Can workplace friendship enhance employees’ innovative behaviour? : A study to the interpersonal relationships that exist in the workplace affecting individual innovative behaviour, looking beyond the predominant relati

(1)

Can workplace friendship enhance

employees’ innovative behaviour?

A study to the interpersonal relationships that exist in the workplace

affecting individual innovative behaviour, looking beyond the

predominant relationships with your supervisor and team members.

Ellen van Diemen (10453253)

Master Thesis

MSc in Business Administration Entrepreneurship and Innovation track

University of Amsterdam

(2)

1 Abstract: Workplace creativity and innovation as exposed by individual employees are crucial to business success, and interaction between people within an organization can foster this desired behaviour. The present research then investigates to what extent different

interpersonal relationships at the workplace affect individual innovative behaviour.

Conducting an one-time self-reported online survey among 134 participants, it was found that workplace friendship has a higher direct effect on innovative behaviour than for high-quality relationships with your supervisor or team members, and together they explained a substantial amount of variance in the preferred behaviour. These findings lend support to the role of workplace friendship in fostering innovative behaviour at work. Practical implications and future research recommendations are also discussed.

Keywords: innovative behaviour, interpersonal relationships, workplace friendship

Statement of Originality

This document is written by Ellen van Diemen who declares to take full responsibility for the contents of this document.

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

references have been used in creating it.

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

(3)

2

1. Introduction, relevance and research purpose

1.1 Employees as those bringing about innovative behaviour

Most organizations in today’s economy are knowledge-based organizations and their survival and success therefore largely depends on the extent to which they use their in-house knowledge in a creative and innovative manner. While enhancing quality and efficiency also is and remains essential for the survival and success of a company, this often does not provide a crucial edge over your competitors (de Pury, 1994). Such competitive advantage is needed to stand out in the more and more dynamic business environment of today, and it is through innovation that an organization generates this advantage, resulting in unique value creation and the achievement of a superior position (Chahal & Bakshi, 2015). So, innovation seems to be the key to any successful business. Since knowledge is embedded in individual employees, managers should recognize that they make a critical difference when it comes to innovation, and thus ultimately business success. Human resource management (HRM) is the practice of overseeing an organization’s employees, and the process of hiring and developing employees so that they become more valuable to the organization. The concept regards people as the most important single asset of the organization (Poole, 1990). It is employees themselves who are valuable to the organization through their creativity, commitment and skills (Druker et al., 1996). As suggested by Sisson (1990), one of the main features increasingly associated with the concept of HRM is the exercise of initiative. Organizations need employees who not only fulfil their formal job requirements, but also exceed their standard work behaviours by engaging in innovative behaviour (Janssen, 2001). In the end, it is individuals who come up with ideas and in a way are responsible themselves for turning those ideas into a reality. As such, organizations benefit from knowing how these processes can be best encouraged.

(5)

4 1.2 Looking beyond the formal interpersonal relationships

The generation, promotion and implementation of ideas inherently contain a social element. Those committed to how to best promote innovative behaviour have hence emphasized the role of interpersonal interaction as a key antecedent variable, which occurs between two or more individuals within the workplace. There has been an increasing interest in the concept of interpersonal interactions as a social phenomenon fostering innovative behaviour. According to Heaphy and Dutton (2008), positive interpersonal connections may result in better individual and work-related outcomes. Positive interactions can encourage interpersonal relationships, and for many working individuals it is from the maturation and maintenance of these relationships they find fulfilment, and an opportunity to fulfil their need to belong (Gardner, Pickett & Brewer, 2000). As argued by Baumeister and Leary (1995), the need to belong is a fundamental human motivation, influencing both involuntary and voluntary behaviours. Examining the features and functions of business relationships, Uzzi (1997) suggest that close relationships, as compared to more transactional market relationships, have positive effects on whether your company is successful. Because of its personal nature trust is created, which allows for a more detailed and tacit information transfer, as well as for more flexible and efficient joint problem-solving behaviour. Closely related to trust, evidence is found that high-quality relationships foster psychological safety (Carmeli, Brueller & Dutton, 2009; Carmeli & Gittell, 2009) and thus enable organizational members to engage in learning from failures, which is thought to be the answer for innovation for established companies (Sosna, Trevinyo-Rodríguez & Velamuri, 2010). In an attempt to understand the diverse social exchanges that occur in the workplace, Cole, Schaninger and Harris (2002) suggest that the exchanges between leaders, subordinates and co-workers together compose an interconnected social system that operates in organizations. However, they only take in mind the (formal) relationships between an individual and the organization, their supervisor and

(6)

5 their work group. By this, they tend to ignore the more implicit, lateral relationships between people in the organization. Thinking more broadly, based on the work of Herman, Dasborough and Ashkanasy (2008), there are three predominate interpersonal relationships to be found within the workplace: those between a superior and its subordinate, between workgroup members, and between co-workers. Still, when examining the antecedents of individual innovative behaviour, most of the existing studies tend to overlook the formal interpersonal workplace relationships between a superior and its subordinate (e.g., Basu and Green, 1997; Ishak, 2005; Janssen & Van Yperen, 2004; Scott & Bruce, 1994) and those between workgroup members (e.g., Hülsheger, Anderson & Salgado; Scott & Bruce, 1994). However, informal relationships (those that emerge without organizational involvement) are also related to positive work-related outcomes. This cohesiveness among co-workers, or workplace friendship, causes increased job satisfaction, job involvement, job performance, team cohesion, organizational commitment and decreased turnover intentions (Reich & Hershcovis, 2011). Though, no research has examined the direct linkage of this type of relationship with individual innovativeness. This scarcity of empirical evidence of direct effects of all the different interpersonal relationships within the workplace makes it valuable to extend the body of research. Moreover, it is likely that supervisor-employee and team member relationships, and workplace friendship, not only contribute independently but also simultaneously to the explanation of innovative behaviour outcomes, as they are all part of the interconnected social system that is present in teams and organizations. This thesis then is the first that collectively discusses the influences and interactions between the three predominate interpersonal relationships as stated above, in relation to employees’ innovative behaviour.

