1
The influence of cultural intelligence on knowledge sharing in multinational
project teams, considering the mediating role of knowledge sharing self-efficacy
and the moderating role of empowering leadership.
Sterre de Geus
University of Amsterdam
Author: Sterre Maria de Geus
Student ID: 11378565
Final version, submission date: June 19, 2018
MSc Business Administration | Track: International Management
University of Amsterdam
Supervisor: dr. Mashiho Mihalache
2 Statement of Originality
This document is written by Sterre de Geus, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Sterre Maria de Geus
Acknowledgements
With a lot of fun I worked on this master thesis. Both in terms of conducting academic research and in terms of content, I have learned a lot and I enjoyed the process. I would like to thank my supervisor Mashiho Mihalache for her support during the process. I also want to thank the Dutch MNE who was willing to serve as a research case for this thesis.
Also, I want to thank my friends Marius Hoogeveen and Celine Amoureus for providing me with advise and support, and my parents for their faith and their good care.
3 Table of contents
1. Abstract ... 4
2. Introduction ... 5
3. Literature review ... 10
3.1 Knowledge sharing in multinational enterprises ... 10
3.2 Cultural intelligence ... 13
3.3 Knowledge sharing self-efficacy ... 17
3.4 Leadership and international knowledge management ... 18
3.4.1 Empowering leadership ... 19
4. Theoretical framework ... 22
4.1 Cultural intelligence and knowledge sharing ... 22
4.2 Knowledge sharing self-efficacy as mediator ... 24
4.3 Empowering leadership as moderator ... 26
4.4 Conceptual model ... 28
5. Methodology ... 28
5.1 Design and sample ... 29
5.2 Measures ... 29 5.3 Reliability ... 32 5.4 Statistical strategy ... 34 5.5 Descriptive statistics ... 35 6. Results ... 36 6.1 Correlation ... 36
6.2 Cultural intelligence and knowledge sharing ... 39
6.3 The mediating effect of knowledge sharing self-efficacy ... 41
6.4 The moderating effect of empowering leadership ... 44
7. Discussion ... 44
7.1 Findings ... 45
7.2 Academic relevance and managerial implications ... 48
7.3 Limitations of this research ... 50
7.4 Future research ... 51
8. Conclusion ... 52
9. References ... 55
4 1. Abstract
Given the global presence of a knowledge economy, sharing knowledge has become a key task for multinational enterprises (MNEs). However, knowledge sharing in multinational and therefore multicultural teams does not occur spontaneously. Cross-cultural differences between employees may lead to frictions and misunderstandings in communication and knowledge sharing. In order to overcome this impediments, cultural intelligence (CQ) might play a crucial role, as employees who are able to deal with situations characterized by cultural diversity are likely to communicate more easily across borders. Studies on CQ have already found a positive influence of CQ on knowledge sharing. However, these studies were focused on a single-country context, neglecting national borders across MNE networks. This research hypothesises that the positive influence of CQ on knowledge sharing counts also within a multinational context. In addition, the mediating role of knowledge sharing self-efficacy, as a crucial determinant to human actions, and the moderating role of empowering leadership, as a crucial enabler to a knowledge sharing environment, are examined and discussed in this research.
Using a sample of 69 IT project team members from multinational project teams in three different countries, this study found no significant influence of CQ on knowledge sharing. Therefore, the expected mediating and moderating effects were rejected as well. However, this study found a significant and positive relationship between empowering leadership and knowledge sharing. Also, in this study CQ is found to have a positive influence on knowledge sharing self-efficacy, and in turn: knowledge sharing self-efficacy is found to positively influence knowledge sharing. Therefore, empowering leadership and knowledge sharing self-efficacy are argued as antecedents to actual knowledge sharing in multinational project teams.
Key words: knowledge sharing, cultural intelligence, knowledge sharing self-efficacy, empowering leadership.
5 2. Introduction
In the age of a knowledge economy, knowledge is considered as the key wealth-creating asset that can provide a firm with a sustainable competitive advantage (Grant, 1996; Cabrera et al., 2006; Xue et al., 2011; Wu and Lee, 2017). In order to perform well, a firm should efficiently develop, leverage and share knowledge embedded in its network (Grant, 2001; Birkinshaw & Sheehan, 2002; Monteiro et al., 2008; Chen, 2017). Knowledge sharing therefore, is a central managerial task within a firm (Birkinshaw & Sheehan, 2002). The concept of knowledge sharing has received a lot of attention by academic researchers for the past decades, seeking for various antecedents that enhance knowledge sharing, such as trust, motivation, efficacy, and leadership (Cabrera et al., 2006; Gagné, 2009; Wang et al., 2010; Wu & Lee, 2017). Most researchers on knowledge sharing focus on a single-country context (Wang et al., 2010), potentially overlooking the expanded presence of multinational companies within a global business environment (Dunning & Lundan, 2008).
Due to the increased presence of a world economy, an increasing number of organizations becomes multinational enterprises (MNEs), operating in multiple countries and consisting of employees with different national and cultural backgrounds (Dunning & Lundan, 2008; Bartlett & Ghoshal, 2009). Today, our 21st century business world is dominated by MNEs: there are more than 100,000 MNEs, operating with more than 900,000 affiliates worldwide. Particularly, more than 40 percent of these foreign affiliates consist of foreign employees, emphasizing the multicultural character of MNEs (UNCTAD, 2016). This multicultural identity of MNE employees possesses challenges for knowledge sharing, since cultural differences between employees may lead to frictions and resistance to knowledge sharing (Ghemawat & Altman, 2014; Hutzschenreuter et al., 2016). In this sense, borders and distance between countries still matter for MNEs (Ghemawat & Altman, 2014).
6 Given that knowledge sharing by individuals is crucial to the performance of an MNE (Grant, 1996; Wu and Lee, 2017), there lies a challenge in understanding how MNEs can improve knowledge sharing within their multicultural network. This research responds to this challenge by examining how cross-border knowledge sharing can be enhanced, applying the concepts of cultural intelligence, knowledge sharing self-efficacy, and empowering leadership to the concept of knowledge sharing in multinational project teams.
In order to overcome the impediments of cross-cultural distance to knowledge sharing, the concept of cultural intelligence (CQ) may play a crucial role. This concept is relatively new to the research on intercultural competences, and refers to an individual’s capability to adapt to situations characterized by cultural diversity (Van Dyne et al., 2017). Various researchers suggest that culturally intelligent employees are more likely to engage in knowledge sharing activities in multinational teams (Thomas et al., 2008; Chen & Lin, 2013). However, most researchers investigate the concept of CQ in a single-country context, potentially overlooking the influence of national borders. This research aims to examine the applicability of the influence of CQ on knowledge sharing to a multiple-country context.
