The Influencing Factors on Employee’s Willingness to Communicate in
Multinational Organizations
Master thesis by Gina Kaulen 10837574
University of Amsterdam
Graduate School of Communication
Master’s Programme Communication Science Supervisor: Christine Liebrecht
2 Abstract
Although much research has addressed the development of Multinational Companies (MNCs) which are the fruit of globalization, not enough attention is dedicated towards language in this context. Even though Willingness to Communicate (WTC) is an extensive researched topic, insights of WTC in an organizational environment are lacking and therefore discussed in this study. By using WTC as the focus of this research, the famous Big Five personality traits and the Management Communication style (MCS) were incorporated in order to investigate on mediation effects with perceived Diversity Climate (DC) and WTC. Besides, a moderation effect with English as a foreign language (EFL) was expected. 110 participants from diverse cultural and work backgrounds were included in the survey. A positive DC proved to enhance WTC. Additionally, a mediation effect was found with extraversion. MCS sell only led to an indication of mediation. For the rest, several direct effects could be detected. Based on the research findings, it can be stated that a beneficially perceived DC is crucial in order to ensure good communication among employees. Additionally, the important roles of EFL, extraversion, and MCS sell will be explained throughout this study.
Introduction
Nowadays, the ongoing process of globalization acts as a crucial determinant for the growth of MNCs (Morgan, 2009). The consequence of this development is that workplace diversity became a worldwide well established matter of fact, which asks for better intercultural
understanding (Gallagher, 2013). When employees with different cultural backgrounds come together, the impact which diversity and in turn the perception of the DC has on an
organization’s performance, is in need of special attention (Homan, Hollenbeck, Humphrey, Van Knippenberg, & Van Kleef, 2008). While more diverse employees are working together, the role of English as a common corporate language (CCL) is gaining more importance. Nevertheless, research in language within MNCs so far is still scarce (Lauring & Selmer, 2012).
This study will put its focus on determinants of WTC, a widely established field of research. It can be described as the probability that a person will communicate with somebody else, being free to choose to do so or not (McCroskey & Baer (1985). WTC is essentially considered a personality trait and its concept was originally elaborated by McCroskey and associates, based on Burgoon’s (1976) work on unwillingness to communicate (McCroskey & Richmond, 1987). However, WTC does not only exist on its own. Second language (L2)
3
communication and WTC are intertwined concepts which are widely discussed. So far, WTC in a L2 is primarily investigated in the context of language education with high school or university students (e.g. Yashima, Zenuk-Nishide, & Shimizu, 2004; Gallagher, 2013;
Baghaei, 2013; MacIntyre, Dörnyei, Clément, & Noels, 1998). As existing literature does not provide sufficient insights into the effects within an organizational setting under the influence of the DC, this will be investigated in the current study.
Along with the changing environment wherein companies act due to diversity, comes employees’ perception about the DC at work. This perception is very personal and can differ from one individual to the other, depending on one’s character. When looking at the heuristic model of WTC (MacIntyre et al., 1998), it can be assumed that DC and WTC are related to each other as the model addresses the intergroup climate. The assumption is made that a positive attitude towards an ethnic group leads to positive interactions with that group and vice versa (MacIntyre et al., 1998). Translated to the current study, a positive perception of a company’s DC would lead to higher WTC and the other way around. However, as no direct prove exists for this conclusion, further research is needed.
Next to perceived DC the famous concept of the Big Five personality traits is incorporated in this study. Earlier research by Oz (2014) already proved that the traits extraversion, agreeableness and openness could predict a person’s WTC in a L2, whereas conscientiousness and neuroticism did not confirm this relationship.
Contrary to a person’s character and one’s perception of a DC which are both
obviously personally determined, a management’s communication style (MCS) represents an external, situational related factor. There are four different types of the MCS which are distinguished in the literature: tell, sell, consult, and join (Richmond & McCroskey, 1979). The MCS has the power to directly increase employee satisfaction and commitment if the communication between supervisor and subordinates leads to decision-making processes where both parties are involved (Rozilah, Muhammad, & Kamaluddin, 2013). The current study aims to go one step further by investigating if also WTC can be affected by MCS. Additionally, the link between MCS and perceived DC will be investigated. The connection between these two variables is deduced from the relationship between leadership
communication and the organizational climate and will be explained more in detail afterwards.
This research aimed to reach employees from many different countries of origin for the purpose of presenting a broader picture of the DC’s influence, as well as personality traits, MCS and speaking a foreign language on WTC. Besides, the relationships between perceived
4
DC and WTC, the mediation effect with MCS, perceived DC, and WTC, and finally the link between the Big Five and perceived DC have never been addressed before in the literature. Only the link between the Big Five and L2 WTC plus the effect of EFL on WTC is a well established part of existing research. Therefore, it is aimed that the outcomes can
serve MNCs as a guide which indicates the importance of the variables discussed. There might be factors which receive more or less attention than necessary in order to be successful in a diversity environment. Based on the knowledge gap and aim of this study, the following research question arises:
What factors influence employees’ WTC in EFL in a multinational organization where English is the CCL?
Literature Review
Before explaining the proposed relationships between the introduced variables, WTC (in a L2) and perceived DC will be discussed in detail.
WTC
The concept of WTC has been widely investigated and proven to be an important variable of interpersonal communication processes (MacIntyre, 1994). According to McCroskey, WTC represents the intention to start a conversation when given the choice (McCroskey & Baer, 1985; Mc Croskey & McCroskey, 1986a; 1986b). WTC is further defined as the result of communication apprehension, introversion, extraversion, reticence, and shyness (McCroskey, 1992; McCroskey & Richmond, 1987).There exist two mayor determinants for WTC which can be distinguished in the literature. The first one is personality traits because due to people’s personality, consistency in their behaviour can be observed (McCroskey & Richmond, 1990; MacIntyre, 1994). These personality traits are divided into extraversion versus introversion, emotional stability versus neuroticism, self-esteem, communication apprehension and finally perceived communication competence (MacIntyre, Babin, & Clément, 1999). The second determinant however focuses on the entire situation a person finds him- or herself in. This means that although a person’s character determines a certain way of action, still differences in this person’s behaviour can occur depending on the kind of situation (McCroskey &
Richmond, 1990). For example, an employee who generally has a very positive attitude and is very sociable would probably communicate very differently if he has just heard his expected job promotion will not go through. This is why the state level, aside from a person’s traits
5
level, does also matter. It is argued that trait WTC determines if a person is likely to seek for situations that ask for communication behaviour in the first place (MacIntyre, Babin, & Clément, 1999). “Within a particular situation, state WTC predicts the decision to initiate communication” (MacIntyre et al., 1999, p. 227).
