MASTER THESIS
ELUCIDATING CREATIVITY IN VIRTUAL TEAMS
A SOCIAL PERSPECTIVE -
By
KARST WESTRA
University of Groningen
Faculty of Management and Organization MScBA Strategy & Non-Technical Innovation
May 2007
Supervisor: Dr. Thijs L.J. Broekhuizen Co-assessor: Dr. Hendrik Snijders
Van Sijsenstraat 43a 9724 NL Groningen +31 (0)6 45166674 karst.w@gmail.com Student number: 1337262
ABSTRACT
This study explores how creativity in high degree virtual teams can be stimulated. A social perspective was adhered to resulting into the discerning of three interrelating antecedents to virtual team creativity. The frequency of communication, two subcomponents of intra-team trust and the degree of shared mental models were included into a conceptual research model.
Furthermore, virtual team age and virtual team experience were incorporated as two important control variables. Derived from extensive literature research, the three antecedents were each hypothesized to exhibit inversely U-shaped relationships to virtual team creativity. Too low or too excessive degrees were expected to inhibited virtual team creativity whereas moderate to moderately high levels were expected to increase creativity levels in virtual environments.
Subsequently, an empirical study was performed directed to testing the avowed hypotheses.
Moreover, a greatly homogeneous set of respondents was examined by means of a multiple choice questionnaire rooted in previous academic studies. The results exhibited by correlation and regression coefficients implied that benevolence as a subcomponent of trust and the degree of shared mental models bear a statistically significant inversely U-shaped relationship to virtual team creativity levels. Furthermore, virtual team age displayed significant impeding results whereas virtual team experience demonstrated a non-significant positive relationship. The relationship between the frequency of communication and virtual team creativity did also not prove sufficient alignment and was accordingly rejected.
Following the outcomes, a number of managerial implications were raised. In brief, concerning the degrees of mental models an initial and moderate overlap of understandings should be obtained, possibly by bringing together the team at the outset. Furthermore, to achieve the required cognitive overlap, it was suggested to explicitly create a common set of norms, context and team purpose. To not exceed moderate levels, a system of alternation was proposed.
Managers should consider altering a team’s composition at times, possibly concurrent with distinct team creativity phases.
To attain early levels of benevolence, initially drawing together team members was also proposed. Moreover, it was recommended to select team members that have a higher capacity to trust and a preference to independent working. Additionally, a method of swift trust building was shown to suit virtual teams seamlessly. To keep the obtained levels of trust in position, the system of alternation also seems of great consequence to raise creativity levels.
Finally, limitations inherent to the study and suggestions for future research were provided.
ABSTRACT 2
CONTENTS 3
CHAPTER 1
INTRODUCTION 4
CHAPTER 2
VIRTUAL TEAMS
2.1 Virtual Teams Opposed to Collocated Teams 7
2.2 Continuous Virtual Teaming Dimensions 7
2.3 Advantages of Virtual Teams 10
2.4 Important Disparities of Virtual Teams 12
2.5 Conceptualization of Social Constructs 14
CHAPTER 3
CREATIVITY IN VIRTUAL TEAMS
3.1.1 Intangible Creativity Clarified 16
3.1.2 Divergent Thinking Fluency as a Creativity Measure 17
3.2 The Significance of Team Creativity 18
CHAPTER 4
INTERLINKING SOCIAL ANTECEDENTS OF VIRTUAL TEAM CREATIVITY
4.1 Virtual Team Communication 19
4.2 Virtual Team Trust: Reliability and Benevolence 21
4.3 Virtual Team Shared Mental Models 23
4.4 Synopsis of the Hypothesized Relationships 25
CHAPTER 5
EMPIRICAL EXAMINATION
5.1 Methods 26
5.2 Sample 27
5.3 Instruments 27
5.3.1 Frequency of Interaction 28
5.3.2 Trust 28
5.3.3 Shared Mental Models 28
5.3.4 Team Creativity 29
5.3.5 Additional Control Measures 29
CHAPTER 6
RESULTS
6.1 Correlation Analyses 31
6.2 Regression Analyses of Virtual Team Creativity 33
6.2.1 Control Variables 33
6.2.2 Regression Analysis 34
CHAPTER 7
DISCUSSION AND IMPLICATIONS 38
ACKNOWLEDGEMENTS 45
REFERENCES 46
APPENDICES 53
1. INTRODUCTION
Creativity is increasingly becoming a key topic for present-day organizations to improve competitiveness. One of the most decisive motives for organizations to induce creativity has been a general shift in terms of market competition. The industrial economy has been succeeded by the information economy while now gradually a creative knowledge economy is starting to dominate the market (Caves, 2002). This growing power of generating meaningful and novel ideas and consequently striving to market these successfully has created a significant need for organizational creative efforts. Furthermore, it has required companies to create and introduce products more rapidly.
In today’s competitive business environment where this speed-to-market and creativity are crucial, teaming also plays a vital business role. Key contributors are brought together to focus on common problems and to develop innovative solutions. Not just at an individual level but also at a team level. Continuous evolving information technology has facilitated the bringing together of these key experts from anywhere across the globe. Moreover, this form of organizing ascertains a variety in cognition and information of team members to develop novel ideas (Wuyts et al., 2005). Valuable teaming opportunities do thus not limit themselves to colleagues in one’s own organization, but stretch to people everywhere.
In line with this, GartnerGroup research (Bell & Frey, 2002) estimated that by 2002 over 100 million people worldwide were working outside traditional offices to a certain degree. Moreover,
“by 2006, in 75 percent of enterprises, project teams will have virtual membership; in more than half of these enterprises, some project team members will be outsiders (80 percent probability)”.
As a consequence of this vogue to meet new market standards and the need for creativity head- on, teams have to communicate with colleagues while lacking face-to-face contact. These distributed work groups are accordingly also labeled ‘virtual teams’. Ideally, they allow organizations to utilize the very best talent to create high quality innovative solutions and rapid responses to market and customer needs. Furthermore, as competition has shifted to a global level, the need for expertise and information has increased. Although not easy, virtual teams can leverage this tacit knowledge by bringing people together on projects without collocating them.