Related to the concept of cultural intelligence, is the psychological mechanism of self-efficacy. Self-efficacy refers to an individual’s belief in his or her own capabilities to accomplish a certain goal (Bandura, 1977). Researchers show that the probability of knowledge sharing behaviour increases when employees are convinced that they are in the position to contribute to the organization’s performance, that is, when they have a high self-efficacy (Hsu et al., 2007;
Wu & Lee, 2017). In addition, various researchers suggest that CQ is positively associated with self-efficacy, so that individuals who are culturally intelligent, feel motivated and self-confident to engage in activities characterized by cultural diversity (Ott & Michailova, 2018), such as knowledge sharing across borders. Therefore, within this research, knowledge sharing self-efficacy is expected to have a mediating role between CQ and knowledge sharing.
7 Next to psychological factors, is effective leadership crucial in order to ensure a knowledge-sharing environment. Leaders are in the position to stimulate and to motivate team members to share knowledge (Wu and Lee, 2017). Since various studies show a positive relationship between transformational leadership and knowledge sharing, researchers call for studies on divergent leadership styles other than transformational leadership (Monteiro et al., 2008; Nemanich and Vera, 2009; Lee et al., 2010; Wu and Lee, 2017). This research responds to this call by analysing the moderating influence of empowering leadership on the relationship between cultural intelligence and knowledge sharing. Empowering leadership refers to a leadership style in which a leader stimulates active participation and self-management by their team members (Conger, 1989; Srivastava et al., 2006; Wu & Lee, 2017).
This research aims to address three gaps in the literature. First, the prevailing literature on the influence of CQ on knowledge sharing potentially overlooks the influence of national borders within an MNE network. Although Chen & Lin (2013) and Froese et al. (2016) found that CQ is positively related to knowledge sharing within a given country, they ask future researchers to focus on context-typical cases and to investigate the influence of CQ on knowledge sharing in other, non-Asian countries (Chen & Lin, 2013; Froese et al., 2016). Moreover, since national borders may play a role in knowledge sharing processes (Ghemawat & Altman, 2014), it is important to also consider the influence of CQ on knowledge sharing in an MNE network in a multiple-country setting. Hence, this research contributes to the literature by using a case of multiple countries in Europe, and it examines whether CQ enhances knowledge sharing also within a multiple-country context.
A second gap within the literature concerns the mediating influence of knowledge sharing self-efficacy. Considering that most research on knowledge sharing is focused on an organizational or team level, the concept of efficacy is mostly considered as a team matter (team-efficacy) (Srivastava et al., 2006; Chen & Lin, 2013). Therefore, efficacy on the individual level might
8 have been overlooked, and various researchers call for more attention to individual characteristics with regard to knowledge sharing (Wang et al., 2010; Michailova & Mustaffa, 2012). For this reason, this research considers the concept of knowledge sharing self-efficacy on an individual level. In addition, this research extends the literature on the underexposed mediating role of self-efficacy with regard to knowledge sharing. Literature shows that there is a growing interest in the mediating role of self-efficacy on work-related issues (Judge & Bono, 2001; MacNab & Worthley, 2012). However, the mediating role of self-efficacy on the relationship between CQ and knowledge sharing is not discussed yet.
A third gap in the literature concerns the moderating role of empowering leadership. Although various authors show the positive influence of empowering leadership on knowledge sharing (Srivastava et al., Xue et al., 2011; Lee et al., 2014; Wu and Lee, 2017), the moderating influence of empowering leadership on the relationship between CQ and knowledge sharing has never been discussed. This underexposed moderating influence of empowering leadership, together with the demand for a consideration of the influence of different leadership styles other than transformational leadership on knowledge sharing (Wu & Lee, 2017), stress the relevance of examining the moderating influence of empowering leadership on the relationship between CQ and knowledge sharing. Moreover, empowering leadership has only been examined in a single country context, potentially overlooking the role of national borders within an MNE network. Hence, this research contributes to the literature by considering empowering leadership as a moderating variable in a multiple country context.
Even though knowledge sharing is a fundamental task within an organizational network, it does not occur spontaneously (Srivastava et al., 2006; Wu and Lee, 2017). Knowledge sharing processes include convoluted human behaviours and depends on the willingness of individuals to actually engage in knowledge sharing (Wu & Lee, 2017). Especially for firms operating in multiple countries and working with multinational organizational units, effective knowledge
9 sharing can be seen as a key challenge (Ghemawat & Altma, 2014). Therefore, this quantitative research, using a survey design, tries to reveal antecedents to knowledge sharing behaviour among different organizational members in multinational project teams situated in multiple countries. Using a case study of a Dutch mid-sized MNE, and a cross-sectional survey design, 69 project team members across three different countries are employed in this research. Considering the mechanisms as described above, the research question within this research is as follows:
What is the influence of cultural intelligence on knowledge sharing behaviour in multinational project teams, and to what extent does knowledge sharing self-efficacy mediate this relationship and does empowering leadership moderate this relationship?
This thesis is structured as follows. In the upcoming section, the literature on knowledge sharing within MNEs, on cultural intelligence, on self-efficacy, and on empowering leadership is discussed in order to provide a theoretical base for this research. In the fourth section, the theoretical framework and the hypotheses of this research are discussed. In the fifth section, the methodology of this research is presented after which the data collection, the data analysis, and the statistical tests are discussed. The sixth section presents the results of the quantitative data analysis and briefly discusses the support or the rejection of the hypotheses. The seventh section discusses the interpretations of the results, including the academic contributions, managerial implications, and limitations of this research. The eighth section comprises a conclusion of this research.
10 3. Literature review
3.1 Knowledge sharing in multinational enterprises
The belief that knowledge is a critical asset for MNEs to achieve a sustainable competitive advantage, is not new. The importance of sharing knowledge within an MNE network can be traced back to Hymer (1960), who introduce the idea that MNEs create value by internalizing their acquired knowledge. In doing so, knowledge sharing contributes to the integration of activities, and to the development of new ideas and business opportunities (Hymer, 1960; Grant, 1996). In this sense, one could even argue that an MNE exists because it can exploit and transfer knowledge more efficiently within its internal network than through the external market (Grant, 1996; Peng, 2001). Knowledge sharing therefore, is crucial to the organizational performance of an MNE (Grant, 1996; Fey & Furu, 2008; Monteiro, 2008).
Essentially, knowledge sharing within MNE networks consists of both knowledge collecting and knowledge donating, and it includes for example team discussions, updating other team members on new tasks, or sharing best practices with other team members (Van den Hooff & De Ridder, 2004; Lee Endres et al., 2007). Within this study, knowledge sharing is defined as team members sharing task-relevant ideas, information and suggestions with each other (Srivastava et al., 2006: 1239; Lee et al., 2014: 367).
While efficient knowledge sharing improves firm performance, does inadequate knowledge sharing impede the development and the transfer of knowledge in organizations. This in turn, may negatively affect organizational performance. Therefore, it is crucial to understand which factors enhance and which factors inhibit knowledge sharing behaviour (Xue et al., 2011).
