Projects and Networks: Understanding Resource Flows and Governance of Temporary Organizations with Quantitative and Qualitative Research Methods

Abstract

Project-based organizing is becoming increasingly common, but projects are challenging for managers because they must coordinate resources, including people and information, under time pressure to achieve a one-off outcome. In this article, we suggest that a network lens is ideal for researching project coordination because it enables the interested party to study the flows of resources and the structures that allow the project to be effectively governed. Both qualitative and quantitative, including graph-theoretic methods for analyzing networks are suitable for examining project network governance and network resource flows. However, governance tends to be studied more with qualitative methods, and resource flows are more commonly researched with social network analysis. The article concludes with considerations regarding multi-method approaches to the study of project networks.

Keywords

project research networks qualitative quantitative network analysis governance

Introduction

Projects represent a form of organizing that is growing in economic significance. More economic activity is organized in and around projects than ever before and the scale of these projects is also increasing, with growing attention not only on ‘megaprojects’ but also on ‘terra projects’ (Flyvbjerg, 2014). Like other forms of organizing and organization, projects exist to coordinate and control the production of goods and services but unlike other forms, projects are temporary with a designed beginning and end (Bakker, DeFillippi, Schwab, & Sydow, 2016), even though they are often not finished in time or—like the new international airport in Berlin—do not seem to end. Projects, however, are rarely set up and carried out in isolation, neither with respect to time nor social relations. For this reason, Engwall (2003) insightfully stated that “no project is an island.” This becomes particularly clear from a relational or network perspective that has increasingly been applied in project management research during the last decade or so. Projects are connected to other projects through space and time, but they are also frequently intertwined with permanent organizations.

Looking at projects as an organizational form invites a network view to understand the dynamics of the systems that are associated with this type of temporary organization. Projects come into existence to combine diverse inputs into a unique output. Projects must therefore coordinate people, materials, and knowledge and the flow of these resources. And these ‘project systems’ (Packendorff, 1995; Shenhar & Dvir, 1996) can be represented as a network. However, the temporary and often unique nature of project organizations also implies a governance challenge that is atypical of most organizational forms; therefore, projects must also control and govern activity and these patterns of governance and power can be described as a network too.

Because network theory or network analytical approaches are so adept at explaining resource coordination and governance (Borgatti, Brass, & Halgin, 2014), there is a natural fit with these core concerns of managing projects or temporary forms of organizing more broadly. The applicability of a network lens to the understanding of these temporary systems has been well-appreciated by the project research community with an increasing number of papers being based upon some form of network concept.

Figure 1 provides support for the rising importance of network approaches and the impact it has achieved over the past 15 years in providing solutions for problems and theory development in the field of temporary organizing. The compound annual growth rate (CAGR) of the number of publications using network approaches was 15.5% and for research combining network approaches and projects of all types it was 14.2%. The two graphs on the right side of Figure 1 show this growth in network approaches and is not yet fully reflected in the field of project management (CAGR of 7.6%), nor in dedicated project management journals (CAGR of 10.4%) for example, International Journal of Project Management, Project Management Journal^®, and International Journal of Managing Projects in Business. The combination of (1) the importance of networks in research across disciplines, (2) the importance of networks in project research, and (3) its potential to be applied more in the field of project management and in dedicated project management journals provides the foundation of the purpose of this article as well as this special issue.

Figure 1

The growing importance of networks in the project management literature.

Indeed, the growing importance of network research to the project management field is reflected in the contributions to this special issue. We were pleasantly surprised by the number and standard of submissions, which cover a range of topics, settings, and research methods. After the review process was completed, there were too many articles for a single volume, so the special issue has been spread across two issues of Project Management Journal^®. The articles by Pryke et al., Klaster et al., Laursen et al., Adami et al., and Åberg et al. will appear in this issue. The remaining articles by Terhorst et al., Lu et al., Takahashi et al., Braun et al., Eriksson et al., and Kaulio will appear in a later 2018 issue.

Researching project networks, however, does not have a singular methodology. A network view means explaining project phenomena by looking at relationships between people, resources, and processes. It examines the relational embeddedness of an actor with another actor through the strength and variety of connections between them as well as the structural embeddedness of an actor with the surrounding network and the connections within this broader network (Grannovetter, 1985). These interactions can be examined by using social network analysis (SNA) (Wasserman & Faust, 1984), a quantitative methodology based on graph theory in mathematics. SNA provides researchers with a wide range of possible measures of network properties that relate to underlying characteristics of the network and thus presents several ways of operationalizing structural embeddedness. Relational embeddedness, however, is more difficult to capture in social network analysis and is usually captured in the strength of ties between actors, which is measured as the frequency, duration, or intensity/importance of the interaction. Research carried out with public and private sector client organizations in construction (Pryke, 2012, 2017) and more recently in infrastructure schemes (Pryke, Badi, Soundararaj, Watson, & Addyman, 2015) has shown that tie strength is important in our understanding of relational embeddedness in project networks.

At least as important, projects are also embedded in networks from a governance perspective, which highlights the collaborative quality of networks, conceived as either a hybrid form of governance between market and hierarchy or as a distinct form beyond market and hierarchy (Jones, Hesterly, & Borgatti, 1997; Sydow, Schüßler, & Müller-Seitz, 2016). The investigation of project networks from a governance perspective is mostly of a qualitative nature, making frequent use of case studies. This form of research is well suited to studying relational embeddedness, where case studies and interview-based data reveal more about the types of connection between the actors and the meanings they ascribe to these interactions. On the other hand, structural embeddedness is harder to capture with qualitative information unless many actors are interviewed about their relationships with others in the network. This is where quantitative network analysis, involving secondary data or surveys of actors and who they are connected to, becomes most useful.

In the following section we will present both perspectives on projects and networks in more detail and sort the contributions included in this issue into either the SNA or the governance perspective on project networks. Both methodological approaches will be evaluated in how they might approach these governance and coordination questions in project research settings. We conclude by considering the possibility of combining qualitative and quantitative research in future studies of project networks.

Projects from a Network Governance Perspective

Project networks as a specific form of governance are to be found in film and television production, in science-based industries such as biotech or optics, but also in the event and construction industry. Sydow, together first with Windeler (Windeler & Sydow, 2001) and later Manning (Manning & Sydow, 2011) conducted a series of studies of the television industry in Germany and found that each interorganizational (multi-party) television project depended on the viability of longer-term, more sustainable networks, which he characterized as ‘project networks.’ In its purest form, project networks embed projects, being a form of temporary organization (Lundin & Söderholm, 1995), into longer-term, eventually even open-ended networks of relationships. Thus, project network relationships may be embedded within such disparate networks of inter-organizational relations as strategic alliances, business ecosystems, regional clusters, and global production networks (Sydow et al., 2016). Our special issue includes project networks embedded in a regional government cluster (Klaster, Wilderom, & Muntslag, 2018) and in the European Cultural Capital ecosystem (Laursen, 2018).

The study of these networks as a distinct organizational form, characterized by latent as well as activated ties among project entrepreneurs and/or organizations, supplements the research on classical project-supported (PSO) as well as truly project-based (PBO) organizations (Lundin et al., 2015). What is more, this organizational form seems on the rise in an economy that increasingly relies on project business (Arrto & Wikström, 2005; Davies & Hobday, 2005; Lundin, Arvidsson, Brady, Eksted, Midler, & Sydow, 2015), perhaps even outpacing the general trend toward projects alluded to earlier. Managing projects in interorganizational networks may be analytically distinguished from managing PSOs and PBOs by the comparison of distinctive governance modes adopted within each type of project organization.