WTC in a L2
Whereas WTC is a straightforward concept, WTC in a second (L2) makes it more complex and therefore cannot simply be related to personality traits (Yashima et al., 2004). “Thus, L2 WTC can be viewed as a behavioral intention determined by a unique person-situation interaction” (Gallagher, 2013, p. 56). This means that personal and situational factors come together and thereby determine the use of L2. Additionally, WTC is defined as the final step in the communication preparation process of language learners (MacIntyre et al., 1998). An essential influencer for L2 behavior however is L2 proficiency (see MacIntyre et al., 1998; Figure 1), as well as perceived communicative competence, and communication anxiety (Dörnyei, 2005; Clément, Baker, & MacIntyre, 2003). Furthermore, Gardner’s model (1985) detected that the higher the level of someone’s integration and the stronger L2 learning motivation is, the better the interaction with an L2 language group will be. Plus, language anxiety can determine the degree of WTC (MacIntyre, 2007). By taking into consideration these findings it becomes clear that L2 communication and WTC are intertwined concepts. Finally, WTC in a L2 cannot be simply deduced by the language proficiency in which an individual is asked to communicate. A person who masters a language perfectly can still be reluctant to communicate, whereas another person who still struggles with the language tries to communicate whenever possible (MacIntyre et al., 1998).
WTC represents the focus of the current study and is therefore used as the outcome variable. Plus, closely related to WTC is the perceived DC as it is supposed to mediate the relationship between the Big Five personality traits or MCS variables and WTC. This is why in the following section the role and meaning of the perceived DC will be addressed in detail.
Perceived DC
Due to the developments provoked by globalization processes, a rising importance of language diversity management and the promotion of positive diversity attitudes need to be ascribed on multinationals’ agendas (Lopez-Duarte & Vidal-Suarez, 2010). A positive diversity climate can be defined as the groups’ openness to linguistic, visible, value and informational diversity (Lauring & Selmer, 2011). Several effects of a positive diversity
6
environment have been detected, such as improved information processing (Phillips, Mannix, Neale & Gruenfeld, 2004).
Previous research has proven that cultural diversity in the work environment could both lead to positive and negative outcomes (Cox & Blake, 1991). Employees’ increased productivity (Richard, 2000) on the one hand, and decreased group performance on the other hand (Jackson, Joshi, & Erhardt, 2003) are just a couple out of many examples. In order to be able to measure both types of attitudes, The Benefits and Threats of Diversity Scale (BTDS) was developed by Hofhuis, Van der Zee, and Otten (2013). This scale has a highly relevant advantage in the sense that positive and negative attitudes towards diversity are approached separately. This is crucial, because a person can perceive cultural diversity in certain ways as a benefit and in other ways as a negative influence (Hofhuis et al., 2013).
Perceived DC & WTC
From a revision of existing literature it appeared that the relationship between perceived DC and WTC has never been discussed before. Earlier studies rather looked at the influence of DC on group satisfaction (Lauring & Selmer, 2011) or organizational loyalty (Jauhari & Singh, 2013). Therefore, the current research will focus on the influence of DC perception which can be either positive or negative, on employees’ WTC.
Allen (1995) addresses the topic of diversity in organizational communication
whereby two conclusions are important. First of all, racially-based differences in attitudes and behaviours might influence interpersonal communication (e.g. Daniel & Smitherman, 1976) and second of all race-ethnicity based stereotypes, expectations, and prejudices influence formal and informal communication. Based on these findings it becomes clear that diversity can directly influence interpersonal communication processes which leads again to the
assumption that a person’s perception of the organizational climate depending on the diversity situation in a multinational influences their WTC. Another study, even though not directly concerned about organizational diversity climate, showed that ethnocentrism and intercultural communication apprehension are both related to intercultural WTC (IWTC) (Lin & Rancer, 2003). As these two phenomena occur because of a cultural diverse environment, one could assume that they influence the perception of a DC. Therefore, it is used in the current context to justify that a relationship between DC and WTC might exist.
More concrete research on the relationship between the two variables unfortunately remains untouched so far, but at the same time makes the necessity and importance of the current topic even more obvious. Based on the findings that the either positive or negative
7
perception of diversity influences organizational outcomes and the link between DC and WTC has been made clear, the following hypotheses arise:
H1a: A DC that is perceived as beneficial positively relates to WTC. H1b: A DC that is perceived as threatening negatively relates to WTC.
EFL proficiency
Studies reveal that attention towards language in multicultural organizations is scarce (Lauring & Selmer, 2012). International business literature for instance either ignores language or only regards it as a part of culture (Peltokorpi, 2007), which means that it is not fully recognized.At least recently, the interrelated concept of culture and language has been acknowledged in the Global PR literature (Inoue, 2003). This is what makes it interesting to integrate English as CCL of MNCs in the current study.Before the moderation effect of EFL on the relation between perceived DC and WTC will be further addressed, the term EFL, as it is defined for the present research, will be explained.
After a profound review of the literature, it became clear that a number of different terms in the field of language acquisition exist. L1 stands for the language that a person learned first in his or her life, also called mother tongue. L2 then represents the second language, which is the language a person learned after acquiring the L1. Additionally, there exist the terms ‘English as a second language’ (ESL), which just further defines L2, and ‘English as a foreign language’ (EFL), which differs from ESL in the sense that English does not necessarily have to be someone’s second language acquired but maybe the third language. Finally, ‘English as a lingua franca’ (ELF) is: “a term focused on the forms of English which emerge when English serves as the default language of communication between speakers of different first languages (L1s)” (Murphy, 2014, p. 259). For the current research, the term EFL will be applied, because the English-speaking environment (MNCs) where employees work will be examined. Doing so, employees’ proficiency in English will be taken into account because it can vary from one person to the other.
EFL, perceived DC, & WTC
So far, there exists research based on the relationship between the management of a CCL and a positive DC (Lauring & Selmer, 2012). Also a clear relationship between EFL and WTC is affirmed by using the socio-educational model by Gardner (1985). Depending on how well the language learning process develops, a person’s WTC in that particular language will be.
8
This process can be influenced by factors such as language anxiety or the level of motivation to learn the language. Emanating from the presented information, it becomes clear that foreign language communication and WTC are intertwined concepts; however studies which do not consider foreign language, or more specific EFL, as a direct influencer but as a moderator of the proposed relationship, is still missing. Only the link between EFL and perceived DC and WTC in each case is proven. This is why it is interesting to investigate on the moderating influence of EFL proficiency. One could assume that a low EFL proficiency constrains the influence of perceived DC on WTC because international communication requires both language proficiency and knowledge of culture (Holden, 2002). This leads to an
interdependency of the two variables and could therefore influence outcomes in WTC. At the same time it could be concluded that a high proficiency in EFL reinforces the effect of a beneficially perceived DC or reduces the effect of a negative perceived DC on WTC. From these assumptions the following hypotheses derive:
H2a: A high EFL proficiency positively influences the relationship between perceived DC and WTC.
H2b: A low EFL proficiency negatively influences the relationship between perceived DC and WTC.
In the following sections, the independent variables Big Five personality traits and MCS and their relation with perceived DC and WTC will be introduced.
Big Five Personality Traits & WTC
The Big Five personality traits serve to illustrate how people differ with regard to their emotional, interpersonal, experiential, attitudinal, and motivational nature (McCrae & John, 1992). These five traits can be seen as broadly abstract domains, whereby each domain consists of more specific characteristics (John & Srivastava, 1999). They are called
extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience. As the five domains capture a person’s whole personality, the definition of them is very broad and could lead to vague interpretations. The following table is meant to provide a general overview of the Big Five in order to get a better impression of their meaning spectrum.
