item explains whether an employee is considered capable, efficient, skilled and intelligent enough to perform the given task (Correll, Shelley, Benard & Paik, 2007). The participants were asked to assess both employees’ competence for the communal task as well as the agentic task. The employees’ competence will indicate whether they will be considered successful or unsuccessful, in the job. Table 2displays the results of the reliability test. These results indicate that the items measuring the employee’s competence on the communal task are overall reliable (α > 0.6), except for the male employee in the Dutch survey. However, employee’s competence on the agentic task has a mixed reliability. Whilst in the Dutch survey, male competence on the agentic task has proven to be reliable, the female’s competence was not reliable. In the German survey, the exact opposite is happening: male’s competence on the agentic task was not reliable, whilst female competence on the

26

same task was reliable (>0.6).

Table 2. Cronbach’s alpha, means (and standard deviations) competence

The Netherlands Germany

Male employee Female employee Male employee Female employee Mean (SD) α Mean (SD) α Mean (SD) α Mean (SD) α Communal task 8.00 (2.01) 0.584 8.82 (2.24) 0.868 9.62 (2.44) 0.856 9.04 (2.32) 0.911 Agentic task 7.98 (2.07) 0.659 7.88 (2.05) 0.375 7.64 (2.40) 0.577 7.83 (2.32) 0.705

Psychological Well-being

Psychological well-being is caused by an employee’s job satisfaction, which in turn leads to happiness. Happiness was measured by a single item on a 7-point Likert scale (Brayfield & Rothe, 1951). An employee is considered happy when he/she finds the tasks interesting, satisfying and pleasant, is enthusiastic about the tasks and finds enjoyment in the tasks. Happiness was measured by the question: “If employee x were to work on task x, to what degree do you think he/she is 1 = unhappy to 7 = happy”. Scoring high on this item means that the participant expects the employee to be happy when performing a certain task. This item was asked for both employees on both tasks.

Comfortableness

The employee’s comfortableness was measured with a single item on a 7-point Likert scale (“If employee x were to work on task x, to what degree do you think he/she is 1 = not

comfortable to 7 = Comfortable”. This question measured whether the employee was

considered by others as fit for the tasks; to be accepted, liked and not considered odd by coworkers. In other words, the employee would be not judged by others and therefore comfortable with performing a certain task. A high score means that the employee is considered to be comfortable with performing a task, whilst a low score indicates that the employee is uncomfortable. This item has also been measured for both employees on both tasks.

27

Physical well-being

Physical well-being was measured using 2 items. These items are based on the absence of negative working conditions, which cause stress and strain (Motowidlo et al., 1986; House & Rizzo, 1972; Wall, Jackson, Mullarkey & Parker, 1996; Ellenbecker, Porell, Samia, Byleckie & Milburn, 2008). The question measuring stress is the following: “If employee x were to work on task x, to what degree do you think he/she is 1 = stressed to 7 = Not stressed”. Strain was measured by asking the following “If employee x were to work on task x, to what degree do you think he/she is 1 = pressured to 7 = Not pressured”. A high score on these items suggests that an employee is considered to be physically well when performing a certain task. A low score, on the other hand, suggests that the task causes stress and strain, and thus decreases physical well-being. Similarly, these items were asked for both employees on the communal task as well as the agentic task.

Table 3 displays the results from the reliability analysis, which indicate that the items for both tasks were overall characterized by a good inter-item consistency (α > 0.6), except for the communal tasks for the female in the Netherlands.

Table 3. Cronbach alpha, means (and standard deviations) physical well-being

The Netherlands Germany

Male employee Female employee Male employee Female employee Mean (SD) α Mean (SD) α Mean (SD) α Mean (SD) α Communal task 7.66 (2.93) 0.819 7.63 (2.57) 0.576 8.07 (2.96) 0.673 7.98 (2.84) 0.724 Agentic task 7.75 (2.92) 0.641 7.27 (3.35) 0.750 7.29 (3.57) 0.713 7.64 (3.78) 0.784

4.5.3 Communality & Agency

Communality and agency were measured using a 9-item bipolar scale adapted from Eagly and Steffen (1984) in combination with Hossiep and Paschen (1998) and Mapstone (1996). The participants were asked to rate how people-oriented, communicative, concerned about others’ feelings, warm and understanding they think the employees are. These traits represent

communal characteristics. Moreover, they rated how analytical, dominant, competitive and rational they perceived the employees. These traits represent agentic characteristics. The results of these ratings will lead to the conclusion if a stereotypical view of gender is being

28

held by participants. All nine items were measured on a 7-point scale (e.g. Please describe what you think employee x is like using the following scales 1= cooperative to 7 =

competitive). The internal consistency of these questions has been tested by means of a

reliability analysis, which has been performed for both the male and female participants, on each task, for each country. This test led to several Cronbach alphas displayed in table 4. From these alphas can be concluded that the items measuring communality were overall internally consistent since they are greater than 0.6. On the other hand, the items measuring agency overall have alphas below 0.6 and therefore have a low internal consistency, except for female’s agency in the Netherlands.

