Abstract
This study examines the influence of project managers’ communication behaviors in integrated project delivery (IPD) projects. This study also considers the impact of team-level information sharing on team-level outcomes (i.e., goal alignment, decision quality, process commitment, and project quality). Results indicate that team information sharing was perceived as beneficial, but the effects of project manager communication behaviors were inconsistent. Interaction analyses indicated that project managers’ communication behaviors hindered team functioning if teams were already engaging in adequate levels of information sharing. This manuscript contributes to the project management corpus by explicating an integrative model whereby project manager and team-level information-sharing behaviors are modeled simultaneously.
Introduction
Well-functioning project teams are indispensable for multiorganizational collaborations. Across all industries, project teams plan schedules, establish procedures, and monitor financial and personnel resources to develop products and improve processes (Baiden & Price, 2011). Where project team collaboration falters, product quality suffers, delays abound in delivery, and costs exceed their budgets (Kappelman et al., 2006; Pinto & Mantel, 1990). External reasons for poor project team quality include owners’ lack of commitment, delays in materials, and a lack of communication with stakeholders (Druskat & Wheeler, 2003; Kappelman et al., 2006; Pinto & Mantel, 1990). Internal reasons derive typically from the lack of trust among members, hostile climates, and weak managerial behaviors (Kozlowski & Bell, 2013; Salas et al., 2018). 1
The responsibility for achieving a high level of collaboration within teams usually rests on the project manager (Chipulu et al., 2013; Kappelman et al., 2006; Müller & Turner, 2010; Sommerville et al., 2010). Organizations seek out individuals with strong intellectual (e.g., critical thinking), managerial (e.g., managing resources), and emotional (e.g., sensitivity, motivation) competencies for these positions (Müller & Turner, 2010). Recent research also points to the importance of critical project manager communication behaviors for multiparty team success, such as monitoring team performance quality, negotiating responsibilities and schedules, spurring creative approaches to problems, monitoring the team's external environment, and challenging team members’ assumptions (Morgeson et al., 2010; Paik et al., 2017; Sun et al., 2015; see also Tjosvold et al., 1999).
At the same time, project managers can only do so much. Team members must also buy into information sharing and contribute to collaboration processes (Manata et al., 2018; Marks et al., 2000). Well-functioning teams share information on task requirements, provide updates, reflect on progress to date, and address problems as they arise (Brill et al., 2006; Sommerville et al., 2010). Perhaps, the most critical team communication behavior associated with performance quality is information elaboration (Marlow et al., 2018), which influences the quality of information shared (Lu et al., 2012; Mesmer-Magnus & DeChurch, 2009). In essence, project managers’ internal and external communicative actions and team members’ within-team information sharing provide the foundation for successful multiorganizational collaborations. Yet, the reality is that project managers’ as well as members’ communication behaviors can vary widely (e.g., Kappelman et al., 2006; Paik et al., 2017), leaving one or the other to pick up the slack to deliver high-quality, on-time, and on-budget products.
In the interest of exploring these dynamics further, this study examines the influence of project team managers’ communication behaviors in integrated project delivery (IPD) teams, a unique form of multiorganizational collaboration, and does so while considering the impact of existing team-level information-sharing behaviors. This study also sets forth the importance of examining teams at the unit level of analysis (Bliese, 2000; Kozlowski & Klein, 2000), a methodological characteristic that has received less attention in the project management corpus. Overall, we believe addressing such issues will contribute to the project management corpus in three ways. First, we explore the intricate nature of communication within IPD teams, as well as the important role of the project manager. Moreover, in considering both project management and team-level dynamics simultaneously, this article develops an integrative model for understanding the effects of project manager communication and team information sharing on team-level outcomes in IPD teams. Ultimately, we contend that this approach will offer greater explanatory power than previous approaches, which have examined both dynamics in isolation. Second, this study provides a quantitative analysis of major assumptions regarding project team managers’ as well as team members’ communication behaviors. Indeed, in examining the dual effects of project management and team-level information-sharing behaviors in a quantitative manner, this article attempts to contribute to the theoretical explanations in the project management literature about when and how project management behaviors can have negative or positive effects on project quality. Considering the increased frequency of IPD (Ghassemi & Becerik-Gerber, 2011) and its costly nature, project management practitioners and scholars may find such findings helpful. Third, we offer a team-level conceptualization and analysis of IPD projects, which can provide unique insights into these multilevel phenomena. That is, investigating IPD dynamics at the project team level of analysis will yield insights regarding how project managers interact with team-level information-sharing behaviors, as well as the extent to which team-level behaviors impact project quality in their own right.