Despite the shared belief that knowledge sharing improves firm performance, not all employees are equally engaged in knowledge sharing activities. Knowledge sharing within an MNE network typically takes place between highly capable employees who are ‘’in crowd’’,
11 compared to an isolated minority which is rarely engaged in knowledge sharing activities. This isolated minority is found to underperform other organizational units, stressing the importance of individual knowledge sharing as crucial asset for an MNE’s performance (Grant, 1996; Monteiro et al., 2008). Thus, even though knowledge sharing is a fundamental task within the international network of an MNE, it does not occur spontaneously (Wu and Lee, 2017). Knowledge sharing is a complex process that involves convoluted human behaviours. Therefore, various studies seek for underlying mechanisms that improve knowledge sharing within MNEs. In most of these studies, the focus is on the organizational or group level (Wang et al., 2010,). Bharadwaj & Saxena (2005) for example, discuss the difficulties of team knowledge management and the role of leadership and communication practices (Bharadwaj & Saxena, 2005). Also Srivastava et al. (2006), consider knowledge sharing as a team process, stressing the influence of leadership on team knowledge sharing (Srivastava et al., 2006).
The extensive attention to knowledge sharing on the organizational or team level, make researchers stressing the lack of attention to how individual characteristics affect knowledge sharing within an MNE network (Wang et al., 2010). After all, effective knowledge sharing depends on the willingness and the ability of individuals to share their own knowledge and to assimilate the knowledge of others (Xue et al., 2001; Lee Endres, 2007; Wu and Lee, 2017). Knowledge sharing therefore, is not a group matter. It is a possession that resides within the individual, and effective knowledge sharing requires that all individuals within the MNE network are aware of the knowledge possessed by themselves as well as by other organizational members. Thus, individuals should understand the behaviour and the messages from other organizational members, and they should be aware of the importance of sharing knowledge within the organization they work for (Grant, 1996; Peng, 2001).
The biggest challenge in advancing knowledge sharing therefore, is to increase the willingness of individuals to engage in knowledge sharing with other team members. Researchers on
12 knowledge sharing find various antecedents that encourage and motivate employees to share knowledge, such as a feeling of trust, tight personal relationships, having interpersonal similarities, and sharing a common vision (Osterloh & Frey, 2000; Makela et al., 2007; Fey, 2008). A crucial psychological determinant to the willingness to share knowledge, is a feeling of confidence to share knowledge and a feeling of trust in one’s own capabilities (Hsu et al.,
2007). This belief in individual capabilities to reach a certain level of performance plays a crucial role in the actual knowledge sharing by employees (Hsu et al., 2007; Lee Endres et al., 2007), and will be touched upon more in depth in this research.
Next to challenges with regard to stimulating individual knowledge sharing behaviour, does the increasing presence of MNEs in multiple countries, and the multicultural network of MNEs cause challenges with regard to cross-cultural communication (Wang and Noe, 2010). Cross-border knowledge sharing is complicated by culture, so that cultural distance between organizational units may pose difficulties to share knowledge, potentially resulting in frictions and misunderstandings between individuals (Ang et al., 2007; Hutzschenreuter et al., 2016). Additionally, the fact that an MNE’s operations and its knowledge sources are spread over geographically dispersed locations, may lead to difficulties to transfer and to coordinate knowledge across national borders (Mihalache et al., 2012). For this reasons, it is important for both academics and practitioners, to understand which individual characteristics and which external drivers are related to cross-border knowledge sharing. After all, organizational members who are likely to share knowledge across borders, will provide an MNE with a valuable resource (Chen et al., 2013). Here the concept of cultural intelligence might play a crucial role, since individuals who can easily adapt to other cultures, might also easily engage in knowledge sharing with team members from other cultures (Chen & Lin, 2013). In the upcoming section this concept will be discussed.
13 3.2 Cultural intelligence
Many 21st century organizations are multinational and therefore multicultural organizations, consisting of people with different cultural backgrounds (Triandis, 2006; UNCTAD, 2016). This makes that individuals who were historically foreign to each other in terms of culture, are forced to cooperate and to share knowledge within the network of an MNE (Montagliani and Giacalone, 1998; Raver and Van Dyke, 2017). Unfortunately, cross-border communication is not easy, since cultural distance between countries may withhold people to share knowledge. Cultural distance refers to the cultural differences between country pairs in terms of language, norms and values, and the way people interact with each other (Hofstede, 1984; Hutzschenreuter et al., 2016). A high cultural distance between people from different countries may create frictions and misunderstandings between people, which impedes knowledge sharing in multinational teams (Chang et al., 2015; Ghemawat & Altman, 2016; Hutzschenreuter et al., 2016). Therefore, there is a crucial need for managers and employees who can handle the complexities of intercultural interactions, and who can communicate across borders easily (Raver and Van Dyne, 2017). Here the concept of cultural intelligence may play a crucial role.
Cultural intelligence (CQ) refers to an individual’s capability to adapt effectively to a context
characterized by cultural diversity (Chen and Lin, 2013; Froese et al., 2016). People who are culturally intelligent have the ability to think and act in intercultural appropriate ways, which makes them capable to communicate across borders easily (Raver and Van Dyne, 2017). Explicitly, culturally intelligent people expect that misunderstandings in cross-cultural interactions will happen, and they postpone their judgement of any situation until they achieve complete understandings (Ott & Michailova, 2018). Thus, not only do culturally intelligent people cognitively understand the culture of their foreign colleagues, they are also able to physically adjust their behaviour and their body language in an unfamiliar cultural context.
14 Cultural intelligence is a relative new concept within the academic literature on cross-cultural communication (Raver and Van Dyne, 2017). However, considering the increasing presence of a world economy, both within academic literature and with regard to management practices, the concept of CQ evolved rapidly. Various researchers prove the vital importance of CQ in cross-border activities and in intercultural communication (Earley and Mosakowski, 2004; Van Dyne et al., 2017). Also, CQ is found to have a predictive validity to various studies discussing the advantages of CQ to firm performance, leadership, knowledge sharing, and other outcomes (Ang & Van Dyne, 2015). Consequently, the acquired evidence of the predictability of CQ, has pushed the concept from a theoretical perspective into a practical framework that organizations worldwide use for selection, training and organizational development (Van Dyne et al., 2017).
Research shows that CQ substantially differs from other personal traits and competencies. Compared to personality traits, CQ is an ability – rather than a trait – targeted at unfamiliar cultural contexts. Unlike other types of intelligence, such as emotional intelligence and general cognitive ability, CQ is specified to certain contexts, and it includes motivational and behavioural elements of intelligence (Ang & Van Dyne, 2015).
In a world where cross-border activities are routine, CQ is a vitally important skill (Earley and Mosakowski, 2004). Having employees with a high level of CQ is advantageous to an MNE in various ways. Raver & Van Dyne for example, state that CQ is positively related to psychological well-being, leadership effectiveness, and firm performance in general (Raver & Van Dyne, 2017). Also with regard to offshoring, CQ is argued to be a valuable capability to possess, since expatriates with a high CQ have strong cross-cultural negotiation skills, and a high intrinsic motivation to cooperate with inpatriates (Froese et al., 2016). Van Dyne et al. (2017) conclude that CQ is vital for most organizations because the world and its cultures are diverse: present-day organizations acknowledge the crucial value of bridging cultures for
15 organizational success. Hence, having employees with a high level of CQ is a valuable and rare resource that provides a firm with a competitive advantage (Van Dyne et al., 2017).