(7)

6 1.3 Research purpose and research question

How one should promote innovative behaviour among employees has been a general concern among managers, since workplace creativity and innovation exposed by individual employees is a key driver of organizational success. Human capital is an important existing resource, and it’s about a firm’s capacity to make full use of this and convert it into a competitive advantage. Other studies questioning how interpersonal relationships at work affect innovative behaviour have a highly narrow scope and only take the formal interpersonal relationships into consideration. The purpose then of this study is to examine how different interpersonal relationships at work explain the variation in employees’ innovative behaviour, so that organizations know how to best foster the individual level of innovation in their organization. Therefore, the primary research question is as follows:

RQ: How do high-quality supervisor-employee relationships, high-quality team member relationships and workplace friendship explain variation in employees’ innovative behaviour?

This thesis is organized as follows. First, a theoretical framework is provided, wherein a definition of the complex concept of innovative behaviour is given and relationships are being distinguished from only momentary interactions or connections. Subsequently, each of the relationships that were taken into account for this study are linked to innovative behaviour and to each other, and hypotheses are set up, made visually in the conceptual model. Next, the sample, procedure, data analysis and each of the measures will be described, after which the findings are given to test the hypotheses. In the discussion the findings are linked to existing literature, after which conclusion of this study are provided. As for all studies, this study holds some limitations, followed by recommendations for future research.

(8)

7

2. Theoretical framework and hypotheses

2.1 Reciprocity and voluntary principles

The linkage between interpersonal relationships and innovative behaviour often is explained by social exchange theory, as it is one of the most powerful conceptual paradigms in organizational behaviour. Even though different perspectives of social exchange theory exist, there is an agreement that it involves a sequence of interactions that generate obligations (Emerson, 1976). Despite the fact that this theory is really useful to understand workplace behaviour, it does not explain the effects of the informal interpersonal relationships that exist at the workplace. Even though exchanges with their supervisor and workgroup members are important for employees, the interpersonal dynamics in organizations are not complete without consideration of another very important daily returning dimension - workplace friendship. Encouraging social interaction within the organization not only is profitable for managers as it serves as a tool to play an indirect role in influencing behaviour and guide your people towards the set goals, these informal relationships moreover function as a support mechanism. An important aspect of working together is that the people involved should be able to trust and respect one another, so that they understand one another’s perspectives and style of functioning, solve differences of opinion, communicate effectively, be open to new ideas and question new ideas. Organizational culture influences the extent to which creative solutions are encouraged, fostered and implemented. A culture that is seen as supportive, encourages innovative ways of describing problems and finding solutions. Now that co-workers consider themselves as friends and inherently trust and respect one another, an environment is created wherein creativity and innovation is improved and promoted (Martins & Terblanche, 2003). Within social exchange theory, one’s behaviour often is explained by the rules and norms of exchange that come with having work related relationships. The nature

(9)

8 of such a relationship creates an environment of reciprocity. However, having informal high-quality relationships (those that emerge between two or more people voluntarily) might have the same positive effects on one’s behaviour but cannot be explained in terms of exchange. Employees then might voluntarily engage in innovative behaviour because they feel comfortable and supported to do so, and not out of reciprocity

Rules often also are negotiated aiming to reach beneficial agreements (Cook, Emerson & Gillmore, 1983). These negotiated agreements tend to be more explicit than reciprocal exchanges and have a higher, formal expectation of doing something in return. The responsibilities and obligations are fairly detailed and understood. However, as found by Molm, Takahashi and Peterson (2000), reciprocity generally produces better work relationships than the formal negotiations do, and allows for individuals to be more committed to one another and have higher trust in one another. Moreover, reciprocal exchanges are associated with a higher level of equality (Molm, Peterson & Takahashi, 1999). Meeker (1971) argues that interpersonal exchanges remain individual decisions. Still, since the relationship is based on an exchange, individuals might feel obliged to do so. As reciprocity was found to generally produce better work relationships and outcomes than formal negotiations do, it might be that informal, voluntary relationships at work - such as friendships - produce even better work outcomes than relationships based on reciprocity do.

2.2 Defining the concept of innovative behaviour

Many have tried to capture the complex construct of innovative behaviour, which holds many definitions. In contrast to innovation at the team or organizational level, individual innovative behaviour is based on an individual’s engagement in generating and applying new ideas and approaches in the workplace and often is linked to creativity. However, individual innovative behaviour is a broader and more complex concept than creative behaviour. While creativity

(10)

9 refers to the generation of novel and useful ideas, innovation also encompasses the successful implementation of them. As such, creativity more is a starting point of innovation (Amabile et al., 1996) and innovation then is the process of bringing any new problem-solving ideas into actual use (Kanter, 1984). Other definitions of innovative behaviour range from highly specific to more broad generalizations. Zaltman, Duncan and Holbeck (1973) define innovation as any idea, practice, or material artefact perceived to be new by the relevant unit of adoption. According to Midgley and Dowling (1978), focusing more on the intentionality of those showing innovative behaviour, individual innovativeness refers to the openness to new ideas and decision making to adopt an innovation, free from the influence of the experiences of other employees. As argued by Myers and Marquis (1969), innovation is a process in which all of those things act together. Therefore, altogether innovative behaviour can be defined as “the intentional introduction and application within a role, group or organization of new ideas, new to the relevant unit of adoption, designed to significantly benefit the individual, the group, organization or wider society”. Several aspects of this definition might need further explanation. First, it does not require absolute novelty of an idea, the idea simply has to be new to where it will be adopted. So, if an individual brings new ideas to an organization from his or her previous job, this would be considered an innovation within the terms of the definition. Second, the anticipated benefits do not only hold economic benefits, other possible benefits might be personal growth, better interpersonal communication, or increased satisfaction. Further, the definition is not restricted to technological change, but also includes new ideas or processes in administration or other departments within the organization. The definition as stated above is relied on throughout this study because it gives the most accurate and comprehensive interpretation of innovativeness.

(11)

10 2.3 Interpersonal relationships at work

Before beginning to examine the effects of positive interpersonal relationships in the workplace, a clarification of how these relationships are defined is necessary. Heaphy and Dutton (2008) made an essential distinction between a connection and a relationship in the workplace. As these authors say, a connection requires that both parties are aware of that an interaction has taken place; yet, it does not per se hold intimacy or that the interaction lasts for a longer time. Meanwhile, a relationship develops from these returning interactions. Hence, both connections and relationships demand the awareness and contribution of two individuals but the difference is that relationships are more than momentary. The authors also note that “people’s subjective experience [emphasis added] of their connections with others has immediate, enduring, and consequential effects on their bodies” (Heaphy & Dutton, 2008, p. 138). So, it might be that only individuals themselves need to appraise a relationship as such in order to experience its effects on behavioural outcomes. For this reason, an interpersonal relationship is defined as an individual’s subjective experience of reoccurring interactions with another individual. Since this study is about employees and whether they individually engage in innovative behaviour, the focus is on an employees’ individual perception of a relationship that exists between themselves and others within the firm.