Additionally, myriad possible benefits can be expected from working virtually, among which;
- Reduced real estate expenses (Cascio, 2000)
- Environmental benefits (Benson-Armer, Hsieh, 1997) - Access to global (talent) markets (Kerber & Buono, 2004) - Increased speed-to-market (Bell & Frey, 2002)
- Increased productivity and higher profits (Cascio, 2000)
- Enhanced innovation and creativity (Bell & Frey, 2002; Nemiro, 2004).
From these possible benefits enhancing creativity is the core of this research. “Virtual team structures may actually lead to higher levels of team creativity as a result of more openness, flexibility, diversity, and access to information than exists in traditional group structures”
(Nemiro, 2004). On the other hand, it may be extremely thorny to develop and maintain a social context that encourages high levels of creativity. In devising virtual teams, one cannot disregard the context of such understanding. “No matter how complicated the next technology may seem it is still the human that is the most complex, flexible and adaptive part of the system” (DeRosa et al., 2004). Moreover, “virtual teaming is 90 percent about people and 10 percent about technology” (Bell & Frey, 2002). The by now frequently researched matter of technology allows for the electronic connection of virtual teams, but it does not necessarily lead to a fitting social context. In this research this social context, consisting of a number of factors, is seen as an important antecedent to creativity as output variable. The fundamental and decisive answers for these problems are thus positioned in the field of human and organizational relations, a nearly blank field regarding virtual teaming. “Research on social factors necessary for fostering creativity in virtual teams is however rare” (Nemiro, 2000).
Virtual teams with a high understanding regarding creativity will facilitate organizations meeting the new market conditions. Accordingly, generating meaningful and novel ideas, innovation and speed-to-market will likely be enhanced. Moreover, organizations and virtual team members themselves must have knowledge of how to induce and attain high levels of creativity in these novel and ever more appearing teams. The key research goal of this unexplored area which is central to this study is:
To provide information to organizations and virtual team members regarding differentiating social virtual team factors, in comparison with collocated teams, that instigate virtual team creativity, as well as to provide specifics concerning the ways these factors interrelate.
As a consequence of limitations of research to social factors, technology and environment factors are not taken into consideration when no direct relationship with the research variables exists. As affirmed, human and social factors essentially account for team and creative outcomes (Bell &
Frey, 2002; DeRosa et al., 2004). Furthermore, technology on the subject of virtual teams has already been extensively researched by a multitude of scholars (e.g. Leenders, 1995; Pauleen &
Yoong, 2001). Linking virtuality, creativity and social dynamics to each other is however a wholly new perspective. Academic literature has, somewhat strangely, yet paid adequate notice to this field and these factors combined.
In addition, the delimitation instigates the fact that all sizes of virtual teams are able to benefit from the understandings provided by this research. Human and social factors are omnipresent in all teams while technology varies largely according to manifold issues.
Research Outline
The first part of the research is ordered as follows; after the introduction, in the second chapter, virtual teams and their most critical distinguishing characteristics vis-à-vis collocated teams are discussed to establish a solid shared comprehension. Subsequently, a conceptual model sketching out hypothesized relationships among the discerned variables is put forth. Next, in the third chapter the output variable creativity is researched and linked with virtual teams. In the fourth chapter communication, trust and mental models are studied. When conducting research on these antecedents of virtual team creativity, several hypotheses are set forth. Moreover, all variables are interrelated while staying connected to a virtual setting.
In the second part of the research, starting from chapter five, the asserted hypotheses are tested empirically. A multiple choice questionnaire is distributed among various virtual teams in which creativity as an output is of importance. Outcomes are statistically processed after which correlations and regressions among variables are scrutinized and explicated. In the final sections, results and deviations are discussed and related back to academic literature. Finally, triangulation strengthened conclusions are drawn up, limitations inherent to the research are discussed and recommendations for future research are presented.
2. VIRTUAL TEAMS
At the outset, it is imperative to gain a thorough and shared understanding of the type of teaming that is the core of this research. Therefore, in the forthcoming sections appropriate and defining facets of virtual teams are closely examined. This is to be regarded as the backbone of the in- depth examination of the social dynamics that are studied in the subsequent chapters.
2.1 Virtual Teams Opposed to Collocated Teams
First and foremost, virtual teams have to be perceived as a team. They comprise numerous shared characteristics with conventional collocated teams. “A virtual team is a group of people who work on interdependent tasks guided by a common purpose” (Kerber & Buono, 2004). Moreover, virtual team members must establish communication, plan, subdivide their work, sequence, and schedule and accomplish coordination (Hertel et al., 2004; Workman, 2005).
However, unlike traditional collocated teams, virtual teams possess a number of unique traits.
Virtual teaming comprises “working across space, time and organizational boundaries that often extend across nations on a global basis” (Montoya-Weiss et al., 2001). Furthermore, multi- organizational teams, different time zones, higher task interdependencies, different cultures and withdrawal of non-verbal cues are included into these differing characteristics. Together, they cause substantial difficulties to virtual teaming when opted for this design. To elucidate these defining characteristics, they are represented in Figure 1 below.
Figure 1.
Distinguishing Characteristics of Virtual Teams and Collocated Teams
Shared Team Characteristics - Multiple individuals
- Shared goals - Organizational setting
Virtual Teams Collocated Teams
- Geographically dispersed - Low visibility
- Communication through technology - High(er) task interdependence - High(er) cultural differences
- Face-to-face - High visibility
- Communication primarily face-to-face - Low(er) task interdependence - Low(er) cultural differences
Uniting all of the above characteristics in a concise manner, I define a virtual team as:
A group of people who work interdependently and with a common purpose across space, time and organizational boundaries through the use of technology to facilitate communication, collaboration and defined end-results.
Participation in these virtual teams may be temporary for certain members and team boundaries will vary according to the specific project conditions. Furthermore, I do not assume that virtual team members under no circumstances meet face-to-face. It is however assumed that nearly all communication occurs through technology-supported types of media. Therefore, a high degree of virtuality is present.
2.2 Continuous Virtual Teaming Dimensions
While at this point the significant key differences between virtual teams and collocated teams are indicated, important commonalities exist too as declared in Figure 1. By reason of these commonalities, I see it as non-rational to depict virtual and collocated teams as complete opposites without continuous axes. Therefore, I propose that virtual teams range on specific virtuality dimensions. Herewith, neither fully collocated nor fully virtual teams can be expected to come about in pure forms.
