Performance Improvement in Megaprojects Through Project Network Characteristics and Transactive Memory Systems

Abstract

This study investigates the effects of project network characteristics (i.e., network density and centrality) and transactive memory systems on project performance. Based on 361 valid questionnaires from megaproject teams, a structural equation modeling (SEM) approach is used for data analysis. The findings discover that high network density is positively associated with specialization, credibility, and project performance, without a significant link with coordination. Meanwhile, a high degree of network centrality negatively affects three dimensions of transactive memory systems, as well as project performance. Regarding the mediating role of transactive memory systems dimensions, specialization and credibility serve as the dominant mediating effects. Interestingly, the interactive relationships among transactive memory systems dimensions are empirically examined. These findings provide a network perspective to integrate and utilize organizational knowledge, thus improving organizational flexibility and resilience.

Keywords

network characteristics transactive memory systems project performance organizational resilience

Introduction

Megaprojects, with their huge investments and lengthy construction cycles, inevitably have significant effects on the economy, environment, and society (Ben Abdallah et al., 2022). Megaprojects are often complex, uncertain, and dynamic, involving multiple relationships among a variety of stakeholders, resulting in significant social risks due to conflicting interests among the various stakeholders (Xia & Xiang, 2022). These projects involve unique technological innovations, complex processes, and multiple teams, including the owner teams, main contractor teams, subcontractor teams, design teams, consulting teams, supervisory teams, and so on (Ma & Fu, 2022). Yet, no single team possesses the full knowledge and capabilities required to deliver a project. As such, megaprojects should establish collaborative relationships and form the collaboration network to achieve successful implementation and delivery of projects, which is a critical channel for project teams to access information, knowledge, and resources (Liu et al., 2022a). Hence, the concept of the project network consisting of numerous stakeholders and network governance was born (Oraee et al., 2017).

In a megaproject network, nodes represent the project participants, and the social connections among all participants are denoted by links (Liu et al., 2022a). Given the diversity and dynamism of the various nodes and links, the attributes of the network they form are also sophisticated and varied. Of the numerous network metrics, network density and network centrality are two basic and essential indicators of network structural characteristics that have been widely applied to address network discrepancies (Havakhor & Sabherwal, 2018; Nyuur et al., 2018). The execution of complex tasks in a project network demands effective collaboration across teams, which in turn can enrich an organization’s knowledge resources, helping them deal with dynamic challenges (Zhang et al., 2021). However, project groups are frequently unable to share and integrate distributed knowledge effectively, and thus fail to take full advantage of potential knowledge resources (Denicol et al., 2020). To address this issue, Wegner (1987) initiated the concept of transactive memory systems as a “shared understanding of who knows what” (p. 198). In transactive memory systems, the cognitive division of labor is established, and the breadth and depth of available collective knowledge are enhanced. The transactive memory systems, which is a social cognitive structure, is an effective approach to collecting, encoding, storing, and integrating information and knowledge from various fields (He & Hu, 2021). The team management system functions most effectively when team members are dependent on each other and strive for shared goals (Bachrach & Mullins, 2019). Such interdependence can help teams maximize the value of knowledge and improve team performance, facilitating the successful delivery of the project that often involves sophisticated, unconventional, and creative assignments (King & Sweet, 2021).

In accordance with the network dynamics theory, there is a dynamic process of knowledge transfer and innovation. During this process, actors in pursuit of new knowledge are required to dominate and broaden information pathways via network dynamics (Cannella Jr. & McFadyen, 2016; Xue et al., 2023). Yet，how to coordinate dispersed knowledge and information resources is still a dilemma for academics and practitioners in megaproject organizations (Hartmann & Dorée, 2015; Liu et al., 2022b). Furthermore, knowledge management has been acknowledged to play an essential role in construction projects (Lin & Lee, 2012; Jia et al., 2022). However, there is still little research to date that investigates the direct effect of project network characteristics on knowledge management capabilities in combination with an integrating analysis of the network characteristics and transactive memory systems effects on project performance in megaprojects. This study will answer the following research questions:

How and to what extent do network density and network centrality affect project performance?

In the context of improving project performance, how do transactive memory systems function as a mediating variable within a project network?

Are there interactions between the three dimensions of the transactive memory systems (i.e., specialization, credibility, and coordination) that affect project performance?

This study makes two contributions to megaproject management. First, this study makes a major contribution to research on transactive memory systems and network attributes by demonstrating their intersections in a review of several literatures. The findings indicate that project network characteristics are important predictors of transactive memory systems. Second, this study presents an explanation of the transactive memory systems as the microfoundation of a critical project resource—social cognition. The findings reveal that the transactive memory system serves as an essential mediator that enables project network attributes to take effect and clarify how the different dimensions of the transactive memory system construct interact to influence project performance.

Theoretical Background

Social capital theory provides a theoretical basis and foundation, and guides this article toward the end (Figure 1). Social capital refers to the resources embedded in a network of social relations (Bourdieu, 1986), which provides a vital perspective for exploring knowledge management and innovation activities. The essence of social capital theory is that social networks possess value, which can affect activities that encompass both the creation of individual knowledge and the transfer of interorganizational knowledge (Yan et al., 2023). Knowledge transfer performance depends on the influence of social capital on information processing patterns, which in turn virtually affects the process of team cognition (Huang & Chen, 2018). To further clarify the components and specific contents of social capital, Grootaert and Van Bastelaer (2001) categorized social capital theory into structural capital (e.g., established roles, social networks, and social structures) and cognitive capital (e.g., shared norms, values, trust, attitudes, and beliefs). As a form of structural capital, network structures are strongly associated with knowledge exchange and team performance (Henttonen et al., 2013). These structures can be beneficial in breaking the barriers of organizational boundaries and facilitating more opportunities for knowledge transfer and sharing (Chen et al., 2017). In addition, the transactive memory system, a type of cognitive capital, contributes to project stakeholders’ understanding of task-specific information, helping them to trust the credibility of that information, and enabling them to utilize that information in an effective manner (Laurent & Leicht Robert, 2019). Hence, social capital theory demonstrates that investments in network capital will yield returns in terms of increased access to information, enhanced communication and coordination, a reduction in opportunistic behaviors, and an improvement in project performance (Cheng et al., 2022).

Figure 1.

Theoretical model based on the social capital theory.

Literature Review

Project Network Characteristics

Network density and network centrality are two crucial parameters used to measure the structural characteristics of a network and for comparing the attributes of different networks and their interactions with each other (Havakhor & Sabherwal, 2018). The construction industry in particular is characterized by specific complexity factors, owing to industrial uncertainties and interdependences. This results in project stakeholders being involved with, and influencing, each other through collaborative networks (Han et al., 2018).

Network density refers to the ratio of the number of actual connected ties to the number of theoretical connected ties in a network. Essentially, network density characterizes the frequency and closeness of connectivity among network actors (Borgatti et al., 2018). High network density indicates that actors interact closely; in such networks, there is a greater likelihood of information and resource dependence and exchange (Wu et al., 2019a). When analyzing the social risks of urban construction projects in China, Yuan et al. (2018) discovered that the stakeholder network with the high density is relatively highly intensive, in which interactions among stakeholders are complex. Moreover, combined with survey data and the social network approach, Xue et al. (2018a) found that the dense collaborative relationship networks and some key nodes positively contribute to the innovation process. Similarly, a higher collaborative network density will also yield higher project efficiency and shorter project duration (Tsai & Huang, 2020). Hence, network density is perceived to be the most widely used metric in measuring network connectivity (af Hällström et al., 2021).

Network centrality denotes that an agent in the network is in a privileged position in terms of information and resources; thus, that agent possesses the ability to affect and control other actors (Wasserman & Faust, 1994). For the three off-site construction cases，Xue et al. (2018b) verified that the network relationship (i.e., density and centrality) has a positive effect on cost performance and suggested that general contractors and manufacturers are playing a direct and significant role in cost management. Meanwhile, Wen et al. (2018) suggested that centralized social capital has a more significant effect on decision-making speed and group collaboration efficiency. Through the centrality analysis of the corporate social responsibility, Zhang et al. (2022) identified the major five activities (e.g., implement safety production management system; constrain consumption of materials and reduce material waste; local and overseas business opportunities; brand, image, and reputation; and reduction of supply risk of building materials) that have the highest centrality scores and play the most critical role in the establishment of the resource collaboration. For the off-site manufacturing projects, project owners are located in the central position and have the largest capability to connect other project parties because of their largest degree of centrality (Hu & Chong, 2022). A number of different kinds of centrality measures exist, but degree centrality, closeness centrality, and betweenness centrality are the most popular at present. Degree centrality describes the numbers directly associated with the nodes (Wang et al., 2020). The closeness centrality measures how close a network node is to other nodes (Zhang & Ashuri, 2018). Betweenness centrality can be described by the degree to which it is located between each pair of remaining nodes, which indicates the potential control and influence of a node in the network (Li et al., 2018).