2.3.1 Supervisor-employee relationships and innovative behaviour

To encourage employees to be innovative, managers can boost their employees’ confidence, for example through idea elicitation. This technique seeks to enable individuals in organizations to talk about ideas they don’t usually talk about, and thus to generate more and better new ideas to tackle problems or meet particular challenges. Getting useful feedback from your supervisor can be helpful for an employee to make improvement on the job, because of the process in which he or she might be stimulated to see things from different

(12)

11 perspectives and come up with new and useful ways of doing things (Scott & Bruce, 1994). Farr (1990) also found evidence for that supportive feedback from supervisors is likely to encourage innovative behaviour. In line with this, De Jong and Den Hartog (2007) argue leaders can influence innovation processes by encouraging their employees to think independently and to think creatively around the particular issue they are addressing. Hence, superiors can provide support (defined as the psychological and physical assistance extended by the leader) and guidance, and encouragement to think outside the box. The support leaders provide in turn affects innovative behaviour (Scott & Bruce, 1994). However, this is not restricted to be a one-way relationship. Liden et al. (1997) posited leadership as a unique dynamic that occurs between each supervisor and employee. The nature of such a relationship creates an environment of reciprocity between the supervisor and the employee. As argued by Adams (1965), people always seek to reciprocate those who benefit them. In order to reciprocate, employees may do so through their job performance. From a social exchange theory perspective, innovative behaviour might be an outlet for these positive feelings. According to Blau (1964), social exchange evokes positive affect, trust and affinity, and thus tends to cause feelings of personal obligation and gratitude; whereas economic exchange may be purely instrumental with the payoff being money. Having a high-quality relationship with your supervisor, which is marked by mutual trust and respect, and commitment (Graen & Uhl-Bien, 1995), reduces the potential loss for an innovative employee. It has been observed that supervisors tend to appraise the employees they like and trust in a more positive way (Wayne & Liden, 1995). Thus, when a supervisor trusts and respects an employee, he or she is more likely to assess this employee’s new ideas favourably (Zhou & Woodman, 2003) and to believe these ideas are meaningful and significant. Moreover, when a supervisor empathizes with or likes an employee, he or she is more likely to ascribe positive behaviour outcomes to the employee’s inherent qualities and negative outcomes to the employee’s

(13)

12 situation (Green & Mitchell, 1979). Therefore, employees who are liked and trusted by their supervisor will feel more secure to engage in innovative behaviour because their supervisors are less likely to hold them responsible for a failed innovative attempt. Besides, given that members in higher quality social exchange relationships are likely to have more autonomy, one can expect them to be more innovative (Pelz & Andrews, 1966). By providing them with autonomy, leaders can create an environment that encourages free thinking, exchange of information, and the freedom to explore and examine new ways of handling problems. Members are not bound that much by ideas passed down from the management team, and have the opportunity to engage in novel thoughts and behaviours. Empowered individuals show more innovative behaviour because as autonomous actors, they feel more efficacious about their tasks, and are willing to introduce change (Spreitzer, 1995). Next to subordinates perceiving greater latitude, supervisory attention and feedback, access to organizational resources, and empowerment, those engaging in high-quality leader-member exchange (LMX) relations have a higher level of work engagement (Gerstner & Day, 1997). Work engagement, conceptualized as the investment of an individual’s complete self into a role (Schaufeli & Salanova, 2005), helps to ensure full support of your people, so that management is more likely to meet the required business outcomes. Research has found that higher levels of engagement are strongly related to higher levels of innovation (Bhatnagar, 2012). LMX quality creates conditions favourable to innovative behaviour. Theory states there is a positive relationship between high-quality superior-subordinate relations and innovative behaviour of employees. Therefore, the following hypothesis is conjectured:

Hypothesis 1: High-quality supervisor-employee relationships are positively related to one’s innovative behaviour.

(14)

13 2.3.2 Team member relationships and innovative behaviour

Work teams, defined as sets of interdependent individuals who share responsibility for the outcomes (Sundstrum, Demeuse & Futrell, 1990), are increasingly being used as a way of dealing with competitive environments that demand flexibility and responsiveness to customer demands, and at the same time reduce costs. Instead of structuring work around individual-level jobs, more decentralized organizational forms tend to be built around teams that are expected to think of effective means for accomplishing the work. Companies nowadays embrace this agile set-up as a new way of working, since it has a number of major benefits. Next to for instance increased team productivity, reduced time on projects and improved visibility (Rigby, Sutherland & Takeuchi, 2016), it is related to innovation as well. Agile team members are relatively autonomous parts within a firm, less subjected to the rules that normally constrain actions in a merely hierarchical organization (Nerur & Balijepally, 2007). Within the set goals, innovation can flourish. Besides, being part of a team is more challenging than having a highly simple or routine job, and requires more creative thinking, which in turn promotes innovation (Oldham & Cummings, 1996). As these authors additionally argue, employees being part of a team are expected to focus simultaneously on multiple dimensions of their work and take on a flexible attitude, as their role within the team may change over time. Flexible persons are more competent to adjust themselves cognitively and behaviourally to new experiences and situations, which stimulates innovative behaviour. Teams are being held collectively accountable, so that everyone as well has responsibilities to their team members. There is the existence of interdependence – which often is the reason teams are formed in the first place (Mintzberg, 1979) - which implies individual team members are dependent on other members of the team for being able to carry out their job, and goals can only be achieved when the goals of the other team members are also met (Van der Vegt & Janssen, 2003). Also, team member relationships would affect innovative

(15)

14 behaviour directly and indirectly through their influence on perceptions of the climate for innovation (Scott & Bruce, 1994). Being part of a team thus motivates to do better than you are expected to do. Not only the existence of teams is increasing. Teams are becoming more diverse, as the demographics of the workforce are changing and organizations increasingly employ people from a greater variety of social groups (Jackson & Ruderman, 1995). Moreover, teams are more and more assembled from individuals varying in knowledge, skills, attitudes and abilities (Bowers et al., 2000). Diverse groups have shown to outperform homogenous groups. Heterogeneity in educational, industry and organizational background is found to have a strong positive effect on a teams’ innovation orientation (Talke, Salomo & Kock, 2011). This group diversity means having certain interactive advantages that allow them to perform better than their colleagues who are not part of a team. Members exploit the benefits of diverse perspectives. Through a process of sharing and exchanging critical information, new ideas and insights are fostered (Kozlowski & Ilgen, 2006), thus enabling the innovative outcomes. From this theoretical reasoning, the following hypothesis is conjectured:

Hypothesis 2: High-quality team member relationships are positively related to one’s innovative behaviour.