9 Table 1. Conceptual definitions based on John and Srivasta, 1999.
Personality trait Definition
Extraversion sociability, activity, positive emotionality Agreeableness altruism, tender-mindedness, trust, modesty
Conscientiousness thinking before acting, delaying gratification, following norms and rules, planning, organizing and prioritizing tasks
Neuroticism feeling sad, nervous and anxious
Openness describes the breadth, depth, originality and complexity of an individual’s mental and experiential life
In this research, extraversion will be considered as outgoing and sociable, agreeableness is linked to a trustful, approving person, conscientiousness is defined as being attentive and motivated to get work done, neuroticism is related to someone being stressed, tensed, or nervous and finally is openness attributed to a person with a lot of artistic interests and an active imagination.
Research shows that three out of the five personality traits can significantly predict WTC in English (Oz, 2014). On the one hand does extraversion have a positive effect on communication, because it reinforces social interaction what in turn decreases language anxiety (MacIntyre & Noels, 1994). Plus, people who score high on agreeableness are more likely to communicate with others from the L2 community (Pavičić Takač & Požega, 2011) and actually enjoy this interaction as well (MacIntyre & Charos, 1996). Besides, a curious and creative mind (openness) influences beneficially WTC in a L2 (Oz, 2014). Other researchers (Ghonsooly, Khajavy, & Asadpour, 2012) found a relation between openness and L2 WTC, mediated by attitude towards the international community. The latter variable seems to be comparable to this study’s variable perceived DC. On the other hand, it could not be confirmed that conscientiousness and neuroticism have a significant impact on WTC in English (Oz, 2014). The current study assumes that employees with a high work motivation and need for control over the situation (conscientiousness) are less willing to communicate with colleagues, clients, or their managers in order to fulfil their work successfully. This is because they fear miscommunication that could influence their work negatively. Even though the other variables of the Big Five might seem to have a stronger association with WTC, it remains important to include conscientiousness in order to be consistent with regard to earlier studies (e.g. MacIntyre et al., 1998; Oz, 2014). Finally, the negative feelings of a neurotic
10
person are expected to be related to an introverted character and therefore affect WTC in a negative way. This assumption is supported by Lester (2004) who found a positive but weak association between neuroticism and introversion.
Big Five Personality Traits & perceived DC
Besides the relationship between perceived DC and WTC, perceived DC will be in a next step examined as the mediator between the Big Five and WTC. Therefore, the link between
personality traits and perceived DC are explained as follows.
From the revised literature it appears that the influence of a person’s character on his or her perception of an organization’s DC has not been addressed in any research so far. However, using common sense, it becomes clear that effects probably do exist. This will be illustrated using examples which assume possible negative or positive consequences for the perception of the DC by employees. First of all, an employee with a sociable character rather has a positive attitude towards diversity in the sense that the DC is more positively perceived (extraversion). Plus, someone who is trusting and subtle probably will not easily feel
threatened by diversity and perceive the organizational climate in that respect probably more positive (agreeableness). Furthermore, a person that likes to try new things probably enjoys a diverse climate of people from different nationalities (openness to experience). On the other hand, a person that appreciates control over a situation and is very organized might feel threatened of the fact that different people with different backgrounds and work styles cause chaos (conscientiousness). The same negative prospect accounts for neuroticism: a person that experiences a lot of negative emotions and suffers from anxiety might as well feel threatened by diversity. Deriving from these assumptions, the following sub question is:
To what extent do personality traits, mediated by the perceived DC, influence WTC? In order to answer this question, several hypotheses are formulated:
H3a: A high score on extraversion, mediated by the perceived DC, is positively related to WTC and vice versa.
H3b: A high score on agreeableness, mediated by the perceived DC, is positively related to WTC and vice versa.
H3c: A high score on conscientiousness, mediated by the perceived DC, is negatively related to WTC and vice versa.
11 H3d: A high score on neuroticism, mediated by the perceived DC, is negatively related to WTC and vice versa.
H3e: A high score on openness, mediated by the perceived DC, is positively related to WTC and vice versa.
MCS & WTC
Except from the Big Five, a MCS is also able to influence the perception of WTC. This means that not only individually determined factors like personality or the perception of diversity, but also context factors like MCS are relevant. There are four different types of MCSs which are named tell, sell, consult, and join. MCS tell means that a manager simply communicates a decision from himself or the top management to subordinates, whereas a sell manager
receives a decision from the top but can still adapt it in a restricted way. On the other hand, a consult manager encourages participation from subordinates in decision-making processes but he still is in control of the final decision. In case of a join manager, this is different in the sense that normally subordinates make a final decision, even if the manager is not present (Richmond & McCroskey, 1979). It becomes clear that the first two MCSs (tell, sell) can be compared with an authoritarian management as the communication flow is one-way directed. Managers resist listening to their employees’ ideas or suggestions (Kane, 1996). This would lead to the conclusion that one-way communication negatively influences WTC by
employees. The MCSs consult and join however can be rather categorized as two-way communication styles whereby employees are comforted to give their input in order to participate in the management processes of the organization. This implies open
communication (Kane, 1996), which probably means a positive effect on WTC.
MCS has already been researched in several contexts such as affective learning (Chory & McCroskey, 1999), employee satisfaction (Richmond, McCroskey, & Davis, 1982), and, most relevant in the current context, communication. Depending on the MCS the effect on employees will be different (Carrison, 2010). Besides, research has proven that the
communication style is an important factor for leadership, because it influences among others knowledge sharing behaviours and subordinates team commitment (DeVries, Bakker-Pieper, & Oostenveld, 2010). Due to the fact that a colleague is always preferred to be consulted before an impersonal source (Levin & Cross, 2004), the exchange of knowledge can be seen as a communication process. This finally means that a team member’s communication style has an effect on knowledge sharing. Apparently, agreeableness and extraversion, integrated in the communication style, increases willingness to share knowledge (De Vries, Van den Hooff,
12
De Ridder, 2006). Willingness to share knowledge obviously asks for WTC, because without communication knowledge sharing cannot take place. Based on the researches which show that the communication style affects employees in several ways, such as in knowledge sharing behaviour, it is assumed that a MCS influences employees WTC as well. However, this relationship has not been investigated so far and therefore its relevance is emphasized.
MCS & Perceived DC
Additionally, a link between MCS and the perceived DC can be suggested whereby the latter represents the mediator variable between MCS and WTC.Studies on the relationship between leadership and the organizational climate do already widely exist (e.g. Haakonsson, Burton, Obel, & Lauridsen, 2008; Grojean, Resick, Dickson, & Smith, 2004). They deal for instance with the effects on performance when the leadership style and organizational climate are not aligned (Haakonsson et al., 2008), or with the establishment of an organizational climate based on a certain leadership style (Grojean et al., 2004). As the term ‘climate’ in an organizational context contains shared perceptions of all employees working in the same company (Schulte, Ostroff, & Kinicki, 2006), an organization’s diversity climate is
considered as being a part of the general organizational climate. Furthermore, even though leadership communication and MCS are not identical concepts, they oftentimes fulfil the same purpose. For instance, just as the leadership style spectrum goes from subordinate centred to (top) management centred, does the same apply for MCS (Tannenbaum & Schmidt, 1985). But not only the communication style, also should the communication quality be taken into account, because communication is crucial in order to retain social relationships and social exchange (Blau, 1964; Thibaut & Kelley, 1959). Additionally, high quality
communication supports information sharing about norms and values what in turn helps to create mutual understanding (Lorenz, 1988). Therefore, high-quality communication leads to positive social relationships and vice versa (e.g. Atuahene-Gima & Li, 2002).
