• No results found

1 CREATIVE POTENTIAL AND OVERALL PERFORMANCE: THE MODERATING

N/A
N/A
Protected

Academic year: 2021

Share "1 CREATIVE POTENTIAL AND OVERALL PERFORMANCE: THE MODERATING "

Copied!
42
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

1 CREATIVE POTENTIAL AND OVERALL PERFORMANCE: THE MODERATING

ROLE OF INTRA-ORGANIZATIONAL SOCIAL NETWORKS Master thesis, MSc BA, Strategic Innovation Management University of Groningen, Faculty of Economics and Business

20-01-15

CHRIS BECHTUM Studentnumber: S2011433

Kleine butjesstraat 62 9712 DA Groningen tel.: +31 (0)639757456 e-mail: chrisbechtum@gmail.com

Supervisors

1

st

; prof. dr. W.A. Dolfsma 2

nd

; dr. ir. M.W. Hillen

Total word count: 10694

(2)

2 ABSTRACT

This study focuses on the relationship between the creative potential of individual

employees and their overall performance. More specifically, we aim to test the moderating

role of two intra-organizational social networks on this relationship. These two social

networks are the so-called advice network and hate network. Data for this research was

collected at a large western manufacturing firm and at one of its R&D subsidiaries using

surveys for both employees and leaders. Using a statistical regression analysis we found

support for the direct positive relationship between the creative potential of employees and

their overall performance. However, we did not find significant support for the moderating

role of either the advice network or hate network.

(3)

3 TABLE OF CONTENT

1. Introduction . . . . . . . . . 5

2. Theoretical framework . . . . . . . . 8

2.1 Overall performance . . . . . . . 8

2.2 Creative potential . . . . . . . . 9

2.3 Creative potential and overall performance . . . . 10

2.4 Network theory . . . . . . . . 12

2.5 Multiple networks . . . . . . . 13

2.6 Favorable position in the network . . . . . 14

2.7 The moderating effect of an advice network . . . . 15

2.8 The moderating effect of a hate network . . . . 17

3. Methods . . . . . . . . . . 19

3.1 Sample and procedure . . . . . . . 19

3.2 Measures . . . . . . . . . 20

3.2.1 Creative potential . . . . . . 20

3.2.2 Advice network . . . . . . . 20

3.2.3 Hate network . . . . . . . 21

3.2.3 Overall performance . . . . . . 21

3.2.5 Control variables . . . . . . 22

3.3 Statistical analysis . . . . . . . 22

4. Results . . . . . . . . . . 24

4.1 Network structure . . . . . . . . 24

4.2 Descriptive statistics . . . . . . . 25

4.3 Hypotheses testing . . . . . . . 27

5. Discussion . . . . . . . . . . 29

(4)

4

5.1 Findings . . . . . . . . . 29

5.2 Theoretical implications . . . . . . . 31

5.3 Practical implications . . . . . . . 32

5.4 Strong and weak points . . . . . . . 33

6. Reference list . . . . . . . . . 36

7. Appendix . . . . . . . . . . 41

7.1 Appendix A; Creative potential measure . . . . . 41

7.2 Appendix B; Overall performance measure . . . . 41

7.3 Appendix C; Overview of the results of the component analysis . 42

(5)

5 1. INTRODUCTION

In this day and age, a firm’s ability to innovate is more important in prolonging the lifespan of the organization than it has been ever before. The increase in globalization has internationalized and intensified competition and therefore forces corporations to come up with novel or enhanced products and services, and to develop improved organizational processes (Schilling, 2010). In numerous cases, the birthplace of an innovation can be found in an organization’s employees (Laursen & Salter, 2004). This undoubtedly emphasizes the importance of the overall and innovation performance of an organization’s employees.

Furthermore, a higher level of performance among a corporation’s employees can also help to sustain the lifespan of the company by being more innovative and by achieving advantages over competing organizations.

According to Amabile et al. (1996), creativity lies at the foundation of all innovations.

Likewise, Im and Workman (2004) argue that it is seen as an important starting point for innovation as well. Moreover, high levels of creativity have also been related to gaining competitive advantage (Oldham & Cummings, 1996) and long term organizational survival (Unsworth, 2001; Scott & Bruce, 1994). The creativity of an employee thus shows great possibilities in promoting his or her (innovation) performance as perceived by their supervisor(s). This possibility is present because a highly creative employee undoubtedly is more likely to generate highly creative and useful ideas that may lead to this higher performance.

Regardless of the fact that high levels of creativity among employees are generally

seen as beneficial to their overall performance (Gong et al., 2009; Oldham & Cummings,

1996; Zhang & Bartol, 2010), not much is known about the influence of intra-organizational

social networks on this relationship. For that reason, this paper aims to assess the influences

of such social networks on the relationship between an employee’s creative potential and

(6)

6 overall performance in an international setting. More specifically, this study examines the effect of the so-called “advice network” (Ibarra, 1993) and “hate network” in a manufacturing company that has outsourced a part of its research and development activities to an overseas location. The advice network is an informal network that does not stick to the formal organizational structure but looks at who turns to whom when they are in need of advice or information that can help the individual employees to enhance their ability at their own job.

On the contrary, the hate network looks at with whom a certain employee prefers not to collaborate with. The influence of the latter network is particularly interesting because, to date, no research has broached this concept yet.

It is especially interesting to explore the role of intra-organizational social networks in the previously described context because such linkages have been associated with several positive outcomes. One of these identified consequences is the sharing of various resources (e.g. knowledge and information) (Chow & Chan, 2008). Likewise, Tsai (2002) also argues that social interaction provides channels for information exchange among the members of an organization. Therefore, social networks may promote (advice network) or, on the contrary, deter (hate network) the dispersion of the individual employee’s creative efforts, which may therefore enforce or weaken his or her overall performance respectively. Moreover, crucial knowledge and information that may enable the creativity of an individual employee may very well be embedded in a social network (Kijkuit & van den Ende, 2007). In addition, the creative ideas of the individual may also complement the creative ideas of other employees or be a source of inspiration to them.