2.3.3 Workplace friendship and innovative behaviour

Many full-time employees spend more of the hours they are awake with co-workers than they do with their families. As such, it is important to build quality relationships with your colleagues. Friendship is a unique dimension of the workplace. As claimed by Mao (2006), the phenomenon includes an individual bond and is completely voluntary. Other workplace relationships are imposed and are aimed for organizational, work related benefits. According

(16)

15 to Wright (1984, p. 119) two criteria have to be met before one can speak of (workplace) friendship: (1) a “mutual concern and interest that partners show for one another as being unique and irreplaceable”, and (2) “the commitment of free time to interact with one another in the absence of constraints or pressures that are external to the relationship itself”. This second criteria of voluntary interdependence is different from the interdependence which exists within teams, which is needed to accomplish for the set goals. The characteristics as stated above distinguish workplace friendship from the other dyadic relationships occurring at work. Managers might initially discourage close friendships because they believe it results in romantic liaisons (by some estimates, one third of romantic relations begin at the workplace, according to Bordwin, 1994), favouritism, strong dependence on others, gossip and reduced loyalty to an organization (Berman, West & Richter, 2002; Zaleznik, 1997). This could distract workers from doing their job. As a result, workers might spend less time on work and more time socializing. However, the positive effects seem to predominate. Studies have shown that workplace friendship motivates employees (which has proved to be a prerequisite of innovative behaviour) and establishes innovative climates (Berman, West & Richter, 2002). Besides, friends together are more active in exploring uncommon situations than with non-friends (Schwarz, 1972). As made clear earlier, friendships are a source of support. Depending on their level of comfort, friends use each other to solve problems, such as by discussing the options (Luo, 1999). Once a worker has come up with an idea, he or she seeks for supporters that can provide the necessary power when implementing the idea (Galbraith, 1982). High-quality personal relationships provide psychological safety, which can be described as “feeling able to show and employ one’s self without fear of negative consequences to self-image, status or career” (Kahn, 1990, p. 708) and is closely related to innovative behaviour. At the same time, friends at work make one feel valued and connected in ways that allow them to overcome the uncertainty that accompanies working through

(17)

16 problems and experimenting with solutions (Chait, 2000). Thus, being in high-quality voluntary relationships can contribute to more innovative behaviour. From this theoretical underpinning, the following hypothesis is conjectured:

Hypothesis 3: Workplace friendships are positively related to one’s innovative behaviour.

2.3.4 Supervisor-employee and team member relationships

As stated above, it is likely that all three of the distinguished relationships not only contribute independently but also simultaneously to the explanation of innovative behaviour outcomes, as they are all part of the interconnected social system that is present in teams and organizations. Because of limited time and resources, leaders often develop high-quality LMX relationships with only a few subordinates (Dienesch & Liden, 1986). The differential treatment of employees belonging to the same teams may be problematic because of people’s natural sensitivity to social comparison and perceptions of unfairness (Gilbert & Miles, 2000). As a result, they might develop negative feelings towards their team members, such as being jealous and resentful of their co-workers who enjoy more beneficial high-quality LMX relationships, and thus be less motivated to engage in teams and strive for innovative behaviour. Vice versa, having a good relationship with your supervisor does not give you any reason to be jealous, and thus motivates to be more cooperative in teams instead of competitive. As noted earlier, high-quality supervisor-employee relationships provide employees with emotional and instrumental support. Besides, they provide employees with the right tools and resources to succeed. Due to the unequal relationships and thus resource differences, power differences exist between subordinates in the workgroup (Salancik & Pfeffer, 1978). This implies that those subordinates having a higher quality relationship with their supervisor have more resources for the implementation of ideas they effectively

(18)

17 elaborated on with their team members. For this reason it is supposable that they are higher in innovative behaviour. Also, employees experiencing a high-quality relationship with their supervisor are found to be more likely to contribute by helping one another and to share information, ideas and feedback within teams (Seers, Petty & Cashman, 1995), which is in favour of innovative behaviour. Moreover, Lee (1997) suggested that the nature of an individual’s own exchange with his or her leader substantially impacts how employees approach and build other dyadic relationships at work, and use them for work related efforts. As argued before, high-quality relationships between supervisors and employees are typically marked by mutual trust and respect, and commitment. Therefore, employees who enjoy a high-quality relationship with their supervisor are more likely to simultaneously develop other positive dyadic relationships within existing groups at work, because they have experienced affective bonding within formal relationships and know the affiliated benefits. Likewise, Mueller and Lee (2002) have shown that employees that enjoyed a high-quality relationship with their supervisor also enjoy greater openness and a higher frequency in communication, and cooperative information sharing, all in favour of innovative behaviour. Such employees may feel motivated to extend their expected role. Furthermore, as argued by Wang et al. (2013), leaders might serve as a role model by demonstrating appropriate and innovative behaviour, resulting in a higher innovation climate in teams and thus stimulate individual team members to be innovative. From this theoretical underpinning, it is hypothesized that high-quality supervisor-employee relationships have a moderating effect on the relationship between high-quality team member relationships and one’s innovative behaviour:

(19)

18 Hypothesis 4a: High-quality supervisor-employee relationships have a moderating effect on the relationship between high-quality team member relationships and one’s innovative behaviour, such that the relationship between high-quality team member relationships and innovative behaviour will be stronger when supervisor-employee relationships are of high quality.