Multiorganizational Collaborations and IPD
Multiorganizational projects are collaborative ventures accomplished through a series of interconnected activities, which are guided by predefined goals and time frames that are supported through resources from contributing organizations (Bakker, 2010). In general, contributing organizations commit their representative(s) based on the organization's specialization and project complexity. These representatives also become a part of the project team, whose involvement could last until the project's completion, which could run for years (Hollenbeck et al., 2012). Completion of large, complex projects can require multiple teams that operate in a hierarchical structure overseeing daily progress and budgets (e.g., Adler, 1995; Dyer & Nobeoka, 2000; Malach-Pines et al., 2009).
One unique form of multiorganizational collaboration is the IPD method. The IPD method imposes contractually shared risks and rewards, promotes early collaboration among key constituencies, and fosters collective decision-making among key team members from different organizations (American Institute of Architects [AIA], 2007; Associated General Contractors of America [AGC], 2010; Baiden et al., 2006). This mandate creates a complex, multitiered project team system composed of a core of members from different organizational entities as well an outer core composed of additional organizational members that contribute to team functioning (e.g., subcontractors, suppliers, consultants; Mollaoglu-Korkmaz et al., 2014). The implementation of IPD is expected to enhance team performance quality (Ghassemi & Becerik-Gerber, 2011), improve team trust (Chiocchio et al., 2011), and eliminate barriers that might impede information sharing in teams (Dyer & Nobeoka, 2000). Such enhancements and improvements, however, are thought to be dependent on successful team information sharing and competent project management communication behaviors (Manata et al., 2018; Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; Paik et al., 2017; Sun et al., 2015). Both of these dynamics are reviewed in greater detail as follows.
Project Managers’ Communication Behaviors
Project team managers oversee teams’ efforts to accomplish predefined project goals (Müller & Turner, 2010), which is especially the case in IPD projects where goals can be multifaceted (Sun et al., 2015). To wit, given the inherent complexity of IPD projects, project team members and managers must emphasize the importance of early goal definition for the purposes of assuring goal alignment, project quality, and overall effectiveness (e.g., see Zhang & Hu, 2018). In addition, project managers coordinate the internal activities of team members and monitor the team's external environment (Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; Paik et al., 2017; Sun et al., 2015). For example, one challenge faced by project managers is managing members from different disciplines and with different technical skills (Ajmal & Koskinen, 2008; Cohen & Levinthal, 1990). As Malach-Pines et al. (2009) note in their analysis of successful multiorganizational projects, project managers must align their actions to match the dynamic nature of organizations and projects, which involves adapting to their members’ respective organizational cultures, business processes, and values. Effective project managers on IPD teams must also emphasize their communication abilities, such as fostering a constructive communication climate, monitoring the behaviors of others, and negotiating the demands of project team members (Paik et al., 2017; Sun et al., 2015). Importantly, these behaviors are expected to increase the overall quality of IPD projects (Manata et al., 2018).
Project managers on IPD teams are also tasked with devoting extraordinary attention to monitoring task-relevant information regularly from stakeholders and the broader external environment (e.g., suppliers, markets) for potential threats to the project (Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; see also Druskat & Wheeler, 2003; Morgeson et al., 2010; Safapour et al., 2019). This is especially critical in IPD projects, where important information must be shared among members from different organizational constituencies. Thus, better performing project managers monitor internal and external information (Kozlowski & Bell, 2013), use pertinent information to influence decision-making outcomes (Dulewicz & Higgs, 2004), and at times serve as an information channel among members from different organizational constituencies (Morgeson et al., 2010). A failure to engage in the aforementioned behaviors may lead project managers to omit critical information about design changes, cost overruns, and project delays to executives and leaders of organizations represented on the project team (Safapour et al., 2019; see also Kappelman et al., 2006; Pinto & Mantel, 1990).