We have illustrated an initial research model in order to give a concise and structured

overview of the expected relationships that we will research throughout this paper. This

research model can be found in the image below (figure 1).

(7)

7 Figure 1. Conceptual Model with the Respective Hypotheses.

Researching the effects of the previously mentioned intra-organizational social networks on the positive relationship between an employee’s creative potential and overall performance can provide us with important theoretical and practical insights. Firstly, from a theoretical perspective, it is important to realize that the social network in which the employee is present might be a vastly relevant contextual influence. From a more practical point of view, this study may also possibly lead to a greater understanding among managers about the influence of intra-organizational social networks, which in turn could lead to the willingness to improve social interactions among their workers (e.g. by means of an intervention). As a final point, this study also has the potential to confirm the important role of creativity in the workplace.

We will start this paper by defining and formulating the different variables, concepts

and hypotheses in the theoretical framework. Subsequently, we will elaborate on the

methodology used for the collection of data, measures and statistical analysis in the methods

(8)

8 section. The data that we use in this study was collected at a large western manufacturing company that outsources a segment of its R&D to a subsidiary in a developing country. The data was gathered at both a location in the organization’s home country and at the subsidiary in a foreign nation. After analysis of the data, the results will be presented. Lastly, the results of the study will be critically discussed and elaborated on in the discussion section.

2. THEORETICAL FRAMEWORK 2.1 Overall performance

The overall performance of employees is of course a very important concept for every organization as it lies at the foundation of organizational performance. It can however be seen as quite a broad concept which creates the need for us to define it concisely. During the course of this research we use the following definition for the overall performance of an employee; “the extent to which individuals succeed (in the eyes of management) in contributing to organizational ends” (Kilduff & Krackhardt, 2008). Hence, this definition indicates that the overall performance of individuals will be higher when their supervisor(s) consider the employee to be a factor that plays a significant role in accomplishing organizational results. Therefore, the overall performance of employees is one of the most important elements that can positively affect the performance of the organization. For that particular reason, it is quite naturally an enormously important concept for organizations and their management.

The aforementioned notion that the overall performance of employees is so vital to an

organization’s management also generates the need to understand what the important

predictors that have a significant impact on it are. Central in this research is that we expect

that the creative potential of an employee may be one of the factors with a positive impact on

(9)

9 his or her overall performance. The reasons why we can expect this positive relationship are elaborated on in the following two sections.

2.2 Creative potential

Creativity has been seen as a predictor of various outcomes that have a positive impact on the organization. In fact, many scholars agree that it can generally be seen as a concept that is contributing to the overall performance of an organization (e.g. Gong et al., 2009; Oldham

& Cummings, 1996; Zhang & Bartol, 2010). However, over time, different researchers have been using a number of different approaches for the measurement of creativity. Whereas some academics have mostly focused on personality traits and characteristics of individuals, others have emphasized the actual outcomes of creative behavior (e.g. creative products and achievements) (DiLiello & Houghton, 2008). Until now, a rather extensive body of literature in the field has led to the realization that creativity levels are likely to be higher when an individual possesses certain characteristics or skills (Tierney & Farmer, 2002). Additionally, intrinsic motivation (Amabile et al., 1994; Cummings & Oldham, 2007) and perceived organizational support (Amabile, 1996; Zhou & George, 2001) have just as well been identified as being positively related to creativity levels. Furthermore, domain specific knowledge has also been identified as one of the vital requirements for a high level of creativity (Weisberg, 1999).

An important distinction was made by Hinton in 1968 as he established an important

difference between the potential creativity of an individual and the creative output that is

actually realized in practice by this person. This distinction by Hinton is based on the

assumption that the realized creativity is not only dependent on the individual, but is also

influenced by his or her environment. In other words, an individual may not be able to

completely exploit its creative potential if his or her creative output is repressed by the

(10)

10 environment (Hinton, 1968). In an organizational setting, this means that the creative potential of employees may remain unused given a certain organizational atmosphere. A more supportive organizational environment may however allow an employee to fully exploit his or her creative potential. This also implies that managers can have a partial influence in how much creative output is realized within a firm.

Throughout this study we look at the creative potential of individual employees because this concept is a prerequisite for the eventually realized creative output. Of course, despite a favorable environment, no creative output can be realized without creative potential.

We apply the concept of creative potential in this study because it best reflects the capabilities of the individual employee and forms the very basis of every form of creativity. Moreover, we believe that social networks are part of the individual’s environment and therefore may be able to help them in exploiting their creative potential. Thus, if we plan on studying the effects a social network may have, it would be best to look at an individual’s creative potential.

The definition for creative potential we apply throughout this study is following; “the creative capacity, skills and abilities that the individual possesses” (Hinton, 1968, 1970).

Hence, the concept concerns the different creative aspects of an individual employee. Creative potential is different from practiced creativity because the latter concept looks more at the perceived opportunity to utilize creativity skills and abilities (DiLiello & Houghton, 2008).

Therefore, the concept creative potential does not necessarily assume anything about actually bringing creative ideas forward and into practice.

2.3 Creative potential and overall performance

The effect of an individual’s creative potential on his or her overall performance is

fundamental in this study. Creativity has repeatedly been valued and identified as a predictor

(11)

11 of various positive and crucial organizational outcomes. For starters, creativity is considered to be at the foundation of innovation (Amabile et al. 1996; Im & Workman Jr., 2004), which on its turn is one of the most important characteristics that help businesses to stay ahead of their competitors and prolong the lifespan of an organization. According to Amabile (1996), the “successful implementation of new programs, new product introductions, or new services depends on a person or a team having a good idea and developing that idea beyond its initial state”. Thus, innovation is very dependent on the generation of creative ideas by individuals or teams.

Additionally, creativity itself has also been related to the long term survival of organizations (Unsworth, 2001; Scott & Bruce, 1994) and the gaining of competitive advantage (Oldham & Cummings, 1996). Moreover, several researchers have stressed that creativity is an important input for the growth of organizations and the sustaining of their competitiveness (Amabile & Khaire, 2008; Cohn, Katzenbach, & Vlak, 2008; George, 2007).