2.3.5 Team member relationships and workplace friendship

Whereas workplace friendship is expected to positively correlate to innovative behaviour, it might as well be related to high-quality team member relationships. Irrespectively of the project you are working on, you often have pre-existing relationships with your team members, if only because of working for the same employer in the same building. Because of working together in close proximity, sharing common ground, and extra-organizational socializing, co-workers, or acquaintances, make their transition to being friends (Sias and Cahill, 1998). Teams can further facilitate this transition. To check about their progress, teams meet on a regular basis, establishing a certain degree of familiarity and camaraderie among its members. They must interact to work things out and friendships often grow from these workgroup interactions through continuity and mutual respect or need (Harter, Schmidt & Keyes, 2003). As explained by Sias and Cachill (1998), communication first becomes broader, yet remains relatively superficial. Making a further transition to close friends or even almost best friends, co-workers start sharing about problems in one’s personal work and work experiences, and update each other with life events. As stated earlier, teams are increasingly being assembled from individuals with different backgrounds. However, these individuals often do not differ in terms of organizational level. This is favourable for the arise of workplace friendships, which exists mostly among peer co-workers because of similar workplace context and organizational level (Sias & Cahill, 1998). Since there are fewer job

(20)

19 positions at higher levels of an organizational hierarchy (Tachibanaki, 1987), and thus more job positions at lower levels of the hierarchy, individuals at lower organizational levels have more numbers of peers and therefore are expected to have more workplace friendships. Teams can be seen as a means to get those peer co-workers together. As argued by Sias and Cahill (1998), friendship then is the intrinsic reward for team members. As such, it is the existence and frequency of team interactions that facilitate workplace friendship, but workplace friendship that enhances the quality of team member relationships. Workplace friendship does not hinder but rather nourishes high-quality team member relationships because individual members can trust and value one another, share common interests, and experience emotional and instrumental support (Berman, West & Richter, 2002). Their motivation to engage in high-quality team member relationships lies in the fact that they see them as friends instead of formal colleagues. Conducting a problem-solving experiment among primary school students, Newcomb and Brady (1982) found that those paired with a friend instead of a non-friend, showed more extensive exploration, were more task oriented, engaged in a more mutually orientated conversation, and showed more positive affect than non-friends did. Friends engaging in tasks, as compared to non-friends, cooperate more efficiently and interact more fluently (Charlesworth & LaFreniere, 1983). Other experiments on problem-solving tasks showed that those working together with preferred partners (i.e., friends) interacted with one another more heavily (Philp, 1940; Shaw & Shaw, 1962) and were observed to be more active and constructive than non-friends while working together (Lebediker & Thompson, 1992). More recent experiments on this subject amplify these results. Azmitia and Montgomery (1993) found that those who collaborated with their friends, as compared to non-friends, demonstrated more conflicts (disagreements) during collaboration, more frequent elaborations of their partners’ suggestions, and more frequent check-ups on their progress. This can be explained by the evidence found that friends are

(21)

20 more free to disagree than non-friends (Hartup, 1992). As one would initially think, conflict is a negative happening and does not lead to task success. However, disagreements between those participating in the experiment were not destructive but facilitated correct conclusions. So, why is it that disagreements occurring between friends have positive outcomes? On average, conflicts between friends are known to be less agitated than conflicts between non-friends and are more likely to be solved equitably (Hartup, Laursen, Stewart & Eastenson, 1988). These conditions enable employees to explore one another’s point of view with some more comfort. Friends together manage conflicts more effectively than with a non-friend, because they want these relationships to last. Next to feeling more secure when elaborating on suggestions, another reason friends may be better collaborators as compared to non-friends arises from the fact that friends know each other better than non-friends, meaning they know both their similarities and differences better than non-friends do (Ladd & Emerson, 1984). Having this knowledge means that friend’s suggestions and criticisms are more likely to be appropriate than those of non-friends. These results show that friends facilitate performance not only through speaking up more and staying focused on the task, but as well by constructive utilization of their disagreements. Most of the experiments cited above were conducted among school-aged children, however the findings as well hold promises for employees having friends at work. Workplace friendship than even may be a necessary condition for the formation of high-quality team member relationships. This leads to the second cross-level hypothesis:

Hypothesis 4b: Workplace friendship is positively related to high-quality team member relationships.

(22)

21 2.4 Conceptual model

Figure 1 as illustrated below maps the direct effect of supervisor-employee relationships, team member relationships and the both direct and indirect effect of workplace friendship on individual innovative behaviour, as well as the moderating effect of supervisor-employee relationships. All correlations are predicted to be positive, regardless whether this is relative or substantial, however negative effects are not expected.

(23)

22

3. Methodology

3.1 Sample

The sample consisted of 134 full-time employees from different service organizations. The respondents ranged from 23 to 62 years in age, with a mean age of 32 years. Regarding their gender, 49 per cent of the respondents were women. Other demographic characteristics of interest were educational level and job tenure. Regarding their educational level, 10 per cent of the sample had a lower vocational degree or less, 35 per cent had a bachelor degree and another 55 per cent had completed a master degree. As for job tenure, the average time participants had been working for the company was 5 years, ranging from 0 to 38 full years.

3.2 Procedure

The one-time survey was administered through the online survey software Qualtrics. An online survey was chosen because of its easy access to geographically dispersed respondents in a national context (Evans & Mathur, 2005). A pre-test did not indicate any ambiguous questions, thus adjustment of the survey was not required. Collection of the data was based on a convenience sample, for which it was send out to friends, family and colleagues, who as well were asked to further distribute the survey. Earlier research on response rate levels when conducting online surveys in organizational research shows an average response of only about 50 per cent (Baruch & Holtom, 2008). However, since the author had (in)direct linkages with many of those who were surveyed, the response rate was expected to be higher because those who received the request to fill in the survey might felt obliged to do so. Prior to filling in the questionnaire, participants were shortly informed about the nature, duration and purpose of it. The questionnaire asked about their own perceptions, and with a self-reported questionnaire there always is the possibility of social desirability in answers. By stressing the need for

(24)

23 honesty and clearly state that all answers would remain completely confidential, this risk was tried to be minimized. Participants as well were asked if they met the requirements of usually working a minimum of 35 hours per week, as this is the threshold to be considered as a full-time employee according to the Statistics Netherlands (2015). To reassure for validation, a Force Response option in Qualtrics required respondents to answer each question before progressing to the next page of the survey. Only surveys that were finished were used for further examination. The research question was answered by applying a multiple regression analysis. Relying on the rules of thumb provided by Cohen (1992), having a set of three independent variables and the goal to detect a medium degree of association in the population (an effect which you can not only see through study, but as well is likely to be visible to the naked eye of a careful observer), the necessary sample size to carry out this type of analysis was 108. This prerequisite was widely met since the number of participants was 134.