Table 4. Cronbach alpha, means (and standard deviations) communality and agency

The Netherlands Germany

Male employee Female employee Male employee Female employee Mean (SD) α Mean (SD) α Mean (SD) α Mean (SD) α Communal task 21.35 (4.28) 0.771 24.86 (3.96) 0.823 20.32 (3.33) 0.606 24.48 (3.73) 0.756 Agentic task 17.53 (3.21) 0.521 16.10 (3.20) 0.620 17.89 (2.75) 0.482 15.64 (2.83) 0.571

29 5. Analysis & Results

5.1 Correlations

Table 5. Means, standard deviations and correlations Germany

Note. *p< 0.05, **p < 0.01.

CT: Communal task; AT: Agentic task; Psych WB: Psychological well-being; Phys WB: Physical well-being.

Variables Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1. Communal Task (CT) 1.44 0.5

2. Agentic Task (AT) 1.56 0.5 -1**

3. Competence Male CT 4.31 1.22 -0.12 0.12 4. Competence Male AT 3.82 1.2 0.32** -0.32** -0.06 5. Competence Female CT 4.50 1.16 0.12 -0.12 0.43** 0.21* 6. Competence Female AT 3.92 1.16 -0.21* 0.21* 0.38** 0.11 0.24* 7. Psych WB Male CT 3.78 1.71 0.07 -0.07 0.12 0.05 0.33** 0.19 8. Psych. WB Male AT 3.93 1.35 0.26* -0.26* 0.02 0.40** 0.26** 0.18 0.10 9. Psych. WB Female CT 4.01 1.25 0.30** -0.30** 0.36** 0.18 0.37** 0.28** 0.29** 0.14 10. Psych. WB Female AT 3.76 1.51 -0.26** 0.26** -0.07 0.03 0.26** 0.37** 0.14 0.23* -0.03 11. Comfortableness Male CT 4.03 1.84 0 0 0.18 0.06 0.43** 0.26** 0.47** 0.22* 0.11 0.34** 12. Comfortableness Male AT 3.55 1.24 0.52** -0.52** 0.07 0.46** 0.13 0.10 -0.03 0.36** 0.31** -0.04 -0.07 13. Comfortableness Female CT 3.95 1.22 0.40** -0.40** 0.28** 0.16 0.28** 0.10 0.22* 0.21* 0.60** -0.20 0.10 0.36** 14. Comfortableness Female AT 3.57 1.49 -0.06 0.06 0.28** 0.01 -0.03 0.52** 0.31** 0.02 0.19 0.24* 0.21* 0.12 0.16 15. Physical WB Male CT 4.04 1.48 0.01 -0.01 0.30** 0.17 0.22* 0.22* 0.05 0.01 0.25* 0.10 0.03 0.24* 0.04 0.23* 16. Physical WB Male AT 3.65 1.79 0.15 -0.15 -0.05 0.15 0.08 0.03 -0.01 0.17 0.05 0.10 0.06 0.23* -0.09 0.10 0.30** 17. Physical WB Female CT 3.99 1.42 0.11 -0.11 0.04 0.13 0.19 0.13 0.03 0.03 0.12 -0.02 -0.01 0.14 0.16 0.16 0.37** 0.24* 18. Physical WB Female AT 3.82 1.89 0.14 -0.14 0.15 0.23* 0.16 0.34** 0.10 0.15 0.17 0.18 0.26** 0.29** 0.09 0.27** 0.28** 0.34** 0.32**

30

To test the hypotheses several analyses have been conducted. Firstly, a Pearson bivariate correlation test has been performed to measure the strength of the relationships between the variables of the German sample. The results are presented in table 5. From this table can be concluded that the relationship between the perception of the male employee’s competence on the agentic task is negatively correlated with the allocation decisions of participants on this task, r (101)= -0.32, p=.0.001. This indicates that the more managers the employee was perceived as competent, the less likely he will be allocated to the agentic task. The same type of relationship has been found for male employee’s psychological well-being and the

allocation to the agentic task, r (100)= -0.26, p= 0.010, namely a negative correlation. This means that the higher the rater anticipates the male employee’s psychological well-being to be, the less likely it is that he will be allocated to the agentic task. On the contradict, a strong positive relation has been found between the male employee’s comfortableness on the agentic task and the allocation to the agentic task, r (100)= 0.52, p= 0.000. This illustrates that participants were more likely to allocate the male employee to the agentic task, when they perceive him to be comfortable in this task. Moreover, no significant correlations have been found between a male’s physical well-being on the agentic task and the allocation decisions on this task.

Additionally, a positive correlation has been found between a female’s anticipated psychological well-being and the allocation to the communal task, r (100)= 0.30,

p= 0.002. This implies that the higher the female’s anticipated competence is, the more likely

she will be allocated to the communal task. Furthermore, a positive relationship was found between a female employee’s comfort in the communal task and the allocation to this task,

r (100)= 0.40, p=. 0.000. This suggests that females are more likely to be allocated to the

communal task, when they are perceived as comfortable in this task. Regarding the other two evaluations, no correlation has been found between the female’s anticipated competence or physical well-being on the communal task and the allocation decisions to this task.