Previous research on project governance in PBOs and PSOs has focused on governance mechanisms at work on the level of single projects to ensure a predictable delivery of projects in time, quality, and cost (Ahola, Russka, Artto, & Kujula, 2014; Müller, 2009) and/or on the level of the organization. While even extremely decentralized project-based or project-supported organizations typically rely on hierarchical fiat (Williamson, 1985), project networks rely on different governance modalities. What coordinative mechanisms substitute for or complement hierarchical governance in project networks is subject to some debate (DeFillippi & Sydow, 2016). While some researchers conceive networks as “hybrids” (Williamson, 1991) combining market as well as hierarchical modes of governance, others see networks as an organizational form “beyond market and hierarchy” (Powell, 1990).

One mechanism often cited in the literature on project networks is relational governance and the associated reliance upon trust and/or reciprocity to extend those relationships beyond a single project contract. Jones and Lichtenstein (2008) proposed that trust evolves out of prior network relations that reduce transactional uncertainty and increase the shared understanding needed for effective coordination. Such trust does not exist in a vacuum but rather it is manifest in specific reciprocal transactions between network partners that shape partner expectations for future interactions (Shazi, Gillespie, & Steen, 2015; Swärd, 2016). Crucially, even specific types of contracts (those of a more long-term and/or strategic partnership type arrangement) may contribute to the emergence rather than the undermining of trust (DeFillippi & Sydow, 2016). Hence, positive experiences in previous interactions (“shadow of the past”) and expectations of possible future collaboration (“shadow of the future”) can shape how relationships are extended beyond temporary project engagements among network partners. This is what is meant by temporal embeddedness of projects.

In addition to relational governance, project networks may also be governed by the distinctive roles played by project participants in networks and the functional interdependence between these roles. For example, complex construction projects have well-defined roles for contractor, subcontractor, and component supplier participants (Pryke, 2004, 2017; Davies & Hobday, 2005; Steen, Ford, & Verreynne, 2017). These roles are project specific, insofar as an organization may be a contractor in one project, a subcontractor in a second project, and a component supplier in a third project, depending upon the respective capabilities required by each project. Moreover, by taking on different roles within different projects, the project participants both widen and deepen their relevant project experience and thus create further opportunities for their selection to participate in an array of future projects. Hence, successful performance as a subcontractor on a large complex project may provide the gateway to subsequent participation as a contractor on either a smaller or less complex project in the future. Hence, a project-based career history of project organizations and their individual and team participants may become evident by examining the historical trajectory of their changing roles within projects of varying size, complexity, and reputational importance or status.

A distinctive set of administrative coordinative roles may govern project networks. For any network, authority to coordinate may be organized in at least three forms (Provan & Kenis, 2008): Network governance can (1) be shared among the participating members, (2) a lead organization may govern the network, or (3) a network administrative organization (NAO) may be at the core of network governance (an NAO is a dedicated organization responsible for coordinating the network or at least supporting such processes). The effectiveness of each of these three governance modes hinges on particular boundary conditions that, according to Provan and Kenis (2008), include the level of trust, the number of participants, the degree of goal consensus, and to what extent there is a need for network-level competencies. To give just one example, shared network governance seems to require more goal consensus and trust and, thus, may be suitable only for networks of smaller size. All three forms of governance may also be used in project networks, including lead organizations. This form, like the NAO form, seems to be particularly appropriate, perhaps even indispensable, for larger and dynamic (project) networks in which shared governance, not least because of a lack of goal consensus and trust, may be difficult to establish (DeFillippi & Sydow, 2016). Our special issue includes a conceptual article comparing the roles and tasks of project management offices (A PMO is an organizational unit that supports projects and the management of project-based firms) and NAOs that coordinate member contributions in a project network (Braun, 2018).

Project networks may also be governed by a set of routines based on successful practices employed previously. Routines are repetitive patterns of interdependent actions (Parmigiani & Howard-Grenville, 2011). Davies and Brady (2000) develop the concept of economies of repetition to show that PBOs can offer “repeatable solutions by recycling experience from one project for others in the same line of business” (Davies & Brady, 2000, p. 932). Crucial to the achievement of economies of repetition is the very development of routines, which may—as interorganizational routines—also be in effect in project networks. Once having undertaken a one-off project, the same participants are involved in successive ones of the same type in order to consolidate routines, which they adapt according to the contingencies of each project (Manning & Sydow, 2011; D'Andrea, 2014). García-Canal, Valdέs-Llaneza, and Sánchez-Lorda (2014) argue that when developing new collaborative projects with the same partner, firms even tend to repeat the same contractual form used in previous projects to take advantage of the governance routines developed in the past.

Project networks as a form of governance, thus, reflect tensions between the temporary and the permanent, not unlike but also different from PSOs and PBOs, which, in the end, can rely on hierarchical fiat to also manage these tensions, for instance with regard to the priority of projects. However, the identification of what is “permanent” and what is “temporary” can be problematic and even arbitrary. The dominant project organizational orthodoxy seems to characterize projects as the temporary phenomena compared with the more permanent organizational structures, institutions, and networks in which it is embedded (Bakker, 2010; Burke & Morley, 2016, Grabher, 2004; Schwab & Miner, 2008; Manning & Sydow, 2011). Bakker et al. (2016) also note that the definitions of temporary versus permanent seem rather arbitrary and context-dependent and they recommend that the label “temporary” be employed to describe only organized activities of a determined duration, in other words, whether at the outset (predetermined) or during the performance of the project (post-determined) (see also Bakker, Boros, Kenis, & Oerlemans, 2013; Bakker & Knoben, 2015; Burke & Morley, 2016). “Permanent,” in contrast, can thus be understood as “indeterminate”— that is open-ended with regard to time horizon (Bakker et al., 2016). Project networks may thus be understood as organized projects of predetermined or post-determined duration whose relationships among the participating network members may be open ended, and thus subject to reactivation in some future yet temporally indeterminate project context. Such a definition draws greater attention to the history of project-specific prior engagements among project network members and the contextual details (e.g., technological, institutional, and regulatory) surrounding these project engagements. The article by Åberg, Bengtson, and Havila (2018), included in this special issue, illustrates the importance of these contextual details in explaining the re-activation of relationships among material suppliers and building project subcontractors working on construction projects separated by up to 17 years between projects.

Qualitative Case Study Research on Project Networks

The burgeoning scholarship on project networks as a distinct form of governance has been based on various forms of qualitative research. One of the earliest and most theoretically significant contributions to project networks as an organizational form illustrates the various modes of qualitative research utilized to study project networks. Hellgren and Stjernberg (1995) were the first to utilize a series of qualitative studies of shopping malls to illustrate their conceptual model of project networks and to support a series of research propositions regarding defining the characteristics of project networks. Their empirical evidence was more in the form of illustrations of findings from their prior studies, which began with an action research investigation that was a longitudinal study of the design and implementation of the Ring shopping mall in Stockholm, Sweden from its intensive design phase in 1977 through to the mall opening in 1983. Their second source of empirical evidence was based on an investigation of two other malls (the Sture Gallery and the Triangle, in Malmo, Sweden), whose designers were interviewed after the opening of the malls in 1989 and 1991, respectively. Lastly, they examined an independent assessment of the three projects conducted by a panel of 40 experts for judging both the outcomes and processes of investments in the three projects. The panel judged the Sture Gallery to be one of the most successful outcomes. The Triangle was judged to have had a highly problematic process, leading to the least successful outcome in their sample of investment processes. The Ring was midway in terms of outcome and process criteria.