Transactive Memory Systems

The transactive memory system was commonly conceived as a social cognitive structure at group level that comprises individuals’ knowledge, directory knowledge, and the process by which information is encoded, stored, and retrieved across knowledge domains (Wegner, 1987). The essence of the transactive memory system is that interdependent groups or individuals are able to identify, locate, and access each other's expertise to replenish their own memorizations. Thus, redundant cognition is reduced, their own professional knowledge is consolidated, knowledge pools are expanded, and ultimately, comprehensive cognition that facilitates rapid problem-solving is produced (Wegner, 1987; Wang et al., 2018). According to one perspective of organizational learning, the transactive memory system is a social knowledge transfer mechanism that facilitates team members in terms of exchanging knowledge, learning from each other's experiences, and enriching the knowledge of the whole team (He & Hu, 2021; Sáiz-Pardo et al., 2021). As a result, the cognitive burden on each member is lessened; each member also has access to more information than any one person could manage alone (Ali et al., 2019). Extant empirical evidence has extensively supported the view that transactive memory systems can boost team performance (Peltokorpi & Hasu, 2016; Yan et al., 2023; Zhang et al., 2023).

For a better understanding of the underlying principles and internal mechanisms through which the transactive memory system affects team performance, several scholars have categorized transactive memory systems into different dimensions, in accordance with various perspectives and backgrounds. Examples include a unidimensional and reflective indicator (Choi et al., 2010), a three-dimensional structure (specialization, credibility, and coordination) (Lewis & Herndon, 2011), a four-dimensional framework (knowledge differentiation, knowledge location, the usage of mailing lists, and knowledge credibility) (Chen et al., 2013), and a five-factor model (integration, differentiation, cognitive interdependence, metaknowledge, and transactivity) (Robertson et al., 2013). The three-dimensional model proposed by Lewis (2003) has been most applied in many studies (Huang et al., 2013; Huang & Chen, 2018; Cheng et al., 2022). This is because the three dimensions comprehensively characterize the essence of the transactive memory system, as well as its cooperative process (Yan et al., 2023). Despite the fact that the three dimensions of transactive memory systems are intertwined due to the presence of transactive memory systems, they remain theoretically distinct because specialization and credibility represent cognitive aspects, whereas coordination represents behavioral aspects (Zhang et al., 2023). Since the three dimensions of transactive memory systems have different effects on team performance, examining their separate effects can unravel the real significance of transactive memory systems on team performance (Asim Shahzad et al., 2022). Consequently, following Lewis’ approach, our study also divides transactive memory systems into a three-dimensional variable, namely specialization, credibility, and coordination.

Project Performance

Project performance is commonly taken by construction organizations as the critical criterion for project success (Zhu & Mostafavi, 2017). Project performance is usually defined from a subjective perspective, where different value claims could provide various understandings (Eskerod et al., 2015). In early construction projects, project performance was heavily dependent on the iron triangle criteria: cost, time, and quality (Wu et al., 2020). Project performance measurement indicators, such as health and safety, owner satisfaction, and environmental influence, are evolving as research in construction project management advances (He et al., 2021). Jiang et al. (2016) argued that stakeholder satisfaction and potential cooperation opportunities should be added to the traditional evaluation of project performance. Moreover, when exploring the association between risk attitude and project uncertainty and project performance, Qazi et al. (2021) included time, cost, quality, organizational benefits, and stakeholders’ benefits into the evaluation criteria of project performance. In our study, project performance indicates completing stakeholders’ requirements within the project budget and schedule and realizing the successful delivery of the project through project quality and project management efficiency (Maqbool et al., 2017; Liu et al., 2022c). From the perspectives of all stakeholders involved, performance of megaprojects should be evaluated based on the contributions of all participating stakeholders as well as the satisfaction of users and the public. Accordingly, this study categorizes megaproject performance into two distinct dimensions: the soft and hard dimensions. The hard dimension encompasses project quality, cost, and schedule. Meanwhile, the soft dimension includes collaboration efficiency among teams, public satisfaction, and sharing of experiences and opportunities for collaboration.

Previous studies have also discussed the effects of project networks on transactive memory systems and team performance and have provided excellent theoretical and practical contributions. But they have not yet clarified and revealed the relationship between network structure characteristics and project execution for megaprojects, as most of these studies merely focus on general construction projects. In addition, few studies have not examined how transactive memory systems embedded in project networks influence project performance. Consequently, in order to examine the effects of project network density, network centrality, and transactive memory systems on project performance, this article uses social capital theory as a theoretical background and combines these variables to conduct an empirical study by using the SEM approach.

Hypotheses Development

Project Network Density, Centrality, and Project Performance

Network density is widely investigated in the construction industry. Tang (2012) concluded that the knowledge transfer within a project is inefficient or even ineffective when the network density is low or when the project network path is long. Mohammadfam et al. (2015) suggested that higher network density can better coordinate the behavior of network members. Li et al. (2016) highlighted that the higher the network density of the risk network in a construction project, the more frequent the stakeholder activities will be. Therefore, an increase in project network density not only promotes the building of trust, but also makes it easier for stakeholders to access the needed information and thus improve project management (Pauget & Wald, 2013). In addition, high network density reduces the cost of knowledge acquisition and accelerates the speed of knowledge flow and spillover; the exchange and updating of resources and information are also boosted, ultimately reducing relationship and process conflicts and improving project performance (Wu et al., 2019a).

Network centrality denotes that an agent in the network is in a privileged position in terms of information and resources; thus, that agent possesses the ability to affect and control other actors (Wasserman & Faust, 1994). Since construction projects’ resources are limited, stakeholders with a high degree of centrality play essential roles to coordinate and control project activities. Harrigan (1986) argued that each node in a network can be viewed as a potential source of leaked information. In addition, a node with high centrality, which holds a large amount of information and resources, can create irreversible damage if an information leak occurs. Once the central node makes a mistake with a decision, other nodes are unable to take proactive and effective responses, ultimately resulting in poor project management performance (Wu et al., 2019b). The high complexity and uncertainty of megaprojects means it is risky to share information, technology, and resources across the network, while highly centralized project stakeholders have been known to utilize these resources to seek opportunism and compete against other stakeholders (Liu et al., 2022a). Therefore, the following hypotheses are proposed:

H1a: High network density is positively associated with project performance.

H1b: High network centrality is negatively associated with project performance.

Project Network Density, Centrality, and Transactive Memory Systems

Given the asymmetry of information in megaprojects, an increase of network density will lead to frequent interaction and effective communication among project stakeholders, thus promoting the emergence of transactive memory systems (Choi et al., 2010). Moreover, having more communication paths in a high-density network facilitates the project members’ sharing of information and the understanding of each other's specialized knowledge, thus contributing to the advancement of know-how across the project team (Moreland, 1999). In the project teamwork process, it is possible to build a trust network if project members believe in each other's specialized abilities. This, in turn, can further reduce the redundancy of knowledge caused by repeated search behaviors, improve project team efficiency, and strengthen the credibility of the pool of knowledge and resources (Argote et al., 2018). Beyond that, the specific inspirational approaches and specialized linguistics derived from high-density networks are more instrumental in transferring sophisticated knowledge. Therefore, credibility and cooperation inside the project network are established and maintained, sense among team members of responsibility and accountability is cultivated, and eventually, the coordination of teamwork is boosted (Zhang et al., 2012). Therefore, the following hypotheses are proposed:

H2a: High network density positively contributes to specialization.

H2b: High network density positively contributes to credibility.

H2c: High network density positively contributes to coordination.

Several scholars have argued that centralization of power in a project can become a principal barrier to knowledge sharing among team members. This is because the limited access to communication in highly centralized networks hinders the direct understanding of each team member's specialization (Argote et al., 2018). The hub individual that is present has an advantage in terms of controlling the flow of information and resources and lowering the autonomous capacity of the peripheral ones. This further undermines the credibility of the entire transactive system and triggers relationship conflicts (Loosemore, 1999). Apart from that, the project network individual with high centrality imposes a specific coordination logic on the other members by forcing them to transfer information to the central authority. This approach can easily result in resistance within the team and negatively affect the coordination of the whole project (Hallett & Ventresca, 2006). In terms of network openness, a network with a high degree of centralization tends to create a clique-like structure, which is caused by strong ties and mutual trust among group members. Such a clique network will result in a lock-in effect and over-embeddedness, eventually bringing about structural closure, network rigidity, and a reduction of organizational resilience (Burt, 1982). Accordingly, the following hypotheses are formulated:

H3a: High network centrality negatively affects specialization.

H3b: High network centrality negatively affects credibility.

H3c: High network centrality negatively affects coordination.