On the other hand, having employees who are less culturally intelligent, possesses challenges and difficulties in managing culturally diverse work teams, which in turn may negatively affect firm performance (Earley and Mosakowski, 2004; Chen and Lin, 2013). Therefore, it is important for both academics and practitioners to investigate the antecedents and the outcomes of CQ in multinational project teams.
Initially, Early and Ang (2003) defined CQ as a trichotomy of cognitive, motivational, and behavioural CQ. This triad still prevails within the literature, referring to the idea that CQ encompasses both the head (cognitive), the heart (motivation), and the body (behavioural) of an individual (Early & Ang, 2003; Earley & Mosakowski, 2004; Van Dyne et al., 2017). However, Ang et al. (2007) deepen the concept of CQ by breaking up the cognitive element. They make a sharper distinction between metacognitive CQ and cognitive CQ (Ang et al., 2007; Van Dyne et al., 2017). This four-factor model, including metacognitive CQ, cognitive CQ, motivational CQ, and behavioural CQ, is widely accepted and referred to within the literature on cultural intelligence (Chen and Lin, 2013; Froese et al., 2016; Raver and Van Dyne, 2017). The four elements together determine to what extent an individual is capable to deal with situations characterized by cultural diversity. In order to be culturally intelligent, one should score relatively high on all four constructs (Ang et al., 2007).
Metacognitive CQ as the first component, is the most abstract factor of cultural intelligence and refers to an individual’s cultural awareness before and during interactions with team members
with a different cultural background. Individuals with a high level of metacognitive CQ are mentally capable both to obtain and to understand cultural knowledge (Ang et al., 2006; 2007; Van Dyne et al., 2017). Moreover, they devote attention to reflecting on cultural beliefs and to considering cultural values and norms before and during interactions (Ang et al., 2006).
16 Cognitive CQ as the second component, refers to an individual’s general knowledge about a
culture and its similarities and differences across other cultures. Other than metacognitive CQ, is cognitive CQ about knowledge of specific cultures rather than about awareness of cultural contexts. Individuals with a high level of cognitive CQ not only possess cultural-general knowledge, they also possess cultural-specific knowledge. They for example know how social, legal or economic systems from other cultures work, and they might also be able to speak the same language as people from the other respective culture (Ang et al., 2007; Van Dyne et al., 2017). Metacognitive and cognitive CQ together determine the mental intelligence of an individual with regard to other cultures (Ang et al., 2006).
Motivational CQ, as a third component, represents the drive and the interest of an individual to put all efforts toward learning about interacting in circumstances characterized by cultural diversity. Individuals with a high level of motivational CQ do their best in functioning in intercultural contexts (Ang et al., 2015). They are sure they can deal with the uncertainties and the stress of adjusting to an unfamiliar culture, they feel confident that they can accustom to working conditions influenced by other cultures, and they feel motivated to socialize with people from other cultures (Ang et al., 2007). Behavioural CQ, as the last component of CQ, refers to the actions individuals undertake in situations characterized by cultural diversity. Rather than about what people think of feel in intercultural situations (motivational CQ), is behavioural CQ about a flexibility to interact appropriately with organizational members with a different cultural background. People with a high level of behavioural CQ are capable to change their verbal and non-verbal behaviour when an intercultural interaction requires it (Ang et al., 2007).
Since both CQ and knowledge sharing are related to the emotional belief of individuals in their own capabilities (Earley & Mosakowski, 2004), the concept of self-efficacy is important to consider more in depth. In the next section, this concept will be discussed.
17 3.3 Knowledge sharing self-efficacy
In all work environments, the actions and the attitude of individuals, can be predicted by someone’s self-efficacy. Self-efficacy refers to a person’s belief of his or her own capabilities to accomplish a certain level of performance (Bandura, 1977; 1986). It is a form of self-reflection that determines the decision about what behaviour one should undertake, and the endeavour and persistence one puts forth when challenged by difficulties. In complex work situations, such as in a multicultural work context, self-efficacy is an important factor that determines whether or not an employee genuinely undertakes action (Hsu et al., 2007)
Self-efficacy is considered as the central cognitive mediator in motivational processes: the higher someone’s self-efficacy, the more he or she feels motivated and puts efforts in order to
reach a certain goal (Lee Endres et al., 2007). In a work context, employees with a high level of self-efficacy consider difficult tasks as challenges to be accomplished instead of as threats to be avoided (Bandura, 1994; Judge & Bono, 2001). They deal more effectively with difficult situations because they are persistent in the face of failures (Judge & Bono, 2001; Hsu et al., 2007). Hence, a high level of self-efficacy is positively associated with work-related issues, such as job motivation, job performance, job satisfaction, and knowledge sharing behaviour (Judge & Bono, 2001; Xanthopoulou et al., 2009; Wu & Lee, 2017).
Self-efficacy is a personal belief, and due to external factors, the level of self-efficacy is not always stable over time. Instead, self-efficacy fluctuates with changes in the external environment of an individual. For instance, when someone is tired or faces personal issues, the level of self-efficacy and the confidence this person feels to reach a certain goal, may decrease. On the other hand, personal confidence, persistence, and past successful experiences will positively influence the development of self-efficacy (Bandura, 1982; Lee Endres et al., 2007). Next to this, are a proactive and ingenuine character, and interpersonal skills requirements for a high level of self-efficacy (Sherer et al., 1982; Frese et al., 1996).
18 In a multinational work context, a self-belief in communication capabilities, possessing interpersonal skills and a proactive attitude, and having past experience in multicultural contexts, will support employees to communicate across borders (Hsu et al., 2007; Lee Endres, 2007; Wu and Lee, 2017). Self-efficacy therefore, is a crucial factor that both researchers and international workers should take into account when involved in situations characterized by cultural diversity such as knowledge sharing activities (Hsu et al., 2007).
3.4 Leadership and international knowledge management
Research on leadership emphasizes that culturally diverse teams differ in their leadership forms and expectations. Leader prototypes are culturally determined, and cross-border management requires an understanding of those differences (Holden et al., 2001). However, certain attributes of positive leadership styles are supported universally, in that team members respond positively to leaders who inspire them, who build on intra-team participation, and who make sure that all ideas are shared and heard (Holden et al., 2001). In order to build on new team-created solutions, leaders should actively encourage team members to engage in a regular cross-cultural exchange of ideas and information (Zakaria et al., 2004). According to Holden et al. (2001), managing international knowledge flows requires a new style of management that is more altruistic: managers should encourage and support employees to engage in knowledge sharing activities, rather than steer or command them (Holden, 2001).