Likewise, George (2007) has also specified that creativity is often perceived as an important method for organizations to generate meaningful and lasting value for their multiple stakeholders.

Because creativity has been associated with such a diverse quantity of positive

organizational outcomes, we can definitely reason that creativity delivers a contribution to

organizational ends. Moreover, due to the fact that the creative potential of individuals forms

the basis of any form of creativity, we can also state that a higher creative potential of an

individual employee may very well be positively related to such organizational ends. Because

we define the overall performance of employees as “the extent to which individuals succeed

(in the eyes of management) in contributing to organizational ends” (Kilduff & Krackhardt,

2008), we can argue that the creative potential of an employee is likely to be positively related

to their overall performance. In conclusion, we can reason that an employee’s creative

(12)

12 potential is likely to have a positive impact on an employee’s overall performance, as perceived by their supervisor, because it may very well deliver a contribution to organizational ends.

Based on the explanations above, we think there is sufficient reason to expect that an employee’s creative potential will have a positive influence on his or her overall performance.

This brings us to our first hypothesis:

H1: An employee’s creative potential is positively related to his/her overall performance.

2.4 Network theory

In this paper, we focus on network theory to assess the impact of an intra- organizational social network on individual outcomes. According to Borgatti and Halgin (2011), “network theory refers to the mechanisms and processes that interact with network structures to yield certain outcomes for individuals and groups”. Applied to this study, this means that we will look at the impact of a social network’s structure on the relationship between the level of creative potential and the level of overall performance of individual employees. This means that the structure of a social network is a crucial factor throughout this study. Therefore, it is of course imperative to completely comprehend what a network structure consists of.

The structure of a network is determined by the set of actors and the relationships among them (Aalbers, 2012). Therefore, in the light of this study, we can think of a network structure as the employees within the organization and the different relationships among them.

The individual actors in the network are usually called “nodes”, and the relationships between

them are called “ties” (Aalbers, 2012).

(13)

13 The definition of a social network that we use throughout this research paper is the following; “a specific set of linkages among a defined set of persons with the additional property that the characteristics of these linkages as a whole may be used to interpret the social behavior of the persons involved” (Mitchell 1969, p.2). During the course of this study, the definition of a social network is applied in an intra-organizational setting. In other words, it focuses on a social network restricted to the borders of the organization itself. Furthermore, we examine the influences of the intra-organizational network on an individual level, and not on a network level.

2.5 Multiple networks

It is important to understand that individual actors may very well be embedded in various social networks at the same point in time. The most simple example to illustrate the presence of multiple networks is that employees can function in both a formal network (e.g.

with whom members of the organization work according to the organizational chart) and an informal network (e.g. with whom employees discuss work related matters next to the coffee machine) simultaneously.

In this research paper, the first intra-organizational social network we will examine is

the advice network. This network is an informal social network that does not stick to the

organizational structure but looks at who turns to whom when they are in need of advice or

information in order to enhance their own ability at the job. The second intra-organizational

social network we will focus on is the so-called hate network. On the contrary, the hate

network looks at with whom a certain employee does not want to collaborate with. This

makes it an especially interesting social network as it can be expected to function in sort of an

opposite manner as compared to other social networks. More specifically, actors in the hate

network are expectedly more likely to avoid each other than to be in a lot of contact with each

(14)

14 other. Despite the fact that it would be enormously interesting to know more about the impact of the hate network, previous studies have not yet researched the possible effects that this social network may have. Therefore, this makes studying the role of this intra-organizational network in this context extra interesting.

2.6 Favorable position in the network

The individual positions of the actors in a social network are of essence to understand the influence of the network structure (Aalbers, 2012). Their individual network position is essential in determining the impact that the structure of the social network could have. Having a favorable position in the social network may enable the individual to benefit from the network to a greater extent than those individuals that do not have such a worthy position.

More specifically, having a favorable position means that the actor has a relatively central position or has a high amount of important linkages with the other actors within this social network. In other words, the more embedded the actor is in the social network, the more favorable his or her position is. The embeddedness of an individual in a social network can be referred to as its “centrality” in this network.

When looking at this network centrality in the light of this study, it means that

individual employees may have a more central location in a social network than some of the

other employees within the organization. This more central location may allow these

employees to benefit more from the social network. To sum up, when we look at what kind of

an impact a social network may have at the individual level, we can look at how central the

position of the individual is.

(15)

15 2.7 The moderating effect of an advice network

The first intra-organizational social network that we will investigate is the advice network. This social network shows us who turns to whom when they are in need of advice.

Thus, this network does not necessarily stick to just the formal interactions that are usually dictated by the structure of an organization. The most import aspect of such a network is that it may allow for the exchange of knowledge and information between the actors within the social network. According to Tsai (2002), social interaction provides channels for information exchange among the members of an organization. Numerous authors have pointed out this positive effect that social networks can have (Chow & Chen, 2008; Hansen, 1999; Rogers, 1995; Tsai, 2001). Hence, we can assume that the intra-organizational social network that is researched in this paper, the advice network, can also allow for the exchange of important information and knowledge.

We believe that the positive effect of an individual’s creative potential on his or her overall performance can be amplified by a relatively high centrality in the advice network because the social network can allow for a better spreading of knowledge and information.

We can expect this because we view the creative potential of an individual as a relatively fixed factor that does not change itself when the environment in which the individual functions changes (e.g. becomes more supportive). Again, creative potential only comprises

“the creative capacity, skills and abilities that the individual possesses” (Hinton, 1968, 1970).

However, a more supportive environment would amplify the positive effect of creative potential because it can be utilized to a greater extent. In other words, the important distinction here is that an employee’s creative potential would be exploited to a greater extent but the creative potential itself would not get higher.