Transactive Memory Systems and Project Performance

In transactive memory systems, specialization suggests that each team member is in charge of a particular area of knowledge; this can lead to a strong correlation between individuals and knowledge (Dai et al., 2016). Meanwhile, such knowledge dependency and connection will reinforce the demands for knowledge transfer, thereby facilitating interactive learning of different tasks among team members, integrating unique specialized knowledge, and improving project performance (Heavey & Simsek, 2017). Credibility is perception-based trust, and such credible sources of knowledge can demonstrate the value, usefulness, and effectiveness of knowledge, thus making that knowledge easier to embrace. This, in turn, influences the behaviors of the information receivers, as well as the actions of the entire project team (Chen et al., 2014). Coordination describes the ability of a project team to appropriately distribute authority and tasks among members, based on each member's professional knowledge. In line with the homogeneity theory, coordination can be viewed as a pathway for knowledge transfer, so the knowledge base continually meets the needs of the team's tasks (Zhang et al., 2023). In addition, coordination imparts a sense of sharing, which in turn can help team members overcome cognitive biases and reduce communication barriers, ultimately implementing knowledge transfer activities more effectively (Yan et al., 2023). Hence, three hypotheses are proposed here:

H4a: Specialization is positively correlated with project performance.

H4b: Credibility is positively correlated with project performance.

H4c: Coordination is positively correlated with project performance.

The Interactive Relationship of Transactive Memory Systems Dimensions

The three dimensions of transactive memory systems, namely specialization, credibility, and coordination, all derive from effective communication among team members. Each dimension facilitates and influences the other (Kruser et al., 2023). For instance, knowledge specialization will increase in line with the interaction of team members. When a member performs well in a specific knowledge area, the other team members will rely on them to solve the problems related to that area of knowledge. This approach makes it easier for team members to identify and trust the core members in a highly specialized team, thus forming effective team cohesion and coordination (Wegner et al., 1985). In turn, the coordination of team members is determined by their understanding of the sources and development of knowledge, which evolves with their perceptions of specialization and credibility (Jiang et al., 2023). What's more, the interaction between specialization and credibility demonstrates the real process, which not only creates collaboration within team members but also facilitates innovation and the sharing of new knowledge (Chatterjee, 2016). Therefore, this study puts forward the following hypotheses:

H5a: In transactive memory systems, a positive correlation exists between specialization and credibility.

H5b: In transactive memory systems, a positive correlation exists between credibility and coordination.

H5c: In transactive memory systems, a positive correlation exists between coordination and specialization.

Methodology

Procedure and Sample

This study employed a mixed sampling approach, including purposive and snowball sampling techniques, to collect survey data. Purposive sampling helped the researcher achieve objectives and control the level of differences between the respondents; snowball sampling was used to recruit samples to maximize the number of qualified respondents (Chi et al., 2022). We initially contacted the owners and main contractors listed in the case study and data center of megaprojects (see http://www.mpcsc.org/), which included a significant amount of project data from China. We then requested each organization provide the contact details of the primary members of the project management team. We asked them to nominate several key senior and middle managers who have been involved in the same megaprojects. To increase sample size, we utilized contact information provided by university alumni and acquaintances to interview professionals participating in megaprojects and send questionnaires. This is because they have a deep understanding of the relationships among project parties that were assigned as the investigation target. Most importantly, we offered a guarantee to participants that they would receive a detailed report of the survey results after the study was completed as an incentive to encourage more respondents. We completed 600 refined questionnaires with project professionals through face-to-face interviews. Some responses were excluded from the sample because they contained missing answers, same answers, and timed-out responses. After carefully examining the results, 316 valid questionnaires were received. A total of 316 project participants from 32 project teams participated in our study, providing an overall response rate of 52.67%. Accordingly, the sample size in our study meets the measurement criteria and can yield reliable estimation results. To test the matching of measurement terms and given constructs, content validity was performed. Five academics and seven project managers participated in the validation process. The scale-level content validity index, averaging calculation method (S-CVI/AVE) was 0.95 and complied with criteria of 0.9 threshold, indicating 95% of the total items were judged content valid (Polit & Beck, 2006).

The detailed framework of the whole methodology is shown in Figure 2, and the profiles of the respondents are listed in Table 1. In terms of professional experience in megaprojects, more than one-half of the respondents had less than 7 years of experience, 22.8% had between 8 and 10 years of experience, and 19.6% had more than 10 years of experience in this field. The common method bias is measured by the Harman single factor analysis (Podsakoff & Organ, 1986). The results indicated that six principal component factors were extracted by the maximum variance method, where the first principal component factor’s contribution rate accounted for 14.946% (less than 20%), and the cumulative contribution rate of the six factors was 74.523%. Therefore, this study did not need to take into account the potential disturbance of common method bias. All supplementary materials are presented in Appendices A through E.

Figure 2.

Flow chart of methodology in this study.

Table 1.

Profile of the Data Sample

Characteristics	Individual	Fre	(%)	Characteristics	Megaproject	Fre	(%)
Gender	Male	210	66.5	Project Type	Hydroelectric	96	30.4
Gender	Female	106	33.5		Energy	32	10.1
Age	30–35 years	67	21.2		Road and bridge projects	92	29.1
	36–40 years	89	28.2		Airport or port	73	23.1
	41–45 years	97	30.7		Others	23	7.3
	> 46 years	63	19.9	Team Size	10–15 members	93	29.4
Education Background	College degree or below	119	37.7		16–20 members	54	17.1
	Bachelor’s degree	150	47.5		21–25 members	84	26.6
	Master’s degree	26	8.2		26–30 members	71	22.5
	Doctoral degree	21	6.6		> 30 members	14	4.4
Designation	Project managers	97	30.7	Investment Size	10–50 billion CNY	94	29.7
	Department managers	90	28.5		51–80 billion CNY	65	20.6
	Project engineers	64	20.3		81–100 billion CNY	99	31.3
	Professional managers	32	10.1		> 100 billion CNY	58	18.4
	Grassroots managers	28	8.9	Project Stakeholder	Owners	85	26.9
	Others	5	1.6		Contractors	76	24.1
Work Experience	5–10 years (1–3)	98	31.0		Subcontractors	63	19.9
	11–15 years (4–7)	84	26.6		Supervision	65	20.6
	16–20 years (8–10)	72	22.8		Suppliers	17	5.4
	> 20 years (>10)	62	19.6		Others	10	3.2

Note. The working experience in megaprojects of respondents is in parentheses. CNY is the abbreviation for Chinese Yuan. Fre is the abbreviation for frequency.

Measurement

With the exception of the demographic information, the other variables are measured using a 5-point Likert scale (where 1 indicates completely disagree and 5 denotes completely agree). Network density adopts the measurements coming from Uzzi (1997) and Xue et al. (2018a), which have been widely used in previous studies (Hansen, 1999; Sparrowe et al., 2001). Network centrality is measured using the items presented by Giuliani (2005) and Xue et al. (2018b), all of which have been applied in other studies, because they provide relatively comprehensive network centrality (Reagans & McEvily, 2003; Gebreeyesus & Mohnen, 2013). Similarly, three dimensions of transactive memory systems measurements are identified from prior articles (Lewis, 2003; Comu et al., 2013) with project performance measurements derived from published studies (Lu et al., 2016; Wu et al., 2019b).

Team size is an implicit resource, and large-sized teams may be thought to perform better than small-sized teams (Rau, 2005). As such, in this study, team size is applied as a control variable, to explore the effect of team size on project performance. In addition, the potential effects of project type and project investment size are also included as control variables. The t-test results illustrate that the impacts of these control variables on the research results are not statistically significant (p > 0.05). Hence, this study need not take into account the bias caused by control variables.

Data Reliability and Validity

To test the reliability of different variables, Cronbach’s α values were examined. All α values were higher than the evaluation criterion of 0.5, demonstrating that these variables are equipped with sufficient consistency and reliability (Finch & Bronk, 2011). Moreover, the composite reliability (CR) and average variance extracted (AVE) were adopted to measure the convergence validity. Based on the results that all CR values were over 0.8, and all AVE values were over 0.6, the convergent validity of each variable performs well (Albright & Park, 2009). The confirmatory factor analysis (CFA) was applied to test the discriminant validity and correlation. As shown in Table 2 and Figure 3, these variables have good discriminant validity for further structural model analysis.

Figure 3.

Discriminate validity and Pearson Correlation Matrix. Note. *p < 0.05; **p < 0.01; ***p < 0.001. In this figure, there are scatter plots with fitting regression lines, data distribution plots, and correlation coefficients.

Table 2.