3.3 Data analysis

First, the dataset was checked for missing data. As explained in the previous paragraph, a Force Response option in Qualtrics reassured for missing data. Before running any analysis, there was a check for counter indicative items in the dataset. Since one of the items of the Workplace Friendship scale was reversed to prevent from habituation bias, it had to be recoded. Next, the reliability of the scales that were used was computed, and scale means were computed so that further analysis could be conducted. Different procedures were used to test the hypotheses. Primarily, a correlation analysis was conducted to see whether there were any relationships between the variables used for this study, as well in order to test the correlation hypotheses. A two-way factorial ANOVA analysis then was conducted to compare the main effects of the significant categorical variables on the dependent variable. Of the provided options Tukey’s post hoc test was chosen to most honestly find out which levels

(25)

24 within each independent variable were significant. The correlation analysis already indicated that there were correlations between the main independent variables and dependent variable but it did not explain the variation in employees’ innovative behaviour. To explain the variation in employees' innovative behaviour, as it is the purpose of this study, a hierarchical multiple regression was performed to investigate the ability of supervisor-employee relationships, workgroup relationships and workplace friendship to predict individual innovative behaviour, after controlling for the significant categorical variables. To test the moderation hypothesis, a moderation analysis was run through Process developed by Andrew F. Hayes (Model Number 1), after having standardized the independent variable (WGrelTOT) and moderator (SSrelTOT) that were numerical. Finally, to give the most confident answer in testing the final hypothesis, a mediation analysis was run again through Process (Model Number 4). Using ordinary least squares path analysis, a simple mediation analysis was conducted.

3.4 Measures

Measures for the independent variables, dependent variable and control variables are written below. For each of the four composed variables, only scales with a high reliability score were used. Aside from the control variables, the variables were measured by a Likert or behavioural frequency scale containing up to 7 items. The survey instrument as a whole as well is included in appendix A.

3.4.1 Supervisor-employee relationship

Based on the recommendations of Graen and Scandura (1987), questionnaires were filled in only by employees themselves, as the authors argue leaders are likely to provide socially desirable answers about relationships with their followers (i.e., that they treat them all the

(26)

25 same) to not harm their ego. This is even more the case when surveyed on one moment, than in case of repeated measures are taken over time. Since the survey was carried out as a one-time measurement, and as stated before only they need to appraise it as such, only the employees’ perspective is taken into account. The Leader-Member Exchange (LMX) scale provided by Scandura and Graen (1984) was employed to measure the quality of exchange between supervisors and employees (Cronbach’s alpha was .84). The LMX scale contained the following seven items: “Do you usually feel that you know where you stand / do you usually know how satisfied your immediate supervisor is with what you do?” Always know where I stand (4), Usually know where I stand (3), Seldom know where I stand (2), Never know where I stand (1), “How well do you feel that your immediate supervisor understands your problems and needs?” Completely (4), Well enough (3), Some but not enough (2), Not at all (1), “How well do you feel that your immediate supervisor recognizes your potential?” Fully (4), As much as the next person (3), Some but not enough (2), Not at all (1), “Regardless of how much formal authority your immediate supervisor has built into his or her position, what are the chances that he or she would be personally inclined to use power to help you solve problems m your work?” Certainly would (4), Probably would (3), Might or might not (2), No chance (1), “Again, regardless of the amount of formal authority your immediate supervisor has, to what extent can you count on him or her to "bail you out" (=giving financial assistance) at his or her expense when you really need it?” Certainly would (4), Probably would (3), Might or might not (2), No chance (1), “I have enough confidence in my immediate supervisor that I would defend and justify his or her decisions if he or she were not present to do so” Certainly would (4), Probably would (3), Maybe (2), Probably not (1), “How would you characterize your working relationship with your immediate supervisor?” Extremely effective (4), Better than average (3), About average (2), Less than average (1).

(27)

26 3.4.2 Team member relationship

The ten-item Team-Member Exchange (TMX) scale provided by Seers, Petty and Cashman (1995) was used to measure individual team members’ perceptions of the quality and effectiveness of the reciprocal exchange relationship that exists between themselves and their team members. The ten items are: “How often do you make suggestions about better work methods to other team members?”, “Do other members of your team usually let you know when you do something that makes their jobs easier (or harder)?”, “How often do you let other team members know when they have done something that makes your job easier (or harder)?”, “”How well do other members of your team recognize your potential?”, “How well do other members of your team understand your problems and needs?”, “How flexible are you about switching job responsibilities to make things easier for other team members?”, “In busy situations, how often do other team members ask you to help out?”, “In busy situations, how often do you volunteer your efforts to help others on your team?”, “How willing are you to help finish work that had been assigned to others?”, “How willing are other members of your team to help finish work that was assigned to you?”. Responses to the items used a seven-point Likert scale that ranged from strongly disagree (scored as 1) to strongly agree (scored as 7). Cronbach’s alpha for this scale was .83.

3.4.3 Workplace friendship

Primarily the Workplace Friendship Scale is built on two aspects of friendships in the workplace; (a) the opportunity for friendship (e.g., I have the opportunity to get to know my co-workers), and (b) the presence of friendship, which one is referred to in this study. Nielsen et al. (2000) assessed the two six item subscales for their reliability regarding internal consistency, with Cronbach’s alpha results of .89 for the friendship ‘prevalence’ sub scale. The six items for the friendship prevalence dimension are: “I have formed strong friendships

(28)

27 at work”, “I socialize with co-workers outside the workplace”, “I can confide in people at work”, “I feel I can trust many co-workers a great deal”, “Being able to see my co-workers is one reason I look forward to my job”, “I do not feel that anyone I work with is a true friend (R)”. The questions were rated on a five-point Likert scale ranged from strongly disagree (scored as 1) to strongly agree (scored as 5). One item was reverse scored.