Figure 2 explicates the degrees which can be told apart according to the aforementioned team characteristics of Figure 1. These individual dimensions thus evolve directly from the set apart features of virtual teams vis-à-vis conventional teams. Moreover, in this figure a novel way is exhibited in which a team’s degree of virtuality is placed onto multiple continuous dimensions, starting from fully collocated at the core and leading to high degrees of virtuality at the tops.
Figure 2.
Continuous Dimensions of Team Virtuality
Temporal &
Physical Distance (High – Low/
Level of
Technical Support (Low – High)
Strength of Team Member Links/
Interaction (High – Low)
This figure is to be viewed as a novel combination of the ideas of a number of scholars. Kratzer et al. (2006) have previously publicized that virtuality can undoubtedly be viewed as a continuum.
This view I combine with the empirical research of Griffith et al. (2003), who explored how virtuality controls the aspects that affect the dissemination of knowledge. They consequently divided virtuality into three measurements: 1) the level of technology support used by a team, 2) the proportion of work the team performs with its members distributed across time or space and 3) the allocation of the locations occupied by team members. The second factor of virtuality is, for the purpose of this research, substituted by a continuum of the strength of team member links and interaction. This dimension directly relates to task interdependence as put forth in Figure 1.
This interdependency, how work of one member impacts another, is for a plurality of reasons habitually higher than in collocated teams. Due to an absence of social control derived from power, interdependencies are made higher to secure a social control mechanism (Sillince, 1999).
Furthermore, it is shown that higher task interdependencies are present since it is seen to increase the transfer of highly diverse tacit knowledge which is inherent to higher degree virtual teams (Nonaka & Takeuchi, 1995). Moreover, the implemented factor team member links can be seen as channels, interactions and relationships that create the cohesiveness of a team unfolding over time. “From the view of a system, the group of independent members operating their independent tasks has the cooperate goal and through multiple media and boundary-crossing interaction, the group members build their mutual trust to finish concrete result” (Lipnack & Stamps, 1997).
Second, the level of technical support as a distinct dimension of virtuality can be seen as directly related to communication as stated in Figure 1. While communication in collocated teams primarily occurs face-to-face, it will primarily occur by means of technology in virtual teams.
This communication technology makes virtual team forming with members that do not essentially work in close proximity viable. Moreover, the working across space, time and organizational boundaries consequently reduce the probability that teams, which are homogeneous regarding culture and knowledge, are created for convenience (Griffith & Neale, 2001). Additionally, group support technology can offer added functionality by constructing a team’s tasks, facilitating analysis of the team’s process, and the storage of the team’s information (Griffith et al., 2003).
However, the functionality of such technologies is ever changing and more importantly, socially constructed.
Third, the continuum of temporal and physical distance between team members is to be seen as related to dispersion and visibility of Figure 1. Lipnack and Stamps (1997) suggest that virtual teaming already becomes reality when team members are physically located over 50 feet away
from each other. The underlying thought of this factor is that possible team members are able to correspond with one another. This correspondence makes it possible to familiarize, to recognize shared interests and assess interpersonal compatibility. “Shared experiences foster liking, mutual understanding and increase the perceived similarity toward each other” (Leenders, 1995).
Figure 3.
Linking Distinguishing Characteristics to the Virtuality Dimensions
Distinguishing Characteristics of V.T. Virtuality Dimensions
Communication through Technology Level of Technical Support
Dispersion,
Visibility Temporal and Physical Distance
Task Interdependence Strength of Team Member Links/
Interaction
2.3 Advantages and Disadvantages of Virtual Teams
Following the explication of the key differences and characteristics of virtual teams, an explanation for the fast increase in numbers is now presented. In the initial sections the possible benefits of virtual teams over collocated teams have already been briefly indicated. These advantages led to the sharp rise of virtual teaming in the final decade as avowed by the GartnerGroup (Bell & Frey, 2002). Furthermore, disadvantages counterbalancing the possible benefits are also outlined in greater detail to establish a deeper understanding regarding the possibly rewarding realms of virtuality.
Reduced real estate expenses are seen as imperative when convening a virtual team. Due to eliminating office spaces for employees, except those who truly need them, expenses can be cut substantially. Cascio (2000) states that, when implemented correctly, estimate savings will tally up to $2 for every $1 invested.
Environmental benefits are not at all times seen as directly connected to feasibility but might for future reference become of paramount importance. Already, organizations are claiming this as a strong purpose when creating an image of environmental awareness. A striking example is provided by Benson-Armer and Hsieh (1997) “A U.S. government study showed that if 20,000 federal workers could telecommute just one day a week, they would save over two million
commuting miles, 102,000 gallons of gasoline, and 81,600 pounds of carbon dioxide emissions each week”.
Access to global (talent) markets is also a key essential, especially crossing the mentioned space, time and organizational boundaries. The pool of resources including critical contributors can be greatly expanded “including people from outside the sponsoring organization, such as consultants, supply chain affiliates or members of partner organizations” (Kerber & Buono, 2004). Furthermore, talent might more easily be included in a virtual team and talent that is already in-house might be sheltered and retained.
Improvement in speed-to-market is seen as of great consequence as well when scrutinizing possible benefits. Empirics show that “cycle times are improved by moving information quickly, improving coordination and accelerating work processes” (Bell & Frey, 2002). Moreover and as indicated, this is regarded a vital organizational requirement to keep pace with contemporary market conditions.
Increased productivity and higher profits are also clearly essential when contemplating whether to convene a virtual team or not. In a study by Cascio (2000) it is demonstrated that
“internal IBM studies show gains of 15 to 40 percent in productivity. USWest reported that the productivity of its teleworking employees increased, some by as much as 40 percent”.
Furthermore, “Hewlett Packard doubled revenue per salesperson after moving its salespeople to virtual workplace arrangements”.
Innovation and Creativity are likely enhanced in virtual settings as stated by Nemiro (2004).
This “as a result of more openness, flexibility, diversity, and access to information than exists in traditional group structures”. Moreover, these processes are enhanced due to “empowering the virtual team members with the authority and processes to work independently” (Bell & Frey, 2002). Among others, these factors as associated to the key research variable of virtual team creativity are incorporated within the current research.
Setup and maintenance costs are to be considered the first to be mentioned disadvantage.
