1
Anniek Schepers University of Amsterdam
10178856 Faculty of Economics and Business
Supervisor: Pauline Vromans Bachelor Thesis
30th of June 2014
A Study towards the impact of National and Sexual Orientation Differences on the
Attitudes towards Diversity and Conflict Resolution Styles.
2 Index 2 - Introduction 4 - Theoretical Framework 24 - Method 30 - Results 45 - Discussion 51 - Conclusion 53 - Bibliography
3 Abstract
This study investigates the effect of national and sexual orientation differences on the attitudes towards diversity. Also, the effects of these attitudes on conflict resolution styles are measured. It is found that sexual minorities have a more positive attitude towards diversity than heterosexuals (=majority). In addition, people with more positive attitudes prefer to choose for the conflict resolution styles of Problem-solving or Compromising. Individuals with more negative attitudes towards diversity prefer to choose for Forcing. No effect between attitudes towards diversity and the Yielding style was found.
4 Introduction
Nowadays, according to the NOS (2014), even more refugees than during the second World War are trying to escape from their home countries, such as Syria, Sudan and the Central African Republic. In addition, people are attracted to work abroad because of growing industries and global trade in countries such as Dubai. This country is appealing for its ties with countries that do not have great opportunities in the workplace and therefore people consider to expatriate (Lakhani, 2014). As well as countries with high (youth-)unemployment invite people to leave the country looking for work (CNN, 2014). These people, inter alia, bring diversity to countries. Meetings are organised to discuss this diversity and ways to incorporate it in organizations. Recently, the 7th International Diversity and Inclusion Seminar was organized to discuss failures, opportunities and models, as many companies still struggle to incorporate diversity (Icon Group, 2014). The seminar promotes diversity and inclusion as they think it could lead to success in the company. In addition to this, diversity of sexual orientation is an still issue. Lately, the Ugandan President signed a law that prohibits homosexuality (CNN², 2014). By this means, life sentence is the punishment you get for being gay (CNN², 2014). Not only in Africa, but also in the United States of America is homosexuality still not normal and for some states prohibited to engage in homosexual relationships. So, discrimination because of diversity differences is still on topic. It is therefore interesting to find what impact this has on the attitudes minorities have towards diversity. It might be the case that sexual orientation has a different influence on the attitudes towards diversity than people with different national backgrounds. These differences in attitudes could result in different norms and values. This could eventually lead to misunderstanding of each other, what people try to solve with conflict resolution styles. However, do the attitudes towards diversity give a prediction of the style what is used to solve a conflict?
Following the arguments in favour of diversity, Cox and Blake (1991) introduce the advantages of costs, resource-acquisition, marketing, creativity, problem-solving and system flexibility. However, managing a diverse team to gain the advantages takes effort. Different people have different views and opinions about the importance of diversity and research is needed to investigate this. The problem of effectively managing a diverse team will be addressed. The problem exists of placing employees together who are, based on individual attitudes, best suited to gain the highest work performance.
Harrison, Price and Bell (1998) state the difference between surface- and deep-level- diversity. Surface-diversity are the observable differences among persons. This could be age and gender, as well as race. It is not yet clear if differences among people have a positive or negative effect on
deep-5
level diversity (Harrison et al., 1998). Deep-level diversity consists of beliefs, norms, values and attitudes. Studies have shown that from a sociological perspective, deep-level similarity has a positive effect on relationships employees have with others (Harrison, 1998). It may have a
favourable impact on communication, as two similar persons understand each other better because of the same reference framework. So, individuals could experience that it gets harder to work with someone, when more differences arise (Harrison, 1998). However, providing training for employees such as awareness and skill-building classes could help employees to understand differences because of diversity better (Cox & Blake, 1991). Also, Cox and Blake (1991) explain that training could provide tools how to react to the problems dealing with diversity. Harrison et al. (1998) have also shown that time could moderate the effects and might help to overcome some of the issues of incorporating employees with different nationalities and sexualities. In addition to this, the findings of Watson, Kumar and Michaelsen (1993) indicate that diverse groups get more effective over time. Both teams (diverse and similar) experience higher performance during the length of the study. Most articles, however, are focusing on higher work performances or ratings, instead of the attitudes towards diversity (Watson et al., 1993; Harrison et al., 1998; Bowers et al., 2000; Cox & Blake, 1991).
In addition to this, what happens when a team does not work well together and problems arise? The way companies should respond to this depends on the attitudes people have. Different strategies to overcome conflicts have been researched and a discussion of these articles is addressed later on. De Dreu et al. (2001) for example, have revised the DUTCH test of Van de Vliert, 1997. This version consists of five components how people could react to conflicts. Past research has shown that group performance decreases when members experience conflicts (De Dreu & Vianen, 2001). As mentioned above, it might happen that dissimilar employees are not able to understand each other, what could eventually result in harmful conflicts. These conflicts are resolved by conflict resolution styles, but it is not clear whether different attitudes towards diversity predict conflict styles. So, it is important to investigate this topic further and try to understand the effects of diversity on conflict resolution. Therefore, the research question of this study is:
‘To what extent do national and sexual orientation differences have an impact on the attitudes towards diversity and the style to resolve conflicts?’
This study examines the attitudes towards diversity in teams and the effect of this on conflict resolution. To answer this question various variables, such as age, gender and nationality; which could have an influence on the attitudes, are studied. Researchers mention the advantage a diverse team could have (Cox & Blake, 1991).
6
Several steps are taken to provide an answer to the research question. First, a literature review presents a basis for arguments about the topic. The method of this study is explained after the theoretical framework. This section describes how the research is set up and how the variables are measured. Furthermore, the results section provides the outcome of the study and an analysis of the results is given. After this, the literature and the results are discussed and further analyzed. Also, limitations and future research are provided.
At last, the conclusion presents a small summary and tries to answer the research question.
Theoretical Framework
First, the effects of diversity are discussed. After this, the variables of nationality and sexuality are explained. Also, do national and sexual differences have an impact on the attitudes individuals have towards diversity? Three possible reasons that might explain the attitudes towards diversity are given. At last, an introduction to the strategies of conflict resolution is given.
Attitudes towards Diversity
Attitudes towards diversity can be explained by several views and theories. Some theories explain specific attitudes from the behaviour of people (Planned Behaviour theory). Other theories like the Social Identity and Self-Categorization theory interpret attitudes by the identities individuals want to acquire. These theories give an insight into how perspectives of people are composed. So, the planned behaviour theory of Ajzen (1991) tells that attitudes entail emotions towards attributes of a behavioural action. Whereas the social identity perspective (Social Identity theory and
Self-Categorization theory) explain that individuals want to be part of a group and adapt the behaviours and attitudes that belong to the group (Hogg & Reid, 2006; Tajfel & Turner, 1975, 1979).