Results of the Measurement Model

Construct	IDs	SFLs	CFA
Network Density α = 0.921 AVE = 0.713 CR = 0.937	ND1	0.800	χ²/df = 3.381 CFI = 0.959 GFI = 0.921 NFI = 0.944 RMSEA = 0.034
	ND2	0.850
	ND3	0.840
	ND4	0.800
	ND5	0.840
	ND6	0.930
Network Centrality α = 0.927 AVE = 0.757 CR = 0.956	NC1	0.810
	NC2	0.830
	NC3	0.800
	NC4	0.820
	NC5	0.900
Specialization α = 0.872 AVE = 0.632 CR = 0.873	SPE1	0.780	χ²/df = 1.031 CFI = 0.999 GFI = 0.971 NFI = 0.977 RMSEA = 0.043
	SPE2	0.800
	SPE3	0.800
	SPE4	0.800
Credibility α = 0.903 AVE = 0.661 CR = 0.886	CRE1	0.790
	CRE2	0.830
	CRE3	0.840
	CRE4	0.790
Coordination α = 0.933 AVE = 0.755 CR = 0.939	COO1	0.830
	COO2	0.920
	COO3	0.840
	COO4	0.830
	COO5	0.920
Project Performance α = 0.920 AVE = 0.620 CR = 0.907	PP1	0.790	χ²/df = 4.739 CFI = 0.970 GFI = 0.955 NFI = 0.962 RMSEA = 0.034
	PP2	0.840
	PP3	0.750
	PP4	0.750
	PP5	0.820
	PP6	0.770

Note. SFLs = standardized factor loadings; χ²/df = relative chi-square; GFI = goodness of fit index; CFI = comparative fit index; NFI = normed fit index; RMSEA = root mean square error of approximation; IDs indicate identifications; CFA indicates confirmatory factor analysis; SFLs presents standard factor loadings; CR means composite reliability; AVE demonstrates average variance extracted.

Results

Results of Hypotheses Testing

Figure 4 presents the empirical results of the hypothetical model. Both network density and centrality are significantly associated with project performance (β = 0.25, p < 0.01, β = −0.17, p < 0.05). Therefore, H1a and H1b are verified. Moreover, high network density positively and significantly contributes to both specialization (β = 0.24, p < 0.01) and credibility (β = 0.24, p < 0.01). Conversely, high network density is insignificantly associated with coordination (β = 0.08, p > 0.05), findings which supported H2a and H2b and reject H2c. In addition, high network centrality has a negative and significant effect on specialization (β = −0.36, p < 0.01), credibility (β = −0.29, p < 0.01), and coordination (β = −0.19, p < 0.01). These results support H3a, H3b, and H3c. Most importantly, this study discovers that specialization (β = 0.18, p < 0.01), credibility (β = 0.16, p < 0.05), and coordination (β = 0.24, p < 0.01) are significantly and positively contributed to project performance. These findings confirm H4a, H4b, and H4c.

Figure 4.

The empirical results of the theoretical model.

The results demonstrate that specialization is positively correlated to both credibility (β = 0.29, p < 0.01) and coordination (β = 0.16, p < 0.01). Moreover, a positive and significant correlation exists between credibility and specialization (β = 0.31, p < 0.01), along with a positive relationship between coordination and specialization (β = 0.13, p < 0.05). However, the most striking result is that the connection between credibility and coordination is statistically insignificant (β = −0.01, p > 0.05). Consequently, H5a and H5c are verified and H5b is not supported.

The Mediating Effects of Transactive Memory Systems

Table 3 presents the mediating effects of the three dimensions of transactive memory systems in the theoretical model. First, the effect of network density on project performance is partially mediated by specialization (0.027, 95% confidence interval [CI] without 0) and credibility (0.050, 95% CI without 0). The effect of network centrality on project performance is also partially mediated by specialization (−0.034, 95% CI without 0) and credibility (−0.048, 95% CI without 0). Most interestingly, for the mediating effect of transactive memory systems, credibility has the largest influence (higher than 35%) and performs the most prominent mediating function, followed by specialization.

Table 3.

Mediating Effects of Transactive Memory Systems

Mediating Variables	Specific Indirect Effects	Total Indirect Effects	Percentage	MacKinnon PRODCLIN 95% CI
Mediating Variables	Specific Indirect Effects	Total Indirect Effects	Percentage	Lower	Upper
Network Density → Project Performance
Specialization	0.027	0.114	23.68%	0.003	0.068
Credibility	0.050		43.86%	0.008	0.110
Coordination	0.010		8.77%	−0.017	0.046
Network Centrality → Project Performance
Specialization	−0.034	−0.131	25.95%	−0.072	−0.008
Credibility	−0.048		36.64%	−0.090	−0.013
Coordination	−0.022		16.79%	−0.054	0.001

Note. 2,000 bootstrap samples; CI = confidence interval; PRODCLIN indicates Distribution of the PRODuct Confidence Limits for INdirect effects

Discussion

Effects of Network Density and Centrality on Project Performance

The higher network density is beneficial to the promotion of project performance. This finding is in line with the conclusions of Xue et al. (2018c). In high-density networks, they argued, there may be well-connected subgroups where communication and coordination among project participants promote rapid information flow and foster innovation via beneficial linkages. When the density is higher, the network ties and collaborations among stakeholders become stronger. Through networking, participants can ensure a steady flow of resources and take advantage of opportunities provided by their partners (Yan et al., 2023). This strong network allows for efficient coordination and cooperation among stakeholders, leading to better resource management and leveraging each other's strengths to seize new opportunities. The ability to establish and maintain close relationships within the network also enables stakeholders to enhance trust and facilitate communication, ultimately benefiting everyone involved (Daget & Zhang, 2023). In addition, in a high-density network, the speed of knowledge diffusion within the network will be accelerated; the cost of knowledge acquisition will also be reduced. Furthermore, the goals and behaviors of project members are always in coordination with each other, thus minimizing relationship and process conflicts and eventually promoting project performance (Wu et al., 2019a).

A higher network centrality will be detrimental to performance. The result is in agreement with the findings of Everett et al. (2013). They suggested that the node at the center plays gatekeeper in a project network, which has the ability to filter information and threaten to disrupt the operation of the network. Moreover, they are also powerful since they can block transmissions and make it difficult for nodes to connect with each other. A high-centrality network has actors that concentrate solely on their own (or their own team's) activities, such that they fail to pay attention to the activities of actors in the other teams (Uddin et al., 2023). In addition, a network with a high-level centrality can also possibly cause errors in performance evaluation, since that centrality will influence the capacity of an actor that relies on the acquisition of information (Guo & Kapucu, 2019). Moreover, according to the social cognition theory, information processing often suffers from errors, because people classify other individuals based on social attributes to simplify processing. However, these simplifications may distort the information being processed in the evaluation (Schunk & DiBenedetto, 2020). Hence, a highly centralized network will affect the accuracy and precision of project information diffusion, and this can be detrimental to project performance.

Effects of Network Density and Centrality on Transactive Memory Systems

A positive correlation exists between network density and the specialization and credibility of transactive memory systems, thus verifying the conclusions of Kanawattanachai and Yoo (2007). That is, a dense network is favorable for the knowledge specialization and credibility of transactive memory systems members. A number of studies have discovered that team member familiarity (Akgün et al., 2005), communication volume, and frequency (Peltokorpi & Manka, 2008) are positively associated with transactive memory systems. The high-density collaborative ties act as channels of communication and provide more opportunities for learning, knowledge transfer, and, hence, innovation (Xue et al., 2018a). Nevertheless, the positive but nonsignificant association between higher network density and coordination is dissimilar to the findings of Zhang et al. (2012). This is because construction management activities are often characterized by a lack of coordination and effective communication. Most importantly, megaproject participants have great mobility, which weakens the storage function of transactive memory systems. In the process of new stakeholders entering and old stakeholders quitting, there will be differences in organizational memory. In such cases, new project stakeholders frequently lack the same attributes and knowledge as prior actors, which makes the replacement of new members a challenging task for these teams. Consequently, in a fully connected network, the coordination of transactive memory systems may not be accordingly enhanced, because the team’s coordination logic is not easily adapted by new members, which, in turn, undermines the new members’ ability to make contributions to the whole team (Argote et al., 2018).

Network centrality is negatively linked with the specialization, credibility, and coordination of transactive memory systems. These findings further confirm the arguments of Hallett and Ventresca (2006). They highlighted that when the centrality is higher, the resource elements of team members will be negatively correlated with knowledge innovation. This is because the combination of knowledge elements with other elements is limited: the core knowledge elements whose science and technology have been exhausted are in excessive concentration and high integration, and further combination will minimize the marginal benefit (Xiao et al., 2022). This finding remains even though Manteli et al. (2014) demonstrated that there is a positive, significant correlation between centrality measures and transactive memory systems processes. It is different in megaprojects in light of the theory of economic and social organization, where the limited communication pathways restrict direct communication among members in a network, and this prevents members from accessing other people's specialized knowledge (Fleming & Spicer, 2014). As a result, the centrality of the members of an organization in social networks negatively affects the knowledge stored in their digital knowledge repository. Moreover, some actors with high centrality also compel other actors to operate in a specific pattern, disregarding their preferences and abilities. Instead of providing opportunities for members to adapt the coordination logic, centralized team members will trigger the resistance of other members, thus aggravating relationship conflicts and affecting the coordination and stability of the whole project (Wu et al., 2019a). The result that coordination does not play a mediating role is in accordance with the findings of Wang et al. (2018), who confirmed that the associations of specialization and credibility with team learning were more significant than with coordination. As a typical feature of project networks, the coordination mechanism reduces transaction costs and enhances the adaptability and creativity of megaproject organizations. Project networks embedded in the construction industry are crucial for this mechanism to function, but they are rarely sufficient to achieve coordination effectiveness. This is because coordination mechanisms become more difficult as projects increase in size and complexity. Networks with high centrality, where some stakeholders may not be able to accurately assess their role in knowledge sharing based on the limited information and sources, make it even more difficult to achieve a coordinated role for these low-centrality stakeholders (Gruchmann, 2022).