The presented insights illustrate the relevance of proper communication by the management in order to ease mutual understanding which could lead to a more positive perception of diversity. On the contrary, an example of a Nordic expatriate president in Japan demonstrates that cultural diversity clashes lead to miscommunication between managers and their employees (Peltokorpi, 2007). This emphasises the importance of a manager’s adapted communication style, which needs to align with the specific diversity context.
13
negatively affect WTC, and the assumptions about the relationship between MCS and
perceived DC were elaborated in the second section, the following sub-question is formulated: To what extent does the MCS, mediated by the perceived DC, influence WTC?
In order to be able answering this question, the following hypotheses are formulated: H4a: A MCS tell, mediated by the perceived DC, is negatively related to WTC. H4b: A MCS sell, mediated by the perceived DC, is negatively related to WTC. H4c: A MCS consult, mediated by the perceived DC, is positively related to WTC. H4d: A MCS join, mediated by the perceived DC, is positively related to WTC.
The model below gives an overview of the hypotheses which were presented throughout the literature review.
Figure 1. Conceptual model
Method
Data collection
A quantitative research design and an online survey methodology were applied in order to collect the data, considering a cross-sectional time frame. During the process of data collection it became clear that aiming 200 respondents would not be feasible within the available time frame because in many MNCs in the Netherlands the CCL was still Dutch. Therefore, respondents located in other countries were addresses as well. Besides, the number of respondents was limited to 100 or more if possible.
In order to gather participants, MNCs were directly contacted by phone or email to ask for their participation. When contacting them they were told that in exchange for their
14
participation the final research article would be made accessible to them. However, they all refused to participate due to time issues. For this reason, the researcher started to approach potential participants mainly via several Facebook expat groups from different cities and countries. The link was also shared on Facebook profiles and relevant private contacts were asked to participate as well via Facebook or email. Finally, a few people were also reached through Twitter and Linkedin.
At the beginning of the survey, participants were informed about the broad topic of the survey. It was clearly mentioned that participation in the survey asked for truthful answers and was voluntary, assuring confidential and anonymous treatment of the data. The duration mean of filling in the survey was 20 minutes; the trimmed mean was 11 minutes. The survey was active for 3.5 weeks (15th November until 9th December 2015).
The Sample
For the present research respondents who work in a MNC where English is spoken as the CCL were targeted. Respondents could be either native speakers of English or speak EFL. 183 people opened or started the Qualtrics survey, but only 107 fully it. This comes down to a response rate of 58%. However, for the data set a total of 110 cases remained because all questions regarding the variables of the model were answered. Only when the demographic questions started they dropped out. As a result, a total of N = 110 cases were included in the study. 37 of the participants were male, 67 were female, and 3 did not want to share their gender (and therefore also could not reply to the following demographic
questions). The participant’s age varied between 21 years and 67 years (M= 32.61, SD= 10.14). The evaluation of the survey questionnaire also revealed that the respondent’s educational level ranged from high school to PhD degree. A detailed overview can be found in Appendix 1.
Besides, participants were also asked for how long they were already working for their company. Only seven out of the 107 participants work less than four months for their
company. According to Gallagher (2013), it is important that employees already have been living and working in the Netherlands for at least four months, due to two different reasons: first of all in order to avoid emotional reactions which could intervene their normal behaviour in the first few months due to migration (Ward et al., 2001), and second of all in order to focus on employees who came to work abroad for a longer period of time, which leads to the conclusion that they are probably more motivated to socialize with other people (Gallagher, 2013). Nevertheless, the seven cases which did not fulfil this requirement were not excluded
15
because this number is still very low in relation to the total amount of participants and did not have any impact on the validity of the given survey answers. Plus, on a 5 point Likert scale it was also checked if the MNCs really complied with the condition of English use at the workplace (Froese, Peltokorpi, & Ko, 2012). The results confirmed that English was overall well implemented as CCL in the companies (M = 4.33, SD = 0.69).
Most of the employees work at a MNC located in Europe, but there were also a few who work in Asia, Africa, or the United States. An overview with all countries can be found in Appendix 2. All other indications were unclear and are not reported, because they named several locations at the same, so it was not clear in which country the person actually works at the moment. After the data collection, 72 different MNCs where the respondents work could be identified. A complete overview can be found in Appendix 3. 26 respondents did not want to share the name of their MNC. However, seven of them revealed the sector in which their company was operating. These sectors were EU, recruitment, insurance, software
development, healthcare, and two times university. The complete survey can be found in Appendix 4.
Instrumentation
Before the final survey was distributed, a pre-test was done (N = 38). All variables were measured on a 5 point Likert scale. Based on the results, the scales were adapted as it follows and used for the final survey.
For the Big Five variable the 10-item short version of the Big Five Inventory (Rammstedt & John, 2007) was used. Within the different Big Five traits, extraversion (r = .56, p ≤ .01), conscientiousness (r = .46, p ≤ .01), and neuroticism (r = .55, p ≤ .01) did significantly correlate. However, agreeableness (r = .18, p = .068) and openness (r = .17, N = 110, p = .074) did not correlate with each other. Considering that the Big Five scale has already been tested and proved its reliability and validity in earlier researches, the items could be included anyway.
For the second independent variable MCS an existing scale was applied again,
developed by Rozilah et al. (2013). The number of the initial 20 items was limited in order to control the number of total items of the survey and to retain a balance between the number of items used to test each sub-variable. Four items remained for tell and join and three for sell and consult respectively, which makes 14 items in total. Two items were re-coded. A principal component analysis (PCA) for the different types of MCSs showed that MCS tell (eigenvalue 2.20; α = .72) and consult (eigenvalue 2.24; α = .83) measure the same construct
16
and were also reliable. Also MCS join revealed only one component with an eigenvalue above 1 (eigenvalue 1.75), but its reliability was not satisfactory (α = .57). However, as an existing scale was used, the variable did not have to be adapted. Finally, MCS sell had to be adjusted as both the PCA (eigenvalues 1.53; 1.02) and Cronbach’s alpha (α = .37) were unsatisfactory. When excluding the first of the three items of MCS sell the correlation between the remaining two items became significant, r = .53, p ≤ .01. Therefore, MCS sell was used for the analysis with only these two items.