Building on this argumentation, we believe that the advice network is a factor in an

individual’s environment that can allow for a higher exploitation of his or her creative

(16)

16 potential because there is a relatively high amount of social interaction among the actors that have a central location in the social network. Therefore, an individual’s creative efforts can spread through the advice network more easily too. Hence, creative ideas can spread and as a consequence, these ideas are more likely to be applied or become more valuable. What this means is that the individual’s creative potential becomes more exploited and thus causes an amplification of its effect on overall performance. Moreover, its effect may also be amplified because an individual’s creative ideas may be combined with knowledge, information or creative ideas of other actors in the network. According to Mumford et al. (1991), various sources of already existing knowledge can also be turned into novel ideas. To sum up the above argumentation, we expect that the advice network can allow for a higher exploitation of an individual’s creative potential and is therefore able to amplify its positive effect on the employee’s overall performance.

Moreover, the creative potential of an employee can also be exploited to a greater extent when he or she has access to certain information that can be rooted in the intra- organization social network. Different scholars have argued that knowledge and information that facilitates the creativity of the individual may very well be embedded in social networks (Kijkuit & van den Ende, 2007; Leenders, van Engelen and Kratzer, 2003). Thus, this is another argument for why having a valuable place in the advice network can cause an individual to exploit his or her creative potential to a greater extent. When this effect is present, creative potential can also influence the overall performance of the employee to a larger degree. Therefore, this may thus also lead to an amplification of the impact of creative potential on the employee’s overall performance.

Based on the arguments above, we believe that there is sufficient reason to believe that

a favorable position in an advice network can cause an amplifying effect on the relationship

(17)

17 between the creative potential and overall performance of an employee. This brings us to the second hypothesis:

H2: Centrality in an advice network will moderate the positive relationship between an employee’s creative potential and his/her overall performance in such a way that this relationship becomes stronger when centrality in the advice network is high rather than low.

2.8 The moderating effect of a hate network

Besides studying the effects that the centrality in the advice network may have, we will likewise focus on the possible moderating role of the hate network on the positive relationship between creative potential and overall performance of employees. The argumentations for the expected effect of the hate network are to a certain extent opposite to those that we have formulated for the advice network and will be discussed below.

The hate network shows us which employees prefer not to collaborate with other actors in the network. Therefore, actors in this social network that are linked by ties are more willing to avoid each other in the workplace and will more likely prefer to turn to another colleague within the organizational department. For that reason, we can expect that the social interaction among the different actors in this network is relatively low. As a result, we believe it is likely that the actors that have a high centrality in this social network will share less knowledge and information with each other. Consequently, this will thus also deter the spreading of an individual’s creative efforts throughout the organization. This makes it less likely that such efforts will be combined with other knowledge, information or creative ideas of other actors in the network. As a consequence, the creative efforts of the individual are less probable to be applied or become more valuable.

What this means is that a central location in the hate network will cause a lower

exploitation of the individual’s creative potential. Therefore, the positive effect of an

(18)

18 employee’s creative potential on his or her overall performance would be weakened. Again, this also implies that we view the creative potential of an employee as determined by the individual and not by its environment. It merely looks at “the creative capacity, skills and abilities that the individual possesses” (Hinton, 1968, 1970). A less supportive environment (lower centrality in the hate network) would however weaken the positive effect of creative potential because it is utilized to a smaller extent.

Moreover, the creative potential of actors with a relatively high centrality in the hate network may also be less exploited because crucial information may very well be rooted in a social network (Kijkuit & van den Ende, 2007; Leenders, van Engelen and Kratzer, 2003).

We believe that it is likely that central actors in the hate network will have a reduced amount of access to such information. Therefore, these individuals might benefit from a smaller amount of possibilities to exploit their own creative potential. Subsequently, this would lead to a weakening of the influence of their creative potential on the employee’s overall performance.

In conclusion, our expectations for the hate network are more or less contradictory to those that we have formulated for the advice network. As opposed to the latter, we believe that having a central position in the hate network causes a lower exploitation of an individual’s creative potential. Therefore, it will have a weakening effect on the relationship between the creative potential and overall performance of individual employees. This leads us to our third, and last, hypothesis:

H3: Centrality in a hate network will moderate the positive relationship between an

employee’s creative potential and his/her overall performance in such a way that this

relationship becomes weaker when centrality in the hate network is high rather than low.

(19)

19 3. METHODS

3.1 Sample and Procedure

Our data was gathered at a large western manufacturing firm and at one of its foreign subsidiaries. This was possible because the focal firm outsources a part of its research and development activities to this overseas branch. Questionnaires were used to collect the data that is necessary to test our three hypotheses later on. We have distributed 90 questionnaires among the employees within a specific department of the organization. Of these employees, 77 responded and provided us with reports regarding their potential creativity, the advice network and the hate network. This means that there was a response rate of 69 percent.

According to Edwards et al., a response rate above 50 percent is sufficient (1997). The overall performance of these 77 employees was not rated by the employees themselves, but by their direct superiors (N=20). This has resulted in us obtaining performance data for 61 of those 77 employees. Thus, this results in a final sample size of N=61.

Of the final sample, 67 percent of the employees are active in firm’s home country, the remainder works at the overseas R&D subsidiary. Furthermore, 41 percent of the respondents stated that they have worked for the organization for less than 3 years, 26 percent of them have over 10 years of working experience at the company. The highest level of education for only 10 percent of the final sample is a bachelor’s degree. However, as much as 65 percent of them have successfully obtained a master’s degree and 25 percent has selected a Ph.D. as their highest level of education.

The employees and their supervisors both separately filled out their designated

questionnaires. Due to the fact that supervisors were asked to rate their employees regarding

their overall performance, both were required to fill out a number in order to make both sets

of data linkable.

(20)

20 3.2 Measures

3.2.1 Creative potential. The creative potential of employees was measured using the

6 item measure introduced by DiLiello and Houghton (2008). The 6 items were rated by the individual employees themselves on a 7-point Likert-scale (Likert, 1932). This scale ranged from 1 (strongly disagree) to 7 (strongly agree). Example items include “I feel that I am good at generating novel ideas” and “I have a knack for further developing the ideas of others”. All 6 items of this measure can be found in appendix A. We also performed a so-called component analysis using IBM’s SPSS in order to check whether these items all measure the same underlying concept. This analysis has resulted in all items loading on the same component (eigenvalue = 3.87; all loadings were above .67). Furthermore, the component explained 64 percent of the overall variance. A complete overview of the outcomes of the component analysis can be found in appendix C. Subsequently, we also checked for the Cronbach’s (1951) alpha for this measure. The alpha was .89. Based on the component analysis and the Cronbach’s alpha we found that it was not necessary to delete any of the items in this measure.