Effects of Transactive Memory Systems on Project Performance

The findings indicate that specialization, credibility, and coordination all have positive effects on project performance, which are compatible with the results of He and Hu (2021). On the one hand, the transactive memory system reduces the effort required to exchange and transfer knowledge by creating a knowledge map within the team. On the other hand, the transactive memory system also helps to improve team efficiency because members’ behavior is expected by knowing each individual's knowledge and expertise (Asim Shahzad et al., 2022). Megaprojects require teams to work on complicated, nonroutine, and innovative tasks, requiring team members to leverage the skills and knowledge of others from a variety of functional backgrounds. The transactive memory system comprises the social process of acquiring, storing, transmitting, and utilizing information to create cognitive products and improve team performance at the group level (Cheng et al., 2022). In light of the team adaptation model, the processes associated with task orientation are an essential component of adaptive and innovative performance. Teams with a high degree of task orientation in megaprojects tend to fare better in addressing barriers in the process of innovation implementation. These teams also more easily translate the chosen new ideas into knowledge improvements (Li et al., 2022). Knowledge of social networks and specialization distribution affects different aspects of team performance, and this is the idea behind transactive memory systems. Such systems reduce the cognitive load of each network actor, provide a larger cross-domain information pool for the team or group, and reduce the waste of cognitive effort caused by overlapping individual knowledge (Zhang et al., 2023).

Interactions of Transactive Memory Systems Dimensions

Except for the insignificant relationship between credibility and coordination, the positive correlations between specialization, credibility, and coordination are verified. This is because as each team member’s expertise deepens, the amount of task-related expertise within the team increases, allowing each team member to rely on their collaborators during task completion (Kruser et al., 2023). Knowledge credibility has a positive influence on both communication quality and knowledge sharing, indicating that trust among members enables them to communicate openly and also reduces unnecessary communication expenses (Jiang et al., 2023). Most interestingly, our results are somewhat different from the conceptual model of Chatterjee (2016), who suggested that the three behavioral indicators of transactive memory systems are positively associated with each other. Nevertheless, when faced with complex and uncertain megaprojects, transactive memory systems can make all the difference. The effective diffusion requires the nodes to be vulnerable to the new idea, technology, and so forth, according to the network theory (Collar, 2007). This is because diffusion using strong ties based on trust will pose a problem. On the one hand, a project network formed by strong ties will hinder or even prevent outsiders from entering the network; strong ties will slow down the diffusion of information, knowledge, and resource exchanges among network actors. Over time, it is difficult for network participants to update knowledge and information, ultimately resulting in higher redundancy of static knowledge and reducing the ability of team innovation (Uzzi, 1997). On the other hand, when facing new emergencies, megaproject stakeholders usually deal with problems with conventional thinking and manners. Thus, they fail to tackle the issues fundamentally, eventually threatening the coordination of the entire project (Cheng et al., 2022). Therefore, redundant knowledge credibility cannot completely promote the coordination of a project network, and other factors should be involved.

Implications for Theory and Practice

The first strength of our study illuminates transactive memory systems as a mediator that transmits the effects of network characteristics to project performance. Although previous studies have observed and examined that network characteristics are associated with high trust and effective cooperation in project management (Badi et al., 2017), the ties between network characteristics and project performance may be connected through knowledge transmission and integration. For example, Lee et al. (2014) clarified the role of transactive memory systems in mediating the network structure attributes and team performance links. In addition, it is unclear how and why network characteristics in a project can promote project performance in the megaproject management field, especially from the transactive memory systems’ perspective of knowledge resource transfer and integration. Our study provides an insight into the transactive memory systems literature by examining how differences in network characteristics require different combinations of specialization, credibility, and coordination. New research avenues are opened up by the network perspective, including the study of how variations in the nature, frequency, or efficiency of interactions among members of the same group affect transactive memory systems functions. Assessment of transactive memory systems using a social network method may enable academics to examine the influence of member knowledge distribution on collective structures and processes, whether it is evenly distributed among team members or concentrated among a few members (Cheng et al., 2022). For example, researchers can investigate the nature of the exchanges between members to obtain or provide information or examine the patterns of information exchange among the network participants most likely to influence transactive memory systems. Accordingly, scholars can pay attention to transactive memory systems–related information through network characteristics such as maintaining the community directory of member expert associations.

The findings of this article also provide several implications for practitioners. First, social capital is critical to the integration of knowledge when teams are engaged in communication via megaproject networks. In turn, knowledge integration directly contributes to team decision quality and relationship quality. This clearly indicates that social capital can affect project performance, in part through improving the team's capacity to integrate knowledge (Yan et al., 2023). For this reason, project stakeholders should appropriately optimize the development of their social network structure. Specifically for highly complex and interdependent tasks, project stakeholders are expected to endeavor to establish a decentralized network, which will create open, decentralized communication patterns. This is because decentralization is a cultural character that supports informal relationships and creates a preference for cooperation (Muurlink et al., 2012). In addition, such decentralized structure network embraces the principle of keeping away from the static or religious, which otherwise have a devastating effect on organizational resilience. Second, the shared vision in megaprojects can be described as a shared understanding of all collective goals and directions, as well as appropriate collaborative methods, formed by stakeholders involved in the project (Wang & Pitsis, 2020). Having a shared vision is a top-level concept that aligns with the deepest values and priorities of megaprojects, including collaboration and cooperation, indicating that the viewpoints of participants are aligned rather than on a single project goal (e.g., delivery and performance). Specifically, collaboration with a shared vision is necessary for stakeholders to set common objectives, and their subsequent cooperation enables them to achieve goals as part of the common objective (Chi et al., 2022). Therefore, given the many contradictions and antagonisms that arise in megaprojects, a shared vision can be seen as a relationship mechanism that increases coordination efficiency and promotes understanding, providing a strong foundation for collaboration and communication. Last, but not least, collaborative project governance networks require matching resources to guarantee the completion of tasks, and those resources are provided by the governance structure as part of the resource supply (Daget & Zhang, 2023). During the configuration process of resources among project participants, resource-dispatching information is transferred from the governance structure to the participants via network ties, and the pattern of the network structure influences the efficiency of the information transfer (Wang et al., 2022). When managing megaprojects, project managers need to establish a knowledge pool of successful governance mechanisms; select and promote effective governance strategies to stabilize collaborative network characteristics; and, ultimately, make the project network more manageable.

Conclusion and Future Work

In megaproject networks, how to integrate dispersed knowledge and information resources is still a challenge for both scholars and practitioners. A total of 361 valid responses from China were received and the data were analyzed through a SEM approach. The results demonstrate that high network density is positively correlated with specialization, credibility, and project performance, but not with coordination. By contrast, high network centrality is negatively related to the three dimensions of transactive memory systems and project performance. Additionally, with the strengthening of transactive memory systems, project performance will be improved. Most interestingly, specialization is significantly positively contributed to credibility and coordination. However, the positive relationship between credibility and coordination is not significant. These findings provide theoretical and decision-making guidance that enriches network theory research and will help to strengthen organizational flexibility and resilience in megaprojects. However, several shortcomings still exist in this study. First, similar to several studies on network indicators, most measures were collected through self-reports. Future studies could assess network indicators with more objective measures such as the network diagram. Second, this study merely investigates the cognitive ability of white-collar workers in project management without including blue-collar workers in the research scope. Blue-collar workers will be involved in the future to reduce cognitive bias. Last, but not least, several additional factors might affect transactive memory systems. Future research thus benefits by integrating more team processes found to influence transactive memory systems.

Footnotes

Acknowledgments

This study is supported by the National Natural Science Foundation of China (71972018), the Fundamental Research Funds for the Central Universities (No. 2021CDJSKJC02 and 2022CDJSKPY13), and the High-level Talent Program of Natural Science Foundation of Hainan Province (720RC574). The authors would like to gratefully acknowledge the editor and anonymous reviewers for their insightful comments and suggestions on the earlier version of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Guangdong Wu

Author Biographies

Zhibin Hu is a doctoral candidate at the School of Public Policy and Administration, Chongqing University, China. His research interests are megaproject management and project performance management. He can be contacted at hzbmail1023@163.com

Guangdong Wu is a professor in the School of Public Policy and Administration at Chongqing University, China. His research interests are conflict management in the construction industry, and urban development and management. He can be contacted at gd198410@163.com

Wei Zhang is a researcher at the School of Public Administration, Sichuan University, China. His research areas are project management, innovation, and knowledge management. He can be contacted at hbue_wb@163.com

Appendix A: Survey Questionnaire of This Study

Appendix B.