Planned Behaviour Theory
Following the theory of Planned Behaviour of Ajzen (1991), when you have a strong intention to perform a certain action, it is more likely that you will actually exert that action. So the higher the level of intention, the more it has influence on the actual behaviour. Intention is the degree of willingness to perform a given behaviour. How much effort the person is willing to put in to execute an action. These intentions rely on the available opportunities and resources. The extent of available opportunities and resources gives the person control over his intentions and therefore his actions. The theory of Planned Behaviour added the ‘perceived behavioural control’ to the theory of
7
Reasoned Action. Perceived behavioural control refers to the feeling a person has towards the difficulty of performing an action. Control and intention can thus predict actual behaviour. First, control includes the feeling a person has towards an action. If one person has more positive emotions than another, and equal intentions, the first might have a better chance of succeeding (Ajzen, 1991). Second, perceived control is often considered as an appropriate substitute for measuring the actual control a person has over an action. This only holds under the condition that perceived control is realistic (Ajzen, 1991). The perceived control might not be realistic when a small amount of information is present, when conditions are changed regarding the activity or when the action demands new prerequisites. When these changes occur, the perceived control might not be such a good predictor of the actual control.
The feelings people have towards a specific behaviour or action are established by the attributes (other actions, emotions, characteristics) individuals hold towards it (Ajzen, 1991). Because of the emotions of the attributes, an attitude is developed (Ajzen, 1991). So, an attitude is a set of beliefs about relating actions that sum up to a positive or negative attitude. According to the expectancy value model of Atkinson, a decision to behave in a specific way, depends on the success it may have. So, the expected success (=value) of behaviours are weighted and the one with the highest expected value is chosen. This contrasts actually with the Theory of Planned behaviour of Ajzen, as this theory depends on the emotion one applies to an attitude and the expectancy model is more like a rational way of choosing behaviours.
So, following the planned behaviour theory, the attitudes of individuals can be determined by the intentions people have to actually perform an action. Available opportunities and resources influence this decision. Also, the added construct of perceived behavioural control suggests that the perceived control could indicate the actual control someone has over an action. In addition, the available opportunities and resources gives a person a degree of perceived control. So, when opportunities, resources, control and intentions are known, one can make a suggestion about the actions a person could take. At last, the feelings attached to attributes of an action define the activity as positive or negative. This reasoning could eventually be applied to try to predict actions and behaviour towards diversity.
Social Identity Theory
The Planned Behaviour Theory tries to understand the attitudes of people from the perspective of intentions, perceived control and the feelings attached to attributes of activities. On the other hand, the Social Identity Theory is build from the group perspective. An in-group (the group you belong to) and out-group (not the group you belong to) exists, where people behave like the persons from the
8
in-group. So the actions revealed by the in-group are the norm and thus attitudes are formed by this.
The Social Identity Theory of Tajfel and Turner ( 1975, 1979) suggests that people want to distinguish
themselves from other groups.This could stimulate performance in a way that individuals want to outperform others by excelling in specific tasks, but this could also have negative consequences on the cohesion of groups. The theory suggests that a person advocates negative attributes to the out-group (ie. other out-groups) so that the in-out-group (ie. the out-group the person belongs to) looks better (Hogg & Reid, 2006). So the behaviour of individuals could emerge from the –expected- behaviours of a group (eg. women) that is the norm for the whole in-group. In addition, attention should be paid to intergroup conflicts and competition (Hogg & Reid, 2006). This behaviour is developed because of the devotion to the in-group. Part of this theory is the Self-Categorization Theory, where persons adapt to an identity including behaviours, attitudes and beliefs applicable for that group (Turner et al., 1987). The group –unconsciously- determines the behaviour of the individuals because of its group norms. Thus, the identities of persons in a group become to look similar (Hogg & Reid, 2006).
Surface vs. Deep-Level Diversity
Mannix and Neale (2005) and Harrison, Price and Bell (1998) make a distinction between ‘surface-level’ diversity and deeper-level diversity, where the last is harder to overcome (eg. age has weaker effects). This means that surface-level differences are less important over time than deep-level differences (Mannix & Neale, 2005; Harrison, Price & Bell, 1998). The effect of differences of surface-level diversity is weakened by time, whereas the effect of differences of deep-surface-level diversity is strengthened over time. Mannix and Neale (2005) also mention the importance of different types of tasks, just as Iles (1995) and Bowers et al (2000). Diversity in project groups is more useful than diversity in groups with routine tasks (Iles, 1995). This is because diversity could lead to creativity because of different perspectives, norms and values, which is demanded for projects for example. Routine tasks might prefer less risk of misunderstanding because of differences and thus require less diversity within a group.
Knowledge-sharing is important for a firm to overcome surface-level diversity, but Mannix and Neale (2005) warn for the mistake of sharing information that is non-unique. Group members have to share individual knowledge that might connect people (ie. ‘bridging’). Focusing on similarities instead of differences work out positively. Relating back to the Social Identity theory; bridging makes it possible that people feel connected and thus behave like a group. When they look similar on some points, people recognize collective ideas and thus feel more connected (Mannix & Neale, 2005). The Social Identity theory makes them feel part of a group and this leads to more cooperation, interaction and
9
producing new ideas/insights (Mannix & Neale, 2005; Tajfel & Turner, 1975; 1979).
Positive results are observed when teams consist of a range of diverse people (Mannix & Neale, 2005; Cox & Blake, 1991). Mannix and Neale (2005) note that, as mentioned above, diverse teams brings a mixture of opinions and ideas and therefore hence this leads to creativity. Therefore, the quality of the performance of a group could increase. Cox and Blake (1991) also mention the advantage of creativity. Resource-acquisition, problem-solving and flexibility are other positive results of diversity (Cox & Blake, 1991). The acquisition of different resources is easier when people differ from perspective than when people are similar to each other. The reasoning for the advantage of problem-solving and flexibility are similar to the previous argument; more people with different mindsets bring other contrasting ideas than people with the same mindsets.
Performance
As mentioned previously, attracting and retaining employees of different nationalities could lead to -competitive- advantage, as organizations would like to have the best performance, it is of
importance to benefit from the advantages of multi-culturalism. Cox and Blake divide competitive advantage in six sections (costs, resource-acquisition, marketing, creativity, problem-solving and system flexibility), and mention five components in order to achieve this advantage (1991). One of the components to realize advantages of diversity, consists of conducting research towards the attitudes towards diversity. Research is important for emerging topics like nationality differences as it remains to have a big influence on teams.
Leonard, Levine and Joshi (2003) did not find positive nor negative effects of diversity on
performance. Context of the tasks is an important issue for not being able to find a significant result (Leonard et al., 2003). If diversity is managed well enough, performance of employees could increase. However, if the diversity of employees is not well understood, it could have negative impacts on the firm (Leonard, Levine & Joshi,2003). Organizations must adapt themselves to be able to handle diverse teams, for example by training. Leonard, Levine and Joshi (2003) state that mainly issues with different nationalities are helped by training and development. In contrast to this study, Bowers, Pharmers and Salas (2000) stated that –a small- significant effect is in favour of groups with different kinds of employees, although homogeneous teams benefit from other advantages (eg. need less integration). Earley and Mosakowski (2000) agree with Bowers et al. (2000) that highly
heterogeneous groups and homogeneous groups perform better than moderately heterogeneous groups. However, Bowers et al (2000) note that the tasks of the job to be done are important and the conclusion of the study could differ when tasks are more difficult. This is in line with the previous
10
study mentioned. Also, Iles (1995) states that the nature of the task is important in making conclusions about diversity. In addition, group members have to communicate in order to create synergy. Understanding of each other’s nationalities helps the group’s performance (Leonard, Levine & Joshi,2003; Iles, 1995; Earley & Mosakowski, 2000).