3.2.2 Advice network. The advice network was “mapped” by asking individual

employees to whom they turn when they are in need for advice. The questionnaire featured questions regarding with whom they thought they had the highest, moderate and lowest strength of relationship. Previously, we discussed that we can investigate the influence of a social network on an individual by looking at how central the position of this particular individual is. To date, various measurements for determining centrality have been developed.

The measure we use in this study is called the “in-degree centrality”. “An actor’s in-degree

centrality is measured as the number of times ego is mentioned by alters in a specific

network” (Aalbers, 2012). Hence, it only looks at how often an individual has been mentioned

by the other actors it the social network. Therefore, the construct is more reliable as it does

(21)

21 not include any self-reporting (Casciaro, 1998; Carleyand & Krackhardt, 1996). In the case of the advice network, we asked the employees to identify low, moderate and high strengths of relationships. However, we do not take this distinction into account during our statistical analysis. Hence, we use the in-degree centrality without tie strength for the advice network.

This means that all relationships that were mentioned by the employees have been assigned a value of 1 in the statistical analysis.

3.2.3 Hate network. The hate network was recorded in a comparable way as the

advice network was. However, in this case we did the mapping by asking individual employees with whom they would prefer not to collaborate with on a project. We did not distinguish between different “levels of dislike” for this measure in the questionnaire. Thus, an employee either does prefer not to cooperate with a co-worker or he or she does not.

Therefore, just like the earlier discussed procedure for the advice network, only one value is used when a relationship is present between two actors. Moreover, for the hate network, we also make use of in-degree centrality in order to measure the embeddedness of an individual employee within this social network.

3.2.4 Overall performance. The overall performance of employees was measured

using the 6 item measure of Mehra et al., (2001). This measure looks at both the innovation performance and overall performance of the employees. The 6 items were rated by the direct supervisors of the employees. Just as the measure for creative potential, overall performance was also rated on a 7-point Likert-scale. The measure includes items such as “What is the overall job performance of the individual?” and “Degree to which employee promoted and championed work-related ideas to others”. The full list of items can be found in appendix B.

A component analysis was also performed for this measure. The analysis has showed us that

all of the 6 items in this measure loaded on the same components (eigenvalue = 3.94; all

loadings were above .77). Furthermore, the component explained 66 percent of the overall

(22)

22 variance. A complete overview of the outcomes of the component analysis can be found in appendix C. Additionally, the Cronbach’s alpha for this measure was found to be .89. Again, we concluded that there was no need to delete any of the items in this measure based on the components analysis and Cronbach’s alpha.

3.2.5 Control Variables. We control for three different variables in this research. First

of all, we control for the total tenure of the employees. This variable measures how long a certain employee has worked for the company in total. Secondly we control for the location of employees. This can be either at the headquarters in the home country or at the overseas research and development subsidiary. Lastly, we also control for the level of education of the employees. The level of education was either a bachelor’s degree, master’s degree or a completed Ph.D. We control for the above three variables because they may also be influencing the performance of the individual employees. Thus, in this way we can make sure that they do not influence the final results of the regression analysis.

3.3 Statistical Analysis

A statistical analysis was performed using IBM’s SPSS in order to test the formulated hypothesis. The network data was put into a network analysis program named “Ucinet 6.0”

(Borgatti et al., 2002) using matrixes which were created from the collected survey data in

“Microsoft Excel”. Next, this network data was exported to another program that was created for the analysis of large networks and is called “Pajek64 4.01a” (Bategelj & Mrvar, 1996).

Subsequently, the in-degree centrality was calculated using the latter software. Furthermore,

we also used this software in order to create images of the structure of the intra-organizational

social network. These illustrations that represent the network structure can be found in the

results section.

(23)

23 For the statistical analysis the two measures (in-degree centrality) for the advice and hate network were centered on their mean. Subsequently, the data set was first explored with a descriptive analysis. Next, we looked if we could find any important correlations among the used variables. Lastly, we used a linear regression analysis to test our hypothesis. In order to compute the interaction terms, we have standardized both the independent and the two moderating variables. The three control variables were inserted as dummy variables with a value of either 1 or 0. There is no need to standardize these control variables because they are applied as dummy variables.

The first dummy variable that we will enter in the regression analysis is the level of education. This dummy represents a high or low level of education. More specifically, employees with either a bachelor’s or master’s degree have been assigned a value of 0 and employees that have completed their Ph.D. a value of 1. The second dummy variable represents the location of employees; here a value of 1 was used for the employees in the home location and 0 for those in the overseas subsidiary. The last dummy variable represents the total tenure of employees at the organization. For this variable, a distinction was made between employees with a high total tenure and a low total tenure. More specifically, individuals with over 5 years of working experience at the firm have been assigned a value of 1 and employees that have been at the firm for a shorter period of time a value of 0.

Furthermore, we have also performed a check in order to find out whether there is a need to be concerned about multicollinearity between the different variables in our dataset or not.

The regression analysis was performed using the following steps; firstly, the dummy

variables for our three control variables were entered into model 1. In the second step, our

independent variable, creative potential, was entered. Next, the centrality in the advice

network was entered which was followed by the interaction term consisting of this centrality

variable and creative potential. After this, the two were removed and centrality in the hate

(24)

24 network was also entered. Subsequently, this variable was followed by the interaction term entailing this variable and creative potential. Lastly, all variables were entered in order to create a final model.

4. RESULTS 4.1 Network structure

In order to provide an overview of the structure of both the advice and hate network, both networks were drawn using the Pajek software. This has resulted in the images below that represent this structure (figure 2). In the image, the actors in the social network are represented by the blue circles; the lines between them represent the relationship among the actors. Furthermore, the sizes of the blue circles in the image represent the in-degree centrality of an actor. Thus, the bigger the blue circle that belongs to an actor, the greater the actor’s in-degree centrality is.