Hence, in this seminal project network publication, we see evidence of diverse forms of qualitative evidence, including a longitudinal case study based on data collection in real time (the Ring shopping mall project), two retrospective case studies (the Surge Gallery and the Triangle projects), and the use of an independent panel assessment of project processes and outcomes utilizing the review of documentary evidence. This section will discuss the forms of qualitative research on project networks represented in this special issue and their contribution to our understanding of project networks.

The most rudimentary form of qualitative research on project networks (and forms of temporary organizing, more generally) consists of case studies in which the investigators collect archival data on a completed project and supplement these data with interviews with participants in the project. Such data are collected at a single time point. This form of qualitative research is often justified as a highly flexible methodology, which can offer some advantages when investigating emergent and dynamic empirical settings (Bakker et al., 2016). Typically, researchers use case studies to develop theory inductively for research questions more or less tightly scoped within an existing theory (Eisenhardt & Graebner, 2007). However, such retrospective single case studies are limited in their ability to capture the evolving and emergent nature of project network governance and this has led to a call for more longitudinal studies that collect data at multiple points of time in the evolution of a project network (Bakker et al., 2016).

In addition to single case studies, there has been the development of qualitative research programs that collect data from multiple cases, for example, of two or more projects in order to compare and contrast salient conditions of project network governance and to identify the factors that may help explain the observed similarities and differences between these project networks. In such studies, the criteria for selecting the comparison cases are explicit in order to control for those factors, which are theorized to drive the evolution or performance of different project networks under study (Ahuja, Soda, & Zaheer, 2012; Majchrzak, Jarvenpaa, & Bagherzadeh, 2015). Evidence from multiple cases is generally more robust, generalizable, and testable compared with single case study findings (Eisenhardt & Graebner, 2007). Researchers can choose cases for comparison that enable replications, contrasts, boundary tests, and the elimination by design of alternative explanations (Yin, 1994). Multiple case studies imply substantially larger research projects, but they promise a much deeper interpretation and understanding of the investigated phenomena, including processes (Bakker et al., 2016). Our special issue features four articles that employ qualitative research methods for understanding the diverse facets of project networks and they are summarized below for their methodological and substantive contributions. One of these articles (Kaulio, 2018) employs a single case study design. The other three articles employ a comparative case study logic, and at least one article (Åberg, Bengtson, & Havila, 2018) utilizes an explicitly longitudinal case study design.

Kaulio (2018) provides a retrospective case history of a supplier–customer project engagement (an interorganizational project) to examine how psychological contracts between project members evolve during the course of various critical incidents that arise during the course of a single project engagement. The distinctive methodological contribution is to examine these critical incidents from the perspectives of each organization's perceptions of these events.

Laursen (2018) examines how value is created in project networks via the theoretic perspective of the Service Dominant Logic. This study employs a comparative case design based on two specific projects from the European Capitals of Culture program, which were funded in 2014 for the same European Capital award recipient. Interviews were conducted with members of each project team over a short period of time and thus the interview data may be interpreted as retrospective in their orientation because there were no follow-up interviews with these subjects to discern how their understanding of project network value creation activities, and outcomes evolved with the evolution of each project. The study reviewed historical documents over a longer project time period and thus may also be seen as consistent with a comparative case study retrospective design. Data analysis followed the standard qualitative research methodological practice of constructing broad coding categories during the course of data collection as an inductive process. These broad categories were then subsequently deconstructed into more narrow subcategories that detailed specific components of the original macro category scheme (Yin, 1994). The coded interview data were then combined with the document-based time line of project activities to construct a general model of the project value creation process whose elements are derived from patterns found in both case studies.

Åberg, Bengtson, and Havila (2018) examine the presence of recurring business relationships between construction building contractors, material suppliers, and subcontractors. They employ a comparative case longitudinal research design by comparing these recurring relationships in three Swedish construction projects that were investigated in 1996, 1997, and 2014. The first project employed a participant observational method of engagement with project leaders and established the initial database of interview and archival documents of initial relationships between the contractor and the project's material suppliers and subcontractors. The second project was completed a year later, and interviews and archival evidence from this investigation provided the data to conduct a comparative assessment of recurring relationships between the contractor, material supplier, and subcontractor participants within both projects. Finally, the third project occurred 17 years later, and the same investigative team collected interviews and archival evidence to provide a second set of comparisons of recurring business relationships between participants in the most recent project with participants in each of the predecessor projects. Data analysis was essentially retrospective in examining interviews referencing earlier stages of the project; however, the comparison of business relationships in three projects separated by space and time qualify this project as longitudinal in design. The analysis focuses on comparing the characteristics of the recurring relationships between each project and the common features of the legal, economic, and social relationships common to each participant. The article concludes with a general model of relationship continuity in the project mode of organizing.

Klaster, Wilderom, and Muntslag (2018) conclude our collection of qualitative research articles with their examination of the characteristics and drivers of overlapping regional networks that emerged as an unintentional result of an amalgamation of central governmental projects. Their empirical evidence is derived from a comparative case study of 11 project networks in four Dutch regions. This article's unique contribution lies in its use of both qualitative and quantitative data in a mixed-methods comparative research design. The qualitative data includes interviews with project participants in all four regional networks supplemented by archival documents to assess the relative success of each project. The project also employed a questionnaire to collect social network data.

Data analysis began with the usage of these data to characterize four geographically identified meta-networks. Four indicators of meta-networks were used to measure the relative strength of a meta-network: density, centrality, size, and congruency. Qualitative data derived from regional network participant interviews and documents were then employed to identify distinctive characteristics of each regional network and possible institutional drivers that might explain the variation in each mega-network's quantitative SNA profile. The article concludes with a detailed assessment of lessons to be learned from these mega networks for both project participants and the government program managers of these projects.

Substantively, these articles provide evidence of the continuing value of qualitative methodologies in broadening our understanding of the dynamics that drive the creation and maintenance of project networks despite the more temporary nature of the projects they sponsor and execute. It is the ability of qualitative methodologies to incorporate such contexts as institutional history, culture, and politics in understanding how these project networks form, evolve, and endure.

Methodologically, the four qualitative articles point to the increasing salience of comparative case research design to enrich our understanding of the project network phenomenon from a governance perspective. The criteria for choice of cases for comparison become a critical research design feature in these study designs and they shape both the relevance and reach of the questions asked for various project network domains. Such comparative designs are by their nature more challenging research and may call for greater collaboration between teams of research scholars who can pool both their expertise and their access to relevant project network research data in order to execute such comparative research endeavors.

The use of mixed methods (qualitative and quantitative), as touched upon by Klaster et al. (2018), offers perhaps the most ambitious research strategy for understanding project networks. The details of such mixed-method design, and in particular the contributions of SNA to our understanding of project networks, are the province of the subsequent section of this introductory article.

Projects from a Network Analytical Perspective

Many researchers have tried to explain why complex projects go wrong (Flyvbjerg, 2014) but very few have attempted to look at the interdependencies that occur between the subsystems through which projects are delivered as a cause of these failures (Pryke, 2012, 2017; Steen et al., 2017). SNA studies gather very precisely defined packages of data and enable data analysis between very different types of data to be compared structurally, using a common method of data gathering and analysis. It is possible, for example, to look at the structure of contractual relationships and compare these with information flows or dispute resolution relationships of a more informal nature. SNA, therefore, enables the quantitative study of the subsystems that need to be integrated and coordinated to create the project management ‘system of systems.’