Information About Experts

Expert	Category	Affiliated Organization	Work Experience (years)	Designation	Interview Time (min.)
1	Academic	Tongji University	11	Professor	35
2	Academic	Chongqing University	10	Professor	34
3	Academic	University of Adelaide	12	Professor	40
4	Academic	University of Queensland	9	Associate professor	42
5	Academic	Sichuan University	7	Associate professor	34
6	Practitioner	China RAILWAY 18TH BUREAU (GROUP) Co., Ltd.	15	Project manager	36
7	Practitioner	China RAILWAY 17th BUREAU (GROUP) Co., Ltd.	12	Vice project manager	40
8	Practitioner	China Construction Eighth Engineering Division Corp., Ltd.	10	Senior engineer	28
9	Practitioner	China Construction Third Engineering Bureau Company Ltd.	22	Deputy chief engineer	30
10	Practitioner	China RAILWAY MAJOR BRIDGE Engineering Group Co., Ltd.	13	Department manager	36
11	Practitioner	China RAILWAY Group Ltd.	12	Safety supervisor	33
12	Practitioner	China Construction Communications Construction Group Co., Ltd.	14	Project manager	35
Average			12.25	-	35.25

Appendix C.

Results of Content Validity

Item	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Expert 9	Expert 10	Expert 11	Expert 12	Number in Agreement	I-CVI
1	4	4	4	3	3	3	3	4	4	3	4	4	12	1
2	4	4	4	3	4	4	4	4	4	4	4	4	12	1
3	4	4	4	4	4	4	4	4	4	2	4	2	10	0.83
4	2	3	4	4	3	3	2	3	4	4	4	2	9	0.75
5	4	3	4	3	3	4	3	3	3	3	3	4	12	1
6	3	3	4	4	3	3	3	3	3	3	3	4	12	1
7	3	3	3	4	3	3	3	4	3	3	4	3	12	1
8	3	3	4	4	4	4	3	4	3	4	3	3	12	1
9	3	4	3	4	3	4	4	4	3	3	4	3	12	1
10	4	3	3	3	4	3	4	3	4	4	4	3	12	1
11	4	4	4	3	4	4	4	3	3	4	4	3	12	1
12	3	3	3	4	3	4	3	3	4	4	3	3	12	1
13	3	4	3	3	4	4	3	4	3	3	4	4	12	1
14	3	3	3	4	3	3	4	2	2	4	2	2	8	0.666
15	4	4	3	3	3	3	4	3	4	3	3	3	12	1
16	3	4	4	4	3	3	3	4	3	3	3	4	12	1
17	4	3	3	4	3	4	3	4	3	4	3	3	12	1
18	4	4	4	4	4	3	4	3	4	4	4	4	12	1
19	4	3	4	4	3	3	3	3	3	4	4	3	12	1
20	4	3	3	4	3	3	4	3	3	3	3	3	12	1
21	3	3	4	3	3	4	3	4	4	4	3	4	12	1
22	3	3	2	4	4	2	4	4	2	4	2	4	8	0.666
23	4	3	4	3	3	3	3	3	4	3	4	3	12	1
24	3	3	4	4	3	3	3	4	3	3	3	3	12	1
25	3	4	3	3	4	3	3	3	3	4	4	4	12	1
26	4	4	4	3	4	4	4	4	3	4	4	3	12	1
27	4	4	4	3	4	3	3	3	3	4	4	4	12	1
28	4	3	2	4	3	4	4	2	2	2	2	3	7	0.583
29	4	3	3	4	4	4	4	3	4	4	4	3	12	1
30	4	4	3	3	4	4	4	3	3	4	3	3	12	1
S-CVI/AVE														0.95
Total Agreement														25
S-CVI/UA														0.833

Note. Rating 1 = Not Relevant; 2 = Somewhat Relevant; 3 = Quite Relevant; 4 = Highly Relevant. I-CVI means item level content validity index; S-CVI/AVE indicates the average value scale level CVI; UA demonstrates universal agreement among experts.

Appendix D. QQ Plot of Normal Distribution for Variables

To confirm the validity of data aggregation, this article calculated the reliability within group (R_wg) of each construct to estimate within-group homogeneity, that is, the consensus on a given problem. Moreover, reliability of score within group (ICC₁) and reliability of mean group score (ICC₂) were introduced to prove the reliability of data aggregation and analysis. When R_wg is greater than 0.7, ICC₁ is larger than 0.05 and ICC₂ is more than 0.5, the data at the individual level are suitable for aggregation to the team level (Bliese, 2000). R_wg ranged from 0.711 to 0.954, ICC₁ varied from 0.617 to 0.755, and ICC₂ ranged from 0.872 to 0.939, thereby meeting the measurement requirements. Therefore, the data deriving from individual level can be aggregated into team level for hypotheses testing analysis.

References

af Hällström

Bosch-Sijtsema

Poblete

Rempling

Karlsson

(2021). The role of social ties in collaborative project networks: A tale of two construction cases. Construction Management and Economics, 39(9), 723–738.

Akgün

A. E.

Byrne

Keskin

Lynn

G. S.

Imamoglu

S. Z

. (2005). Knowledge networks in new product development projects: A transactive memory perspective. Information and Management, 42(8), 1105–1120.

Albright

J. J.

Park

H. M.

(2009). Confirmatory factor analysis using Amos, LISREL, Mplus, SAS/STAT CALIS. The Trustees of Indiana University: Indianapolis, IN, USA.

Ali

Wang

Khan

A. N

. (2019). Mechanism to enhance team creative performance through social media: A transactive memory system approach. Computers in Human Behavior, 91(Feb), 115–126.

Argote

Aven

B. L.

Kush

. (2018). The effects of communication networks and turnover on transactive memory and group performance. Organization Science, 29(2), 191–206.

Asim Shahzad

Iqbal

T.,

Imad Ud Din Akbar

Bakhsh

Ahmad

(2022). Linking authentic leadership to transactive memory system, team innovativeness, and selling performance: A multilevel investigation. Frontiers in Psychology, 13, 884198.

Bachrach

D. G.

Mullins

. (2019). A dual-process contingency model of leadership, transactive memory systems and team performance. Journal of Business Research, 96, 297–308.

Badi

Wang

Pryke

. (2017). Relationship marketing in Guanxi networks: A social network analysis study of Chinese construction small and medium-sized enterprises. Industrial Marketing Management, 60, 204–218.

Ben Abdallah

El–Boukri

Floricel

Hudon

P. A.

Brunet

Petit

M. C.

Aubry

(2022). A process-oriented framework to measure development performance and success of megaprojects. International Journal of Project Management, 40(6), 685–702.

10.

Borgatti

S. P.

Everett

M. G.

Johnson

J. C

. (2018). Analyzing social networks. Sage Publications Limited.

11.

Bourdieu

(1986). The forms of capital. Greenwood.

12.

Burt

R. S

. (1982). Toward a structural theory of action. Academic Press.

13.

Cannella

Jr., A. A.

McFadyen

M. A

. (2016). Changing the exchange: The dynamics of knowledge worker ego networks. Journal of Management, 42(4), 1005–1029.

14.

Chatterjee

. (2016). Enhancing team performance through transactive memory. Development and Learning in Organizations: An International Journal, 30(4), 13–16.

15.

Chen

C.-J.

Hsiao

Y.-C.

Chu

M.-A

. (2014). Transfer mechanisms and knowledge transfer: The cooperative competency perspective. Journal of Business Research, 67(12), 2531–2541.

16.

Chen

Huang

Davison

R. M

. (2017). The role of website quality and social capital in building buyers’ loyalty. International Journal of Information Management, 37(1, Part B), 1563–1574.

17.

Chen

Clark

J. G.

Dietrich

G. B

. (2013). Knowledge sharing in open source software project teams: A transactive memory system perspective. International Journal of Information Management, 33(3), 553–563.

18.

Cheng

Zhang

Chen

Jiang

Lin

. (2022). How does relational governance affect knowledge collaboration in integrated project delivery? A transactive memory system process. Engineering Management Journal, 35(3), 299–312.

19.

Chi

Chong

H.-Y.

. (2022). The effects of shared vision on value co-creation in megaprojects: A multigroup analysis between clients and main contractors. International Journal of Project Management, 40(3), 218–234.

20.

Choi

S. Y.

Lee

Yoo

. (2010). The impact of information technology and transactive memory systems on knowledge sharing, application, and team performance: A field study. Management Information Systems Quarterly, 34(4), 855–870.