When we first look at the image of the structure of both social networks it becomes

evident that the advice network is considerably more “elaborate” than the hate network. In

other words, the first social network has both a lot more actors and relationships (ties) among

these actors. Also, we can quite clearly perceive two separate “clusters” within the advice

network. This distinction can likely be explained by the different geographic location of the

actors in this network. On the other hand, the hate network does not feature many actors and

relationships among them. Apparently, only very few employees within the organization have

specified in the questionnaire that they prefer not to collaborate with a specific actor on a

project.

(25)

25 Figure 2. Structure of the advice network (left) and hate network (right).

4.2 Descriptive statistics

All the results of the descriptive analysis can be found in table 1. This table includes the means, standard deviations and correlations of all the previously discussed variables that are used in this research.

Here, we will highlight some of the more striking correlation among the different variables in the statistical analysis. For starters, we have found an interesting and substantial correlation between the creative potential of employees and total tenure (r = .28, p < .05).

Moreover, we have also found two substantial correlations of the advice network with both

the location (r = .26, p < .05) and total tenure (r = .49, p < .001) of employees. The other

intra-organizational social network, the hate network, has a significant correlation with the

level of education (r = .34, p < .01). Furthermore, we have also identified a substantial

correlation between the two social networks (r = .29, p < .05). Lastly, we found a correlation

between the overall performance of employees and the centrality in the advice network (r =

.32, p < .05) too. As we mentioned earlier, the complete overview of correlations and

descriptive statistics can be found in table 1.

(26)

26 Table 1. Descriptive Statistics and Correlations among the Variables

Variables Mean SD 1 2 3 4 5 6

1 Level of education .25 .43

2 Location .67 .47 .32*

(p = .013)

3 Total tenure .49 .50 -.11

(p = .421)

.20

(p = .126)

4 Creative potential 5 Advice network 6 Hate network

7 Overall performance

5.79 .24 -.01 3.64

.70 5.26 .34 .79

-.12

(p = .361)

.11

(p = .392)

.34**

(p = .007)

-.10

(p = .457)

.13

(p = .312)

.26*

(p = .048)

-.04

(p = .765)

-.12

(p = .346)

.28*

(p = .028)

.49***

(p = .000)

.20

(p = .121)

.17

(p = .201)

.19

(p = .150)

.09

(p = .470)

.32*

(p = .014)

.29*

(p = .026)

.15

(p = .259)

.12

(p = .342)

N = 61. Level of education, location and total tenure at the firm are represented by dummy variables with a value of 1 or 0.

*p < .05, **p < .01, ***p < .001

(27)

27 4.3 Hypotheses testing

An overview of the outcomes of the regression analyses that was performed in order to test our hypothesis can be found in table 2.

First of all, we did not find indications for concern about multicollinearity issues when performing the regression analysis despite finding various correlations in the correlation matrix; all VIF-values were under 10 and the corresponding tolerances above .10.

Furthermore, it is hard to say which exact model provides us with the best overall fit.

However, models 2 and 4 seem to stand out as both their R

2

-value and F-value are relatively high (model 2; (R

2

= .14), (F = 2.26)) (model 4; (R

2

= .19), (F = 2.06)). Despite the better fit of models 2 and 4, model 7 remains the main model because all the expected relationships are tested in this final model.

A first look at the regression table shows us that our analysis has not resulted in many significant coefficients. We only find these for our independent variable creative potential in most of the 7 models. It is however interesting to see that the coefficient for this same variable becomes insignificant when the moderating effect of centrality in the advice network is added (model 4). Moreover, when looking at this model, this interaction term has a significance level that could almost be considered scientifically valid (b = -.23, p = .114). However, it is surprising that the coefficient is negative. A negative effect of the advice network would be contradictory to our initial expectations.

Subsequently, we can also look at our hypotheses in the light of the regression results.

Our first hypothesis (H1) stated that the creative potential of individual employees is

positively related to their overall performance. This first hypothesis was supported in most of

the models of our regression analysis. More precisely, we found several significant

coefficients (e.g. (b = .25, p < .05) in model 2) in the results of the regression. However, it is

(28)

28 Table 2. Results of Regression Analysis.

Overall performance

Variables 1 2 3 4 5 6 7

Level of education Location

-.05

(p = .838)

-.26

(p = .283)

.02

(p = .924)

-.32

(p = .166)

-.00

(p = .991)

-.34

(p = .145)

-.05

(p = .842)

-.33

(p = .157)

-.04

(p = .880)

-.29

(p = .231)

.03

(p = .924)

-.28

(p = .252)

-.06

(p = .848)

-.30

(p = .221)

Total tenure .30

(p = .154)

.19

(p = .375)

.11

(p = .656)

.10

(p = .670)

.15

(p = .484)

.17

(p = .452)

.10

(p = .693)

Creative potential Advice network Hate network

.25*

(p = .022)

.25*

(p = .026)

.08

(p = .463)

.17

(p = .140)

.15

(p = .226)

.25*

(p = .028)

.06

(p = .591)

.21†

(p = .081)

.08

(p = .503)

.16

(p = .180)

.13

(p = .329)

.05

(p = .698)

Creative potential X Advice network Creative potential X Hate network

R

2

.05 .14 .15

-.23

(p = .114)

.19 .14

-.18

(p = .291)

.16

-.21

(p = .216)

-.07

(p = .709)

.19

Adjusted R

2

.00 .08 .07 .10 .07 .07 .07

F 1.07

(p = .367)

2.26†

(p = .074)

1.90

(p = .109)

2.06†

(p = .074)

1.84

(p = .120)

1.73

(p = .132)

1.53

(p = .172)

∆R

2

.05

(p = .367)

.09*

(p = .022)

.01

(p = .463)

.04

(p = .114)

-.04

(p = .591)

.02

(p = .291)

.03

(p = .401)

N = 61. Unstandardized regression coefficients are reported for the respective regression steps.