Building on the work of Nohria and Eccles (1992), the following points provide further justification for the use of SNA in research on temporary forms of organizations. All human activity is ‘social’ in that it involves interaction and it makes sense, therefore, to classify project organizations in social terms and to analyze that social interaction and/or its product: social structure in order to understand the project functions (Pryke, 2012). Project entrepreneurs and project organizations affect and are affected by their social environment (Steen, Coopmans, & Whyte, 2006). If we really want to understand behavior in the project environment, we need to understand the network environment in which individuals or organizations operate (Emirbayer & Goodwin, 1994). Structural positions within networks can both effect and constrain action. Actors also have the power to change their structural place or position and, increasingly, we see that network position (prominence, for example) must be won or commanded by an individual, rather than allocated by ‘superiors’ in a hierarchical sense (Padgett & Ansell, 1993; Macdonald, Steen, & Shazi, 2016).

Networks comprise nodes and connections between those nodes. The node is described as an individual or collective actor in the network and might be, for example, people in a group, departments within a firm, or businesses within an industry. Actors can perform the role of transmitters and receivers in a network; an actor who performs both functions at once is defined as a carrier. The number of incoming connections to any given actor is measured as the in-degree for that actor (expressed as a number of other nodes sending to the given actor). A similar principle applies to the term out-degree, in relation to an actor.

Traditionally, in SNA, actors were linked to others by social ties. Increasingly, as SNA research explores new applications, these ties have included links that are not strictly defined as ‘social.’ Examples of this include technology alliances between firms and supply chain networks in a manufacturing industry (Verspagen & Duysters, 2004; Lomi & Patterson, 2006). Project network research has also expanded beyond ‘human to human’ ties to consider other forms of ‘actors’ and the relationships between them. Some useful examples of the most common types of non-social ties are the transfer of material resources (for example, along the supply chain), association, or affiliation (for example, belonging to the same department), behavioral interaction (talking together, sending messages), and formal relations (for example, organizational hierarchies).

Additional examples of the application of both social and non-social ties in project network research include:

•

Payments between actors—reflecting compensation or ‘consideration’ associated with contractual project relationships (Pryke, 2012)

•

Incentives to perform—reflecting the compensation or ‘consideration’ associated with contractual relationships; extra-contractual incentives allocated to modify behavior or create roles not envisaged in the original contract for the project (Pryke, 2005b, 2006)

•

Contractual relationships—reflecting the formal contractual relationships allocated at contract stage (Pryke & Smyth, 2006; Loosemore, 1998)

•

Instructions issued—important be cause many contracts place emphasis on ‘directing’ and ‘instructing,’ which reflect the presumption for hierarchical relationships in many project contracts (Pryke, 2012)

•

Information sent and/or received—fundamental to the analysis of the post contract project implementation stage of projects, in particular (Kastelle & Steen, 2014; Pryke, Badi, Soundararaj, Watson, & Addyman, 2015)

•

Risk transferred—the transfer of risk, particularly where non-legitimate and post contract, is an important and undesirable source of adversarial behavior leading to disruption of relationships (Pryke & Ouwerkerk, 2003; Loosemore, 1998)

Using these forms of network tie, SNA also has many ways to visualize networks, which show the organization of a project or a sequence of projects in a different way compared to organizational charts and project plans (Cross, Borgatti, & Parker, 2002). Importantly, networks can show the organization of a project as being different from how it is formally organized, and this can result in outcomes that are not intended by the project managers (Kastelle & Steen, 2010).

Beyond the descriptions of projects as networks of actors that can be linked in many ways, social network analysis includes a body of theory that makes predictive statements on how network structures influence outcomes from network processes. Borgatti and Halgin (2011) identify two main groupings of theory within the SNA literature that can be classified as models. In model-based theorizing, outcomes are the results of unseen processes that are specified by the model (Lave & March, 1993). Network measures are used to operationalize certain parameters of the model and relate them to dependent variables. The first of these generic network theories is the ‘flow model’ where networks move some property between actors within that network and this causes a particular outcome that the researcher is interested in (Borgatti & Halgin, 2011).

Within the flow model, centrality is an important measure to operationalize the dynamics of movement through a network. Freeman (1978) referred to three main groups of centrality measures: degree of points, betweenness, and closeness. The degree of points, or extent to which a given point is connected to other points, provides, a measure used in the analysis of communication or interaction activity of some sort. It is a measure of connectivity and reflects a likelihood that an actor will be able to access resources that are moving through the network or distribute resources to others in the network. High degree centrality of a given actor within a network implies a high level of prominence in an information exchange, or some other type of communication network. Point centrality would provide a measure of the importance of an actor, either because the actor was responsible for the very wide dissemination of information (out-degree) or was responsible for gathering information from a large number of other actors (in-degree).

Betweenness centrality is a measure of potential control over communication. Actors with high betweenness indices can restrict or enable flow of information in a network, which has implications for the performance projects (Kastelle & Steen, 2010). It therefore relates to the incidence with which a given node falls between two other nodes. Finally, closeness centrality, involves the measurement of path lengths between two given points and can also be a measure of efficiency of transmission of information or some other resource that is important for project management. There are many different forms of centrality that can be used to operationalize a wide variety of theoretical mechanisms within the flow model of networks (Borgatti, 2005)

Compared to flow models of networks, ‘bond’ models, in which ties govern the behavior of actors are much rarer in the literature (Borgatti & Halgin, 2011). Theoretically, actors within the network can have power over others because of their position in that network. An example of this is the study of power in exchange networks pioneered by Cook and Emerson (1978), where the ability to negotiate favorable terms with other actors in the network is a function of dependency rather than flows. Being connected to structurally weak actors makes one strong, and being connected to structurally strong actors makes one weak (Bonacich, 1987).

Network structures can also confer solidarity between actors. The sociologist, Georg Simmel, proposed that a triadic network configuration that consists of three actors, with each actor having ties with the other two, are qualitatively unique (Wolff, 1950). Dyads embedded in such triads are fundamentally different from non-embedded dyads due to regulation dynamics and stability granted by the presence of the third party. In short, triads form a basis for non-contractual governance that are frequent in densely connected networks, such as clusters in construction coalitions (Gray, 1996; Holti, Nicolini, & Smalley, 2000). Whereas qualitative research is more prevalent for the study of project governance, SNA has its own measures of power and control that are conferred by the structure of the network around the actors.

This area of network theory is currently underexplored by project researchers but the potential can be seen in the contribution from Adami and Verschoore (2018) to this special issue. Using network data from wind-farm projects they model the flow of information, governance and supply chains from both the ‘network ties as bonds for governance’ and ‘network ties as channels for the flow of resources’ perspective. Being clear about what phenomena the network variables are actually operationalizing is important because centrality can be a measure of network control or network flow. In this case, high beta centrality can give actors a greater level of governance over a project, while high degree centrality indicates a high degree of informational control in the flow model of the project. If network research in projects is going to contribute to improved theories of project performance, researchers will need to move beyond purely descriptive studies of project network attributes.

Quantitative Social Network Analysis in Project Research

The special issue contains four articles that use quantitative SNA to investigate projects in different ways and provide a substantive overview of how SNA techniques are being applied to project research. Two of these articles use traditional SNA measures to describe the organization of projects (Pryke, Badi, Almadhoob, Soundararaj, & Addyman, 2018; Lu, Wang, Söderlund, & Chen, 2018), whereas two others use statistical models for network data to test theoretically derived hypotheses (Terhorst, Lusher, Bolton, Elsum, & Wang, 2018; Takahashi, Indulska, & Steen, 2018).