The mediator variable perceived DC was measured with the BTDS by Hofhuis et al. (2013). This scale contains originally of five dimensions concerning benefits of diversity and four dimensions concerning threats of diversity. On the one hand the dimensions for benefits were ‘Understanding Diverse Groups in Society’, ‘Creative Potential’, ‘Image of Social Responsibility’, and ‘Social Environment’. On the other hand, the dimensions for threats were ‘Realistic Threat’, ‘Symbolic Threat’, Intergroup Anxiety’, and ‘Productivity Loss’. In order to keep a balance between the two, the benefit dimension ‘Job Market’ got dismissed. Both sides of diversity consisted of 20 items, which is four items per dimension. For the present survey, two items respectively were chosen in order to keep a balance with the amount of items chosen for the other variables. Additionally, three items were re-coded.
The scale for DC benefits did not require any changes (Cronbach’s Alpha = .80), even though two eigenvalues above 1 (eigenvalues 3.58; 1.04) were detected. The same accounts for DC threats (eigenvalues 2.60; 1.35), however Cronbach’s Alpha was only moderate (α = .64). After a reduction of the items 1, 2 and 8, an eigenvalue of 2.36 and a sufficient reliability (α = .72) was the result, which is why the variable DC threats was adapted.
Question items for EFL proficiency were self constructed, based on the global common references scale from the Common European Framework of References for
languages (Council of Europe, 2001). Based on the descriptions of the language skills which a person needs to be proficient in, in order to earn a certain language certificate ranging from A1 to C2, six essential skills could be identified. No changes were necessary (eigenvalue 4.16, α = .88).
For the last variable WTC were existing scales not applicable when taking into consideration an organizational environment. They rather described everyday life situations; hence questions were adapted in a way that they fit the context based on the Willingness to communicate in a foreign-language scale (WTC-FLS) by Baghaei (2013). Items 1, 3, 5, and 7 were dedicated to WTC with native speakers, and items 2, 4, and 6 were about WTC with non- native speakers. One item was re-coded. Two components revealed an eigenvalue above
17
1 (eigenvalues 3.61; 1.14). However, the scale is reasonably reliable, Cronbach's alpha = .76. Besides, the items were adapted from an existing scale, such as in the case of perceived DC and MCS, which already proved its reliability.
Results
Hypothesis 1 – Relation between Perceived DC & WTC
The first hypothesis of the research aimed to discover the link between the perceived DC and WTC. H1a stated that a DC perceived as beneficial would positively relate to WTC. The regression model with WTC as dependent variable and perceived DV as independent variable is significant, F(1, 108) = 19.63, p < .001. The regression model can therefore be used to predict WTC, but the strength of the prediction is very low: 15.4% of the variation in WTC can be predicted on the basis of perceived DC (R² = .15). Perceived DC,
b = 0.37, t = 4.43, p < .001, 95% CI [0.20, 0.53] has a significant, moderate association with WTC. H1 can be accepted.
As a next step, a DC that is perceived as threatening was expected to negatively relate to WTC (H1b). In contrast to the regression model with DC benefits in H1a, this time it was not significant, F(1, 108) = 1.44, p = 0.233. Therefore, H1b needs to be rejected. Both results are summarized in the table below.
Table 2. Simple regression model to predict WTC for DC threats & benefits
WTC DC F b S.E. R² Benefits 19.63 0.37*** 0.80 0.15 Threats 1.44 -0.14 0.12 0.01 df 1, 108 Note. N = 110 *** p < .001
Hypothesis 2 – Relation between EFL, DC, & WTC
With regard to the second hypothesis, it was assumed that EFL proficiency moderates the relationship between perceived DC and WTC in a way that the relation between perceived DC, for both benefits and threats, and WTC would increase when EFL proficiency is high. At the same time it was assumed that a low proficiency in EFL would decrease the relation
18
between perceived DC and WTC. To test these hypotheses a multiple regression analysis was conducted with WTC as dependent variable, perceived DC as independent variable, and EFL as moderator variable. The moderation analysis was executed twice, once with perceived DC benefits, F (3, 106) = 8.94, p < .001, and the other time with perceived DC threats, F (3, 106) = 2.95, p < .05. In both cases, no interaction effect could be found to prove the existence of a moderation effect. This means that EFL cannot predict the relationship between perceived DC, benefits or threats, and WTC. Therefore, H2 has to be rejected. However, several significant direct effects could be found which are presented in the tables 3 and 4. Table 3. Multiple regression model to predict WTC (with DC benefits)
WTC b S.E. Constant 29.44*** 0.34 DC benefits 1.29*** 0.35 EFL 0.78* 0.34 Interaction term 0.40 0.32 F 8.94 df 3,106 Note. N = 110, R² = 0.202 * p < .05, *** p < .001
The main effects of DC benefits on WTC, b = 1.29, t = 3.71, p < .001, 95% CI [0.60, 1.98], and EFL on WTC, b = 0.78, t = 2.27, p < .05, 95% CI [0.10, 1.46] revealed to be significant, while the interaction effect is not significant, b = 0.40, t = 1.24, p = .220, 95% CI [-0.24, 1.03]. The strength of the prediction is low: 20% of the variation in WTC can be predicted on the basis of DC benefits, EFL and the interaction term of both (R² = .20).
DC benefits and EFL both have a significant association with WTC. The one with DC benefits is moderate, whereas the one with EFL is weak, while the interaction term of both variables has a weak association as well. For each additional point on the scale of DC, WTC increases by 1.29, whereas for EFL WTC only increases by 0.78. For all these effects other independent variables are assumed to be held constant.
19 Table 4. Multiple regression model to predict WTC (with DC threats)
WTC b S.E. Constant 29.52*** 0.36 DC threats -0.29 0.37 EFL 0.99** 0.37 Interaction term 0.02 0.37 F 2.95 df 3,106 Note. N = 110, R² = 0.278 ** p < .01, *** p < .001
In this case, the main effect of DC Threat on WTC, b = -0.29, t = -0.80, p = .428, 95% CI [-1.01, 0.43], is not significant, whereas the effect of EFL on WTC, b = 0.99, t = 2.68, p < .01, 95% CI [0.26, 1.72] is significant. Again, the interaction effect is not significant, b = 0.02, t = 0.06, p = .951, 95% CI [-0.72, 0.77]. The strength of the prediction is also low: 28% of the variation in WTC can be predicted on the basis of DC threats, EFL, and the interaction term of both (R² = .28). EFL, b = 0.78, t = 0.06, p = .951, 95% CI [0.10, 1.46] has a significant but weak association with WTC. For each additional point on the scale of EFL, WTC increases by 0.99, whereas with DC threats WTC decreases by -0.29. For all these effects other
independent variables are assumed to be held constant.
In summary it can be stated that even though the relations between DC benefits and DC threats with WTC are the same as in H1, and no moderation could be found, EFL proficiency still plays a role.
Hypothesis 3 – Relation between the Big Five, Perceived DC, & WTC
The third question was dedicated to hypotheses 3a to 3e. The five different personality traits which were discussed earlier were supposed to influence WTC, mediated by the perceived DC. All analyses were again executed separately with the variables DC benefits and DC threats. Only hypothesis 3a (extraversion) revealed significant results and could therefore be
20
accepted. Therefore, hypotheses 3b to 3e had to be rejected. Their results are summarized in the tables 5 to 7 in Appendix 5. Separate regression analyses were executed in a first step order to check if the direct relations between each Big Five variable and WTC and perceived DC were both significant. This was never the case. Therefore, further steps of the mediation analysis were not applied. However, direct links could be detected between openness and DC benefits, b = 0.62, t = 2.50, p < .05, 95% CI [0.13, 1.11] as well as between neuroticism, b = -0.85, t = -4.35, p < .001, 95% CI [-1.24, -0.46] and WTC, and agreeableness and WTC, b = 0.62, t = 2.10, p < .05, 95% CI [0.04, 1.21].