According to statistics by Cascio (2000), the additional costs for individual employees required to prepare a mobile office varies from about €2,500 to €4,000 plus approximately €750 in upgrades and supplies every subsequent year. Additionally, to be viable, virtual teams call for a large deal
of online content that is reachable from distant locations. Technology must thus function without a flaw and technology support should be broadly available. For managers this implies that incremental setup and maintenance costs need to taken into account.
Loss of cost efficiencies concerns the second disadvantage of virtual teams. In collocated teams, equipment and services are typically concentrated in one site to which as a result multiple team members can utilize them. In virtual teams however, this equipment and these services are physically distributed to which cost efficiencies may be lost. Contradicting the mentioned cut of real estate costs, it is more cost effective to have all team members collocated at one single site.
Cultural issues also play a large blocking role in virtual environments. High degree virtual teams frequently have to transfer their business strategies and culture to collaborate with physically and temporally dispersed members. This might bring about potential clashes of business and national cultures, which can destabilize the complete association. Moreover, “if the members of a virtual organization or a virtual team are not empowered to make decisions, the technology that enables their collaboration will add little value, and the competitive advantage associated with rapid responses to demands in the market place will be lost” (Cascio, 2000).
Feelings of isolation is regarded the final important disadvantage of virtual teaming. Certain levels of social interaction with superiors and other team members can be considered imperative.
Lacking it, team members will feel isolated and out of the sphere of essential communication and contact with decision makers who can make or break them.
Connecting the above advantages and disadvantages, a comprehension is set out in which it is shown that virtual teams may offer ample organizational rewards but also imply shortcomings.
2.4 Important Disparities of Virtual Teams
Completing the above is a multitude of disparities which must be carefully studied prior to deciding whether and how to convene a virtual team. Moreover, due to virtual teams increasingly receiving positive notice, managers may not offer this adequate consideration. The following disparities are a direct follow-up and are directly related to the virtuality dimensions of Figure 2.
Trust, mental models and communication can be seen to differ and influence each other in dissimilar manners in virtual settings compared to conventional collocated settings. Furthermore, these three factors combined fittingly possibly induce the research output variable creativity.
Below, trust, mental models and communication are briefly delineated whereas in the subsequent chapters a further in-depth exploration is executed and relationships are explored and established.
Figure 4.
Linking Social Virtual Team Factors
Distinguishing
Characteristics Of V.T.
Virtuality Dimensions Important and
Differing Social Virtual Team Factors Communication through
Technology
Level of Technical Support Trust,
Mental Models, Communication Dispersion,
Visibility
Temporal and Physical Distance
Trust, Mental Models, Communication Task Interdependence Strength of Team Member
Links/ Interaction
Trust, Communication
Building trust and a trusting environment are seen as crucial and differing elements to team work. Conventional social control derived from power lacks, so as a consequence, self-direction and self-control is vital. Moreover, virtual team members need to be certain that everyone complies with their responsibilities and behaves in a consistent, foreseeable way. In line with this, trust can be defined as “the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustee, irrespective of the ability to monitor or control that other party” (Mayer et al., 1995). However, due to the inherent characteristics and virtual team dimensions, getting hold of trust will be even harder, and the importance is even greater, than in collocated teams.
Alongside this lack of social control “the lack of face-to-face interaction with its associated verbal and non-verbal cues and the synergies that often accompany face-to-face communication”
(Cascio, 2000) is an additional strong downside.
Shared mental models are a second important element to establish an efficient and creative virtual team. “Creation of a common understanding is necessary for group performance” (Yen et al., 2006). These common understandings, relate to “the overlapping, parallel, compatible, or dispersed knowledge that involves attitudes, beliefs, ideas and expectations that is used to coordinate behavior” (Mathieu et al., 2000). Or they can be seen as “The extent to which members of a team have a similar understanding of task demands, environmental contingencies,
or appropriate behavior for team members” (Eby et al., 1999). Moreover, sharing mental models should be emphasized due to a strong relationship between shared mental models and a higher team performance. Although this element is of significance to any team, due to virtual teams’
differing communication, dispersion and limited visibility this is however more difficult and more important to obtain.
Good communication being the third social factor, acts as a facilitator for team performance.
Communication broadly refers to the “the formal as well as informal sharing of meaningful and timely information between persons” (Anderson & Narus, 1990). Moreover, managers and virtual team members face a great challenge when different cultures, and therefore likely also differences of language and mental models, are present in a team. Sending and understanding information will clearly be more difficult by reason of such conditions. Furthermore, compared to conventional collocated teams, several more distinguishing facets have to be considered when looking at virtual team creativity.
As acknowledged, on the positive side, working across space, time and organizational boundaries lowers the likelihood that teams which are homogeneous regarding culture and knowledge are created for convenience (Griffith & Neale, 2001). Additionally, group support technology may provide further functionality by facilitating analysis of the team’s process, shaping a team’s tasks, and the collecting of the team’s information (Griffith et al., 2003). However, such technologies are socially constructed and thus utilizing available information and knowledge differs largely from conventional collocated teams. Moreover, when a team moves further upon the virtual dimensions, this difference will become yet larger.
2.5 Conceptualization of Social Constructs
Following the exploration of multiple virtual team facets and a build-up toward differing social virtual team factors in appearance and importance, a conceptualization is now exhibited. A conceptual model is the visual representation of predicted functional relationships among components of a system or process. The following conceptual representation encompasses the socially constructed variables; communication, trust, shared mental models, creativity and innovation as components of a process. The first three are depicted as part of a network model each containing reciprocal relationships. When one variable is influenced positively, the linked variable will experience a similar positive influence. Conversely, when a variable is influenced in a negative manner, the connected variables will undergo a similar negative influence.
Although an initial starting point in the model can thus not be indicated, the research line of Anderson and Narus (1990) is followed. They conducted valuable empirical research regarding connections amid trust and interaction declaring the latter is to be seen as the antecedent of trust.
Following the three antecedents, the output variable virtual team creativity is positioned. It is assumed that, when combined fittingly, the antecedents will lead to greater creativity. However, if they are trapped in a negative spiral, virtual team creativity is assumed to lower significantly.