Both the studies by Pryke et al. (2018) and Lu et al. (2018) use important network measures to describe project networks in construction, but the ideas they explore in the organization of projects are quite different. Pryke and colleagues start with the view that projects need self-organizing properties to be successful. While there may be a formal organizational chart, this is different from the network behind the organization that allows the project to distribute information and solve emergent problems in a timely way. Using ideas from graph theory and ‘small world’ networks, they compare the properties of project networks to other self-organizing networks from biology, engineering, and scientific research collaborations to show that their project network case has similar self-organizing properties. Furthermore, project networks can be proactively analyzed to reveal gaps within the network, which can lead to problems in the project.

The article from Lu and colleagues (Lu et al., 2018) examines the project optimization problem in a different way. By comparing the informal communication network with the formal project management network, they calculate how well the informal network ‘fits’ the formal network and then infer a level of informal optimization based on this measure. This presents a contrasting view from Pryke et al. (2018) who propose that a self-organizing informal network is important for project success. Unlike the assumptions of Lu et al., this informal network may well diverge from the formal network for good reasons and be positive for project performance. These counterpoint studies are an example of the need for theories of project networks that are sensitive to contingencies. Rather than one optimal structure, different types of project networks may support projects under varying circumstances. Thus, a further avenue for research in this area is to be more specific about these contingencies and attempt to test these ideas by relating network structures to project performance.

Network data present a special challenge for quantitative researchers who typically follow a pattern of using theory to develop hypotheses and then build regression models to test these hypotheses. A key assumption of regression models is that the independent variables are, in fact, independent from each other. Variables taken from the same network violate this assumption because network dynamics are largely endogenous. The properties of the network or the attributes of an actor at T = 0 will affect the network in later time periods. This means that when we test a hypothesis relating to a network, the comparison (or null hypothesis) must be against a set of simulated random networks to determine if the network effect observed in the data is significant or something that could have arisen through chance.

Two techniques for significance testing with network data have evolved since the 1980s and these are now commonly used for empirical testing of network theory. The study from Terhorst et al. (2018) uses exponential random graph modeling (ERGM) on open innovation research projects in an R&D organization. The importance of tacit knowledge in successful innovation projects is well-understood but the authors' use of ERGM to examine the combined effects of actor characteristics and network effects on the transfer of tacit knowledge. They find that actors who display a high level of autonomous motivation are more likely to both seek and share tacit knowledge. In other words, people who are more engaged with ideas because they find them interesting or exciting are important for R&D projects that are dispersed across organizations and have a large component of tacit knowledge. Interesting, they also find that triadic closure also supports the transfer of tacit knowledge. As discussed earlier, actors in a triad have a level of informal self-governance. What this shows is that knowledge flows and governance are very related to each other and researchers basing their studies in either perspective should be aware that flows affect governance and vice versa.

In a similar empirical setting of technology transfer projects, Takahashi et al. (2018) use another method called quadratic assignment procedure (QAP) to compare the observed network against a large set of randomly simulated networks. In the context of early stage R&D projects with a high-level of ambiguity, they find that tie-strength is important for knowledge transfer and there is also a positive interaction with knowledge codifiability. This result suggests that in these projects, relational governance is more important than structural embeddedness for the transfer of knowledge.

Mixed Methods in Project Network Research

The debate about whether and how qualitative and quantitative research methods should be combined in social sciences is not new, although it has gained in importance over the past 10 years, shown not in the least by the creation of the Journal of Mixed Methods Research in 2007 and has also been discussed in the area of project management by Cameron, Sankaran, and Scales (2015). With mixed-methods research designs scholars usually refer to the combination of qualitative and quantitative data analyses (rather than a mix of different quantitative or qualitative methods). Especially from the view of primarily qualitative research approaches this is not new at all, and is often referred to as data triangulation, contributing to higher levels of reliability and substantiation of constructs (Barratt, Choi, & Li, 2011). In project management research this could, for instance, include triangulating relational data derived from coding interviews (types of interdependencies between projects and perceived project efficiency) with quantitative data derived from organizational charts and performance outcomes such as meeting cost targets of project units.

In the area of networks and project research, scholars have argued not only for combining qualitative and quantitative research methods but even for a need to bridge positivist and interpretivist approaches with more qualitative methods. Lin (1998) and Loosemore (1998) argued that SNA is a quantitative tool capable of being applied within an interpretative context in construction research. In this respect, SNA could be used in a post-positivistic manner (Kilduff & Tsai, 2003). In line with such thinking, Loosemore questions the association of quantitative and qualitative methods with causality and the production of universal models, but feels that both quantitative and qualitative methods (jointly) have a part to play in understanding social roles, positions, and behavior in the project environment.

Others have argued that qualitative and quantitative approaches can be integrated using critical realism, and combining extensive and intensive methods is more important than a quantitative–qualitative dichotomy. The main critique of positivism from the critical realist viewpoint is that explanations are both general and particular/context dependent; this position may be appropriate when considering issues associated with the complexity of projects (Smyth & Morris, 2007).

If we aim at improving the quality of research in project network research by combining the analysis of qualitative and quantitative data, the question is: How exactly should this be done? A valuable categorization is given by Dominguez and Hollstein (2014, pp. 11–18), which is partially adapted and shown graphically by Mühlenhoff (2016). We will discuss some of these approaches in light of how these could be implemented in project network research and provide examples of studies that have used these approaches successfully (see Table 1).

Table 1

Categorization of major mixed-methods SNA designs applied to project research.

Type	Description	Sample Applications in Project Research
Sequential design	Qualitative data analysis, followed by quantitative data analysis or vice-versa	Lu et al. (2018); Eriksson et al. (2018)
Parallel design	Qualitative data analysis and quantitative data analysis simultaneously and parallel and independently (i.e., explaining the same phenomenon in different ways)	D'Andreta, Marabeli, Newell, Scarbrough, & Swan (2016)
Fully integrated design	Qualitative data and quantitative data analyzed simultaneously in an integrated way (i.e., qualitative analyses depend on quantitative analyses and/or vice versa)	Obstfeld (2005)

Source: Adapted from Dominguez and Hollstein (2014); Mühlenhoff (2016)

First, the sequential design would simply mean that after a phase of qualitative data analysis, a second phase of quantitative data analyses is used. Berthod, Grothe-Hammer, and Sydow (2017), in the context of a high-reliability network, showed that the reverse sequence can also be extremely beneficial. Sequential usage would in any case mean that one method confirms the findings of another or builds on the findings of the other. The parallel design means undertaking both methods simultaneously, but independently. A fully integrated design is achieved, when the phenomenon that the researcher tries to explain can only be explained if quantitative and qualitative data are integrated and analyzed in an integrated way.

The case study in this special issue, presented by Eriksson, South, and Levitt (2018) of Californian public–private partnerships, is also an example of sequential mixed-methods design. Interviews were used at the beginning of the study to identify critical project events and then document analysis was used to convert these events into network data. This is a good example of how report data that contains information about stakeholders and the connections between them can be represented as a network. Another feature of this article is the longitudinal nature of the data collection and the tracking of network changes through the different stages of the project. Eriksson et al. (2018) show a surprising level of dynamism in these stakeholder networks, with many stakeholders entering and leaving the network, including ‘anchor’ stakeholders. By classifying ties as formal or informal, the study also shows the use of different institutional logics by the stakeholders over time, with a tendency to more formal ties over the life of the project.

It becomes quickly visible from Table 1 that data triangulation for the purpose of reliability and robustness of constructs (Barratt, Choi, & Li, 2011) is not the only purpose of mixed methods in project network research. Choosing the right method is a key to high quality research designs, which is why we introduce and discuss the key reasons that could lead to choosing mixed-method designs (see Table 2).