Level of education, location and total tenure at the firm are represented by dummy variables with a value of 1 or 0.

† p < .10, * p < .05, ** p < .01, *** p < .001

(29)

29 important to mention that no support was found for this relationship in the model that entails all of our variables (model 7).

Hypothesis two (H2) predicted that the centrality in an advice network would moderate the positive relationship between an employee’s creative potential and his/her overall performance in such a way that this relationship becomes stronger when centrality in the advice network is high rather than low. We did not find support for this hypothesis in the results of the regression analysis (b = -.21, p = n.s.). Despite this, it is still interesting to see that adding the interaction term containing centrality in the advice network causes the coefficient that represents creative potential to become insignificant. Hence, some kind of effect is apparently present.

Our third and last hypothesis (H3) stated that centrality in the hate network will moderate the positive relationship between an employee’s creative potential and his/her overall performance in such a way that this relationship becomes weaker when centrality in the hate network is high rather than low. We also did not find support for the last hypothesis in the results of the regression analysis (b = -.07, p = n.s.).

5. DISCUSSION 5.1 Findings

The overall aim of this research was to investigate the effects of intra-organizational social networks on the relationship between the creative potential of individual employees and their overall performance. We focused on both the so-called advice network and hate network as moderators on this specific positive relationship.

Our look at the illustration that represent the structures of both social networks (figure

2) shows us that the advice network is far more elaborate than the hate network is. It features

a greater amount of actors and ties. Furthermore, the descriptive statistics have shown us the

(30)

30 means and standard deviations of the variables in the analysis. More importantly, it has also presented us with some interesting correlations among these variables.

The results of the regression analysis have not provided us with many significant coefficients across the various models. These results do however allow us to conclude whether there is support for our hypotheses or not. Despite the fact that the findings of the regression analysis do largely support the expected positive relationship between an individual’s creative potential and overall performance (H1) in almost all of the 7 models, we did not find support for a moderating role of either the advice network or hate network (H2, H3) on this first positive relationship.

A different aspect that we want to highlight here is that our independent variable creative potential has insignificant regression coefficients in the models where the interaction term between centrality in the advice network and creative potential is added. Moreover, this interaction term is associated with a significance level that is almost smaller than p = .100 in model 4. The reason for highlighting this is that apparently some kind of effect of the interaction term is present despite it not being significant. What is also interesting is that this interaction term is associated with a negative coefficient in model 4 (b = -.23, p = .114). Our expectations for the moderating role of the advice network where the opposite.

A possible reason that might explain why we did not find support for the moderating

effect of both intra-organizational social networks is the issue of multicollinearity among our

predicting variables. Despite the fact that we did find some significant correlations among our

variables (table 1), we did not detect indications for concern when looking at the collinearity

diagnostics during the regression analysis. Hence, we do not think it is likely that it is the

cause of not finding many significant results. Another possible explanation for the

insignificant results for the second and third hypotheses may be explained by the rather small

sample size of this study. However, this also implies that the effects which we do find to be

(31)

31 significant are quite meaningful. We will elaborate further on this specific issue in the “strong and weak points” section.

5.2 Theoretical implications

First of all, this study confirms the positive effects of an employee’s creative potential on his or her overall performance at the job. For theory, this once more stresses the absolute importance of creativity in the workplace and its effect on performance outcomes. Hence, it supports the view that creativity can be seen as one of the crucial factors that have a positive effect on the perceived performance of employees.

Secondly, we did not find support for the moderating effects of the advice and hate network on the relationship between the creative potential and overall performance of employees. Hence, the outcomes of this study suggest that the role of intra-organizational social networks is not as influential as we expected beforehand. For theory, this might mean that the advice and hate network are not to be considered very relevant contextual factors within organizations.

It is especially deplorable that we did not establish the expected moderating effect of the hate network because previous research has not really focused on this concept. If we would have found a significant role for the hate network in the researched context this could have been more convincing to also explore its role in additional settings. However, despite the fact that this paper has not been able to identify any significant impact of both researched intra-organizational social networks, we would not recommend to completely ignore the potential effects of such social networks.

Due to the aforementioned notion, we believe future research is still necessary to

explore a wide variety of paths related to (intra-organizational) social networks. For starters,

novel research could focus on other social networks that may exist within an organization

(32)

32 (e.g. the innovation network) and explore their influences in the context of creativity and performance among employees. Moreover, we would also encourage more research on other effects of the hate network because it is still a rather uninvestigated concept. We believe that it will probably be extremely valuable to learn more about the, expectedly negative, consequences of an individual’s centrality in a hate network. The potential impact of the hate network can of course still be explored in numerous settings. Lastly, another direction for future research would be to focus on the network level effects instead of only focusing on the impact of social networks on the individual employees.

5.3 Practical implications

The research has shown the creative potential of employees has a positive impact on their overall performance. Therefore, this paper also brings forward a more practical point of interest. Because the results of this study show us that the creative potential of individual employees has a positive impact on their overall performance, it is naturally quite important to make as much use of an individual’s creative potential as possible. Thus, the results of our study emphasize the need for management to exploit the creative potential of individual employees to the largest extend possible because it helps them in contributing to organizational ends. Managers can possibly achieve this by creating an organizational environment in which employees feel free to share and bring forward their creative ideas and efforts. In other words, it is important for management to build a supportive environment in which creativity can flourish.

Furthermore, in this study we expected that a high embeddedness in the advice

network would be one of the factors that helps employees to exploit their creative potential

and thus increasing its effect on their overall performance. However, as previously mentioned,

we unfortunately did not find support for this expectation. The same applies to our

(33)

33 expectations for the hate network. Hence, a practical implication of this research could be that managers do not have to make an attempt to improve the social networks and interactions of their employees. For example, organizing an intervention that would make employees closer to one another might be redundant.

5.4 Strong and weak points

In retrospect, we believe that several strong and weak points of this research can be identified. A first weakness that we would like to mention is the size of the sample that was used for this research. The final sample size of 61 respondents is not extremely low when performing a network analysis. However, a larger sample could have made this research more statistically valid and may have yielded more significant results. It is nonetheless something that is quite customary when performing a social network analysis because doing this on an even larger scale would be particularly time consuming.