21.

Collar

. (2007). Network theory and religious innovation. Mediterranean Historical Review, 22(1), 149–162.

22.

Comu

Iorio

Taylor

J. E.

Dossick

C. S

. (2013). Quantifying the impact of facilitation on transactive memory system formation in global virtual project networks. Journal of Construction Engineering and Management, 139(3), 294–303.

23.

Daget

Y. T.

Zhang

. (2023). Influence of collaborative relationships on the performance of design-construction efficiency of industrialized construction. Journal of Civil Engineering and Management, 29(5), 418–438.

24.

Dai

Roundy

P. T.

Chok

J. I.

Ding

Byun

. (2016). ‘Who knows what?’ in new venture teams: Transactive memory systems as a micro-foundation of entrepreneurial orientation. Journal of Management Studies, 53(8), 1320–1347.

25.

Denicol

Davies

Krystallis

. (2020). What are the causes and cures of poor megaproject performance? A systematic literature review and research agenda. Project Management Journal, 51(3), 328–345.

26.

Eskerod

Huemann

Ringhofer

. (2015). Stakeholder inclusiveness: Enriching project management with general stakeholder theory. Project Management Journal, 46(6), 42–53.

27.

Everett

M. G.

Borgatti

S. P.

Johnson

J. C

. (2013). Analyzing social networks. Sage Publications Ltd.

28.

Finch

W. H.

Bronk

K. C

. (2011). Conducting confirmatory latent class analysis using M plus. Structural Equation Modeling, 18(1), 132–151.

29.

Fleming

Spicer

. (2014). Power in management and organization science. Academy of Management Annals, 8(1), 237–298.

30.

Gebreeyesus

Mohnen

. (2013). Innovation performance and embeddedness in networks: Evidence from the Ethiopian footwear cluster. World Development, 41, 302–316.

31.

Giuliani

. (2005). The structure of cluster knowledge networks: Uneven and selective, not pervasive and collective. Danish Research Unit for Industrial Dynamics.

32.

Grootaert

Van Bastelaer

(2001). Understanding and measuring social capital: A synthesis of findings and recommendations from the social capital initiative. World Bank, Social Development Family, Environmentally and Socially Sustainable Development Network .

33.

Gruchmann

. (2022). Theorizing the impact of network characteristics on multitier sustainable supply chain governance: A power perspective. International Journal of Logistics Management, 33(5), 170–192.

34.

Guo

Kapucu

. (2019). Examining stakeholder participation in social stability risk assessment for mega projects using network analysis. International Journal of Disaster Risk Management, 1(1), 1–31.

35.

Hallett

Ventresca

M. J

. (2006). Inhabited institutions: Social interactions and organizational forms in Gouldner’s Patterns of Industrial Bureaucracy. Theory and Society, 35(2), 213–236.

36.

Han

Taylor

J. E.

Zhong

. (2018). Characteristics and evolution of innovative collaboration networks in architecture, engineering, and construction: Study of national prize-winning projects in China. International Journal of Construction Engineering and Management, 144(6), 04018038.

37.

Hansen

M. T

. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44(1), 82–111.

38.

Harrigan

K. R

. (1986). Managing for joint venture success. Simon and Schuster.

39.

Hartmann

Dorée

. (2015). Learning between projects: More than sending messages in bottles. International Journal of Project Management, 33(2), 341–351.

40.

Havakhor

Sabherwal

. (2018). Team processes in virtual knowledge teams: The effects of reputation signals and network density. Journal of Management Information Systems, 35(1), 266–318.

41.

H. W.

Y. S

. (2021). The dynamic impacts of shared leadership and the transactive memory system on team performance: A longitudinal study. Journal of Business Research, 130, 14–26.

42.

Wang

Chan Albert

P. C.

. (2021). Developing a list of key performance indictors for benchmarking the success of construction megaprojects. International Journal of Construction Engineering and Management, 147(2), 04020164.

43.

Heavey

Simsek

. (2017). Distributed cognition in top management teams and organizational ambidexterity: The influence of transactive memory systems. Journal of Management, 43(3), 919–945.

44.

Henttonen

Janhonen

Johanson

J. E

. (2013). Internal social networks in work teams: Structure, knowledge sharing and performance. International Journal of Manpower, 34(6), 616–634.

45.

Chong

H. Y

. (2022). Integrated frameworks of construction procurement systems for off-site manufacturing projects: Social network analysis. International Journal of Construction Management, 22(11), 2089–2097.

46.

Huang

C.-C.

Chen

P.-K

. (2018). Exploring the antecedents and consequences of the transactive memory system: An empirical analysis. Journal of Knowledge Management, 22(1), 92–118.

47.

Huang

Liu

Zhong

. (2013). The impact of transactive memory systems on team performance. Information Technology and People, 26(2), 191–212.

48.

Jia

Xiang

Chen

. (2022). Prioritizing the operation and maintenance complexity of mega transportation projects based on systems thinking. International Journal of Construction Engineering and Management, 148(2), 05021014.

49.

Jiang

Yang

Zhou

. (2023). The impact of social media usage on expertise coordination and team creative performance in distributed agile software development. Kybernetes, ahead-of-print(ahead-of-print).

50.

Jiang

. (2016). Trust and project success: A twofold perspective between owners and contractors. Journal of Management in Engineering, 32(6), 04016022.

51.

Kanawattanachai

Yoo

. (2007). The impact of knowledge coordination on virtual team performance over time. Management Information Systems Quarterly, 31(4), 783–808.

52.

King

Sweet

. (2021). Using social network analysis to measure transactive memory systems. Team Performance Management: An International Journal, 27(1/2), 80–98.

53.

Kruser

J. M.

Solomon

Moy

J. X.

Holl

J. L.

Viglianti

E. M.

Detsky

M. E.

Wiegmann

D. A

. (2023). Impact of interprofessional teamwork on aligning intensive care unit care with patient goals: A qualitative study of transactive memory systems. Annals of the American Thoracic Society, 20(4), 548–555.

54.

Laurent

Leicht Robert

. (2019). Practices for designing cross-functional teams for integrated project delivery. International Journal of Construction Engineering and Management, 145(3), 05019001.

55.

Lee

J.-Y.

Bachrach

D. G.

Lewis

. (2014). Social network ties, transactive memory, and performance in groups. Organization Science, 25(3), 951–967.

56.

Lewis

. (2003). Measuring transactive memory systems in the field: Scale development and validation. Journal of Applied Psychology, 88(4), 587.

57.

Lewis

Herndon

. (2011). Transactive memory systems: Current issues and future research directions. Organization Science, 22(5), 1254–1265.

58.

A. N.

Sherf

E. N.

Tangirala

. (2022). Team adaptation to discontinuous task change: Equity and equality as facilitators of individual and collective task capabilities redevelopment. Organization Science, 34(3), 1287–1314.

59.

C. Z.

Hong

Xue

Shen

G. Q.

Mok

M. K

. (2016). Schedule risks in prefabrication housing production in Hong Kong: A social network analysis. Journal of Cleaner Production, 134(Part B), 482–494.

60.

H. X.

Patel

Al-Hussein

Gül

. (2018). Stakeholder studies and the social networks of NetZero energy homes (NZEHs). Sustainable Cities and Society, 38, 9–17.

61.

Lin

Y.-C.

Lee

H.-Y

. (2012). Developing project communities of practice-based knowledge management system in construction. Automation in Construction, 22, 422–432.

62.

Liu

Cao

Duan

. (2022a). Effect of network position on inter-team conflict and project success in megaprojects. Engineering, Construction and Architecture Management, ahead-of-print(ahead-of-print).

63.

Liu

Cao

Zhao

Zuo

. (2022b). Interenterprise collaboration network in international construction projects: Evidence from Chinese construction enterprises. Journal of Management in Engineering, 38(2), 05021018.

64.

Liu

Houwing

E.-J.

Hertogh

Yuan

Liu

. (2022c). Explorative learning in infrastructure development megaprojects: The case of the Hong Kong-Zhuhai-Macao Bridge. Project Management Journal, 53(2), 113–127.

65.

Loosemore

. (1999). Responsibility, power and construction conflict. Construction Management and Economics, 17(6), 699–709.

66.

Qian

Chu

. (2016). Role of opportunism and trust in construction projects: Empirical evidence from China. Journal of Management in Engineering, 32(2), 05015007.

67.

. (2022). A governance framework for the sustainable delivery of megaprojects: The interaction of megaproject citizenship behavior and contracts. International Journal of Construction Engineering and Management, 148(4), 04022004.

68.

Manteli

van den Hooff

van Vliet

(2014).The effect of governance on global software development: An empirical research in transactive memory systems. Information and Software Technology, 56(10), 1309–1321.

69.