Finally, as outlined in the following sections regarding creativity and innovation, a strong relationship exists between creativity and innovation (Nemiro, 2004). These variables can to some extent be seen as intertwined processes. The concept of innovation, however, extends one step further and beyond the scope and core of this research. Moreover, by examining relationships among the discerned antecedents of creativity, successful product introductions into the marketplace cannot be inferred. Innovation is nevertheless a frequently sought after organizational end result and thus ought to be included conceptually to generate a comprehensive view. Accordingly, amid virtual team creativity and innovation a dashed borderline is placed.
This line represents the fact that innovation as a factor is not included in the research scope and, although a strong link with creativity is present, finds itself in the research environment.
The social constructs and predicted relationships, which come forth from the preceding figures and which are visually represented below, are further commented upon in the chapters three through four. Correspondingly, hypotheses regarding relationships which are put forth in these chapters are tested empirically in various high degree virtual teams.
Figure 5.
Conceptual Model of the Social Antecedents of Virtual Team Creativity
3. CREATIVITY IN VIRTUAL TEAMS
In the preceding chapters a shared understanding regarding the differentiating facets of virtual teams in relation to collocated teams was ascertained. Alongside, a model conceptualizing the assumed relationships of the stated antecedents to virtual team creativity was elucidated. Now following in this third chapter, an important foundation of creativity, the key research variable, is set out. Moreover, details are provided on various facets of creativity and divergent thinking fluency is presented as a team level creativity assessment measure.
3.1.1 Intangible Creativity Clarified
Although a large deal of scholarly attention has by now been devoted to clarify creativity, a multiplicity of definitions and measures is present. Even though many scholars find consensus in several facets including novelty and appropriateness (e.g. Amabile, 1988; West & Farr, 1990), operationally defining creativity remains elusive. Furthermore, there is no consent reaching agreement upon appropriate instruments and methods regarding this concept. “Measuring team level creative performance is notoriously difficult and no generally accepted measure exists”
(Kratzer et al., 2006).
Among the large array of definitions, Kristensson and Norlander (2003) consider creativity as an aspect of competitive advantage for organizations; “it is a resource for organizations to innovate new products that meet the needs of customers”. Amabile (1988) moreover reasons that creativity is shown when a product or service is made which is both novel and useful relating to the organization. Furthermore, Woodman et al. (1993) propose that creativity refers to “the creation of a valuable new product, service, idea, procedure, or process by individuals working together in a complex social system”. Additionally, King and Anderson (1990) state that creativity at the team level explicitly incorporates the interpersonal discussion among team members.
While some scholars (e.g. Houtz & Krug, 1995) deem this multiplicity of definitions and measures as indicative of the creative field, the definition of Woodman et al. (1993) will be adhered to throughout the research. This description comprehensively encompasses all relevant facets together and ties well with social factors as the core of this research. Furthermore, the above proposition by King and Anderson is seen as important while discussing communication, trust, and mental models. Each of these discerned variables compellingly relates to interpersonal communication among virtual team members.
Additional to the array of definitions, as delineated conceptually by means of an interconnecting arrow, a link between creativity and innovation exists (Nemiro, 2004). In a macro perspective as often adhered to by sociologists, the influence of the environment on creativity is often scrutinized (Ford, 1996). Moreover, this macro perspective has also been associated with a significant notice to innovation. “The intentional introduction and application…of ideas, processes, products or procedures, new to the relevant unit of adoption, designed to significantly benefit” (West & Farr, 1990). Since innovation entails novelty and appropriateness (benefit), it can be perceived to encompass creativity, however not counting successful adoption. Creativity is thus a necessary but not sufficient criterion to innovation (Amabile, 1988).
3.1.2 Divergent Thinking Fluency as a Creativity Measure
The second focal issue of which scholarly consent is yet reached upon concerns appropriate measuring instruments. Besides numerous definitions, a plurality of measures exists. The lion’s share of the contemporary creativity instruments herein is identified as divergent thinking or ideational fluency measures (Paulus, 2000). Contrary to traditional intelligence tests, which concentrate on converging on one right answer to a question or problem, divergent thinking focuses on generating multiple responses. Moreover, it is thinking that moves away in diverging directions involving a variety of aspects which may lead to novel ideas. As clarified by Runco (1999), “Because some of the resulting ideas are original, divergent thinking represents the potential for creative thinking”.
Scrutinizing the fluency or quantity of the resulting ideas is thus a frequently applied instrument to measure creativity performance. As however avowed by Kratzer et al. (2004), to date hardly any researches are completed in an innovation context and not any of these on the level of teams measuring a team its creative performance. Furthermore, the intricate virtual team characteristics pose an even greater deviation from existing research. Due to this, it is opted to utilize a measure taken from Cohen and Cohen (1991) closely linked to ideational fluency and which is perfectly able to be extended to virtual settings. Research subjects are inquired to rate their virtual team’s creative performance regarding the generation of new ideas, methods, approaches and inventions or applications.
Thus, as far as divergent thinking fluency is reliable and valid, it can be correctly utilized to estimate the potential for creative thinking. Moreover, it encompasses the research variable creativity, untying it from innovation which is regarded convergent thinking. In addition, the discerned variables each have a measurable influence on virtual team creativity.
Accordingly, the measure used to assess virtual team levels of creativity is to be perceived as a near self-report measure since the assessments are provided by the team members themselves.
These appraisals are however at times condemned for claims of inflated scores and subjects not being capable to report personal performance precisely. Inadequate processes of introspection are cited by Locke et al. (1988) as the foremost grounds for this criticizing. This involves the observation of one's own mental processes, thinking and emotional states. Since the research and measures however concern a team level output, self-reporting measures and inflated scores are not considered a problem. This given that respondents’ self-knowledge and thinking processes are to a lesser extent of significance and due to full anonymity and no existing personal rationales.
3.2 The Significance of Team Creativity
Creativity measured by means of divergent thinking fluency is thus exclusively included as output variable. Moreover, and as just declared, it concerns team level creativity. Prior scholarly research on this topic distinguished between individual and team level creativity. Whereas the former has to a great extent been researched, the subject matter of this study has not. This foremost, as declared by Kratzer et al. (2006), since measurement of team level creativity is awfully complex.
More importantly though, “the quality of new products and services is heavily dependent on the creative performance of individuals working in teams” (Kristensson & Norlander, 2003).