Table 2

Reasons/categories why mixed-methods SNA in project research could be beneficial.

Reason	Description
Triangulation	Confirming the same results with a mixed-methods approach
Additive	In order to add findings that could not have been added with the single method design, due to inappropriateness of this design
Interdependency	Results that could not been found without the use of both methods

Triangulation is usually understood as confirming research results in a mixed-methods design that have been developed with the other type of data or method. As described previously, this might be especially important for qualitative analyses as it leaves higher potential for “subjective” interpretations, but could as well be used to (re-) confirm findings from quantitative analyses. We would see triangulation almost as a standard for good quality SNA research in project research. Within this special issue, the article by Lu et al. (2018) is an example of this, because the authors explore coordination and flow models of networks in complex wind farm projects. While the SNA results may indicate the presence of a phenomenon, the researchers use interviews to verify what is happening within the network (see also Berthod et al., 2017). Importantly, they find that both flow and coordination networks exist in the projects, but flow models are more appropriate for information networks, and bond/coordination models are more able to investigate formal contractual ties.

The second usage is labeled as “additive” design and should be chosen, when through the use of a mixed-methods design, additional findings could be developed that could not be developed with only quantitative or only qualitative data (analyses).

We see the third and final type of reason (“interdependency”) as the most promising in SNA research, because SNA data and projects are often complex in nature and single method approaches simply do not allow for drawing the required conclusions. A typical example could be the analysis of all email exchanges during a project from beginning to end), aggregated on a subproject level. Relevant information on these subprojects might be lacking and could be collected in qualitative interview designs and then analyzed jointly.

Whereas a mixed-methods approach might be strong from an empirical point of view, there are costs involved. Additionally, for practitioners collecting quantitative and qualitative data may not be an option as it is too time consuming and costly, and the project under investigation may be over before this sort of research design leads to results and recommendations.

But even in the academic setting aimed at theory development, we believe mixed-methods designs should be considered in a more complete view of quality indicators in project network research. Given the developments in quantitative SNA over the past 20 years, we would like to integrate the mixed-methods view with two additional key variables that may determine the quality of a project network research design (see Figure 2).

Figure 2

Important dimensions in SNA, especially for project research.

Probably the most important shift that we currently see is a shift from measurements at one point in time to longitudinal designs. Networks, projects and project networks are never stable by definition; they rely on the creation, activation, and discontinuation of ties. A dynamic network view has become a key criterion whether SNA research should be accepted in high quality journals—and for a good reason in our opinion. While case study research is often longitudinal, quantitative SNA has also developed longitudinal analytical techniques to test hypotheses about network dynamics. QAP can be extended to include the analysis of network matrices over time and SIENA is a longitudinal version of ERGM (Borgatti, Everett & Johnson, 2013). This provides an untapped method for project researchers to build and test theories about the evolution of project networks.

Another important aspect is the level where measurements take place. Previous research in SNA and project research suggested that multi-level interactions and the measurement at the smallest unit available (e.g., individual level), which can then be aggregated to higher levels is important for capturing outcomes in project networks (Michelfelder & Kratzer, 2013). If feasible, measuring always at the smallest unit available and then aggregating to higher levels and considering interactions between these levels is clearly another aspect to be considered. But more often than not, the aggregation of individual level data may not be appropriate so that a real multi-level approach is needed, considering in the case of projects networks, for instance, that not only individuals but the organization (represented by individuals) interact.

Relating back to the benefits of mixed methods including its drawbacks as it adds to costs and complexity, a (1) dynamic (longitudinal) (2) mixed-methods design (3) measured at multiple levels would certainly lead to methodological rigor for project research. If these criteria cannot be achieved simultaneously, however, we need to make informed decisions in which we reduce rigor or why and how a mixed-methods design can truly contribute to the project management literature.

Conclusion

The level of global economic activity coordinated through projects is increasing in response to the need for organizations and industries to adapt to new conditions. New products, services, infrastructure, and strategies are created by projects that deliver these outcomes. As we have argued, a network lens enables a better understanding of projects as organizational forms. The temporality of projects means that they cannot be understood as pure hierarchies or markets because of the dynamic relationships between actors within the project who contribute different inputs over the life of the project. As such, projects are organized as network forms that are between the extremes of hierarchy and market coordination.

Projects becoming more prevalent phenomena is one reason network research will become more prevalent in the project management literature. Another reason is the increasing sophistication of network research methods and the awareness of these methods among the project management research community. A criticism of the project management research literature is that it can lack theoretical rigor compared with other disciplines within the management field. Not only does the network literature provide an empirical basis for studying projects, it also has an indigenous set of theories that can be applied to project contexts (Borgatti & Halgin, 2011). The potential for network-based research in project management is substantial and there are still many empirical and theoretical avenues for project researchers to explore.

Footnotes

John Steen is an Associate Professor in Strategy and Innovation at UQ Business School, the University of Queensland, Brisbane, Australia. His research focus on innovation extends to complex capital intensive businesses, such as mining and oil and gas. He has published in a range of peer-reviewed journals and collaborates in research projects with a range of well-known companies. He can be contacted at j.steen@business.uq.edu.au

Robert DeFillippi is Professor of Strategy and International Business at the Sawyer Business School, Suffolk University, Boston, Massachusetts, USA, and a Fulbright specialist scholar in project-based organizing. He is the co-editor of the “Business Innovation and Disruption in Creative Industries” book series (Edward Elgar). Dr. DeFillippi has published widely on project-based organizing in such journals as California Management Review, Journal of Media Business Studies, Journal of Organizational Behavior, Organization Studies, and Project Management Review and has contributed chapters on project organizing to scholarly books, including Advances in Strategic Management, Projects and Temporary Organizations, and The Oxford Handbook of Creative Industries. He can be contacted at rdefillippi@suffolk.edu

Dr. Jörg Sydow is a Professor of Management and Chair for Inter-firm Cooperation at the School of Business & Economics at Freie Universität Berlin, Germany; he is also the director of the Research Unit “Organized Creativity,” sponsored by the German Research Foundation (DFG). Dr. Sydow serves or has served on the editorial boards of Organization Science, Organization Studies, Academy of Management Journal, Academy of Management Review, Journal of Management Studies, and The Scandinavian Journal of Management, and recently co-authored Managing Inter-organizational Relations—Debates and Cases (Sydow, Schüßler, & Müller-Seitz, 2016, Palgrave-Macmillan). For more information visit . He can be contacted at Joerg.Sydow@fu-berlin.de

Stephen Pryke is Professor of Supply Chain and Project Networks in the School of Construction and Project Management at University College London, United Kingdom. Stephen is a Fellow of the Royal Institution of Chartered Surveyors and Director of the Centre for Organisational Network Analysis at UCL. His research deals with relationships and networks in the field of construction projects. Stephen has published five books and a large number of journal articles in these areas and recently finished a two-year research project with Transport for London, using SNA to identify the self-organizing aspects for a large, complex infrastructure engineering project. He can be contacted at s.pryke@ucl.ac.uk

Ingo Michelfelder is a postdoctoral fellow at the MIT Sloan School of Management, Cambridge, Massachusetts, USA. His joint research with Jan Kratzer on how a network and project-based organizational design called “semi-open organization” can be used to increase innovation outcomes was awarded with the Christer Karlsson Best Paper Award. A further developed version was published in the Journal of Product Innovation Management. Ingo has a PhD from the Technical University Berlin, practical work experience through his employment with the Boston Consulting Group, and is currently working at the MIT Sloan Sustainability Initiative. He can be contacted at ingo.michelfelder@tu-berlin.de

References

Åberg

Bengtson

, & Havila

(2018). Beyond project closure: Why some business relationships recur in subsequent projects. Project Management Journal, 49(2), 89–104.