In sum, project managers on IPD teams are tasked with managing information internally and externally so that project team members’ goals are in alignment when making informed decisions (Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014). For example, project managers are dealt the important task of managing personnel from different organizations, who may have different project goals, work procedures, or assumptions regarding project quality (Firth et al., 2015). Ultimately, this management of information is expected to contribute to team members’ commitment to the project and its processes (Ancona & Caldwell, 1992; Morgeson et al., 2010); it is also expected to promote overall effectiveness in IPD teams, for example, project quality and decision effectiveness (Manata et al., 2018). Indeed, project managers on IPD teams are critical to facilitating communication among diverse team members from different organizations, as well as traversing myriad organizational boundaries so that they are better equipped to promote project success (see Manata et al., 2018; Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; Paik et al., 2017; Sun et al., 2015).
Project Team Members’ Information Sharing
In addition to utilizing project managers, organizations also use project teams to draw upon team members’ unique knowledge to address problems, develop products, or improve processes, all of which are complex in nature (Earl, 2001; Whyte et al., 2008). Drawing upon team members’ unique knowledge is decidedly applicable to the IPD method, where members of diverse organizational affiliations come together to share their unique expertise (Sun et al., 2015). Manata et al. (2018), for instance, argued for the importance of ensuring that IPD team members monitored their information-sharing environment and exchanged information so that comprehensive decisions could be made.
Other investigations of the IPD method have corroborated this general line of reasoning (e.g., Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; Paik et al., 2017; Sun et al., 2015). Moreover, numerous meta-analytic investigations in other related fields have confirmed these general claims. Mesmer-Magnus and DeChurch's (2009) meta-analysis, for example, shows that information sharing in teams is a strong predictor of both decision-making quality (ρ = .45) and perceptions of project quality (ρ = .51). Marlow et al.'s (2018) meta-analysis show likewise that communication quality, specifically information elaboration, is related to team performance (ρ = .52) and that organizations “should thus ensure that teams understand the importance of effective communication to performance” (p. 155). Finally, Lu et al.'s (2012) recent meta-analysis of the hidden profile literature also indicates that sharing unique information in groups is critical for fostering decision-making quality (see also Miller et al., 1994; Reimer et al., 2010). Overall, these findings point to the importance of information sharing among all team members on key aspects of team processes and project quality (see also Salas et al., 2018), a conclusion that is largely consonant with the project management and IPD corpora (e.g., see Manata et al., 2018; Mollaoglu-Korkmaz et al., 2013; Mollaoglu-Korkmaz et al., 2014; Paik et al., 2017; Sun et al., 2015).
When combined with the influence of project manager communication, we contend that both dynamics are expected to have beneficial effects on myriad IPD team-level outcomes such as overall project quality, decision-making quality, commitment, and goal alignment. As it will be argued, however, we contend further that the effects attributed to either project management or team-level information-sharing behaviors may be contingent on one another. That is, this article proposes an integrative model that investigates the conditions under which project management and team-level information-sharing behaviors are either beneficial or harmful to myriad team-level outcomes (e.g., project quality). Overall, such an integrative model will provide a valuable contribution to the project management corpus because both dynamics are usually examined in isolation. Before this integrative model is explicated fully, however, we offer a rationale to justify the importance of examining multilevel phenomena (e.g., Kozlowski & Klein, 2000).
Project Team-Level Outcomes
Many phenomena of interest in organizational studies are intrinsically multilevel, meaning that they are contextually bound with hierarchies whereby influences can be bottom-up or top-down (Klein et al., 1994; Kozlowski & Chao, 2018; Kozlowski & Klein, 2000; Rousseau, 1985). Multilevel designs enable researchers to compare among levels of analysis and provide choices for analyzing phenomena evidenced beyond a single level, leading to more holistic explanations of communication activities in organizations (e.g., Manata et al., 2016). Alternatively, ignoring differences between levels and influences on a focal unit of analysis underreports organizational complexity, removes the illuminating potential of contexts, and invites the misspecification of a study's generalizability (Klein et al., 1994). Investigations of communication behaviors and resulting attitudes at the unit or organizational level explore the possibility that theories and their constructs present themselves more forcefully among a collectivity rather than at the individual level of analysis.