Research indicates that teams of individuals should be appointed to develop a new product or service, from idea generation to market introduction (Brown & Eisenhardt, 1995). Moreover, “the development process of new products and services has provided to be a complex and iterative process and is often performed by groups of people. However, product development is not just any organizational work process; it is a process highly dependent on creativity”.
Teaming is thus seen as critical to establish creative outcomes. I however require placing a caveat concerning the above. Empirical research has also implied that teaming to establish creativity is not at all times a best requirement as just suggested (Diehl & Stroebe, 1991; Pinsonneault et al., 1999). A number of these studies empirically show that a team can be surpassed by an identical amount of non-interacting individuals, often identified as nominal groups. It is yet commonly assumed that to bring about high levels of creativity, teaming is a vital prerequisite. Accordingly, this line of research is ensued. “Although the critical role of teams in fostering creativity has long been established, the question of how to configure such groups remains a critical problem of any new product development effort” (Leenders et al., 2003).
4. INTERLINKING SOCIAL ANTECEDENTS OF VIRTUAL TEAM CREATIVITY
Throughout the initial chapters three socially constructed variables have been discerned which conceptually are related to one another and virtual team creativity. Clarity is provided on this elusive research variable which stands out as particularly important for virtual teams to correspond to contemporary market conditions. In this chapter, communication, trust and mental models as antecedents to virtual team creativity are interrelated to establish a further understanding of their interactions. Moreover, hypotheses regarding these interrelations are constructed which are subsequently empirically tested.
4.1 Virtual Team Communication
Communication as a first discerned variable is in teaming literature habitually seen as the frequency of interaction (e.g. Anderson & Narus, 1990; Jarvenpaa & Shaw, 1998; Mohr, 1999;
Kratzer et al., 2004). Anderson and Narus define this factor as “the number of formal as well as informal times in which meaningful and timely information is shared between persons”. Bleeke and Ernst (1993) even put that interaction among team members is the foremost important factor to bring about successful cooperation; “the most accurately designed relations will crumble without good, frequent interaction”. In addition, within the research at hands it is assumed that a full balance exists between team members to reach creativity. Interacting team members therefore cooperate expecting to achieve positive outcomes and likewise convey information timely and accurately.
The relationship involving both communication and trust is earlier depicted as a reciprocal one within the research model. In line with Mayer et al. (1995) and Jarvenpaa and Shaw (1998), team members are to a smaller amount prepared to cooperate when low trust is found within a team.
Jarvenpaa and Shaw (1998) assert that frequent communication is the best manner to maintain trust which is deemed to occupy a central position in the communication process and collaboration. Nonetheless, establishing virtual trust is considered exceptionally thorny and important due to the previously discerned virtual team characteristics. Moreover, “the development of trust among virtual team members is slower than in collocated teams” (Wilson et al., 2006). Handy (1995) even issues whether virtual teams can operate successfully in the absence of frequent face-to-face contact. Besides Jarvenpaa and Shaw (1998), Mayer et al. (1995) empirically found that several factors, among which repeated interaction, are seen to facilitate interpersonal trust amid team members. Hence, a positive relationship is expected to be present.
Hypothesis 1a: Frequency of interaction is positively correlated to reliability and benevolence as elements of trust
Furthering to creativity, Anderson and Narus (1990) empirically affirm that creativity, with given personal expectation levels, show a significant positive correlation with frequency of interaction.
A clarification for this is reciprocal facilitation, whereby reaching outcomes that surpass expectation levels facilitate good communication between team members. Consequently, communication will facilitate achieving outcomes that surpass the expectation level. It can thus be said that; the number of formal and informal times in which meaningful and timely information is shared, will contribute in an iterative manner to the surpassing of expectations and to virtual team creativity.
This positive relationship amid frequency of communication and creativity is however not as straightforward as appears at first glance. As indicated by Pinsonneault et al. (1999) creativity in team communication can be either improved or decreased regarding quality and quantity. Factors that improve and enhance a team’s idea generation productivity are referred to as process gains.
On the other hand, one can witness process losses, factors that decrease and inhibit a team’s idea generation and creativity. Scholarly attention regarding these process losses has argued the positive influence of communication on team creativity. Leenders et al. (2003) have encompassed this strand of attention concisely.
First, team member distraction is developed upon. “The presence of others can serve as a distraction and either energize people or disrupt their performance…Distraction can lead to attention conflict, which may mediate social facilitation effects and cause attention overload”.
Virtual team members will thus rely on cognitive shortcuts such as heuristics and pre-existing schemata. Consequently, a negative spiral-effect is presumed to be set off within teams when members distract one another. Therefore, high frequencies of interaction should significantly reduce individual levels of creativity and thus virtual team creativity.
Second, creative blocking is brought forth as impeding team creativity. “High levels of interaction may carry teams along by the momentum of their enthusiasm for an innovative idea, rather then by clear understanding of its real value…First, it can decrease the levels of critical thinking through which ideas are evaluated…Second, high frequency of interaction can push team members towards mutual beliefs”.
High levels of interactions are for these reasons assumed to impede virtual team creativity in a sloping manner. Furthermore, very low frequencies of team interaction are not functional to creativity either since the sharing of tacit knowledge will be outstandingly low. Therefore, it is assumed that;
Hypothesis 1b: Frequency of interaction will have will have an inversely U-shaped relationship to virtual team creativity
4.2 Virtual Team Trust: Reliability and Benevolence
Next, trust as a second distinguished variable with regards to collaboration between team members has been researched by a plethora of scholars (e.g. Bosch-Sijtsema & Rispens, 2003;
Cascio, 2000; Mayer & Davis, 1999). It implies the willingness to be vulnerable to another party which in turn means that a particular risk element is present; otherwise trust as a factor is not of importance (Mayer et al., 1995). When a team member trusts a colleague, chances are higher that the first mentioned will undertake actions of which the risks are higher when no trust would be present. Ring and Van de Ven (1992) state that team members are prepared to undertake these actions since the perceived risks appear smaller when interpersonal trust exists. Additionally, trust is all the more important in virtual settings since self-control is fundamental in these and situations are therefore perceived high of risk. Members of virtual teams need to be convinced that all members comply with their responsibilities and behave in a consistent, foreseeable way.