Attitudes towards Diversity
The Planned Behaviour Theory, Social Identity Theory and the effect of diversity on performance provide a overall perspective about the main elements of behaviour and attitudes of humans. Now, the results of several articles are discussed to present a view about the general attitudes people have towards diversity. First, the attitudes towards persons with different nationalities that do not belong to the in-group (so different nationality than the participant) are presented. Then, the general attitudes towards people with different nationalities that do not belong to the in-group are discussed. This discussion is mainly about the difference between majority (heterosexuals) and minority (gay, lesbian, bisexual). After this, the attitudes of people with different nationalities are discussed. The nationalities are divided by Hofstede’s dimensions to compare the results of the attitudes. Then, an evaluation of some articles and its concluding hypotheses are presented. At last, the attitudes of sexual orientation minorities are examined. This part also includes an evaluation and concluding hypotheses.
Attitudes towards Nationality Differences
Nationality differences exist among a lot of employees on the work floor. These differences could lead to misunderstanding because of different cultures. To gain advantage of diversity, it is important to be aware of the effects diverse employees could have on the business.
National similarity is according to Riordan and Shore (1997) an essential aspect to create a situation where an individual would cooperate in. However, it differs among nationalities to what extent this condition holds (Riordan & Shore, 1997).
Hanassab (2006) found that students with a Middle-East or African nationality still experience negative attitudes from university staff, students and others. In addition, this discrimination was worse off-campus than on-campus (Hanassab, 2006). Williams, Yu and Jackson (1997) even found that discrimination because of different nationalities lead to health issues.
11 Attitudes towards Different Sexual Orientations
Diversity does not only exist among nationalities, it is also present among people with different sexual orientations. Nowadays, the abbreviation LGBT is used. LGBT is an abbreviation for Lesbian, Gay, Bisexual and Transgender. People who belong to the LGBT generally experience more negative attitudes towards them than heterosexuals do (Cotten-Huston & Waite, 1999). Some factors that could help towards more positive attitudes are education or training and exposure to homosexuals (Horvath & Ryan, 2003; Schellenberg et al., 1999; Herek, 1987; Cox & Blake, 2001). On the contrary, religion and gender role beliefs are stated as leading towards negative attitudes towards LGBT (Horvath and Ryan, 2001; Cotten-Huston & Waite, 1999)
The presentation of Cavanaugh and Capell (Icon Group, 2014) revealed some factors to pay attention to. First, it is important to build a trustworthy relationship between the manager and the employees. This relationship is of importance for example when employees want to come out. Elements of trust are transparency, care, competency and results. Religiosity, beliefs about traditional gender roles and previous exposure to homosexuals also influence the attitudes people have towards LGTB-individuals (Horvath & Ryan, 2003). So, it might be interesting to provide trainings for certain religions as Cox and Blake (2001) suggest. As Herek (1987) mentions; when religious groups have tolerance towards different groups (religious or sexual), than they have a more positive view towards homosexuals. On the other hand, if the religion does not promote tolerance, followers do not have tolerance or even have negative feelings towards LGTB (Herek,1987). Cotten-Huston and Waite (1999) also found out that people following a religion, had a more negative attitude towards homosexuality. This is explained by the conservative approach most religions have. Furthermore, Berkman and Zinberg (1997) did find an effect of religiosity on homophobia, but they did not find an effect of education positive attitudes in the long-run. The results might conflict with studies mentioned before, but this could be the case because of non-experimental studies, no comparison groups and the fact that most studies focus on undergraduate students who were more aware of the direction of the studies.
A possible reason of the positive results of education and training is that people get in contact with homosexuality. As Berkman and Zinberg (1997) did find a negative correlation between the
familiarity with homosexuals and homophobia. This means that people who spend time with LGTB-people, generally have more positive attitudes towards them. This is in line with the finding of Horvath and Ryan (2003). They found a small, but positive correlation between the contact people had with homosexuality and the attitudes these participants had towards homosexuals. In addition, Cotten-Huston and Waite (1999) found a positive effect of contact with homosexuals on the
12
attitudes towards them. They do however indicate that this contact should be long-term to be effective. A limitation of the results of these study is that people choose their friends voluntarily and might already have more positive attitudes to diversity. So, this could be an indication why people who hang around with homosexuals have more positive attitudes towards them. A Canadian study by Schellenberg, Hirt and Sears (1999) also shows that only the attitudes of male students improved over time. However, this effect might even come from more education, life experience the students gained or exposure to LGTB, instead of only the effect of time.
The last construct of attitudes towards LGTB-people following Horvath and Ryan (2001), are the beliefs about traditional gender roles. Children are raised with stereotypes in movies and stories about relationships of men and women. In this way, a specific belief system is built that being with someone of your own gender is different than normal. Also, being feminine may be a signal that a male is homosexual (Horvath & Ryan, 2001). Information about one trait could lead to stereotype-thinking and does not necessarily become neutral when other characteristics prove the opposite (Deaux & Lewis, 1984). So, one feature could implicate other characteristics, which then initiate stereotyping. However, Horvath and Ryan (2001) found that male applicants who thought that negative consequences would arise when hiring homosexuals, rated feminine heterosexual lower than the males who did not think about negative consequences. So, males indicated females lower when negative thoughts existed about homosexuals. This fits in the gender role beliefs, as
homosexuals were seen as ‘different’. In line with this, females were given characteristics of being ‘less’ than males, as stereotypical discrimination happens on the work floor. Whitley and Aegisdottir (2000) confirm stereotypes against homosexuals through the social dominance theory. The study suggest that the higher people score on social dominance, the more traditional their beliefs are about gender roles. This means that people with higher scores on social dominance generally hold a more negative attitude towards homosexuals (Whitley & Aegisdottir, 2000). Cotten-Huston & Waite (1999) did not find a significant result from their test about the gender-role theory, but they base the negative impact of gender-role orientation on previous findings of Whitley. This means that gender role beliefs could have negative impact on the feeling people have towards LGTB-people. The conservative idea exist that females and males have a specific role in the community (Horvath & Ryan, 2001; Whitley & Aegisdottir, 2000). A different trait could lead to stereotyping and defining an individual to a different ‘gender role’ and that might make it harder to be accepted (Deaux & Lewis, 1984). Gender role beliefs are also related to religiosity because of their conservative view. It was not accepted by society for a while that people would date someone of the same gender.
All together, it means that exposure homosexual people, as well as education or training probably helps to understand and accept differences in sexuality (Horvath & Ryan, 2003; Schellenberg et al.,
13
1999; Herek, 1987; Cox & Blake, 2001). This is in line with the second conclusion; studies have shown that familiarity with homosexuals has a positive effect on the attitudes people have towards LGTB (Schellenberg et al., 1999; Berkman & Zinberg, 1997; Horvath & Ryan, 2003; Cotten-Huston & Waite,1999). If people have more contact with different sexualities (personal contact, but also training or education), it may help them to accept these differences. At last, gender role beliefs have a important impact on the attitudes people have towards LGTB. These beliefs are built from the social dominance theory.