Studies of work units and teams at the unit level of analysis to date are becoming more common in part due to the explication of research methods and reviews (e.g., Kozlowski & Klein, 2000). One reason for the earlier relative dearth of unit-level investigations stems from the difficulty of collecting responses from intact teams or being able to code research participants by units within their organization. Other reasons pertain to (1) the munificence of individual-level studies and theorizing, which can monopolize researchers’ focus (Klein et al., 1994); and (2) when considering the group level of analysis, the need for many groups or units participating, as 100 individuals in 10 units equates to a group-level N of 10. Yet, when the responses of all unit members are combined, the emerging data can provide a team-by-team comparison, revealing insights into how the team project manager or information adequacy among all team members relate to other variables. In other words, unit-level analyses can uniquely reveal how project team members’ interactions are shaped by the demands of each individual team's work contexts and personnel (Kozlowski & Chao, 2018).
The issues highlighted in this study are how project managers’ communication behaviors and the team's level of information sharing interact to influence IPD team outcomes. Though numerous ways to assess a team's effectiveness exist (e.g., Kozlowski & Bell, 2013; Mathieu et al., 2008; Salas et al., 2018), four outcomes are particularly relevant to IPD project teams: team goal alignment, decision-making quality, commitment to the team's task processes, and overall project quality. Of note, these outcomes were chosen based on theory and tradition. Specifically, elements of project quality, decision-making effectiveness, and commitment are treated commonly as outcomes of interest in organizational literatures (e.g., Kozlowski & Bell, 2013). Moreover, and as will be argued, goal alignment was considered an important theoretical outcome given our independent variables of interest (i.e., project managers are expected to align goals, and team information sharing should also contribute to goal alignment). In essence, the extent to which goals are aligned is analogous to investigating a group's level of cohesion, which is critical to team effectiveness (e.g., Manata, 2020). Each of these four outcomes is reviewed in greater detail as follows.
Team Goal Alignment
Goal alignment at the team level refers to the extent to which team members’ aims and objectives are in sync with each other. The extent to which individuals are not working at cross-purposes and have agreed upon overall goals is a foundational element of team effectiveness (Kozlowski & Bell, 2013) and of IPD teams in particular (Sun et al., 2015). Having coherent goals that team members follow lessens conflict among members and contributes to the productive use of team members’ efforts (Colbert et al., 2008; Gibson & Birkinshaw, 2004). Project manager boundary spanning contributes to team goal alignment by procuring information from the external environment, ascertaining threats (e.g., shipment delays, changes in material pricing) to the project, and protecting “the team from unnecessary interruptions” (Morgeson et al., 2010, p. 21). At the same time, internal team information sharing enables members to develop shared mental models of their challenges, resources, and options (Marlow et al., 2018; Mathieu et al., 2008; Salas et al., 2018). Rich information-sharing environments enable teams to maintain goal alignment through all phases of their project, which in turn increases their adaptability to unplanned developments and exigencies (Salas et al., 2018).
Quality of Project Team Decisions
The ability of IPD project teams to make apt decisions depends on their ability to acquire and process relevant information (Manata et al., 2018; see also Amason, 1996). In this regard, the project manager's ability to identify and relay stakeholders’ orientations and acquire information (e.g., budget, market assessments) that may not be readily available to team members is essential (Ancona & Caldwell, 1992; Koschman et al., 2012; Nelson-Marsh, 2017). Relaying information about what is politically feasible or acceptable among stakeholders can also steer team members to invest their time in options that will have broader external support (Marrone, 2010). In turn, good decisions require both explicit and tacit knowledge (Earl, 2001; Whyte et al., 2008), which are best made possible by information sharing among IPD team members (Manata et al., 2018). In essence, high levels of information sharing increase the likelihood of obtaining unique insights from team members (Lu et al., 2012).