Adami

V. S.

, & Verschoore

J. R.

(2018). Implications of network relations for the governance of complex projects. Project Management Journal, 49(2), 71–88.

Ahola

Russka

Artto

, & Kujula

(2014). What is project governance and what are its origins? International Journal of Project Management, 32, 1321–1332.

Ahuja

Soda

, & Zaheer

(2012). The genesis and dynamics of organizational networks. Organization science, 23(2), 434–448.

Artto

, & Wikström

(2005). What is project business? International Journal of Project Management, 23(5), 343–353.

Bakker

R. M.

(2010). Taking stock of temporary organizational forms: A systematic review and research agenda, 12, 466–486.

Bakker

R. M.

Boros

Kenis

, & Oerlemans

L. A. G.

(2013). It's only temporary: Time frame and the dynamics of creative project teams. British Journal of Management, 24, 383–397.

Bakker

DeFillippi

Schwab

, & Sydow

(2016). Temporary organizing: Promises, processes, problems. Organization Studies, 37(12), 1703–1719.

Bakker

R. M.

, & Knoben

(2015). Built to last or meant to end: Intertemporal choice in strategic alliance portfolios. Organization Science, 26, 256–276.

10.

Barratt

Choi

T. Y.

, & Mei

(2011). Qualitative case studies in operations management: Trends, research outcomes, and future research implications. Journal of Operations Management, 29(4), 329–342.

11.

Berthod

Grothe-Hammer

, & Sydow

(2017). Network ethnography: A mixed-method approach for the study of practices in interorganizational settings. Organizational Research Methods, 20(2), 299–323.

12.

Bonacich

(1987). Power and centrality: A family of measures. American Journal of Sociology, 92(5), 1170–1182.

13.

Borgatti

S. P.

(2005). Centrality and network flow. Social Networks, 27(1), 55–71.

14.

Borgatti

S. P.

Everett

, & Johnson

(2013). Analyzing social networks. London, England: Sage.

15.

Borgatti

S. P.

, & Halgin

D. S.

(2011). On network theory. Organization Science, 22(5), 1168–1181.

16.

Borgatti

S. P.

Brass

D. J.

, & Halgin

D. S.

(2014). Social network research: Confusions, criticisms, and controversies. Research in the Sociology of Organizations, 40, 1–29.

17.

Braun

(2018). Configurations for interorganizational project networks: The interplay of the PMO and network administrative organization. Project Management Journal (Forthcoming).

18.

Burke

C. M.

, & Morley

M. J.

(2016). On temporary organizations: A review, synthesis and research agenda. Human Relations, 69(6), 1235–1258.

19.

Cameron

Sankaran

, & Scales

(2015). Mixed methods use in project management research. Project Management Journal, 46(2), 90–104.

20.

Cook

K. S.

, & Emerson

R. M.

(1978). Power, equity and commitment in exchange networks. American Sociological Review, 721–739.

21.

Cross

Borgatti

S. P.

, & Parker

(2002). Making invisible work visible: Using social network analysis to support strategic collaboration. California Management Review, 44(2), 25–46.

22.

D'Andrea

(2014). Projects, routines and economies of repetition. In Wang

(Ed.), Encyclopedia of business analytics and optimization (pp. 1945–1953). Hershey, PA: IGI Global.

23.

D'Andreta

Marabelli

Newell

Scarbrough

, & Swan

(2016). Dominant cognitive frames and the innovative power of social networks. Organization Studies, 37(3, SI), 293–321.

24.

Davies

, & Brady

(2000). Organizational capabilities and learning in complex product systems: Towards repeatable solutions. Research Policy, 29(7-8), 931–953.

25.

Davies

, & Hobday

(2005). The business of projects: Managing innovation in complex products and systems. Cambridge, UK: Cambridge University Press.

26.

DeFillippi

, & Sydow

(2016). Project networks: Governance choices and paradoxical tensions. Project Management Journal, 47(5), 6–17.

27.

Domínguez

, & Hollstein

(2014). Mixed methods social networks research: Design and applications. Structural analysis in the social sciences. New York, NY: Cambridge University Press.

28.

Eisenhardt

K. M.

, & Graebner

M. E.

(2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1), 25–32.

29.

Emirbayer

, & Goodwin

(1994). Network analysis, culture, and the problem of agency. American Journal of Sociology, 99(6), 1411–1454.

30.

Engwall

(2003). No project is an island: Linking projects to history and context. Research Policy, 32(5), 789–808.

31.

Eriksson

South

, & Raymond

(2018). How infrastructure public–private partnerships change over project development phases. Project Management Journal (Forthcoming).

32.

Flyvbjerg

(2014). What you should know about megaprojects and why: An overview. Project Management Journal, 45(2), 6–19.

33.

Freeman

L. C.

(1978). Centrality in social concepts: Conceptual clarification. Social Netowerks, 1(3), 215–239.

34.

García-Canal

Valdés-Llaneza

, & Sánchez-Lorda

(2014). Contractual form in repeated alliances with the same partner: The role of inter-organizational routines. Scandinavian Journal of Management, 30(1), 51–64.

35.

Grabher

(2004). Temporary architectures of learning: Knowledge governance in project ecologies. Organization Studies, 25, 1491–1514.

36.

Granovetter

M. S.

(1985). Economic action and social structure: The problem of embeddedness. The American Journal of Sociology, 91, 481–510.

37.

Gray

(1996). Value for money. Berkshire, UK: Reading Construction Forum and The Reading Production Engineering Group.

38.

Hellgren

, & Stjernberg

(1995). Design and implementation of major investments: A project network approach. Scandinavian Journal of Management, 1(4), 377–394.

39.

Holti

Nicolini

, & Smalley

(2000). The handbook of supply chain management. London, England: CIRIA.

40.

Jones

Hesterly

W. S.

, & Borgatti

S. P.

(1997). A general theory of network governance: Exchange conditions and social mechanisms. Academy of Management Review, 22, 911–945.

41.

Jones

, & Lichtenstein

B. B.

(2008). Temporary inter-organizational projects: How temporal and social embeddedness enhance coordination and manage uncertainty. In Cropper

Ebers

Huxham

, & Ring

P. S.

(Eds.), The Oxford handbook of inter-organizational relations (pp. 231–255). Oxford, England: Oxford University Press.

42.

Kastelle

, & Steen

(2010). Are small world networks always best for innovation? Innovation: Management, Policy and Practice, 12, 75–87.

43.

Kastelle

, & Steen

(2014). Networks of innovation. In Dodgson

Gann

, and Phillips

, The Oxford handbook of innovation (pp. 102–120). Oxford, UK: Oxford University Press.

44.

Kaulio

(2018). A psychological contract perspective on project networks. Project Management Journal (Forthcoming).

45.

Kilduff

, & Tsai

(2003). Social networks and organizations. London, England: Sage.

46.

Klaster

Wilderom

, & Muntslag

(2018). Beyond the network border: The emergence of regional “meta-networks” and their effects on Dutch public-policy projects, Project Management Journal, 49(2), 42–55.

47.

Laursen

(2018). Project networks as constellations for value creation. Project Management Journal, 49(2), 56–70.

48.

Lave

C. A.

, & March

J. G.

(1993). An introduction to models in the social sciences. Lanham, MD: University Press of America.

49.

Lin

A. C.

(1998). Bridging positivist and interpretivist approaches to qualitative methods. Policy Studies Journal, 26(1), 162–180.