Maqbool

Sudong

Manzoor

Rashid

. (2017). The impact of emotional intelligence, project managers’ competencies, and transformational leadership on project success: An empirical perspective. Project Management Journal, 48(3), 58–75.

70.

Mohammadfam

Bastani

Esaghi

Golmohamadi

Saee

. (2015). Evaluation of coordination of emergency response team through the social network analysis. Case study: Oil and gas refinery. Safety and Health at Work, 6(1), 30–34.

71.

Moreland

R. L

. (1999). Transactive memory: Learning who knows what in work groups and organizations. Shared cognition in organizations (pp. 3–32). Psychology Press.

72.

Muurlink

Wilkinson

Peetz

Townsend

. (2012). Managerial autism: Threat-rigidity and rigidity's threat. British Journal of Management, 23(S1), S74–S87.

73.

Nyuur

R. B.

Brecic

Debrah

Y. A

. (2018). SME international innovation and strategic adaptiveness. International Marketing Review, 35(2), 280–300.

74.

Oraee

Hosseini

M. R.

Papadonikolaki

Palliyaguru

Arashpour

. (2017). Collaboration in BIM-based construction networks: A bibliometric-qualitative literature review. International Journal of Project Management, 35(7), 1288–1301.

75.

Pauget

Wald

. (2013). Relational competence in complex temporary organizations: The case of a French hospital construction project network. International Journal of Project Management, 31(2), 200–211.

76.

Peltokorpi

Hasu

. (2016). Transactive memory systems in research team innovation: A moderated mediation analysis. Journal of Engineering and Technology Management, 39, 1–12.

77.

Peltokorpi

Manka

M.-L

. (2008). Antecedents and the performance outcome of transactive memory in daycare work groups. European Psychologist, 13, 103–113.

78.

Podsakoff

P. M.

Organ

D. W

. (1986). Self-reports in organizational research: Problems and prospects. Journal of Management, 12(4), 531–544.

79.

Polit

D. F.

Beck

C. T

. (2006). The content validity index: Are you sure you know what's being reported? Critique and recommendations. Research in Nursing & Health, 29(5), 489–497.

80.

Qazi

Daghfous

Khan

M. S

. (2021). Impact of risk attitude on risk, opportunity, and performance assessment of construction projects. Project Management Journal, 52(2), 192–209.

81.

Rau

. (2005). The influence of relationship conflict and trust on the transactive memory: Performance relation in top management teams. Small Group Research, 36(6), 746–771.

82.

Reagans

McEvily

. (2003). Network structure and knowledge transfer: The effects of cohesion and range. Administrative Science Quarterly, 48(2), 240–267.

83.

Robertson

Gockel

Brauner

. (2013). Trust your teammates or bosses? Differential effects of trust on transactive memory, job satisfaction, and performance. Employee Relations Law Journal, 35(2), 222–242.

84.

Sáiz-Pardo

Haro Domínguez

M. C.

Molina

L. M

. (2021). Transactive memory systems mediation role in the relationship between motivation and internal knowledge transfers in a military environment. Journal of Knowledge Management, 25(10), 2396–2419.

85.

Schunk

D. H.

DiBenedetto

M. K

. (2020). Motivation and social cognitive theory. Contemporary Educational Psychology, 60, 101832.

86.

Sparrowe

R. T.

Liden

R. C.

Wayne

S. J.

Kraimer

M. L

. (2001). Social networks and the performance of individuals and groups. Academy of Management Journal, 44(2), 316–325.

87.

Tang

. (2012). Knowledge transferring features in traditional construction project team in China: Based on SNA. Technology and Investment, 3(4), 230–235.

88.

Tsai

M.-H.

Huang

S.-R

. (2020). Team efficiency estimation for construction process considering the collaborative behaviors. Journal of Applied Science and Engineering, 23(1), 79–92.

89.

Uddin

Ong

Matous

. (2023). Stakeholder engagement variability across public, private and public-private partnership projects: A data-driven network-based analysis. PLOS ONE, 18(1), e0279916.

90.

Uzzi

. (1997). Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly, 42(1), 35–67.

91.

Wang

Pitsis

T. S

. (2020). Identifying the antecedents of megaproject crises in China. International Journal of Project Management, 38(6), 327–339.

92.

Wang

Müller

Zhu

. (2022). Network governance for interorganizational temporary organizations: A systematic literature review and research agenda. Project Management Journal, 54 (1), 35–51.

93.

Wang

Huang

Davison

R. M.

Yang

. (2018). Effect of transactive memory systems on team performance mediated by knowledge transfer. International Journal of Information Management, 41, 65–79.

94.

Wang

Thangasamy

V. K.

Hou

Tiong

R. L. K.

Zhang

. (2020). Collaborative relationship discovery in BIM project delivery: A social network analysis approach. Automation in Construction, 114, 103147.

95.

Wasserman

Faust

. (1994). Social network analysis: Methods and applications. Cambridge University Press

96.

Wegner

D. M

. (1987). Transactive memory: A contemporary analysis of the group mind: Theories of group behavior (pp. 185–208). Springer.

97.

Wegner

D. M.

Giuliano

Hertel

P. T

. (1985). Cognitive interdependence in close relationships. Compatible and incompatible relationships (pp. 253–276). Springer.

98.

Wen

Qiang

M. S.

Gloor

. (2018). Speeding up decision-making in project environment: The effects of decision makers’ collaboration network dynamics. International Journal of Project Management, 36(5), 819–831.

99.

Zheng

Zhao

Zuo

. (2019a). Effects of structure characteristics of project network on conflicts and project success. Engineering, Construction and Architecture Management, 28(1), 101–124.

100.

Liu

Zhao

Zuo

Zheng

. (2019b). Effects of fairness perceptions on conflicts and project performance in Chinese megaprojects. International Journal of Construction Management, 22(5), 1–17.

101.

Zheng

Zhao

Zuo

. (2020). How does strength of ties influence project performance in Chinese megaprojects? International Journal of Conflict Management, 31(5), 753–780.

102.

Xia

Xiang

. (2022). Dynamic network analysis of stakeholder-associated social risks of megaprojects: A case study in China. Engineering, Construction and Architecture Management, ahead-of-print(ahead-of-print).

103.

Xiao

Cen

Hao

. (2022). Transactive memory system and green innovation: A cross-level mediation of social network. Industrial Management & Data Systems, ahead-of-print(ahead-of-print).

104.

Xue

Zhang

S. J.

Y. K.

Z. Z.

Yang

R. J

. (2018b). Effect of stakeholder collaborative management on off-site construction cost performance. Journal of Cleaner Production, 184, 490–502.

105.

Xue

Zhang

Wang

Fan

Yang

R. J.

Dai

. (2018a). Collaborative innovation in construction project: A social network perspective. Korean Society of Civil Engineers Journal of Civil Engineering, 22(2), 417–427.

106.

Xue

Zhang

Wang

Skitmore

Wang

. (2018c). Analyzing collaborative relationships among industrialized construction technology innovation organizations: A combined SNA and SEM approach. Journal of Cleaner Production, 173, 265–277.

107.

Xue

Shen

G. Q.

Deng

Ogungbile

A. J.

Chu

. (2023). Evolution modeling of stakeholder performance on relationship management in the dynamic and complex environments of megaprojects. Engineering, Construction and Architecture Management, 30(4), 1536–1557.

108.

Yan

Peng

Zha

. (2023). Transactive memory system (TMS) and knowledge sharing: The effects of social capital and task visibility. Library & Information Science Research, 45(2), 101233.

109.

Yuan

Chen

Wang

Skibniewski

M. J

. (2018). Social network analysis for social risks of construction projects in high-density urban areas in China. Journal of Cleaner Production, 198, 940–961.

110.

Zhang

C. H.

Hong

D. C.

Ling

. (2012). A study on the mechanism of tacit knowledge integration: The role of social ties and transactive memory systems. International Journal of Innovative Computing, Information and Control, 8(8), 5847–5857.

111.

Zhang

Ashuri

. (2018). BIM log mining: Discovering social networks. Automation in Construction, 91, 31–43.

112.

Zhang

B. L.

Lim

B. T. H

. (2022). Linking corporate social responsibility (CSR) practices and organizational performance in the construction industry: A resource collaboration network. Resources, Conservation and Recycling Advances, 179.

113.

Zhang

Wei

Tang

de Pablos

Ordonez

, P. (2023). The establishment of transactive memory system in distributed agile teams engaged in AI-related knowledge work. Journal of Knowledge Management, ahead-of-print(ahead-of-print).

114.

Zhang

Wei

H.-H.

Zhao

Han

Chen

. (2021). Understanding innovation diffusion and adoption strategies in megaproject networks through a fuzzy system dynamic model. Frontiers of Engineering Management, 8(1), 32–47.

115.

Zhu

Mostafavi

. (2017). Discovering complexity and emergent properties in project systems: A new approach to understanding project performance. International Journal of Project Management, 35(1), 1–12.