Commitment to Team's Processes
Critical to the success of any team is its acceptance of and commitment to its structure and the procedures used to accomplish its goals (London & Howat, 1978; Salas et al., 2018), and this is also the case for IPD teams (Manata et al., 2021; Sun et al., 2015). Indeed, facilitating a positive bond with a team's internal processes appears especially vital where team members do not have a history of working together or have traditions of distrust between their specialties (Baiden & Price, 2011; Ebrahimi & Dowlatabadi, 2018; Pinto & Mantel, 1990). When IPD project managers facilitate member coordination as well as minimize interference from outside forces, team members can see the benefits of their collaborative processes (Mollaoglu-Korkmaz et al., 2013). Processes that foster team members’ information sharing are also likely to be well-received (Kozlowski & Bell, 2013).
Project Quality
The ultimate concerns for owners and IPD project participants alike are the quality, expediency, and cost of their efforts (Mollaoglu-Korkmaz et al., 2013; Paik et al., 2017; Sun et al., 2015; see also Ibbs et al., 2007; Rojas & Kell, 2008). Project managers’ ability to manage factors internal and external to the team appears consistently related to project quality and budget concerns (Marrone, 2010; Müller & Turner, 2010). In the same way, the extent to which team members elaborate on issues and extrapolate on their implications is directly related to the quality of team performance (Marlow et al., 2018; Mesmer-Magnus & DeChurch, 2009).
As project team management and information sharing in IPD teams should have positive effects on IPD teams’ overall decision-making quality, the extent to which team members’ goals are aligned, commitment to their internal processes, and project quality, we hypothesize:
Nonadditivity
Project managers’ internal and external communication behaviors and levels of IPD team information sharing do not occur in isolation from each other. In the best of all worlds, team-level dynamics mirror project managers’ well-developed communication behaviors (Mollaoglu-Korkmaz et al., 2014; Nofera et al., 2011). At other times, those in managerial positions face considerable challenges in creating unified goals, facilitating cooperation, encouraging follow-through, and other critical team-building behaviors (e.g., Yukl et al., 2002). Nevertheless, where IPD team members’ information sharing is substandard, project managers’ active involvement in promoting interaction among team members and in monitoring the external environment will likely make the difference between successful and mediocre or unsuccessful project outcomes (Mollaoglu-Korkmaz et al., 2013; see also Morgeson et al., 2010).
In contrast, project managers at times are inexperienced and impede team functioning, harming project quality (Iorio et al., 2012; Müller & Turner, 2010). Project managers’ attention may also be spread too thinly due to multiple responsibilities (Sherf et al., 2019). Notably, how to develop project managers’ competencies is an underlying or specific theme in research (e.g., Brill et al., 2006; Chipulu et al., 2013; Müller & Turner, 2010; Paik et al., 2017). Skills related to fostering team members’ communication behaviors are typically at the forefront of the list of ideal project manager competencies (e.g., Müller & Turner, 2010). Despite these potentially dire circumstances, IPD team members may actively engage in information sharing independently of project manager behavior and generate high-quality outcomes. Their successful, independent collaboration may stem from a combination of factors, including previous, positive working relationships with many or all of the team members (DeRue et al., 2010; Salas et al., 2018); mutual respect and contagion of cooperation (Eisenberg, 1990); and spurring the sharing of tacit information (Dossick & Neff, 2011; Earl, 2001; Whyte et al., 2008). In such cases, team members initiate information sharing among themselves and with stakeholders, resulting in the project manager's role being more symbolic and less instrumental (Tjosvold et al., 1991). Stated differently, if IPD team members already engage actively in information sharing without the help of a project manager, then the role of a project manager may be of less consequence and may even impede team functioning.
Taken together, these materials imply nonadditive, or interaction, effects of a competing nature. In essence, these materials suggest that project manager communication will be critical to promoting beneficial IPD team-level outcomes when team information sharing is low. In addition, when team information is already at optimal levels, project manager involvement may either (1) enhance IPD team functioning above and beyond what the team is able to accomplish on its own (i.e., a positive nonadditive effect) or (2) impede what is otherwise considered optimal IPD team functioning (i.e., a negative nonadditive effect). In the interest of investigating this matter further, two competing hypotheses are offered (Figure 1):

Integrative model with the predicted moderation effects.
Method
Procedure
In acquiring a sample of extant project teams working on IPD projects in the construction sector, two of the authors (1) searched the internet for viable IPD projects that were under contractual obligations to implement the IPD method and (2) contacted members from each of these project teams to confirm that they were working on an IPD-by-contract construction project. Upon confirmation, researchers solicited responses to an online survey from all the team members on each of the teams. Participants received reminders for up to 10 months.