Leadership is a crucial predictor of knowledge sharing as leaders have a powerful influence on their team members, who ought to be willing and able to share their knowledge with others. Thus, leaders are in the position to establish a knowledge-sharing environment (Oliver & Kandadi, 2006; Nemanich & Vera, 2009; Wu and Lee, 2017; Xiao et al., 2017). Most studies on the relationship between leadership and knowledge sharing focus on transformational leadership and its influence on knowledge sharing (Gagné, 2009; Nemanich & Vera, 2009; Lin & Hsiao, 2014; Han et al. 2016; Wu & Lee, 2017). This leads to a consensus on the idea that
19 leaders should use a positive leadership style, in which participation and personal relationships are essential in order to stimulate knowledge sharing behaviour. Using an autocratic leadership style on the other hand, is found to inhibit knowledge sharing (Srivastava et al., 2006). In addition, the use of market elements, such as reward systems, may also lead to withholding knowledge (Osterloh and Frey, 2000; Monteiro et al., 2008; De Vries et al., 2010). Considering the focus on transformational leadership, previous studies are potentially overlooking other leadership styles that have a significant impact on knowledge sharing. Therefore, researchers ask future research to investigate more in detail the relationships between other positive leadership styles and knowledge sharing behaviour (Monteiro et al., 2008; Nemanich and Vera, 2009; Lee et al., 2010; Wu and Lee, 2017). As in current organizations work autonomy and is becoming more important, and because managing international knowledge flows requires a new style of management that is more altruistic (Holden, 2001; Wu and Lee, 2017), the empowering force of leaders and their influence on knowledge sharing deserves more attention. And since a feeling of empowerment is found to be crucial with regard to individual knowledge sharing (Xue et al., 2014; Wu and Lee, 2017), it is interesting to consider the concept of empowering leadership in relationship with knowledge sharing.
3.4.1 Empowering leadership
The concept of empowering leadership is introduced by Conger in 1989 and refers to the art of empowering others. Empowering leaders are social and organizational architects who build and sustain trust with their followers; they are not team members, but they are responsible for providing leadership to others. They typically share their power with subordinates and they stimulate active participation and self-management within the organization (Conger, 1989). The definition of empowering leadership used in this research is ‘when leaders share power with
subordinates and when they engage in actions that raise the intrinsic motivation of team members’ (Srivastava et al., 2006: 1240; Wu and Lee, 2017: 478).
20 Within the literature on leadership, the concept of empowerment has deserved a continuous interest and development (Conger, 1989; Ahearne et al., 2005; Vecchio et al., 2010). Previous studies have shown that empowerment is a crucial driver for organizational effectiveness and job satisfaction (Ahearne et al., 2005; Vecchio et al., 2010). An empowering leadership style therefore, is associated with various advantages to organizational performance. In the first place, as empowering leaders share their power with their team members, team members have a greater capacity for self-control and work involvement. This in turn, lead to higher levels of job performance and job satisfaction (Vecchio et al., 2010). Furthermore, as empowering leaders provide team members with care and support, team members are likely to encourage themselves to set higher objectives, to experiment with new methods, and to handle work situations with optimism (Wu & Lee, 2017). Moreover, as empowering leaders allow team members to participate in decision making, team members feel intrinsically motivated and confident to share their opinion and ideas with others (Ahearne et al., 2005; Wu & Lee, 2017).
Arnold et al. (2000) operationalize the concept of empowering leadership as consisting of five components. The first component is leading by example, which refers to the behaviour of a leader that shows commitment to the leader’s own work as well as to the work of his or her
team members. The second component is coaching, referring to the education and the support of a leader to help his or her team members to become self-reliant. The third component is participative decision making, alluding to a leader’s use of the information and the input given by team members when making decisions. The fourth component is showing concern, referring to behaviour that shows a general regard and respect for the well-being of team members. The last component is informing, referring to a leader’s dissemination of important company-wide information, such as the company’s vision, its mission and its strategy (Arnold, 2000).
From 2006 until 2017, four groups of authors discuss empowering leadership with regard to knowledge sharing. In 2006, Srivastava et al. introduce a research into the influence of
21 empowering leadership on effective knowledge sharing processes in hotel properties in the US. They find that empowering leadership is positively associated with knowledge sharing (Srivastava et al., 2006). In 2010, Xue et al. elaborate on the research done by Srivastava et al. and they suggest that empowering leadership will strongly increase the extent of knowledge sharing behaviour among students at a US university. Both Srivastava et al. and Xue et al. ask for future research in order to validate their findings in a real-world organizational setting. In 2014, Lee et al. give a response to this call and investigate the influence of empowering leadership on team performance in IT project teams, using knowledge sharing as a mediating variable. They conclude that empowering leadership facilitates knowledge sharing within IT project teams in Korea (Lee et al., 2014). Recently, Wu and Lee (2017) extend the literature by conforming that empowering leadership in a work group has a cross-level impact on knowledge sharing by organizational members in Taiwan. All previous studies mentioned above use the same Empowering Leadership Questionnaire developed by Arnold et al. (2000) to measure empowering leadership.
An empowering leadership style, implying participative decision-making, shared leadership, and a leader who shows respect for the well-being of team members, might make team members willing to engage in knowledge sharing activities (Arnold, 2000; Wu & Lee, 2017). Within the literature, there is a consensus on the idea that empowering leadership positively affects knowledge sharing within an organization that operates in one country (Srivastava et al., 2006; Xue et al., 2011; Lee et al., 2014; Wu & Lee, 2017). However, the respective American, Korean and Taiwan cultural background of the organizations considered in previous studies might significantly influence knowledge sharing behaviour. Moreover, previous authors focus on knowledge sharing within one country, potentially overlooking the difficulties in knowledge sharing across national borders. Therefore, this comparative research on different – geographically dispersed – project teams within an MNE network will extend the literature.
22 4. Theoretical framework
In this section, the theoretical framework is developed and the corresponding hypotheses are discussed. The first part discusses the direct influence of CQ on knowledge sharing, after which in the second part the mediating influence of knowledge sharing self-efficacy is discussed. The third part provides a clarification of the moderating role of empowering leadership, and in the fourth and last part, the conceptual model is presented.
4.1 Cultural intelligence and knowledge sharing
Given that knowledge sharing by individuals is crucial to the performance of an MNE (Grant, 1996; Wu and Lee, 2017), and given the idea that cross-cultural knowledge sharing requires communication capabilities that diminish the negative effects of cultural distance, their lies a challenge in understanding what individual characteristics make individuals likely to become knowledge sharers within an MNE network. Considering individual competencies in cross-cultural communication, the concept of cross-cultural intelligence, referring to an individual’s capability to adapt effectively to contexts characterized by cultural diversity, is likely to play a crucial role in the actual knowledge sharing within an MNE network (Chen and Lin, 2013). In this study, a high level of CQ is expected to lead to more knowledge sharing by individuals.
According to Chen & Lin (2013), CQ is the primary driver of knowledge sharing in culturally diverse teams. Building on the idea that the impediments of cultural distance in cross-cultural communication diminish when someone knows how to deal with situations characterized by cultural diversity, CQ is likely to enhance knowledge sharing in multinational teams. Metacognitive CQ to start with, refers to a person’s cultural awareness before and during interactions with organizational members with a different cultural background (Ang et al., 2003; Van Dyne et al., 2017). People with a high level of metacognitive CQ are aware of the cultural preferences of others and they adjust their mental status during and after knowledge sharing
23 interactions (Chen & Lin, 2013). Moreover, they are able to abstract knowledge acquired from a specific experience into broader principles which they can use in future interactions (Thomas et al., 2008). The second aspect of CQ is cognitive CQ, referring to the general knowledge an individual has of a given culture and about its similarities and differences across other cultures (Ang et al., 2006; 2015). Cognitive CQ facilitates the self-confidence about knowledge sharing, and this specific intelligence is crucial to effective knowledge sharing (Chen & Lin, 2013).