Differences by Nationality in Attitudes towards Diversity
The results of Riordan and Shore (1997) show that attitudes of individuals towards diversity are not the same. Different nationalities could have different opinions about working in diverse teams and these differences might lead to misunderstanding of each other. Dividing nationalities by the dimensions of Hofstede makes it easier to compare the attitudes of people with different nationalities because of the rules what is set to range the countries by criteria (Masculinity, Indulgence, Power Distance, Uncertainty Avoidance and Individualism).
Joshi, Liao and Jackson (2006) found that women and coloured people performed better when they worked in a diverse team, not by making a homogeneous group of women and coloured employees. So by making diverse teams, competitive advantage could be acquired, just as Cox and Blake stated. The study of Riordan and Shore (1997) show that most people are more positive towards people who are more like them, but this affects demographic groups in different ways. The Hispanic participants in the study felt less comfortable in groups that were dominated by White or African-American participants. White respondents were more negative towards diversity when a majority of a
nationality group existed (Riordan & Shore, 1997). So, when there is an equal ratio of nationalities, a positive attitudes towards diversity is remained. Promotions of African-Americans themselves are interpreted less positively than others. This means that the nationality of a person determines the attitudes towards diversity to a great extent.
Hofstede’s Dimensions
Following Hofstede’s Dimensions (2010) of culture, all countries can be divided into certain groups with specific characteristics. The six categories of Hofstede are as follows: Power Distance (PDI), Individualism vs. Collectivism (IDV), Masculinity vs. Femininity (MAS) and Uncertainty Avoidance (UAI). In 1991 the dimension Pragmatic vs. Normative was added. The last dimension,
14
added in 2010, is Indulgence vs. Restraint. Hofstede collected data from 1967 to 1973 from more than 70 countries (Hofstede’s Centre, 2010). Analyses from these countries contributed to the dimensions nowadays.
Hofstede’s dimensions are connected with other authors’ findings about culture and values. For example Schwartz (1994) divided values into ten aspects. He mentions the importance of interconnecting the variables, as some oppose each other. Also, he points out that some variables are related in a sense that one action leads to another. So for example a strong desire for power is in relation with striving for achievement (Schwartz, 1994).
Power Distance
The first of Hofstede’s dimensions includes the way in which inhabitants accept a hierarchical order. A country having a low score on power distance generally wants to have equal power distribution among people. Countries scoring high on power distance accept a hierarchical order, where everybody accepts his or her place in the society (Hofstede’s Centre, 2010). For example, Bangladesh scores relatively high (80) on power distance and thus this means that the people do not equalize power as much as Ireland does (score of 28). Schwartz (1994) also supports the power dimension in his article.
Individualism vs. Collectivism (Self-Direction vs. Universalism)
A high score on the dimension of Individualism indicates that people should take more care of themselves rather than of their community. A country scoring low on Individualism (Collectivism) appears to have more social control (Hofstede’s Centre, 2010). The society takes care of each other and ‘we’ exists more than ‘I’. Colombia is an example which scores low (13) on this
dimension. This dimension is titled as ‘Self-Direction’ by Schwartz (1994). This value also focuses on how an individual chooses to be independent or not. ‘Universalism’ is the opposite of this value and includes how people take care of all people, just as Hofstede defines. An addition Schwartz (1994) made, is the value ‘Benevolence’. This is the degree of increasing the well-being of persons around you, with who you like to socialize with. It is different from Universalism as it focuses only on the personal network.
Masculinity (Achievement) vs. Femininity
15
masculine country are rewards, competition, success and heroism (Hofstede’s Centre, 2010). Countries scoring high on masculinity are for example, the United States of America (score of 62) and Venezuela (score of 73). A country scoring low on this dimension generally characterizes itself by caring, cooperation and quality instead of quantity. The Netherlands (score of 14) is a good example of a country scoring low on masculinity. The ‘achievement’ value of Schwartz (1994) is similar to the masculinity dimension of Hofstede’s model. It also focuses on the striving of success of an individual and competition among people.
Uncertainty Avoidance
Countries who score high on uncertainty avoidance are less inclined to take risks. Also, the society feels more uncomfortable with uncertainty (Hofstede’s Centre, 2010). This means that the country tries to set rules and have planning to coordinate ambiguity. Chile, for example, scores high (86) on the dimension of uncertainty avoidance, signalling that Chile prefers to avoid uncertainty. Bhutan has a score of 28, meaning that this country has a low preference of evading uncertainty and therefore typically is more open to innovative ideas (Hofstede’s Centre, 2010) One aspect of Schwartz (1994) that would be similar to the dimension of uncertainty avoidance might be ‘Security’. This aspect entails that people strive to get stable relationships with themselves, others and society (Schwartz, 1994).
Pragmatic vs. Normative
Countries with a high score (Pragmatic) on this dimension are inclined to accept things around them (Hofstede’s Centre, 2010). The inhabitants do not look for explanations as much as countries with higher scores on this dimension. Scoring low (Normative) on this dimension entails that countries like to know the truth. Bulgaria scores relatively high (69) on this dimension, compared to Burkina Faso (score of 27). Comparing it to the values of Schwartz (1994), the value of ‘Tradition’ has almost the same meaning as being pragmatic. Schwartz explains it as accepting your environment including its norms, values, ideas and beliefs.
Indulgence vs. Restraint
The last dimension focuses on how ‘free’ people in a country feel. A high score indicates indulgence, this means that the society has a preference to enjoy life (satisfaction) and having fun (entertaining) instead of constrict their impulses. Also, freedom of expressions and less strict
16
norms are part of a high score on indulgence (Hofstede, 2010). Restraint (low score) countries are more strict and keep themselves to norms and values. New Zealand scores quite high (75) on Indulgence, whereas Hungary has a relatively low score (31). Hedonism, Stimulation and
Conformity of Schwartz (1994) are values which correspond with the Indulgence dimension. The value of Hedonism includes how people enjoy life and look for pleasure (Schwartz, 1994). In line with hedonism is Stimulation; looking for challenges and novelty in life. It indicates how free a person is to live the life he or she wants to. At last, people who are inclined to keep themselves to norms because they do not want to hurt anyone, are probably part of the Restraint side of Hofstede. This value, Conformity, honours the parents and people are generally inclined to be polite (Schwartz, 1994).
Evaluation of articles and Hypotheses about the effect of Nationality Differences
The studies of Lewis (2003) and Ng and Burke (2004) show that even differences among people within a nationality could emerge. However, in this study a separation, by Hofstede’s dimensions, is attempted to make so that a general conclusion can be made about the attitudes towards diversity. Ng and Burke already attempt to divide the sample into Hofstede’s Dimensions.