A regression analysis with DC benefits was used to investigate the hypothesis that perceived DC mediates the effect of extraversion on WTC. Results indicated that extraversion was a significant predictor of perceived DC benefits, b = 0.75, t = 3.55, p < .01, 95% CI [0.33, 1.17], perceived DC benefits was a significant predictor of WTC, b = 0.37, t = 4.43, p < .001, 95% CI [0.20, 0.53], and finally that extraversion was a significant predictor of WTC, b = 0.50, t = 2.47, p < .05, 95% CI [0.10, 0.90]. These results support the hypothesis. Extraversion was no longer a significant predictor of WTC after controlling for the mediator, perceived DC benefits, b = 0.25, t = 1.25, p = .214, 95% CI [-0.15, 0.65], consistent with full mediation (b = .50, p < .05; b’ = .25, ns; Sobel’s Z = 2.59, p < .01).
Approximately 41% of the variance in WTC was accounted for by the predictors (R² = .41). For each additional point on the scale of extraversion, WTC increases by 0.25, whereas with DC benefits WTC increases slightly more by 0.33. For all these effects other independent variables are assumed to be held constant.
Hypothesis 4 – Relation between MCS, Perceived DC, & WTC
Finally, hypotheses 4a to 4d concerned the last question about the influence of the MCS, mediated by perceived DC, on WTC. Hypotheses 4a, 4c, and 4d had to be rejected, because each relationship between one of the four MCS types and the mediator perceived DC, both threatening and beneficial, was already not significant. Therefore, chances of finding a mediation effect are nearly zero. All corresponding results are listed in the tables 8 to 10 in Appendix 6. However, MCS sell is a significant predictor of perceived DC benefits, b = 0.46, t = 2.15, p < .05, 95% CI [0.04, 0.89], perceived DC benefits was a significant predictor of WTC, b = 0.37, t = 4.43, p < .001, 95% CI [0.20, 0.53], and finally was MCS sell a significant predictor of WTC, b = 0.43, t = 2.12, p < .05, 95% CI [0.03, 0.8]. These outcomes are
contradictory to the hypothesis, as the proposed relationship was negative, and not positive as it results here. Therefore, H4b can be rejected. MCS sell was no longer a significant predictor
21
of WTC after controlling for the mediator, perceived DC benefits, b = 0.27, t = 1.40, p = .165, 95% CI [-0.11, 0.65], consistent with full mediation (b = .43, p < .05; b’ = .27, ns).
Nevertheless, the Sobel’s calculation just failed to provide significance, Z = 1.90, p = .058, ns). This will say that the indirect effect just failed to be significantly different from zero. Approximately 17% of the variance in WTC was accounted for by the predictors (R² = .17). For each additional point on the scale of MCS sell, WTC increases by 0.27, whereas with DC benefits WTC increases again slightly more by 0.34. For all these effects other independent variables are assumed to be held constant.
Discussion
The purpose of this study was to examine three different kinds of relationships. First of all, a direct relationship was assumed between perceived DC and the dependent variable of the model, which is WTC. In a next step, a moderation effect of EFL proficiency on the
relationship between perceived DC and WTC was proposed. Finally, two different mediation effects were expected. The first one assumed that perceived DC would mediate the relation between the different five personality traits and WTC. Additionally, the same mediation effect was expected with the independent variable MCS, which in turn consisted of four different variables.
In line with our assumptions it could be confirmed that the perceived DC benefits predicts WTC, even though no concrete literature about this relationship was found in the first place. Therefore, a valuable contribution to research in that field could be made. There exists prove that diversity in the workplace can have a positive influence on the organizational outcome in different ways, such as with higher profits (Herring, 2009). A more direct approach to explain the significant relationship between perceived DC benefits and WTC would be that the reason for why participants of this research perceived the DC as beneficial was because the cultural minority members already adapted themselves to the majority group (Hofhuis, Van der Zee, & Otten, 2015). Therefore, the perception of threat and anxiety related to cultural differences lowers, as well as communication problems and interpersonal conflicts (Fiske, 1998; Milliken & Martins, 1996). One could conclude that less communication
problems and interpersonal conflicts are linked to more WTC, which would in turn clarify the positive relationship between perceived DC benefits and WTC.
Contradictory to hypothesis 1b does a DC perceived as threatening not directly lead to less WTC which would be the opposite effect of what is described in hypothesis 1a. However, plenty of executed research supports that cultural diversity can have a negative effect on
22
individuals and organizations. The most occurring problems caused are among others misperception, miscommunication, increased tension, and conflict (Adler & Gundersen, 2008). Using common sense, it could be concluded that these problems would lead to less WTC. An explanation for this non significant outcome could be that the variable perceived DC threats only consisted in the end of five items instead of eight, because both factor analysis and reliability test showed insufficient results. As the existing scale was already reduced to eight items in the first place, it might not have captured the full concept of this variable, and therefore led to a result which is not significant. On the contrary, it could also be that DC threats is simply less influential than DC benefits. Even though DC threats can have negative impacts as describes earlier, in matters of WTC it has no effect because WTC could be perceived as a more subtle, subconscious variable.
Also the next two hypotheses regarding the moderation effect of EFL on the relationship between perceived DC and WTC had to be rejected. However, EFL showed significant effects with both benefits and threats. When using the variable benefits, EFL and perceived DC benefits (already addressed in hypothesis 1a) showed both significant direct relations with WTC. The association with WTC was weaker for EFL than for perceived DC benefits, which highlights the superior role of DC in comparison to EFL in this specific situation. This is an interesting result in the sense that existing literature showed clear indications for the association between these two variables, whereas no direct prove existed for the link perceived DC and WTC. However, when using the variable DC threats, the relationship with WTC is no longer significant, while the one between EFL and WTC obviously still is. The latter result confirmed what was already expected, namely the
intertwined relationship between language and WTC, referred to as L2 WTC. As a conclusion of these results it can be stated that the influence of a threatening perceived DC on
employee’s WTC is not strong enough to show a meaningful effect. Probably the reason therefore is that the implementation of diversity management strategies in MNCs is already successful which already made the employees accept the situation of a diverse work force (see Hofhuis et al., 2013). Due to the fact that generally spoken diversity within companies is nothing new anymore, threat perception of diversity could have been decreased in the course of time (Morgan, 2009). Besides, the respondents evaluated themselves with an English level that lies between ‘advanced’ and ‘expert’, which means that the overall EFL proficiency was high. Therefore they did not struggle with the language which leads to the conclusion that language anxiety (Alemi, Daftarifard, & Pashmforoosh, 2011), caused by not speaking one’s mother tongue or a language one already masters nearly perfectly, could not affect WTC in a
23
negative way and even had a small positive effect on WTC.