In teaming literature a number of distinct aspects of trust are often cited. Among these, reliability and benevolence have been identified as key when developing interpersonal cooperation based on trust (Kumar et al., 1995; Selnes & Gønhaug, 2000). A team member can be seen as reliable when “delivery is according to contract, when relevant information is provided timely and accurately, when members of the team are well informed concerning running business etcetera”.
Subsequently, the benevolence of team members can be seen as “the perceived willingness of a team member to act in a manner to which another party gains advantage”.
Both reliability and benevolence are by Anderson and Narus (1990) found to be empirical antecedents to interpersonal cooperation which in turn is related to expectations. It is assumed that when trust is augmented, for instance by communication, cooperation will also increase.
Cooperation is defined as “similar or complementary coordinated actions taken by persons in interdependent relationships to achieve mutual outcomes with expected reciprocation over time”
(Anderson & Narus, 1990). When trust is ascertained, team members perceive that coordinated, joint efforts will bring about results that surpass outcomes compared to members acting solely in their own concern. Perceptions and expectations thus occupy an important position in virtual teaming. Consequently, when members comply with their responsibilities and behave in a consistent, foreseeable way, expectations are met and tacit knowledge is shared. According to Wuyts et al. (2005) sharing this tacit knowledge among team members leads to a higher understandability of one another and thus to a greater overlap of mental models. As a result, it is assumed that;
Hypothesis 2a: Reliability is positively correlated to the degree of virtually shared mental models
Hypothesis 2b: Benevolence is positively correlated to the degree of virtually shared mental models
Turning to the conceptual linkage concerning reliability and virtual team creativity, a positive relationship is also expected. When team members are perceived reliable, the sharing of tacit knowledge will augment. Likewise, as a result of the great team heterogeneity, the tacit knowledge input will be diverse which is expected to bring about a greater novelty of ideas and creativity. Moreover, the team is more easily able to frame break, be less path dependent and make linkages between different domains. In effect, it can be said that the team members would intellectually cross-pollinate easily. Therefore, it is assumed that;
Hypothesis 2c: Reliability will have a positive relationship to virtual team creativity
Conversely, benevolence as a second differentiated antecedent of virtual team creativity is expected to take on another relationship vis-à-vis reliability. “Benevolence may block creativity through suppression of emotion, value of humility, silence ethic, conformity, and stigmatized eccentricity” (Kim, 2005).
Reasoning along similar lines as maintained in the paragraphs handling communication and creativity, very low levels as well as very high levels are seen as to impede creativity. It is thus assumed that;
Hypothesis 2d: Benevolence will have an inversely U-shaped relationship to virtual team creativity
4.3 Virtual Team Shared Mental Models
As a third discerned variable, mental models concern knowledge that involves attitudes, beliefs, ideas and expectations that is used to coordinate behavior. Although “creation of a common understanding is necessary for group performance” (Yen et al., 2006) virtual teams, contrary to collocated teams, will not possess these shared perceptions or expectations but they include diverse expectations (Bosch-Sijtsema & Sorge, 2003). As affirmed virtual teams are usually largely heterogeneous. They thus encompass members that stem from diverse organizations and cultures and moreover frequently possess dissimilar knowledge, this influencing the shaping of mental models. Due to this inherent heterogeneity, a virtual team’s cognition develops under mainly diverse circumstances. Furthermore, team members often have little history of previously having worked together and possess distinct experiences of working in teams (Bosch-Sijtsema, 2003). In addition, due to the temporal and physical distance as virtuality dimension, creating shared expectations or mental models becomes yet more difficult and critical. Along similar lines, Jarvenpaa and Shaw (1998) empirically observed that virtual team members are less committed and do not possess a shared culture or organizational identity.
Speaking in terms of Wuyts et al. (2005) the cognitive distance of virtual team members is very large and sharing tacit knowledge to achieve creativity is therefore extremely difficult. Moreover and as confirmed, shared perceptions or expectations differ substantially inside higher degree virtual teams. This discrepancy brings about a low understandability among team members which in turn impedes creativity.
Concurrent with a constructivist point of view, mental models however come into being through interaction with one’s physical and social environment. “Interaction allows for the creation of relationship-specific heuristics and shared mental models that reduce cognitive distance, and hence facilitate the transfer of the tacit components of knowledge” (Uzzi, 1997). It can thus be assumed that to bridge the significant gap of understandability amid virtual team members, the variable frequency of communication is fundamental. To substantiate the linkage in the research model between virtually shared mental models and the frequency of communication, a positive relationship is assumed to exist.
Hypothesis 3a: Degree of virtually shared mental models is positively correlated to the frequency of interaction
Inherent to this hypothesis an important trade-off should however be witnessed. In the first chapter, a large diversity in skills, tacit knowledge and mental models was brought forth as a core rationale to convene a virtual team. The novelty and newness value of ideas within virtual teams will increase due to a higher difference in cognition among team members (Wuyts et al., 2005).
Bridging the gap of understandability and mental models will thus inherently decrease this value of ideas and therefore the level of virtual team creativity. Moreover, McAllister (1995) empirically finds that cognitive distance decreases as interaction is more frequent. Therefore, it is assumed that;
Hypothesis 3b: Degree of virtually shared mental models will have an inversely U-shaped relationship to virtual team creativity
Figure 6.
Relationship between Shared Mental Models and Virtual Team Creativity Adjusted model based on: Wuyts et al., 2005
Generally expected Optimum
in V.T Degree of shared mental models
At very low degrees of shared attitudes, beliefs, ideas and expectations among virtual team members, the understandability will be very low whereas the novelty and newness value of ideas will be very high. Consequently, knowledge flows will be insufficient to obtain high levels of virtual team creativity. Moreover, this large organizational focus yields a risk of myopia or group think (Kratzer et al., 2006; Wuyts et al., 2005), each team member attempts to conform his or her opinions to what they believe to be the consensus of the group.
Conversely, at high degrees of shared mental models, the interpersonal understandability of members within a heterogeneous group will be very high. The consequential organizational myopia or groupthink will result in significantly lowered levels of virtual team creativity.
Novelty Value Understandability
Creativity
4.4 Synopsis of the Hypothesized Relationships
To provide a concise summation, all relationships between the variables, as acknowledged by means of hypotheses throughout the chapter, are outlined below in the research model. Bold, one- headed lines represent one-way relationships with creativity whereas the thin, two-headed lines represent inter-antecedent relationships. Furthermore, the connection amid trust and virtual team creativity is split up into the subcomponents benevolence and reliability since these are expected to exhibit different relationships. Conversely, the other relationships concerning trust are persistently avowed to show positive associations.