A study of Lewis (2003) indicates that the opinions about laws to prohibit anti-homosexuality of African-Americans (blacks) and Americans (whites) do not differ much. In the study of Lewis (2003), the author characterizes African-Americans as blacks and Americans as whites. As mentioned above, education does have a positive effect on the ideas of people (Lewis, 2003; Horvath & Ryan, 2003; Schellenberg et al., 1999; Herek, 1987; Cox & Blake, 2001), but it has less impact on African-Americans than on American people. Lewis did find that it had a slightly more positive effect for Americans. Also, age has a negative relationship with attitudes towards
diversity; older African-American people have a more negative opinion about LGBT (Lewis, 2003). Religion and culture might be an explanation of this outcome. Religion decreases the possibility of a positive attitude towards gay and lesbian for both African-American and American
(Lewis,2003). Older people are generally more conservative towards homosexuality, partially because of religion and partially because of gender role theory (Horvath & Ryan, 2001; Whitley & Aegisdottir, 2000; Berkman and Zinberg, 1997). However, to sum up; religion, age, gender and education has less an impact on African-American opinions than on American opinions. This means that the attitudes of African-American people generally are more negative towards homosexuality compared to the attitudes of American people.
17
inhabitants. They divided the sample in two groups; Non-Caucasians and Caucasians. The study indicated that Caucasians generally were more individualistic, are less inclined to avoid
uncertainty, were less masculine and equalize power more than Non-Caucasians do (Ng & Burke, 2004). To further research the topic, Ng and Burke (2004) broke down the sample into two other categories, namely, born in North-America (Canada and the USA) and born in non-North
America. This leads to the outcome of non-North Americans scoring higher on Masculinity (indicating a more feminine preference of cooperation and caring), as well as higher scoring on Power Distance (meaning that a hierarchical order is accepted) than people born in North-America.
Low scores on Masculinity and low scores on Power Distance were related to more favourable attitudes towards diversity, following Ng and Burke (2004). An explanation for this is that masculine countries try to focus on achievement and (material) success, instead of cooperation (Feminine). People scoring low on masculinity (thus high on the femininity part) are looking for solidarity and cooperation and therefore are more inclined to be open to other people/diversity (Ng & Burke, 2004). As a result, a high score on masculinity is expected to have more negative attitudes towards diversity. This leads to:
Hypothesis 1a: People from countries scoring high on the dimension of Masculinity have more negative attitudes towards diversity than people from countries scoring low on Masculinity
In addition, Ng and Burke (2004) explain that high scores on Power Distance could have a negative impact on the attitudes towards diversity. High scores on PD emphasize that people accept that power is distributed in a hierarchical order. Therefore, inequalities among individuals in a society arise because people do not expect fair treatment. Therefore, a high score on power distance is expected to have less positive attitudes towards diversity. Thus:
Hypothesis 1b: People from countries scoring high on the dimension of Power Distance have more negative attitudes towards diversity than people from countries scoring low on Power Distance
If we check the scores on the website of Geert Hofstede (The Hofstede Centre, 2010 ), a pattern arises among countries. Developed countries like The Netherlands, United States of America , Australia and New Zealand score quite high on the indulgence dimension. Respectively; 68, 68,
18
71 and 75. This means that these countries are focused on pursuing passions and ambitions (The Hofstede Centre, 2010) In addition, African countries also score high on this dimension, although Africa is not characterized as a developed continent. For example: South Africa (63), Angola (83) and Mozambique (80). The website LGBTnet (2012) states that the minister of Justice of
Mozambique declared in 2011 during the Universal Periodic Review of the UN that
homosexuality is not an offence. In Angola however, it is illegal to be homosexual (LGTBnet, 2012). So it is not clear what these high scores implicate for the dimension of indulgence to the attitudes towards diversity. Following The Hofstede Centre (2010), it seems that people in countries with high scores on the dimension of Indulgence, tend to have a positive attitude and more optimism than countries with lower scores and probably have more positive attitudes towards diversity than people in countries with a low score on Indulgence.
In addition to this, Asian countries such as Thailand (45), Indonesia (38), China (24), South Korea (29) and Vietnam (35) have a low score on indulgence. As well as Middle-East countries such as Iran (40), Iraq (17), Lebanon (25), Turkey (49) and Pakistan (0). At last, Eastern-European countries such as Estland (16), Latvia (17), Russia (20), Romania (20), Slovakia (29) and Poland (29) also score low on the indulgence dimension. These societies are more restraint and tend to be more pessimistic and cynical than the ones scoring high on indulgence (The Hofstede Centre, 2010; Hofstede, 2011). Some of these countries are considered to have a more controversial thought about sexual orientation for example. This might be an explanation for the low scores and therefore its constraint and pessimistic view. As The Hofstede Centre (2010) and Hofstede (2011) mention: societies with a low score on indulgence are inclined to be restraint and generally have a more negative attitude. The low scores of societies of these countries could therefore indicate that they have less favourable attitudes towards diversity than people in countries with high scores on indulgence. Therefore,
Hypothesis 1c: People from countries scoring high on the dimension of Indulgence score have more positive attitudes towards diversity than people from countries scoring low on Indulgence
Sexual Orientation Differences on Diversity
The sexuality of people could be divided in Heterosexuality, Homosexuality and Bisexual as in this study. Following Berkman and Zinberg (1997), a distinction is made between the majority and the minority of sexual orientation. Homosexuality and Bisexual people are grouped as minority, as only a small percentage of the world population defines him- or her-self to this category. It is
19
hard to give an exact number of how large this group is, because most individuals want to be anonymous and therefore information about these people is not present. This is mainly because a lot of LGBT are still discriminated against, homophobia –fear of homosexuality- exists. Few studies have focused on the attitudes of homosexuals towards diversity. Most research studied the effect of attitudes of heterosexuals towards homosexuality. Therefore, hypotheses are built on the theory of Social Identity which explains that people prefer to be part of a group and discriminate out-groups (Tajfel & Turner, 1979; Hogg & Reid, 2006).
An interesting finding of Ragins, Cornwell and Miller (2003) is that sexual tension is felt less present in same-gender group than in mixed-gender groups. Horvath and Ryan (2003) studied the ratings of participants where people rated homosexuals less positively than males, but more positively than females. So, differences in the ‘hierarchy’ of popularity still exist, where some groups of people are considered to be ‘better’ than others. This could mean that diversity is not leading to a competitive advantage as thought. One has to note that the finding of Ragins et al. (2003) is reversed for homosexual employees; a lesbian employee working with other lesbians might experience sexual tension. These tensions consist of uncomfortable feelings and insecurity. The outcome of these studies are important, as it presents an insight why some teams do no work perfectly together. The reason of sexual tension might be forgotten, or not taken into account when some persons did not come out as homosexual. So homosexuals feel less sexual tension when they are working with heterosexuals and this could have a positive impact on their attitudes.
Relating back to the theory of Social Identity of Tajfel and Turner (1979), people want to be part of a group and advocate the out-group (including people who are not like the group the person belongs to) with negative aspects. In this way, the in-groups looks more positive. In addition to his, Jellison, McConnell and Gabriel (2004) show that homosexuals have a strong preference for the in-group. This, however, also applies to heterosexuals. So, no difference can be made between gay and straight people concerning their preference of the in- or out-group. However, Jellison, McConnell and Gabriel (2004) state that negative attitudes can be changed to positive when people are exposed to homosexuality. This reciprocal relationship is also found by Berkman and Zinberg (1997), Horvath and Ryan (2003) and Cotten-Huston and Waite (1999). In addition to this preference of in-group by homosexuals, Banse, Seise and Zerbes (2001) show that
homosexuals have more positive attitudes towards homosexuality than heterosexuals have. Could it be that homosexuals do not have such favourable attitudes towards diverse individuals (so, heterosexuals) and prefer to be with LGBT? Or is this evidence that homosexuals are more
20
open to minorities than heterosexuals? Research does not give much solutions to these issues.