50.

Lomi

, & Patterson

(2006). Manufacturing relations: An empirical study of the organization of production across multiple networks. Organization Science, 17(3), 313–332.

51.

Loosemore

(1998). Social network analysis: Using a quantitative tool within an interpretative context to explore the management of construction crises. Engineering, Construction and Architectural Management, 5(4), 315–326.

52.

Wang

Söderland

, & Chen

(2018). The interplay between formal and informal institutions in projects: A social network analysis. Project Management Journal (Forthcoming).

53.

Lundin

Arvidsson

Brady

Eksted

Midler

, & Sydow

(2015). Managing and working in project society: Institutional challenges of temporary organizations. Cambridge, England: Cambridge University Press.

54.

Lundin

, & Söderholm

(1995). A theory of temporary organization. Scandinavian Journal of Management, 11(4), 437–455.

55.

Macdonald

Steen

, & Shazi

(2016). Aiming for excellence: Reflections on the advanced institute of management research and its elite. British Journal of Management, 27(2), 438–454.

56.

Majchrzak

Jarvenpaa

S. L.

, & Bagherzadeh

(2015). A review of interorganizational collaboration dynamics. Journal of Management, 41(5), 1338–1360.

57.

Manning

, & Sydow

(2011). Projects, paths, and practices: Sustaining and leveraging project-based relationship. Industrial and Corporate Change, 20(5), 1369–1402.

58.

Michelfelder

, & Kratzer

(2013). Why and how combining strong and weak ties within a single interorganizational R&D collaboration outperforms other collaboration structures. Journal of Product Innovation Management, 30(6), 1159–1177.

59.

Mühlenhoff

(2016). Applying mixed methods in entrepreneurship to address the complex interplay of structure and agency in networks: A focus on the contribution of qualitative approaches. In Berger

E. S. C.

, & Kuckertz

(Eds.), FGF studies in small business and entrepreneurship: Complexity in entrepreneurship, innovation and technology research (pp. 37–61). Cham, Germany: Springer International Publishing.

60.

Müller

(2009). Project governance. Aldershot, UK: Gower.

61.

Nohria

, & Eccles

R.G.

, 1992. Face-to-face: Making network organizations work. Boston, MA: Division of Research, Harvard Business School.

62.

Obstfeld

(2005). Social networks, the Tertius Iungens orientation, and involvement in innovation. Administrative Science Quarterly, 50(1), 100–130.

63.

Packendorff

(1995). Inquiring into the temporary organization: New directions for project management research. Scandinavian Journal of Management, 11(4), 319–333.

64.

Padgett

, & Ansell

(1993). Robust action and the rise of the Medici, 1400–1434. American Journal of Sociology, 98(6), 1259–1319.

65.

Parmigiani

, & Howard-Grenville

(2011). Routines revisited: Exploring the capabilities and practice perspectives. Academy of Management Annals, 5(1), 413–453.

66.

Powell

W. W.

(1990). Neither market nor hierarchy: Network forms of organization. Research in Organizational Behavior, 12, 295–336.

67.

Provan

K. G.

, & Kenis

(2008). Modes of network governance: Structure, management and effectiveness. Journal of Public Administration Research and Theory, 18(2), 229–252.

68.

Pryke

S. D.

, & Ouwerkerk

(2003). Post completion risk trabnsfer audits: An analytical risk managemnt tool using social netwrk analysis. Proceedings of 2003 Construction and Building Research Conference of the Royal Institution of Chartered Surveyors (COBRA 2003)

69.

Pryke

S. D.

(2004). Analysing construction project coalitions: Exploring the application of social network analysis. Construction Management and Economics, 22(8), 787–797.

70.

Pryke

S. D.

(2005a). Towards a social network theory of project governance. Construction Management and Economics, 23(9), 927–939.

71.

Pryke

S. D.

(2005b). Using social network analysis to understand innovation in construction procurement. Research Paper Series, Royal Institution of Chartered Surveyors, London, England.

72.

Pryke

S. D.

(2006). Legal issues associated with emergent actor roles in innovative UK procurement: Prime contracting case study. Journal of Professional Issues in Engineering Education and Practice, 132(1), 67–76.

73.

Pryke

S. D.

, & Smyth

H.J.

(2006). Introduction. In Pryke

S. D.

, and Smyth

H. J.

(eds) The management of complex projects: A relationship approach (1–20). Oxford, UK: Wiley-Blackwell.

74.

Pryke

S. D.

(2012). Social network analysis in construction. West Sussex, UK: Wiley Blackwell.

75.

Pryke

S. D.

Badi

Soundararaj

Watson

, & Addyman

(2015). Managing complex infrastructure projects: A social network perspective. In COBRA 2015. University of Technology, Sydney, Australia.

76.

Pryke

S. D.

(2017). Managing networks in project-based iorganisations. West Sussex, UK: Wiley Blackwell.

77.

Pryke

S. D.

Badi

Almadhoob

Soundararaj

, & Addyman

(2018). Self-organizing networks in complex infrastructure projects. Project Management Journal, 49(2), 18–41.

78.

Schwab

, & Miner

(2008). Learning in hybrid-project systems: The effects of project performance on repeated collaboration. Academy of Management Journal, 51(6), 1117–1149.

79.

Smyth

, & Morris

(2007). An epistemological evaluation of research into projects and their management: Methodological issues. International Journal of Project Management, 25(4), 423–436.

80.

Shazi

Gillespie

, & Steen

(2015). Trust as a predictor of innovation network ties in project teams. International Journal of Project Management, 33(1), 81–91.

81.

Shenhar

A. J.

, & Dvir

(1996) Toward a typological theory of project management. Research Policy, 25(4), 607–632.

82.

Steen

Coopmans

, & Whyte

(2006). Structure and agency? Actor-network theory and strategic organization. Strategic Organization, 4(3), 303–212.

83.

Steen

Ford

, & Verreynne

(2017). Symbols, sublimes, solutions and problems: A garbage can theory of megaprojects. Project Management Journal, 48(6), 117–131.

84.

Swärd

(2016). Trust, reciprocity, and actions: The development of trust in temporary inter-organizational relations. Organization Studies, 37(12), 1841–1860.

85.

Sydow

Schüßler

, & Müller-Seitz

(2016). Managing inter-organizational relations: Debates and cases. London, England: Palgrave Macmillan.

86.

Takahashi

Indulska

, & Steen

(2018). Collaborative research project networks: Knowledge transfer at the fuzzy front end of innovation. Project Management Journal (Forthcoming).

87.

Terhorst

Lusher

Bolton

Elsum

, & Wang

(2018). Tacit knowledge sharing in open innovation projects. Project Management Journal (Forthcoming).

88.

Verspagen

, & Duysters

(2004). The small worlds of strategic technology alliances. Technovation, 24(7), 563–571.

89.

Wasserman

, & Faust

(1984). Social network analysis: Methods and applications. Cambridge, UK: Cambridge University Press.

90.

Williamson

O. E.

(1985). The economic institutions of capitalism. New York, NY: Free Press.

91.

Williamson

O. E.

(1991). Comparative economic organization: The analysis of discrete structural alternatives. Administrative Science Quarterly, 36(2), 269–296.

92.

Windeler

, & Sydow

(2001). Project networks and changing industry practices—Collaborative content production in the German television industry. Organization Studies, 22(6), 1035–1061.

93.

Wolff

(1950). The sociology of Georg Simmel. Glencoe, IL: Free Press.

94.

Yin

R. K.

(1994). Case study research: Design and methods (2nd ed.). Thousand Oaks, CA: Sage.