Next to cognitive abilities, is motivational CQ essential to knowledge sharing. Successful intercultural communication requires the drive and the interest to put all efforts toward learning about interacting in circumstances characterized by cultural differences (Ang et al., 2015; Van Dyne et al., 2017). Organizational members with a high level of motivational CQ are open and eager to adapt to new situations, and to collect and to share knowledge with other organizational members across borders (Chen & Lin, 2013). The last component of CQ is behavioural CQ. People with a high level of behavioural CQ are flexible to interact appropriately with people from other cultures and they are provided with the skills and the competencies to engage in cross-border knowledge sharing (Ang et al., 2007; Chen and Lin, 2013).
Chen and Lin (2013) conclude that CQ positively influences knowledge sharing behaviour. Communication difficulties that occur due to cultural distance are likely to decrease when individuals adjust their mental status during cross-cultural interactions, when they are able to abstract knowledge to broader principles, and when they possess general knowledge about other cultures (Ang et al., 2007; Chen & Lin, 2013). Moreover, cross-cultural knowledge sharing will actually improve when individuals are open and eager to adapt to situations characterized by cultural diversity, and when they are provided with the skills to interact with people from other cultures, such as speaking foreign language(s) (Chen & Lin, 2013; Froese et al., 2016). Hence, for organizational members with a high level of CQ, cultural distance is not immediately a barrier to interact. Thus, project team members who are culturally intelligent are expected to be
24 more engaged in knowledge sharing than organizational members who are less culturally intelligent (Chen and Lin, 2013). This leads to the following hypothesis:
H1: Cultural intelligence will positively influence knowledge sharing in multinational project teams.
4.2 Knowledge sharing self-efficacy as mediator
Self-efficacy serves as personal belief that influences decisions about what actions an individual undertakes (Bandura, 1986; Hsu et al., 2007). Previous research show that employees with high self-efficacy are beneficial to organizational processes and outcomes such as job performance, work engagement, and knowledge sharing (Judge & Bono, 2001; Hsu et al., 2007; Wu and Lee, 2017). With regard to knowledge sharing, knowledge sharing self-efficacy is argued to influence the decision whether or not to share knowledge (Cabrera et al., 2006; Hsu et al., 2007). Therefore, in this research knowledge sharing self-efficacy is considered as a mediating variable on the relationship between CQ and knowledge sharing.
With regard to knowledge sharing, Wu & Lee (2017) conclude that organizational members with high self-efficacy are more confident in facing new challenges and that they are more likely to collect knowledge from other organizational parts (Wu & Lee, 2017). Also, and specifically relevant to knowledge sharing, is the research done by Hsu et al. (2007), who include the concept of knowledge sharing self-efficacy. This concept refers to the confidence an individual possesses to share knowledge with other team members. Hsu et al. find that knowledge sharing self-efficacy is a key driver to knowledge sharing. This is because when people think they contribute to the performance of an organization by sharing knowledge, they will be more willing to actually share knowledge (Cabrera et al., 2006; Hsu et al., 2007). Furthermore, given that individuals with high self-efficacy are confident in dealing with complex work situations, such as in a multicultural work context, self-efficacy is likely to be
25 an important factor that determines whether or not an employee genuinely shares knowledge (Hsu et al., 2007). Therefore, knowledge sharing self-efficacy is likely to positively influence knowledge sharing in multinational teams.
With regard to cross-cultural communication, MacNab & Worthley (2012) and Ott & Michailova (2018) argue that is crucial for employees to possess cognitive, motivational and behavioural abilities to interact and to cooperate effectively with team member of other cultures. That is, when they are culturally intelligent (MacNab & Worthley, 2012; Ott & Michailova, 2018). Various authors have shown that the concept of CQ is positively related to self-efficacy, so that individuals who are culturally intelligent are likely to have a high self-efficacy to engage in cross-cultural activities. The underlying mechanism is that these individuals are provided with the awareness, the knowledge, and the skills that make them feel confident to interact with people from other cultures (Van Dyne et al., 2017). Specifically, CQ is described as a mental ability that facilitates one’s self-confidence to share knowledge (Chen & Lin, 2013). The belief
that CQ and self-efficacy are related concepts (Ott & Michailova, 2018), and the idea that culturally intelligent people feel confident to interact with people from other cultures (Van Dyne et al., 2017), lead to the expectation that culturally intelligent people have a high efficacy to share knowledge. Thus, that CQ positively influences knowledge sharing self-efficacy in multinational teams.
Building on the belief that self-efficacy is a fundamental psychological mechanism that determines whether or not an individual genuinely undertakes action, the concept of knowledge sharing self-efficacy is likely to play a mediating role towards knowledge sharing. Since this specific research is focused on knowledge sharing in multinational and multicultural teams, the ability to deal with situations characterized by cultural diversity is likely to influence the self-efficacy of team members to share knowledge. Considering that culturally intelligent employees have a higher self-efficacy to share knowledge with people from other cultures (Van
26 Dyne et al., 2017), and considering that in turn, this knowledge sharing self-efficacy is likely to lead to actual knowledge sharing (Hsu et al., 2007), the concept of knowledge sharing self-efficacy is treated as a mediating variable. In other words: employees with a high level of CQ, who are provided with the awareness, the skills and the confidence to interact with people from other cultures, are likely to have a high knowledge sharing self-efficacy. This in turn, will encourage and motivate them to actually engage in knowledge sharing with other team members. This leads to the following hypotheses:
H2: The positive relationship between cultural intelligence and knowledge sharing is mediated by knowledge sharing self-efficacy;
H2a: There is a positive relationship between cultural intelligence and knowledge sharing self-efficacy;
H2b: There is a positive relationship between knowledge sharing self-efficacy and knowledge sharing.
4.3 Empowering leadership as moderator
Leadership is found to be an essential predictor of knowledge sharing as leaders are in the position to shape a knowledge sharing environment (Wu and Lee, 2017). In order to create such an environment, leaders should actively encourage team members to engage in a cross-cultural exchange of ideas and information (Zakaria et al., 2004). Given the idea that managing knowledge requires a leadership style that is mostly altruistic and focused on steering rather than commanding, empowering leadership play an essential role in knowledge sharing (Holden, 2001; Wu and Lee, 2017). Empowering leadership therefore, is treated as a moderating variable, strengthening the expected relationship between CQ and knowledge sharing.
Building on the concept as introduced by Conger (1989), various authors have investigated the origins and the consequences of empowering leadership. One of these consequences is that
27 empowering leadership, often referred to as the art of empowering others, is found to be positively related to knowledge sharing within organizations in a given country (Srivastava et al., 2006, Xue et al., 2011; Lee et al., 2014; Wu and Lee, 2017).