Day and Schoenrade (1997) show that homosexuals who are ‘openly’ gay, had more positive attitudes compared to homosexuals who are still ‘in-the-closet’ and heterosexuals. It could be that ‘openly’-gay are more satisfied because they are accepted by the people around them. At least, they could open up and trust people around when they announce that they are
homosexual. In line with this study, Meyer (1995) found an adverse relationship between the mental health of homosexuals and minority stress. Thus, minority stress (ie. stress derived from being part of a minority group) is related to the mental health of gay men (also applicable to other minority groups) (Meyer, 1995). In addition, Mirowsky and Ross (1995) also found that minority groups experience more stress than majority groups .This might be a reason for negative attitudes towards diversity of LGTB.
Evaluation and Hypotheses of Sexual Orientation Differences
So, following the argument of sexual tension, homosexuals generally have a positive attitude towards diversity because they prefer to work with heterosexuals (Ragins et al., 2003). Diversity in workplaces for example is considered to be a good option as homosexuals feel less sexual tension. However, according to the Social Identity Theory of Tajfel and Turner (1979), people attribute more positive attitudes towards the in-group and thus diversity is not preferred. At last, due to minority-stress (stress of belonging to a minority-group), homosexuals could have
negative attitudes towards diversity as they belong to a minority-group. These groups experience that their mental health is worse because of the stress (Meyer, 1995; Mirowsky & Ross, 1995). To conclude, the expectation is that people belonging to a sexual-minority group (LGBT) have less favourable attitudes towards diversity. Thus, following the previous arguments and theories:
Hypothesis 2: Individuals belonging to a sexual-orientated minority group, have negative attitudes towards diversity
Conflict Resolution Styles
The way people are dealing with conflicts is called conflict resolution style. Cultural differences could lead to different ways of solving issues (Holt & De Vore, 2005). The DUTCH (Dutch Test for Conflict Handling) by De Dreu, Evers, Beersma, Kluwer and Nauta (2001) and made by Van Vliert (1997) tested the five contemporary styles to solve conflicts: Avoiding, Compromising, Yielding,
21
Forcing and Problem Solving. After an explanation of the different style of conflict resolution, an example is given to provide a insight into how the hypotheses are composed.
Blake, Mouton, Barnes and Greiner (1964) invented the Managerial Grid. This grid explained what types of managerial behaviour existed. It included the of concern for production and concern for people. The concern for production focuses on the output, whereas the concern for people tries to improve the relationship with employees (Blake et al., 1964). Another theory which is used together with the findings of Blake and Mouton is the theory of Cooperation and Competition of Deutsch (1949; 2006). Deutsch (1949; 2006) mentions the difference of conflict resolution by a cooperating and a competition style. He found for example that people with a cooperation style of conflict solution had qualitative better discussions, had more coordination of efforts and incorporated the opinion of the opponent better (Deutsch 1949; Deutsch 2006). These conflict theories were seen as the basics for other authors to build further on. Thomas and Kilmann (1978) based the five conflict styles mentioned above, on differences in degree of cooperation and competition. Later on, the concerns of production and people were transformed into concern for self and concern for others (Rahim & Bonoma, 1979). To summarize, most authors base their conflict ability matrix or variables on the two principles of ‘concern for self’ (ie. ‘Competition’) and ‘concern for others’ (ie. ‘Cooperation) (Thomas & Kilmann, 1978; Rahim & Bonoma, 1979; Rahim, 1983; Thomas, 1992; De Dreu et al., 2001).
The concern for self, or competition, is the degree to which a person prioritizes his/her own goals, without involving the others’ concern. On the other hand, cooperation, concern for others, is aimed at including the goals of the opponent. Both parts, cooperation and competition, have an influence on which option an individual chooses to handle a conflict. Options are shaped by degrees of competition and cooperation. This study focuses mainly on the dimensions of De Dreu et al. of 2001, but tries to incorporate other theories that suit those elements.
Avoiding
The first ability to solve conflicts focuses on low concern for the self and low concern for others. This means that there is low motivation from the individual to express a preference (De Dreu et al., 2001). This competence is seen as more ‘laid back’, and leaving the problem as what it is. As De Dreu and Vianen (2001) mention; people are not actively motivated to look for a solution. Thomas (1992) classifies this ability as ‘uncooperative’ and ‘unassertive’. So, low cooperation- and low competition- effort.
22 Yielding
A low self and a high concern for others indicate ‘yielding’ (De Dreu et al., 2001). This ability focuses more on what others want and thus has a high concern for others and a low concern for self. It is focused on the other as it accepts their goals and strives to help (De Dreu et al., 2001). Thomas (1992) also calls this aspect ‘Accommodating’, or ‘Obliging’ as Rahim (1983) mentions it. Rahim (1983) also notes that employees generally use obliging more when communicating with superiors than with peers. This is because they are inclined to say what the manager is stating.
Forcing
Forcing is as it says: forcing your will on someone else. It has a high concern for self, and low concern for others (De Dreu et al., 2001). Also called ‘Competition’ by Thomas (1992) and ‘Dominating’ by Rahim (1983). It is found by De Dreu et al. (2001) to have a negative relation by rating with the other conflict strategies. Following Thomas (1992), people could start forcing when they have limited time, limited abilities to solve problems or when parties do not trust each other.
Problem Solving
When someone is looking to solve problems, he or she wants to agree on both parts, the other and the self, high concern for self and others exists. This ability looks for insights and making trade-offs between issues (De Dreu et al., 2001). ‘Collaboration’ and ‘Integrating’ are the terms respectively for problem-solving by Thomas (1992) and Rahim (1983). Rahim (1992) shows that people were more inclined to solve problems with co-workers than with their bosses. According to Thomas (1992) individuals are inclined to use this approach in the long-term, as in the short-term a possibility to achieve a person’s own goals still exists. When this possibility diminishes in the long-term, people start to look for other solutions. Problem-solving still gives the opportunity to gain something of the goal of yourself.
Compromising
Compromising is added by some authors as the mediocrity concern for self and others was missing (De Dreu et al., 2001). The aspect of compromising can be seen as a bit of problem-solving or actively looking for an ‘in-the-middle’ solution. It consists of moderate concern for
23
others and moderate concern for self. Rahim (1983) note that a compromising approach is reached when both (or multiple) parties have equal power and that employees are prone to compromise with peers than with their superiors.