Figure 7.
Overview of the Hypothesized Relationships
+ +
+ + + +
Rel. + Ben.
5. EMPIRICAL EXAMINATION
From academic literature appears a need to empirical research regarding socially structured facets of virtual teaming. Communication, trust and shared mental models have step by step been discerned as important and differing variables in these. Their relationships as depicted in the conceptual research model have in the former chapter been substantiated by literature. To scrutinize these findings empirical examinations are conducted.
5.1 Methods
The utilized method regarding the gathering of empirical research details is the cross-sectional or social survey design. This involves a research design which encompasses the collection of data on more than one case and at a single point in time. This is executed so as to obtain an ample body of quantitative data in connection with the variables as represented conceptually. Moreover, this form of design is particularly appropriate and widely used for the purpose of descriptive analysis (de Vaus, 2001). In addition, it is supported with extensive sampling theory and multivariate analysis by which subsequently patterns of association including correlations can be examined.
Regarding the cross-sectional design adhered to; this study includes a multiple choice questionnaire developed to encompass the discriminated variables and their subcomponents.
What is more, included teams and their members were approached in a fully standardized manner. An information sheet containing all necessary information was sent alongside the questionnaire. By way of this standardization, social inference and a consequential distortion of results is attempted to be excluded. Moreover, common methods bias, when the instruments employed enter into or affect the scores or measures that are being gathered, is also attempted to be excluded by this means.
The employed method is thus rooted in a combination of academic literature combined with primary data and individual understandings. In this manner, various data sources are adhered to attributing to the research’s triangulation value. This is to be seen as “two or more methods of data collection in the study of some aspect of human behavior” (Burns, 1990). Moreover, by reason of these multiple ways of data collection, the reliability and validity are enhanced (Yin, 2003). Deviant or off-quadrant dimensions of a problem may be uncovered, leading to enriched explanation of the research problem and the synthesis and integration of theories is facilitated (Jick, 1976).
5.2 Sample
For this study data is employed on a total of 78 respondents encompassing 14 different teams distributed over 6 distinct organizations and institutions. With consent, a detailed directory of partaking organizations is to be found as a second appendix.
All fourteen teams which are included into the research were cautiously selected to fit the strictly bounded research criteria. First, each team had as a specific objective to establish creativity in their defined end-results. Specifically, data is collected on New Product Development teams, Research and Development teams and Innovation teams. Second, solely teams which act on higher degrees of the dimensions of virtuality were approached. Especially the degree to which technical support was utilized and the temporal and physical distance of team members was perceived vital in the selection process. Intra-team face-to-face contact has been tried to cut out as completely as possible. As a result of adhering to these research boundaries, the incorporated set of respondents is very homogeneous contributing highly to the data validity.
The rate of return of the survey, including follow-ups and reminders, entails approximately 48 percent. A total of 78 questionnaires were successfully completed and returned whereas the 14 teams comprise, somewhere in the order of, 162 members. Note that approached organizations and teams which did not collaborate are not included into this figure. Also, the completion rate is an estimate for reasons inherent to virtual teaming. Exact member quantities of teams cannot be inferred due to virtual team membership extending beyond boundaries. Teams overlap frequently which influenced respondents’ answers concerning team size. Consequently, averages of these replies were computed trying to approach the actual rate of return.
This rate is an important indicator of how much confidence can be placed in the survey results and statistical analyses conducted. Generally, in these types of research a response rate of 48 percent is perceived moderately high; therefore it does not pose true setbacks. Moreover, the non- respondent sample bias is expected to be low. These members’ answers are predicted not to diverge from the actual respondents their answers due to the survey design and standardized approach.
5.3 Instruments
Of great consequence, all variables and their subcomponents have been previously empirically validated in scholarly research. Below, the use of instruments regarding each factor is expounded upon in greater detail. Furthermore, Cronbach's as a reliability measure and a number of control variables which are expected to influence team level creativity are outlined.
5.3.1 Frequency of Interaction
The frequency of interaction among virtual team members as a component of intra-team communication is based on a five-point Likert scale structured as follows. 1= not at all, 2=
monthly, 3= weekly, 4= several times a week, 5= daily. Reliability with regards to this measure has been extensively proven by, among others, Anderson & Narus (1990), Mohr et al. (1996) and Mohr (1999). The one-tailed correlations coming forth from this, measure the intensity of interaction reported by the team members varying from 0 (no communication at all) to 1 (daily communication). To exclude extreme values, two items were posed concerning the frequency of interaction with the two persons with whom respondents have the most and least contact with.
Finally, the reliability measure Cronbach's , or the extent to which the set of test items can be treated as measuring a single latent variable, is 0.65. As a rule of thumb, a reliability of 0.60 is often referred to as desired.
5.3.2 Trust
To ascertain the overall level of virtual intra-team trust, this measure is subdivided into two separate subscales, benevolence and reliability. These variables were measured on a five-point Likert scale, ranging from 1= strongly disagree to 5= strongly agree. Reliability regarding this scale in conjunction with subscales of trust has been empirically confirmed by, among others, Kumar et al. (1995) and Mayer et al. (1995). Both variables were tested by three modified items each as applied by Mayer and Davis (1999). These items however describe a general construct of trust whereas the nature of this research is to explore virtual teams. Therefore, it was imperative to alter the applied items of trust to a certain extent to fit research objectives. Consequently, benevolence exhibited a good Cronbach's of 0.69, whereas reliability showed a relatively low Cronbach's of 0.56.
5.3.3 Shared Mental Models
To assess the elusive variable shared mental models among virtual team members, the measure is subdivided into four subscales which combined represent individual-level shared mental models.
The subscales as used are; planning and organization, coordination, communication and analysis and are derived from Eby et al. (1999). These variables were as well measured on a five-point Likert scale, ranging from 1= strongly disagree to 5= strongly agree whereas the original scale was anchored in a three-point scale. Every subscale encompasses three items tallying to a total of twelve items. Reliability and content validity have been strongly established with regards to the used items (Eby et al., 1999) where as a rule card scoring or concept mapping tools are applied.