Considering that empowering leaders are participative decision makers - using the input given by team members when making decisions - empowering leaders can create a climate in which team members feel receptive and motivated to create and to share knowledge. This is because when team members know they are included in the decision-making process, and that they have a say, they are likely to feel motivated to contribute to the decision making process by sharing their knowledge (Lee et al., 2014). Moreover, empowering leaders provide their team members with relevant information – leading by example - so that in turn, team members may feel encouraged to also share their ideas within the team. Furthermore, the support and the concern empowering leaders show regarding the well-being and the self-reliance of their team members, might make team members feel free to share new ideas and initiatives (Lee et al., 2014; Wu & Lee, 2017). In doing so, an empowering leader helps team members to elicit behaviour that is consistent with a culture of innovation and knowledge sharing. Given that certain attributes of positive leadership styles are universally recommended, such as supporting team members, assuring intra-team participation, and inspiring team members (Holden et al., 2001), both practitioners and academic researchers may endorse an empowering leadership style as stimulating knowledge sharing, irrespective of the country. This universal assumption might be applicable not only to a single country, but to multiple countries and cultures at the same time. Therefore, when a leader is perceived as empowering, the relationship between CQ and knowledge sharing in multinational teams is expected to be stronger. Thus:
H3: An empowering leadership style will positively moderate the relationship between cultural intelligence and knowledge sharing.
28 4.4 Conceptual model
In the figure below the concepts and the expected relationships are presented. The first hypothesis predicts a direct and positive relationship between CQ and knowledge sharing. The second hypothesis predicts that the direct relationship between CQ and knowledge sharing is mediated by knowledge sharing self-efficacy, so that CQ leads to a higher knowledge sharing self-efficacy, which in turn leads to more knowledge sharing. The third hypothesis at last, predicts a moderating effect of empowering leadership, so that when a leader is perceived to be empowering, the relationship between CQ and knowledge sharing is stronger.
Figure 1. Conceptual model
5. Methodology
In this section, the methods used in this research are discussed. First, the design and the sample are explained, after which the measurements of the various variables are discussed. After that, the reliability of the variables is presented, and the statistical strategy employed in this research is discussed. In the last part, the descriptive analysis of all variables analysed in this research is presented and interpreted.
Cultural intelligence Knowledge sharing
Empowering leadership Knowledge sharing self-efficacy H2a (+) H2b (+) H3 (+) H1 (+) H2 (+)
29 5.1 Design and sample
Using a case study of a Dutch mid-sized MNE and a cross-sectional survey design, the relationship between CQ and knowledge sharing, and in addition the mediating influence of knowledge sharing self-efficacy, and the moderating influence of empowering leadership are investigated. The population of interest for this research includes employees from project teams from MNEs. The sample of this research consists of employees working in multinational project teams from an MNE. The employees of the project teams are situated in Romania, France, and The Netherlands, working together across borders. The sample was collected through non-probability judgment sampling. The sample was drawn from employees from international project teams from MNEs, and the respondents were recruited by asking the manager of both projects to request the team members to fill out the survey. This team manager was reached through e-mail contact.
Through an online questionnaire via e-mail (see Appendix A), the project team members filled out questions related to the variables used in this research. This questionnaire was created by the online Qualtrix program, and in the short introduction the guarantee of anonymously and a confident handling of the data was ensured. During the last weeks of April, and the first week of May 2018, the questionnaire was distributed and filled out by the respondents.
5.2 Measures
Knowledge sharing
The dependent variable in this research is knowledge sharing. According to literature, the concept of knowledge sharing consists of both knowledge donating (i.e. a person shares knowledge with or without being asked for it), and knowledge collecting (i.e. a person seeks and asks for knowledge) (Van den Hooff & De Ridder, 2004). In order to measure knowledge donating and knowledge collecting, respectively six and four propositions from an individual
30 perspective are acquired from the study by Van den Hooff & De Ridder (2004). On a seven-point Likert scale, respondents were asked to assess ten propositions (totally agree – totally disagree). Examples regarding knowledge donating are: ‘I share the information I have with colleagues within my project team’, and: ‘I share my skills with colleagues within my project team’. Examples of propositions regarding knowledge collecting are ‘colleagues within my project team tell me what their skills are, when I ask them about it’, and: ‘colleagues outside of my project team tell me what they know, when I ask them about it’.
Cultural intelligence
The independent variable in this research is cultural intelligence. The way in which this concept is most often measured is through a survey, consisting of four parts, comprising the four constructs of CQ: metacognitive CQ, cognitive CQ, motivational CQ, and behavioural CQ. All constructs together determine the extent to which an individual is culturally intelligent (Ang et al., 2007; Chen & Lin, 2013; Froese et al., 2016; Van Dyne et al., 2017). The questions in the survey are indicators of how able a team member is in to think and to act in intercultural appropriate ways. In the survey, respondents were asked to rate four propositions with regard to each factor of CQ, on a seven-point Likert scale (totally agree – totally disagree). This sixteen statements were appointed from an individual perspective, such as: ‘I am conscious of the cultural knowledge I apply to cross-cultural interactions’ (metacognitive CQ), and: ‘I enjoy learning about culture that are unfamiliar to mine’ (motivational CQ). These statements are
derived from the research done by Ang et al. (2007) and Chen & Lin (2013).
Knowledge sharing self-efficacy
Self-efficacy is a form of self-reflection through which an individual’s belief of his or her own capabilities to accomplish a certain level of performance is mirrored (Bandura, 1982). Therefore, a questionnaire using propositions of one’s own perceived capabilities is the most common used measurement to measure self-efficacy (Hsu et al., 2007; Lee Endres et al., 2007).
31 Considering the wide-ranging character of self-efficacy (i.e. the applicability to multiple work-related areas), the concept of self-efficacy is specified to knowledge sharing self-efficacy. This concept is elaborated by Hsu et al. (2007), and refers to the confidence an individual has to share knowledge with other team members (Hsu et al., 2007). Respondents were asked to assess seven propositions with regard to their knowledge sharing self-efficacy on a seven-point Likert scale (totally disagree – totally agree), which are derived from the questionnaire as used by Hsu et al. (2007). These propositions included for example: ‘I feel confident in providing my experiences, insights or expertise as an example’, and: I feel confident in articulating myself in written, verbal or symbolic forms’ (Hsu et al., 2007).
Empowering leadership
The concept of empowering leadership has been related to knowledge sharing in various studies. In this studies, the questionnaire as introduced by Arnold et al. (2000) was mostly used. Therefore, within this research, the propositions as used by Srivastava et al. (2006), Xue et al. (2011), Lee et al. (2014), and Wu and Lee (2017) are used. This propositions are divided by the five components of leadership: leading by example, participative decision making, coaching, informing, and showing concern. Examples of statements are: ‘my team leader works as hard as anyone in my project team’ (leading by example); ‘my team leader listens to my project team member’s ideas and suggestions’ (participative decision making); ‘my team leader explains
rules and expectations to my project team’ (informing), and: ‘my team leader shows concern for project team members’ success’ (showing concern).