Evaluation and Hypotheses about the Conflict Resolution Styles
Morris et al. (1998) make a distinction into two styles of conflict resolution; Avoiding and Competing. Their results suggest that Chinese managers prefer to avoid, as American managers choose to compete. Holt and de Vore (2005) add to this statement that cultures with a collectivistic approach (eg. China, The Hofstede Centre, 2010 prefer to withdraw or compromise, as cultures with an
individualistic approach (eg. USA; The Hofstede Centre, 2010) prefer to force. China had a score of 20 on the individual dimension of Hofstede, meaning that they prefer to act as a group ( The Hofstede Centre, 2010). The United States of America had a score of 91 on the dimension of Individualism (The Hofstede Centre, 2010). This implicates that America’s society looks after themselves more than for the group. The cultural difference between the two countries makes it that the two different
nationalities make different choices. Chinese managers rely on conformity and tradition, whereas the Americans strive for individual accomplishments. However, research conducted by Mohr and
Spekman (1994) suggest that joint problem solving is the answer to conflicts.
Looking at the concern for others and concern for self, expectations could be made towards the attitudes towards diversity (Rahim, 1983; Rahim & Bonoma, 1979; Thomas, 1992; Thomas & Kilmann, 1978; De Dreu et al., 2001). A high concern for others implicates that these persons take opinions of others into consideration (Rahim, 1983; Rahim & Bonoma, 1979; Thomas, 1992; Thomas & Kilmann, 1978; De Dreu et al., 2001). In addition, positive attitudes towards diversity generally holds that you are open to others and thus have a high concern for others. So, the prediction is that positive attitudes towards diversity positively relates with high concern for others. Another prediction is that high concern for self means that you are more individualistic, and as Holt and De Vore (2005) state: cultures with an individualistic approach prefer to force. Individualism is based on personal
achievements, taking care of yourself and less interdepence in a society as collectivists (The Hofstede Centre, 2010). In addition, collectivists are, to start with, already inclined to work in groups (The Hofstede Centre, 2010), whereas individualists have a desire to work separate. This is in line with negative attitudes towards diversity, as working in groups is not preferred. Therefore, individualistic persons, with high concern for self, are predicted to have a negative attitudes towards diversity, because of the starting point of their preference to work not in groups. Thus, people with negatives attitudes towards diversity prefer to choose conflict style ‘Forcing’.
24
The conflict style ‘Avoiding’ has a low concern for others and a low concern for self, therefore, attitudes towards diversity could be neutral as the opinions of others are not rejected but the self is also not taken into account. The person might want to avoid interaction with group members because a low dual concern exists (De Dreu et al., 2001). Therefore, it is expected that negative attitudes towards diversity leads to the use of conflict style Avoiding. However, a weak relationship is expected as people do not care much about others or self and thus it might be that neutral attitudes occur.
As ‘Compromising’ , ‘Problem-Solving’ and ‘Yielding’ have a high concern for others, it is predicted that positive attitudes towards diversity lead to these conflict styles. Although the problem-solving style also has a high concern for self, it does not undermine the effects of the concern for others as the opinions of others are still tried to be satisfied.
Hypothesis 3a: More positive attitudes towards diversity lead to less use of Conflict Resolution Style ‘Avoiding’
Hypothesis 3b: More positive attitudes towards diversity lead to more use of conflict resolution style ‘Compromising’
Hypothesis 3c: More positive attitudes towards diversity lead to more use of conflict resolution style ‘Yielding’
Hypothesis 3d: More positive attitudes towards diversity lead to less use of conflict resolution style ‘Forcing’
Hypothesis 3e: More positive attitudes towards diversity lead to more use of conflict resolution style ‘Problem-Solving’
Method
Analysis Level
The analysis level of research is conducted on an individual (i.e. person) level. This is because the problem focuses on personal characteristics which are expressed in attitudes towards diversity. The research does not focus on a specific firm, but on different people, so that the results are applicable to multiple firms and individuals.
Research Strategy and Design
25
(´To what extent do national and sexual different characteristics influence the personal attitudes towards diverse teams and therefore affect individual conflict resolution?´) was looking for causal relationships, from a theoretical starting point (deductive). First, some explanatory models and theories are discussed. After this, a survey tests these models and theories. To prevent mistakes and inaccuracy, a pilot study is carried out. The pilot study offered implications how to adjust the
questionnaire. After this, an invitation was sent to the respondents, including a link for access to the questionnaire, whether they would like to fill in the survey. The questionnaire was available in the English language. This is because participants of different nationalities were asked to fill in the survey. Emphasis is placed on the anonymity of the information of the respondents so that they feel more comfortable in expressing themselves. The nature of the study was not revealed, however, the general idea behind the questions was given.
In order to generalize statements to a population, it is important to have a large sample. A
questionnaire is a useful tool to collect data from a sizeable group, as well as it is easy to understand (Saunders, Lewis & Thornhill, 2012). Furthermore, the research strategy is collected in a quantitative way and therefore analysis can be done with a statistical programme (i.e. SPSS). This leads to a more reliable way of generalizing the results to a population.
Restricted time and financial resources requires this study to be cross-sectional (i.e. a snapshot). Within a time range of three months it is not possible to test participants multiple times with a questionnaire (ie. longitudinal). To bring forth a reliable conclusion, a large sample is taken to justify for the cross-sectional method.
The main goal of this study is not to find explorative answers, but to test whether a relationship exists between few variables. Therefore, a questionnaire is appropriate tool as it provides a conclusion to the problem instead of an explanation.
Data Sources
First, models and theories are derived from academic articles from Google Scholar and the electronic library of the University of Amsterdam. This is documented in the theoretical framework and this is used as a basis for the conclusion. Also, current sources from newspapers are used to get a up-to-date view about the topic. After this, the models and theories are tested via questionnaires (survey).
Research Sample (Respondents)
Some questions like demographics were answered 366 times, however, following questions were answered less. This is probably because of the large amount of questions in the survey, because the
26
survey included variables for other students’ research as well. The questions after demographics were generally answered 286 times. This shows that a lot of respondents quit after a while. Some respondents were deleted because they did not answer the questions on all the variables relevant for this study. Three participants were deleted because of being an outlier. When a respondent is considered as an outlier, it has extremely low or high answer on all the questions, or, in this case, all the answers are the same (Saunders et al., 2012). In the end, the sample consisted of 227
participants.
Respondents consisted of people from personal networks. Surveys were distributed on Facebook and by e-mail. Personal networks are easy to access and are more likely to reply as they know the
researcher. This means that a large sample is reached quicker. Also, external validity is secured by using five personal networks instead of one network of distributing the questionnaires in one specific firm. This is called convenience sampling.
The sample consisted of 227 respondents who voluntarily participated to answer the questionnaire. University was the highest education level of most participants, with 35,7%. After this, respectively, the master level, HBO, secondary school and MBO. Emphasis is placed on the anonymity of the information of the respondents so that they feel more comfortable in expressing themselves. The ages of the participants ranged from 14 to 81, the highest percentage was 16,7%; 22 years old. 69,6% of the participants is student or recently graduated. The sample consisted of 149 females, 65,6% (so 34,4% male). 11,5% of the participants indicated that they belonged to a sexual minority
(Transgender, Homosexual, Lesbian or Other). The sample consisted of 30 nationalities, the majority representing Dutch, 60,4% with 137 respondents. Only 16,7% of the respondents were bicultural, this is defined as someone who incorporates multiple nationalities. This might be because the parents have different nationality. Hence, the person is probably raised with several nationalities and therefore he/she is called ‘bicultural’.