31 Table 6. Means, standard deviations and correlations The Netherlands

Note. *p< 0.05, **p < 0.01.

CT: Communal task; AT: Agentic task; Psych. WB: Psychological well-being; Phys WB: Physical well-being.

Variables Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1. Communal Task (CT) 1.62 0.49

2. Agentic Task (AT) 1.38 0.49 -1.00**

3. Competence Male CT 4.04 1.03 -0.30* 0.30* 4. Competence Male AT 3.99 1.04 0.38** -0.38** 0.12 5. Competence Female CT 4.41 1.12 0.10 -0.10 0.26 0.10 6. Competence Female AT 3.94 1.02 -0.52** 0.52** 0.38** -0.09 0.14 7. Psych. WB Male CT 3.62 1.26 -0.24 0.24 0.52** -0.14 0.15 0.15 8. Psych. WB Male AT 3.96 1.17 0.39** -0.39** -0.03 0.24 0.56** -0.12 -0.02 9. Psych. WB Female CT 4.02 1.30 0.26 -0.26 0.03 -0.12 0.66** -0.17 0.26 0.54** 10. Psych. WB Female AT 4.00 1.44 -0.11 0.11 0.28* -0.03 0.28* 0.50** 0.52** 0.18 0.30* 11. Comfortableness Male CT 3.82 1.47 -0.15 0.15 0.60** 0.14 0.14 0.23 0.49** 0.13 -0.02 0.48** 12. Comfortableness Male AT 3.92 1.15 0.23 -0.23 -0.01 0.18 0.52** -0.12 -0.04 0.85** 0.55** 0.06 0.02 13. Comfortableness Female CT 4.40 1.39 0 0 0.34* 0.16 0.68** 0.24 0.25 0.36** 0.33* 0.38 0.24 0.38** 14. Comfortableness Female AT 3.94 1.57 -0.11 0.11 0.24 -0.02 0.29* 0.29* 0.45** 0.12 0.31* 0.78** 0.28* 0.09 0.39** 15. Physical WB Male CT 3.85 1.46 0 0 0.31* 0.21 0.19 0.26 0.12 0.09 0.04 0.17 0.18 0.06 0.16 0.11 16. Physical WB Male AT 3.82 1.50 0.06 -0.06 -0.11 0.10 0.37** 0.08 0.02 0.45** 0.20 0.09 -0.13 0.44** 0.44** 0.05 -0.03 17. Physical WB Female CT 3.81 1.28 -0.15 0.15 0.16 -0.03 0.35* 0.16 0.11 0.36* 0.15 0.16 0.16 0.32* 0.41** 0.10 0.28* 0.35* 18. Physical WB Female AT 3.68 1.69 0.03 -0.03 0.23 -0.09 0.13 -0.24 0.29* -0.12 0.21 0.55** 0.30* 0.01 0.19 0.40** 0.04 -0.13 0.12

32

To test the strength and the direction of the relationship between the variables of the Dutch sample, a second bivariate Pearson correlation test has been conducted. The results are shown in table 6. From this table can be concluded that the male employee’s anticipated competence on the agentic task and the allocation decision on this task are negatively correlated,

r (51)= -0.38, p= 0.005. This means that the more competent the participants anticipated the male employee to be, the less likely it is that he will be allocated to the agentic task. Likewise, a negative relationship has been found between the male employee’s psychological well-being on the agentic task and the allocation to this task, r (51)= -0.39, p= 0.005. This indicates that when the male employee is perceived to be psychological well in an agentic task, it will be less likely that he is allocated to this task. With regard to the male employee’s comfort or physical well-being in relation to the agentic task allocation, no significant correlation was found. Furthermore, no significant relationships have been found between the female’s evaluations and the allocation decisions on the communal task, meaning that the evaluations are not related to the allocation decisions managers make.

5. Hypotheses testing

The hypotheses have been tested by different datasets. Due to the expectation that masculine countries are more likely to make stereotyped decisions, and thus have strong allocation decisions and evaluations for men on agentic tasks and women on communal tasks, the German dataset has been used to test hypotheses 1 and 2. Hypothesis 3 (3A, 3B and 3C) on the other hand, makes a comparison between a feminine culture, The Netherlands and a masculine culture, Germany. Hence, the Dutch sample has been added to test these hypotheses.

5.2.1 Testing hypothesis 1

In order to test whether men are more likely to be allocated to agentic tasks and women to communal tasks (H1A & H1B), a Chi-square test has been conducted. This test led to no significant association between men and agentic tasks, neither between women and communal tasks (χ2 (1) = 1.673, p= 0.196). As a result, hypothesis 1A and 1B are rejected.

Although no significant results have been found, a trend has been detected in the results. The majority of the participants allocated the male employee to the communal task (57), whilst the minority allocated the female employee to the communal task (44). In line with the aforestated, the female employee was overall allocated to the agentic task (57),