Continuing with the Big Five personality traits, only extraversion presented a full mediation. The link between extraversion and perceived DC benefits was exceptionally based on common sense, as no literature on this topic exists so far. It seemed reasonable that a sociable personality would tend to positively influence WTC. In line with existing research (Oz, 2014; Pavičić Takač & Požega, 2011), extraversion confirmed the relation with WTC. Interesting here is that studies on that topic always included a foreign language component with regard to WTC (L2 WTC), which has not been the case in the current study. This and especially the full mediation effect represent valuable findings. That controlling for the mediator perceived DC benefits removes all of the effect of extraversion, reflects the important role which the DC plays in the current study.
For the rest, the only significant direct effects which could be found were between
neuroticism and WTC, agreeableness and WTC, and finally openness and perceived DC. As already mentioned before, a distinction between trait and state levels of WTC exists
(McCroskey & Richmond, 1990). Whereas state level refers to a certain situation which influences one’s WTC, do the trait antecedents of WTC refer to a person’s character. Among these antecedents are extraversion and neuroticism (MacIntyre et al., 1999). Besides, Oz (2014) already found extraversion and agreeableness to be influencers of WTC in English. Hereby a valuable contribution is made in the sense that these three variables are also able to affect WTC without the important influence of L2. In contrast to Oz’s (2014) findings, openness is not related to WTC. Instead, a significant effect on perceived DC benefits could be detected. This outcome is supported by Hofhuis et al. (2013) who claim that openness towards diversity is followed by a better, more beneficial perception of diversity. Finally, the remaining personality trait conscientiousness is the least related to WTC, as it was already expected when reviewing the literature (see Oz, 2014; Pavičić Takač & Požega, 2011). The second mediation effect which was suggested for MCS had to be rejected for the styles tell, consult and join. Also, no direct effects could be detected. However, after MCS sell was controlled for the influence of perceived DC benefits, it appeared that a full mediation was found, as proposed by Baron and Kenny (1986). Nevertheless, Sobel’s Z could not confirm this finding. Therefore, one should only speak of an indication of mediation. So far, in the context of MCS, primarily employee satisfaction has been discussed (Richmond, McCroskey, & Davis, 1982; Richmond, McCroskey, Davis, & Koontz, 1980). This result leads to the conclusion that MCS, being an external factor in contrast to personality characteristics, has not enough power to influence one’s perception of diversity or WTC,
24
except for the style sell, but rather stimulates satisfaction. This assumption is supported by Donovan and MacIntyre (2004) who discovered a significant effect on WTC of the intrinsic determinants age and gender. This supports again that the intrinsic aspects are probably more influential than the extrinsic ones. Nevertheless, it appears that MCS sell seems to have direct effects on both WTC and perceived DC benefits, as well as a indication of mediation. As MCS Sell is a downward communication style, it is surprising and therefore even more
interesting that it affects WTC, as one would rather think that a two-way communication style (consult, join) between an employee and his supervisor stimulates WTC.
Limitations
After discussing the findings of the study, several limitations need to be mentioned. First of all, the participants’ job positions are not taken into consideration. This could have an impact when thinking about the MCS. A respondent who is a manager himself, obviously is less or not affected at all by any kind of MCS. Therefore, no effects on other variables could be found in this case. However, the chances that the majority of the respondents hold manager positions are proportionally small.
Additionally, it has to be considered that the sample size was still relatively small (N = 110) and the actual number of participation (N = 200+) could not be achieved. A higher amount of participants also might have had an effect on the research results in general. Especially in the case of the non-significant value for Sobel’s Z, it could be that the sample size was too small (Watson, 2015). Plus, as never more than four employees from the same company filled in the survey, it is impossible to give these MNCs concrete feedback for their company, taken into consideration how their employees perceive the MCS and diversity climate. The high diversity of countries where the MNCs are located, as well as the different sectors these companies act in, could also affect the generalization of the research outcomes. Finally, it was not explicitly asked in the survey if English was the participant’s native language or not. However, because their English proficiency was determined, ranging from ‘beginner’ to ‘expert’, and the outcomes were not significant anyway, this did not affect the research outcomes in the end. However, it could be that participants ranged themselves too high or too low with regard to their language proficiency. If this was the case, then this could have had an influence on perceived DC and WTC. Not only could have participants
misjudged their language skills, it also cannot be excluded that respondents have answered questions, such as about DC perception, in a way that would be favorably evaluated by the social environment.
25
With regard to the surprising effect of MCS sell, it has to be taken into consideration that only two items tested this variable in the survey in contrast to the other styles, which consisted of three or four items. This could have influenced the statistical outcomes.
Theoretical Implications & Future Research
This study provides several insights which lead to implications for future research.
First of all, it was confirmed that a beneficial DC positively relates to WTC whereas it could not be confirmed that a DC perceived as threatening negatively relates to WTC. This
emphasizes the power of a beneficial DC for MNCs in order to improve employee’s WTC. For future research about antecedents of WTC, a positive DC should be considered as an important factor. Besides, a focus on the variable perceived DC benefits should be made in relation to WTC in the future, as the present study consisted of several variables which led to the unfortunate effect of a limited attention towards that variable. Only a handful items were available to test it.
EFL, or more in general L2, is a widely discussed concept in matters of WTC. When revising existing literature of EFL again, it appeared that publications are limited to the field of linguistics and language education, whereby one focus lies for example on emotional intelligence and foreign language anxiety (Shao,Yu, & Ji, 2013). Additionally, diversity is also taken into account in the education process as language and culture are interrelated (Gulko Morse & Torres, 2008). Nevertheless a link to DC in companies is an effect which is now discovered for the first time. A high EFL proficiency for expatriates therefore gains additional relevance. Not only is language proficiency important for executing an employee’s work properly, it also influences to some point the relevance of a MNC’s diversity
management. Thus, future research should make sure to always test the proficiency in a foreign language and not just rely on the variable itself.
Another important aspect is that in general a person’s character can only moderately predict the perception of a DC (except for the cases of openness or extraversion – DC benefits), which emphasizes the external role of a well established diversity management instead. Contrary to the little impact of personality on the DC did the traits extraversion, agreeableness and neuroticism lead to effects in WTC, which reveals a rather strong influence of a person’s personality as a whole on WTC. Therefore, less attention on further researching the relationship between personality and WTC (in a L2) should be given. Considering that conscientiousness did not show any effect, the suggestion can be made that in future research conscientiousness will be excluded when examining effects on WTC (in L2). It is far more
26
reasonable to relate it with academic achievement (Oz, 2014; Komarraju et al., 2011). Finally, MCS tell, consult, and join did not show any effect on WTC, which is surprising in the sense that this means that both a straight top-down communication style (Tell) and a two-way communication style (Consult, Join) do not influence WTC. Probably, MCS rather has an effect on employee satisfaction, as mentioned earlier. Nevertheless, one mediation effect was found with the sub-variable ‘Sell’, also being a top-down
communication style. These contradictory results ask in the first place for further research, especially, because only an indication for mediation was found. Besides, one would expect to also find a mediation effect for ‘Tell’. Both styles are top-down oriented; however does MCS ‘Sell’ ascribe a bit of authority to the managers who receive decisions from the top
management. Future research should investigate more specific on the mediation effect of the four styles by measuring more items respectively. This is especially important as the
outcomes in the current situation for ‘Sell’ seem to result by chance.