Operationalization
Different variables were measured in this study. First, gender and age are easily measured. For gender a distinction was made between female and male. Age was measured by the years since a person is born.
Nationality is measured by the nationality stated in the passport of the participant. The nationalities were divided by their score on the dimensions of Hofstede into a high or low ranking. After this, it was easier to compare two different categories (high or low) of nationalities. To measure sexuality the following options are given: Homosexual/Lesbian, Bi-Sexual, Transgender, Other and
27
Heterosexual. To measure whether the participants were bicultural or not a few questions were asked. First, the respondent answers whether his/her father has a different nationality than his/her mother. The following question asks whether the nationality of the participant is different than one of both of his/her parents. After this, the period of living in the Netherlands is asked. The following options are given: ‘never’; ‘less than 1 year’; ‘1-4 year’; ‘5-9 years’; ’10-14 years’; ’15-20’; ’20 years or more’; ‘all my life’.
Attitudes towards Diversity (11 items). The attitudes towards diversity is measured by the Workplace Diversity Survey of De Meuse and Hostager (2001). In addition, four items of Homan et al. (2010) were added (adapted from Homan et al. 2007). The ‘Diversity Beliefs’ of Homan et al. measure the attitudes towards diversity. Seven out of twenty items of the survey of De Meuse and Hostager (2001) are used in this study. This is because three items were already mentioned by Homan (2010) and ten items were measuring the actual behaviour -in the workplace-. The items of workplace behaviour do not measure attitudes towards diversity and therefore considered to be redundant. The items were rated (Likert) on a scale ranging from 1 (completely disagree) to 5 (completely agree). For this study a seven point Likert-scale is used. An example of an item of De Meuse and Hostager is: ‘I feel hopeful about diversity’. An example of one of the items of Homan et al. (2010) is: ‘I enjoy working in diverse groups’. A high score on the items indicated a positive reaction towards diversity. Four items of the survey were counter-indicative.
Conflict Resolution style (20 items). Conflict resolution style is measured by using the Dutch Test for Conflict Handling (DUTCH) designed by Van de Vliert (1997) and presented in De Dreu, Evers, Beersma, Kluwer and Nauta (2001). The questionnaire consists of 20 items, divided in five subcategories, all containing 4 items (ie. Yielding, Compromising, Forcing, Problem Solving and Avoiding). An example item of the questionnaire is ‘I try to realize a middle-of-the-road solution’. The items were rated on a scale ranging from 1 (not at all) to 5 (very much). A high score on one category indicated preference for this conflict resolution method.
A factor analysis was conducted to check whether the variables of the conflict resolution styles belonged to the groups they were put into. Only two variables did belong to other groups following the factor analysis. First, ‘I emphasize that we have to find a compromise’, should belong to Problem-Solving instead of Compromising. Second, ‘I stand for my own and other’s goals and interests’ should belong to Forcing instead of Problem-Solving. However, these variables and dimensions are tested multiple times before and proved to be reliable (De Dreu, Evers, Beersma, Kluwer and Nauta, 2001). In addition, the Cronbach alpha’s of the variables were tested by reliability testing. Reliability tests
28
are conducted to be sure that the variables mentioned above could be used for analyses. The Cronbach’s alpha gives an indication of the reliability. Table 1 shows the outcome of the reliability tests and correlations between the variables. The alpha of conflict resolution dimensions Problem Solving (4 items) and Yielding (4 items) are a bit below 0.70 (respectively 0.668 and 0.666). Deleting items of the dimension would not increase the reliabilities of the problem solving dimension. By deleting one item of yielding, the reliability would increase to 0.730. However, Pallant (2002) mentions that the reliability of a variable is allowed to be lower when there are less than ten items. The variable yielding is tested by the authors and is accepted to be a bit below 0.70.
The reliabilities of Forcing (4 items), Compromising (4 items), Avoiding (4 items) and Diversity DM (7 items) are reliable with respectively 0.759, 0.793, 0.729 and 0.769. Diversity HO (4 items) is highly reliable with an alpha of 0.829. Overall, the reliabilities of the variables are considered to be reliable. After this, the reliability of the two Diversity variables together was tested. Due to the high reliability of Diversity Ho and Diversity DM together (α=0,823), the two variables are combined as they both measure diversity properly.
30 Results
First, missing values of the variables were resolved with the Hotdeck method. This program
calculates what the missing values could be, with help of the other answers given. Next, the variables were tested for normal distribution. Not all variables were normally distributed, but because of the large sample it is accepted to proceed with testing. After this, the two groups of sexuality were compared with a T-test. After this, hierarchical regression was conducted to the variables. To test variables with a hierarchical multiple regression it is necessary to build ‘models’ with multiple variables. These variables are tested first, and every next model includes another variable to be tested (including the previous variables).
The outcome of the reliability testing, correlations, means and standard deviations can be found in table 1. This table compromises all variables and will be explained further. In addition, table 2 consists of the results of the hierarchical regressions which are completed.
Results explanation
T-test of Sexuality groups Heterosexuality vs. Minority.
The group of heterosexuals (M= 5.54, SD= 0.847, N=201) appeared to be significant differently from the Minority group on the attitudes towards diversity (M=5.9, SD= 0.69, N=24), t= -1.968, p=0.166 (so equal variances are assumed). Because the p-value is higher than 0,05, the assumption of
homogeneity of variances is not violated. The differences of the two groups is statistically significant with a p-value of 0,05, t=-1.968, df= 223. The T-test controlled for age and gender.
Regression Analysis
Regression Hofstede’s Dimensions on Attitudes towards Diversity
A hierarchical multiple regression was used to test if the cultural dimensions of Hofstede
(Masculinity, Indulgence and Power Distance) significantly predicted attitudes towards diversity. Table 2 shows the results of this regression. It was expected, with hypothesis 1a, that people from countries scoring high on the dimension of Masculinity, have a more positive attitude towards diversity than people from countries scoring low on Masculinity. Hypotheses 1b stated that people from countries scoring high on the dimension of Power Distance also had a more positive attitude towards diversity than people from countries scoring low on this dimension. At last, hypothesis 1c
31
also suggested that people from countries scoring high on the Indulgence dimension have a more positive attitude towards diversity than people from countries scoring low on Indulgence. The model controls for age, gender, being bicultural and the sexual orientation. The results of the regression indicated that the model including the dimensions of hofstede did not significantly explain attitudes towards diversity. It was found that no dimension explained the attitudes towards diversity. A summary of the results can be found in Table 2. So, the dimensions of Hofstede, unexpectedly, did not significantly predict scores of the attitudes towards diversity.
Another regression tested whether Hofstede’s dimension significantly predicted the use one of the five conflict resolution styles (controlling for diversity). This regression showed an interesting finding; a direct link between the dimension of Power Distance existed with the conflict style Avoiding (β=0.206 and p=0.050, F(8,213)=2.527). Also, Hofstede’s dimension Masculinity appeared to have a significant relation with the conflict style Problem-Solving, β=-0.153, p=0.040, F(8,213)=5.329).