3.4.4 Innovative behaviour

When examining individual-level innovation, it often is assessed as the number of creative suggestions submitted by an employee (Buech, Michel & Sonntag, 2010; Oldham & Cummings, 1996). However, submitting suggestions only is one dimension of innovative behaviour. In an attempt to capture the richness of the construct of individual innovation, Kleysen and Street (2001) developed and tested a multi-dimensional measure of individual innovative behaviour. They identified five dimensions that seemed to best categorize behaviours associated with individual innovation, namely opportunity exploration, generativity, formative investigation, championing, and application. Based on the behaviour definitions the authors provided, opportunity exploration can be described as looking for and recognizing opportunities to innovate. Generativity then is actually the generation of ideas and solutions to those opportunities. The third dimension of innovative behaviour is concerned with giving form to and elaborate on ideas, solutions and opinions, and trying them out through investigation. The fourth dimension, championing, is about bringing creative ideas to life. The socio-political behaviours involved in this process include mobilizing resources, persuading and influencing, pushing and negotiating, and challenging and risk-taking. The fifth and final dimension involves working at making innovations a regular part of business as usual. Three basic behaviours regarding application are implementing, modifying and routinizing. The authors created a more far reaching concept of innovative

(29)

28 behaviour, that gives the most comprehensive interpretation of innovative behaviour, with Cronbach’s alpha results of .945. Each dimension was measured with items related to the behaviours as described above. Additionally, items were prefaced by, “In your current job, how often do you…?”. The fourteen belonging items are: “Look for opportunities to improve existing process, technology, product, service or work relationship?”, “Recognize opportunities to make a positive difference in your work, department, organization, or with customers?”, “Pay attention to non-routine issues in your work, department, organization or the marketplace?”, “Generate ideas or solutions to address problems?”, “Define problems more broadly in order to gain greater insight into them?”, Experiment with new ideas and solutions?”, “Test-out ideas or solutions to address unmet needs?”, “Evaluate the strengths and weaknesses of new ideas?”, “Try to persuade others of the importance of a new idea or solution?”, “Push ideas forward so that they have a chance to become implemented?”, “Take the risk to support new ideas?”, “Implement changes that seem to be beneficial?”, “Work the bugs out of new approaches when applying them to an existing process, technology, product, or service?”, “Incorporate new ideas for improving an existing process, technology, product or service into daily routines?”. Answers were rated on a six-point behavioural frequency scale ranged from never = 1, almost never = 2, sometimes = 3, fairly often = 4, very often = 5, and always = 6.

3.4.5 Control variables

Sociodemographic variables of employees could account for variance in innovative behaviour when testing the hypothesized model. To rule out any spurious relationships, educational level (1 = Lower vocational degree or less, 2 = Bachelor degree, 3 = Master degree or more), job tenure (in full years), gender (0 = Male, 1 = Female) and age (in years), were entered as control variables in the analysis. Level of education, job tenure and age have been

(30)

29 consistently related to innovative behaviours. As argued by Mumford and Gustafson (1988), level of education can be critical for enhancing job performance, especially innovative performance. Highly-educated people would be more creative and innovative than those with lower levels of education (Rasmussen, 2012). For job tenure, employees who have been affiliated with an organization for a longer period of time might feel stuck or plateaued in their career, in the sense that their job no longer challenges or motivates them to come up with new ideas and do more than what the job requires them to do, therefore job tenure influences the readiness to create and innovate (Suliman & Al-Shaikh, 2007). As well as for employees who have been affiliated with an organization for a longer period of time, senior employees have a higher chance of having reached a plateau in their career progression. Moreover, younger employees are found to be more creative and innovative at the workplace than their older colleagues (Binnewies, Ohly & Niessen, 2008). Gender was controlled for since previous research points out some differences between men and women regarding their levels of creativity, which is closely linked to innovation. For example, Amabile et al. (2005) noticed that women are rated lower in displaying creativity than for men.

(31)

30

4. Results

Checking for reliability, the Individual Innovative Behaviour scale has high reliability, with Cronbach's Alpha = .956. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Also, none of the items would substantially affect reliability if they were deleted. Second, the Leader-Member Exchange scale to measure individual perceptions’ of the quality of the relationship they have with their direct supervisor, has high reliability, with Cronbach’s Alpha = .888. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Again, none of the items would substantially affect reliability if they were deleted. Then, the Team-Member Exchange scale to measure individual team members’ perceptions of the quality of their team member relationships, has high reliability, with Cronbach’s Alpha = .846. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Again, none of the items would substantially affect reliability if they were deleted. The Workplace Friendship scale to measure individual perceptions’ of the presence of friendship at work, has high reliability, with Cronbach’s Alpha = .935. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Again, none of the items would substantially affect reliability if they were deleted. For each of the scale that was used for this study high reliability was confirmed, with two of them (Individual Innovative Behaviour scale and Leader-Member Exchange scale) even exceeding the results of the authors who originally created the scales (respectively .956 over .945 and .888 over .840). See Table 1 for the reliability results.

(32)

31 Table 1: Reliability Scores

Cronbach's Alpha N of Items

Individual Innovative Behaviour .956 14

Leader-Member Exchange .888 7

Team-Member Exchange .846 6

Workplace Friendship .935 6

Means and standard deviations of the numerical descriptive variables job tenure and age, and the scales are shown in Table 2.

Table 2: Descriptive Statistics Numerical

N of Items Minimum Maximum Mean Std. Deviation

Job tenure 134 0 38 4.96 7.324 Age 134 23 62 31.78 9.028 IBehTOT 134 2.14 6.00 4.3022 .91983 LMExTOT 134 1.57 4.00 3.1301 .58770 TMExTOT 134 2.17 7.00 5.7239 .80907 WFrTOT 134 1.00 5.00 3.6157 7.00795

Descriptive statistics of the categorical variables level of education and gender are shown in Table 3.

Table 3: Descriptive Statistics Categorical

Valid N of Items Percent

Education 1 14 50.7 2 47 49.3 3 73 100.0 Total 134 10.4 Gender 0 68 35.1 1 66 54.5 Total 134 100.0

Table 4 holds an overview of the results of the correlation analysis. Obviously, the relationships that were of highest interest are the ones between the dependent variable and the

(33)

32 independent and descriptive variables. To reassure strong evidence for correlations, only the 0.01 significance level was taken into account.

Table 4: Correlations

Education Job tenure Gender Age IBehTOT LMExTOT TMExTOT WFrTOT Education Pearson’s r 1

Sig. (2-tailed)

Job tenure Pearson’s r -.260** 1 Sig. (2-tailed) .002 Gender Pearson’s r -.001 -.044 1 Sig. (2-tailed) .988 .613 Age Pearson’s r -.235** .836** -.063 1 Sig. (2-tailed) .006 .000 .466 IBehTOT Pearson’s r .281** -.164 -.243** -.069 1 Sig. (2-tailed) .001 .058 .005 .431 LMExTOT Pearson’s r .084 .118 -.212* .122 .387** 1 Sig. (2-tailed) .332 .174 .014 .160 .000 TMExTOT Pearson’s r .045 .214* -.014 .238** .315** .556** 1 Sig. (2-tailed) .604 .013 .869 .006 .000 .000 WFrTOT Pearson’s r .070 -.093 .087 -.066 .437** -.021 .147 1 Sig. (2-tailed) .423 .287 .316 .451 .000 .807 .090 ** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

A positive correlation was found between individual innovative behaviour (M = 4.30 SD = .92) and leader-member exchange (M = 3.13 SD = .59), r = .387, p = < .001. Also, there was a positive correlation for team-member exchange (M = 5.72 SD = .81), r = .315, p = < .001, and workplace friendship (M = 3.62 SD = 1.01), r = .437, p = < .001. All independent variables had a medium effect size that was significant, of which workplace friendship was most close to a large effect. By this, hypothesis H1, H2 and H3 all are supported. Despite of the fact that the correlation analysis only indicates correlations, and thus not show any

(34)

33 asymmetric relationships between variables, it is plausible that employees that report higher quality of their relationship with their supervisor and team members, and higher quality of friendships at work, perform more innovative behaviour. This will be further explained in the findings from the multiple regression analysis.