Conclusion
In conclusion, the model of this research delivered several insights on the relationship between the variables presented. The relevance comes with the establishment of MNCs worldwide. Even though nowadays MNCs are nothing new anymore, it should not be underestimated that challenges come along with this development. Diversity and its positive perception are very important in order to ensure proper communication among employees. That was illustrated by the positive link between DC and WTC. The Big Five proved to have a predominant influence on WTC, especially with the mediation effect of extraversion. However this should not be overestimated when recruiting new employees. There are other antecedents, such as age and gender for example (Donovan & MacIntyre, 2004), which influence WTC as well. With regard to EFL it can be concluded that a high proficiency positively contributes to WTC. MNCs should be advised to ensure that their (future) employees master the language well. This can be tested through personal job interviews or even a language test as part of an assessment center. Finally, MCS should be considered with caution, as only one out of the four styles showed significance. Overall, it should definitely be ascribed less relevance to MCS with regard to DC and WTC. After the different variables which were supposed to affect directly or indirectly WTC were investigated, it should be acknowledged that perceived DC must receive most of the attention by MNCs. Not only was it used to derive relationships which have not been studied so far, it actually proved to be a strong influencer.
27 References
Adler, N. J., & Gundersen, A. (2008). International dimensions of organizational behavior. Cincinnati, OH: South-Western College Publishing.
Alemi, M., Daftarifard, P., & Pashmforoosh, R. (2011). The Impact of language anxiety and language proficiency in EFL context. Cross-Cultural Communication, 7(3), 150–166. Allen, B. (1995). “Diversity” and organizational communication. Journal of Applied
Communication Research, 23(2), 143–155.
Atuahene-Gima, K., & Li, H. (2002). When does trust matter? Antecedents and contingent effects of supervisee trust on performance in selling new products in China and the United States. Journal of Marketing, 66(3), 61–81.
Baghaei, P. (2013). Development and psychometric evaluation of a multidimensional scale of willingness to communicate in a foreign language. European Journal of Psychology of Education, 28(3), 1087–1103.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.
Blau, P. L. (1964). Exchange and power in social life. New York: John Wiley. Burgoon, J. K. (1976). The unwillingness-to-communicate scale: Development and
validation. Communication Monographs. 43(1), 60–69.
Carrison, D. (2010). Does your management style inhibit communication? Industrial Management, 52(5), 6.
Chory, R., & McCroskey, J. (1999). The relationship between teacher management
communication style and affective learning. Communication Quarterly, 47(1), 1–11. Clement, R., & Baker, S.C., & MacIntyre, P. D. J. (2003). Willingness to communicate in a
second language: The effects of context, norms, and vitality. Journal of Language and Social Psychology, 22(2), 190–209.
28
Council of Europe. (2001). Common European framework of reference for languages:
Learning, teaching, assessment. Cambridge, U.K: Press Syndicate of the University of Cambridge.
Cox, T.H., & Blake, S. (1991). Managing cultural diversity: implications for organizational competitiveness. The Executive 5(3), 45–56.
Daniel, J., & Smitherman, G. (1976). How I got over: Communication dynamics in the black community. Quarterly Journal of Speech, 62(1), 26–39.
De Vries, R. E., Bakker-Pieper, A., & Oostenveld, W. (2010). Leadership = communication? The relations of leaders’ communication styles with leadership styles, knowledge sharing and leadership outcomes. Journal of Business and Psychology, 25(3), 367– 380.
De Vries, R. E., Van den Hooff, B., & De Ridder, J. A. (2006). Explaining knowledge sharing: The role of team communication styles, job satisfaction, and performance beliefs. Communication Research, 33(2), 115–135.
Donovan, L. A., & MacIntyre, P. D. (2004). Age and sex differences in willingness to communicate, communication apprehension, and self‐perceived competence. Communication Research Reports, 21(4), 420–427.
Dörnyei, Z. (2005). The psychology of the language learner. New York: Erlbaum.
Fiske, S. T. (1998). Stereotyping, prejudice and discrimination. In D.T. Gilbert, S.T. Fiske, & G. Lindzey, (Eds.), Handbook of social psychology (pp. 357– 411). Boston, MA: McGraw-Hill.
Froese, F.J., Peltokorpi, V., & Ko, K.A. (2012). The influence of intercultural communication on cross-cultural adjustment and work attitudes: Foreign workers in South Korea. International Journal of Intercultural Relations, 36(3), 331–342.
Gallagher, H.C. (2013). Willingness to Communicate and Cross-cultural Adaptation: L2 Communication and Acculturative Stress as Transaction. Applied Linguistics, 34(1), 53–73.
Gardner, R. C. (1985). Social psychology and second language learning: The role of attitudes and motivation. London: Edward Arnold.
29
Ghonsooly, B., Khajavy, G. H., & Asadpour, S. F. (2012). Willingness to communicate in English among Iranian non-English major university students. Journal of Language and Social Psychology, 31(2), 197–211.
Grojean, M. W., Resick, C. J., Dickson, M. W., & Smith, D. B. (2004). Leaders, values, and organizational climate: Examining leadership strategies for establishing an
organizational climate regarding ethics. Journal of Business Ethics, 55(3), 223–241. Gulko Morse, K., & Torres, R.L. (2008). The Cultural Dimension in Focus: Promoting Awareness of Diversity and Respect for Difference in a Finland–Swedish EFL Classroom. The Modern Language Journal, 92(2), 336–337.
Haakonsson, D., Burton, R. M., Obel, B., & Lauridsen, J. (2008). How failure to align organizational climate and leadership style affects performance. Management Decision, 46(3), 406–432.
Herring, C. (2009). Does diversity pay? Race, gender, and the business case for diversity. American Sociological Review, 74(2), 208–224.
Hofhuis, J., van der Zee, K. I., & Otten, S. (2013). Measuring employee perception on the effects of cultural diversity at work: development of the Benefits and Threats of Diversity Scale. Quality & Quantity, 49(1), 177–201.
Hofhuis, J., van der Zee, K. I., & Otten, S. (2015). Dealing with differences: the impact of perceived diversity outcomes on selection and assessment of minority candidates. The International Journal of Human Resource Management, 20, 1–21.
Holden, N. J. (2002). Cross-cultural management: A knowledge management perspective. London: Financial Times/Prentice Hall.
Homan, A. C., Hollenbeck, J. R., Humphrey, S. E., Van Knippenberg, D., Ilgen, D. R., & Van Kleef, G.A. (2008). Facing differences with an open mind: Openness to experience, salience of intra-group differences, and performance of diverse work groups. Academy of Management Journal, 51(6), 1204–1222.
Inoue, T. (2003). An overview of public relations in Japan and the self-correction concept. In K. Sriramesh, & D. Verčič (Eds.), The global public relations handbook: Theory, research, and practice. Mahwah: Erlbaum.