Secondly, when we look closer at the data that comprises both intra-organizational

social networks, it becomes evident that especially the hate network is rather limited. A

reason for this could be that employees in this organization generally do not have a lot of co-

workers with whom they do not like to cooperate on a project. However, this data could also

be biased because individuals may be reluctant, despite the anonymous questionnaire, to write

down those collogues with whom they prefer not to collaborate with. On the other hand, the

amount of collected data for the advice network is much higher, possibly because employees

do not have negative associations with this type of social network. This is something that of

course might be the case for the hate network. A strong point concerning our network data is

that we used the in-degree centrality as a measure. This makes the use of this data more

reliable as it does not rely on self-reporting (e.g. an individual does not obtain a high

centrality by simply mentioning a high amount of actors).

(34)

34 Another limitation of this research is that it does not entail any effects on a network level. We have only looked at what the impact of an intra-organizational social network can be on the individual. However, interesting effects on the larger network scale may also exist.

Hence, future research regarding this may also be interesting.

A possible strong point of this research is that our data was collected in an international setting. This is a positive aspect as it entails respondents from more than one cultural background and may thus make our findings more generalizable. However, the scope is still rather limited because the data was collected in only two nations and within one specific organizational department. Hence, the findings could have been more generalizable if more organizations and countries had been included in the sample. For instance, an important contextual factor that might stand out when looking at social networks in a more international setting could be the degree of collectivism (Hofstede, 1983). Perhaps, the centrality in the advice network is far more important in a highly collectivistic culture. Moreover, this, and other cultural dimensions can vary greatly across nations.

Another possible strong point of this research is the use of a multi-source method.

What this means is that we obtained data through more than one source. Most important here is that the centrality of an individual in one of the social networks is determined by the actors in that specific network. Moreover, the overall performance of employees was rated by their supervisor. Both measurements result in scores that are more objective because they do not rely on self-assessments. Hence, these results are less biased. On the contrary, the creative potential of individual employees does rely on self-assessment.

To sum it all up, this research has enforced the evidence for a positive relationship

between the creative potential and overall performance of individual employees. We were

however not able to find support for the moderating role of centrality in either the advice or

hate network. The finding may convince managers of the importance of creativity in the

(35)

35 workplace and may thus push them towards creating a more supporting work environment.

Furthermore, this research is characterized by several strong and weak points and we have

formulated various ideas for future research.

(36)

36 6. REFERENCE LIST

Aalbers, H. L. (2012). Organizing intra-organizational networks for innovation.

Amabile, T. M. (1996). Creativity in context.

Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of management journal, 39(5), 1154-1184.

Amabile, T. M., & Khaire, M. (2008). Creativity and the role of the leader. Harvard Business Review, 86, 100– 109.

Bategelj, V., Mrvar, A. (1996). Pajek64 4.01a.

Borgatti, S., Everett, M.G., and Freeman, L.C. (2002). Ucinet 6 for Windows. Analytic Technologies.

Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization Science, 22(5), 1168-1181.

Carley, K. M., & Krackhardt, D. (1996). Cognitive inconsistencies and non-symmetric friendship. Social networks, 18(1), 1-27.

Casciaro, T. (1998). Seeing things clearly: Social structure, personality, and accuracy in social network perception. Social Networks, 20(4), 331-351.

Chow, W. S., & Chan, L. S. (2008). Social network, social trust and shared goals in organizational knowledge sharing. Information & Management, 45(7), 458-465.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of

tests.psychometrika, 16(3), 297-334.

(37)

37 Cummings, A., & Oldham, G. R. (1997). Enhancing Creativity: Managing Work Contexts for

the High Potential Employee. California Management Review, 40(1).

DiLiello, T. C., & Houghton, J. D. (2008). Creative potential and practised creativity:

Identifying untapped creativity in organizations. Creativity and Innovation Management, 17(1), 37-46.

Edwards, J.E., Thomas, M.D., Rosenfeld, P., & Booth-Kewley, S (1997). How to conduct organizational surveys: a step-by-step guide. New York: Sage Publications.

George, J. M. (2007). Creativity in Organizations. The academy of management annals, 1(1), 439-477.

Gong, Y., Huang, J. C., & Farh, J. L. (2009). Employee learning orientation, transformational leadership, and employee creativity: The mediating role of employee creative self- efficacy. Academy of Management Journal, 52(4), 765-778.

Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative science quarterly, 44(1), 82- 111.

Hinton, B. L. (1968). A model for the study of creative problem solving. The Journal of Creative Behavior, 2(2), 133-142.

Hinton, B. L. (1970). Personality Variables and Creative Potential. The Journal of Creative Behavior, 4(3), 210-217.

Hofstede, G. (1983). National cultures in four dimensions: a research-based theory of cultural differences among nations. International Studies of Management &

Organization, 13(1/2), 46-74.

Referenties

GERELATEERDE DOCUMENTEN

Sensitivity of the simulated mean diurnal patterns of latent (a) and sensible (b) heat fluxes to the parameter m in the modified root water uptake function and the total soil

As important third cornerstone towards a continuous improvement process in companies, the machine list - in terms of power, time and the estimated energy consumption - has to

Economic performance is defined as income, whereas artistic performance is set up according to the selection system theory, divided in market, peer and expert performance.. This

Het gebrek aan publieke erkenning heeft Plooij ongetwijfeld dwarsgezeten, maar uit zijn laatste - nu postuum gepubliceerde - roman Het landgoed blijkt dat hij de roem ook met

With regard to our last hypothesis, an expected positive relationship between supervisory support, in the form of verbal persuasion, and employees’ creative self-

Hypothesis 1b: Respondents who participated in a chronic physical exercise programme will demonstrate higher increases in measured skills in creative thinking from

This hypothesis states that motivation has an interacting effect on cognitive style and satisfaction, such that brainstorming group consisting of a high average level of

The present study contributes to what is still unclear, and examines the influence of the regulatory focus of leaders, the leader’s emotional expressions