For the descriptive variables, level of education was found to have a positive small effect (r = .281, p = < .001) on individual innovative behaviour, thus significant. For job tenure and gender negative correlations were found. Job tenure (M = 4.96 SD = 7.32) and gender both had a small effect on individual innovative behaviour, respectively r = -.164, p = .058 and r = -.243, p = .005. Only for gender a significant effect was found. For age, no correlation with individual innovative behaviour was found. Logically, length of job tenure increased with age (r = .836, p = < .001). Other significant effects were found between level of education and job tenure, r = -.260, p = .002. Level of education and age also negatively correlated, r = -.235, p = .006. Age then positively correlated with team-member exchange, r = .238, p = .006. These were are small effects sizes. Finally, after the inevitable correlation between job tenure and age, team-member and leader-member exchange was found to have the second highest positive correlation, r = .556, p = < .000.

To further explore the significant effects of the independent categorical variables on the dependent variable, a two-way factorial ANOVA analysis was conducted to compare the main effects of level of education and gender, and the possible interaction effect between level of education and gender on the individual innovative behaviour that was performed. As can be seen in Table 5, there was a significant main effect of level of education on individual innovative behaviour performed, F(2,128) = 4.81, p < .05, η² = .07. In other words, those who had obtained a higher educational level reported higher individual innovate behaviour. And, there was a non-significant interaction effect between level of education and gender on the performed individual innovative behaviour, F(2,128) = .22, p = .80, η² < .01.

(35)

34 Table 5: Test of Between-Subjects Effects

Dependent Variable: IBehTOT

Source Type III Sum of Squares df Mean Square F Sig. Partial Eta Squared

Corrected Model 15.895ª 5 3.179 4.211 .001 .141 Intercept 1254.403 1 1254.403 1661.556 .000 .928 Education 7.267 2 3.633 4.813 .010 .070 Gender 2.872 1 2.872 3.805 .053 .029 Education*Gender .334 2 .167 .221 .802 .003 Error 96.634 128 .755 Total 2592.770 134 Corrected Total 112.53 133

a. R Squared = .141 (Adjusted R Squared = .108).

The Tukey post hoc tests revealed that the performed individual innovative behaviour was significantly higher in the group with a master degree or more (M = 4.54 SD = .96) compared to the group with a lower vocational degree or less (M = 3.86 SD = .74), p = .02. This difference was even more significant for those with a master degree or more compared to the group with a bachelor degree (M = 4.07 SD = .80), p = .01, see Table 6 below.

Table 6: Multiple Comparisons Educational Level

Dependent Variable: IBehTOT. Tukey HSD 95% Confidence Interval (I) Educational Level (J) Educational Level Mean Diff. (I-J) Std. Error Sig. Lower Bound Upper Bound

1 2 -.2143 .26455 .698 -.8416 .4130 3 -.6791 .25351 .023 -1.2802 -.0779 2 1 .2143 .26455 .698 -.4130 .8416 3 -.4648 .16250 .014 -.8501 -.0795 3 1 .6791 .25351 .023 .0779 1.2802 2 .4648 .16250 .014 .0795 .8501

* The mean difference is significant at the .05 level.

Next, there was a non-significant, although barely, main effect of gender on individual innovative behaviour performed, F(1,128) = 3.81, p = .053, η² = .03. The Tukey post hoc tests revealed that men (M = 4.52 SD = .84) showed a tendency to perform more innovative

(36)

35 behaviour compared to women (M = 4.08 SD = .95). Means and standard deviations are shown in Table 7.

Table 7: Descriptive Statistics Educational Level and Gender Dependent Variable: IBehTOT

Educational Level Gender Mean Std. Deviation N

1 Men 3.9214 .83057 10 Women 3.6964 .48926 4 Total 3.8571 .73753 14 2 Men 4.4206 .74509 18 Women 3.8547 .76824 29 Total 4.0714 .80109 47 3 Men 4.7179 .81312 40 Women 4.3160 1.08914 33 Total 4.5362 .96223 73 Total Men 4.5221 .83539 68 Women 4.0758 .95354 66 Total 4.3022 .91983 134

In the plot below the scores of the different categories within educational level and gender on the individual innovative behaviour performed are shown. As it is clearly visible, men perform (slightly) higher on individual innovative behaviour compared to women, and the performance of individual innovative behaviour increases with educational level.

(37)

36

Figure 2: Plot Educational level versus Gender on Individual Innovative Behaviour

Results from the multiple regression analysis to answer the research question of this study are shown below in Table 8. The higher the correlation found, the better the prediction made by the regression model. Among the three independent variables of interest, workplace friendship was found to have the largest effect on one’s innovative behaviour, and thus is expected to be the best predictor of the behaviour of interest in the regression analysis. In the first step of the hierarchical multiple regression, only two predictors were entered: educational level and gender. For a model to be significant, it had to be below the .05 significance level. As shown below in Table 8, this model was statistically significant, F (2,131) = 10.51; p < .001, and explained 13,8% of variance in individual innovative behaviour. After entry of supervisor-employee relationships, team member relationships and workplace friendship at step two the total variance explained by the model as a whole was 44,2%, F (5,128) = 20.29; p < .001. The introduction of supervisor-employee relationships, team member relationships and workplace friendship explained an additional 30,4% variance

3.5 3.75 4 4.25 4.5 4.75 5 1 2 3 Men Women

Estimated marginal means of IBehTOT

Can workplace friendship enhance employees’ innovative behaviour? : A study to the interpersonal relationships that exist in the workplace affecting individual innovative behaviour, looking beyond the predominant relati