Control variables
Within this research, six control variables are used. The first control variable concerns the international work experience a team member has. Among other researchers, Shannon & Begley (2008) and Ng et al. (2012) find that international work experience is positively related to CQ. The longer an employee has worked in another country than their home country, the
32 more likely this person is to be culturally intelligent (Shannon & Begley, 2008; Ng et al., 2012). Within this research, this variable was measured on a three-part scale, using the following proposition: ‘I worked in another country than my home country’. Possible answers were:
yes/no/only on project base for no longer than one month. Later, the variable was made a dummy variable to distillate the respondents with actual international work experience.
The second control variable is the number of years a respective team member is working for the focal organization. Lee (2001) find that the longer employees work for an organization, the more easy they engage in knowledge sharing. This is because long-serving employees are likely to easily find their way through communication canals and to be more confident in approaching colleagues to collect knowledge (Lee, 2001). This variable is measured on an interval scale, asking respondents for how many years they work for the focal organization.
The third control variable is the perceived level of English by a team member. Shannon & Begley (2008) find that in case an employee is proficient in the shared language within an organization, he or she is likely to be more engaged in cross-cultural interactions. This is because a language proficiency facilitates cross-national communication and an understanding of other cultures (Shannon & Begley, 2008). Within this case study, the common language of the MNE is English, so respondents were asked to assess their English proficiency on a seven-point Likert scale (very poor – very good). At last, the fourth, fifth, and sixth control variables concern demographic issues, which are: nationality, age, and gender.
5.3 Reliability
Since both the dependent variable, the independent variable, the mediating variable, and the moderating variable are measured on a seven-points Likert scale, various linear regression analyses are used to test the relationships between the variables. In order to check the assumptions for performing linear regressions, a Kolmogorov Smirnov test is conducted in
33 order to know if the response on the dependent variable (knowledge sharing) is distributed normally. From this test can be concluded that the data is not distributed normally (sig. = 0,025). Therefore, parametric test techniques are less suitable for this dataset. However, it is still possible – yet less likeable to be significant – to use linear regression analyses (Pallant, 2013).
In order to check the reliability of the variables, a statistical reliability check is done. In doing so, Cronbach’s alpha shows the internal consistency of items in a survey to assess the reliability
of a summed variable. In other words, Cronbach’s alpha shows whether the items of a variable actually represent the variable (Cronbach, 1951). A Cronbach’s alpha above 0,7 can be considered as reliable (Santos, 1999). Within this dataset, all variables can be considered as reliable. For knowledge sharing α = 0,885, for cultural intelligence α = 0,912; for knowledge
sharing self-efficacy α = 0,906, and for empowering leadership α = 0,953.
In order to test whether the variables are indeed constructed according to the factors stated in the literature, a factor analysis is performed. According to the Kaiser criterion, the retaining factors with an eigenvalue >1 are most appropriate to represent a variable (Williams et al, 2010).
For knowledge sharing, the factor analysis points out that two factors have an eigenvalue above 1,000. Zooming in on the specific survey questions, we can conclude that SPSS extracts a difference in the factors ‘knowledge sharing inside the project team’ (eigenvalue is 5,051) and ‘knowledge sharing outside the project team’ (eigenvalue is 1,924). For CQ, it is interesting to
see that behavioural CQ has an eigenvalue of 7,1 and that this factor explains 44,375% of the variance in CQ. The other three factors of CQ have an eigenvalue of 1,650 (metacognitive CQ), 1,522 (cognitive CQ), and 0,990 (motivational CQ). Self-efficacy consist of only one factor, which is reflected in an eigenvalue of 4,651. Empowering leadership consists of five factors (Arnold et al., 2000). Interestingly, the factor analysis shows that that one factor has an eigenvalue of 10,681 and that this factor explains 55,883% of the variance in empowering
34 leadership. Hence, it could be possible that the coherence between all five factors is so strong that it is hardly impossible to strictly distinct five factors for empowering leadership.
5.4 Statistical strategy
In order to test the hypotheses, several linear regressions are performed in the statistical programme SPSS. Since all variables in the model are measured on interval scale, linear regressions are most appropriate to examine relationships between two variables (Fields, 2013).
For the first hypothesis, predicting a direct relationship between the independent and the dependent variable, a linear regression analysis is performed while checking for the control variables. In several statistical models, the influence of both the control variables and the independent variable are presented in order to see whether the variance in the dependent variable changes when including the independent variable (Fields, 2013).
For the mediating variable, the method as described by Baron & Kenny (1986) is used. First, the direct effect of the independent variable on the dependent variable should be measured in order to prove that there is an effect to mediate. After that, one should examine whether the mediating variable has a significant effect on the dependent variable while controlling for the independent variable. Third, one should test if the direct influence gets non-significant when controlling for the direct effect. Only if all tests are significant, mediation exists (Baron & Kenny, 1986).
In examining a moderating influence, one is interested in how a given variable affects the direction and/or the strength of a relationship between an independent and a dependent variable (Baron & Kenny, 1986). In order to measure a moderating effect while using interval variables, various steps should be taken. First, the direct influence of the independent variable on the dependent variable should be measured. Second, the moderating variable should be multiplied by the dependent variable in order to create a new variable that is predicted to significantly
35 influence the dependent variable. This variable should be added to a next model in order to see whether the relationship between the independent and the dependent variable changes when considering the moderating variable (Baron & Kenny, 1986).
5.5 Descriptive statistics
In the table below, the descriptive analysis of all variables included in this research is presented.
Table 1. Descriptive statistics
From this table we see that 69 respondents have filled out the questionnaire and that the vast majority of the respondents is male (N = 59). The average age of the respondents is 41,88 and the nationalities of the respondents are distributed quite equally. However, most of the respondents are Dutch. Besides this demographic information, we can extract that the average score on knowledge sharing is 58,88, which is equal to 5,88 on a scale from 1 till 7. Looking at the scores on CQ, we see that the average score on CQ is 78,98, which is equal to 4,936 on a scale from 1 till 7. According to Earley & Mosakowski (2004), an average score of less than 4,2 out of 7 indicates a low level of CQ, whereas a score of greater than 6.3 out of 7 reflects a strong CQ. Therefore, we can conclude that within this research, the average score on CQ (4,936
Variable Min Max Mean S.D. N Freq.
1 Cultural intelligence 37 110 78,98 14,82 69 2 Knowledge sharing 32 70 58,88 8,09 69 3 Self-efficacy 18 49 42,38 6,012 69 4 Empowering leadership 56 133 116,0 15,59 68 5 6 7
International work experience (1 = yes)
Years working for this company
Perceived level of English
0 1 4 1 35 7 0,26 14,34 5,68 0,44 8,14 0,99 69 62 68 18 8 9 10 Dutch nationality French nationality Romanian nationality Age Gender 0 0 0 24 0 1 1 1 63 1 0,39 0,35 0,26 41,88 0,855 0,488 0,434 0,475 9,39 0,355 66 66 66 65 69 26 23 17 59