Sample
One hundred and ninety-one responses were collected from IPD project team members dispersed across 21 IPD-by-contract projects. The average team size was 9.10 (SD = 7.23). Subjects were generally male (90.8%), middle-aged (M = 46.82, SD = 10.51), and had been working with the company for an average of 13.90 years (SD = 8.57). Participants reported taking on the role of contract representatives (35.1%), as well as a host of other relevant roles (designer's representative = 16.2%, MEP subcontractor = 15.7%, owner's representative = 15.2%, consultant = 6.3%, and other = 11.5%). Most projects were classified as new construction (80.1%) and were nested within the healthcare sector (69.6%). Most projects were either in their construction phase (21.5%) or already complete (60.7%). The remaining projects were in earlier stages (e.g., validation phase, conceptual design, and detailed design) of development.
Instrumentation
The main variables measured in this investigation are described as follows. Unless otherwise noted, all items were arrayed in a Likert-type manner with 1 = strongly disagree and 5 = strongly agree. All responses were also aggregated to the team level of analysis.
Project Manager Communication
This instrument, which assessed the extent to which the team's project manager buffered the unit from unnecessary interruptions and coordinated team activities, was adapted initially from Morgeson et al. (2010) and then Manata et al. (2018). This survey used the term project coordinator to direct participant's focus to the person managing the project. The more ubiquitous project manager is a job title and is often interchangeable with project coordinator (i.e., same person). Survey items included both internally and externally focused items such as “the IPD project coordinator helps us communicate with one another,” and “the IPD project coordinator buffers the team from the influence of outside forces or events.” This dual focus is in keeping with Ancona and Caldwell (1992), who identify boundary spanners as serving ambassadorial, task coordinator, and scout roles. 2
Information Sharing
Adapted from Miller et al.'s (1994) No One Ever Tells Me Anything (NETMA) scale, items on this scale were reverse coded and assessed the adequacy of information sharing in IPD teams. Example items included, “I feel like no one ever tells me anything about what's going on in the IPD project” (reverse coded) and “the people who know what's going on here in this IPD project do not share enough information with me” (reverse coded). The reverse coding of the items changed the nature of the scale to emphasize the adequacy instead of the inadequacy of the information environment. Overall, this instrument assesses the extent to which information is being shared among project members, which constitutes one of our focal independent variables of interest.
Goal Alignment
The goal alignment scale was created to measure the extent to which team members emphasized similar priorities, moved toward a common goal, and was in sync. Item creation was modeled after earlier measurements by Colbert et al. (2008) and Gibson and Birkinshaw (2004), which focused on capturing elements of goal importance and alignment, respectively. Examples of created items included, “all members on your IPD team worked toward the same overall goals” and “everyone's goals were generally in sync.” As is apparent from the items’ content, this measure captured the extent to which members’ goals were aligned, which constituted one of our main outcome variables of interest.
Decision Quality
Based initially on Amason's (1996) three-item decision quality scale, eight Likert-type items were used to assess the extent to which team members felt the team was making good decisions. Of note, this measure was based on the work of Amason (1996) because of its conceptual alignment with our construct of interest, in other words, decision quality. Example items included, “the quality of decision-making on your IPD team is very good” and “the decisions made by your IPD team have positive effects on team performance.”
Process Commitment
Adapted from London and Howat (1978), this scale measured the extent to which team members were committed to the general process responsible for project implementation (i.e., IPD). London and Howat (1978, p. 2) define commitment as a “positive, emotional bond between employees and elements of their jobs,” which was consonant with our conceptual treatment of commitment, as well as the conceptual definition of others who conduct research in this arena (e.g., Batra & Hyde, 2020). Example items included, “you are committed to the IPD process” and “if given the opportunity, you would continue working on IPD projects.”
Project Quality
Tjosvold et al.'s (1999) five-item group product quality measure was adapted to assess the extent to which team members believed project-level outcomes were adequate relative to other project delivery methods. Of note, this measure formed the basis of our measure given its conceptual alignment with our construct of interest, in other words, project quality, which is considered one of the many critical elements of overall team effectiveness or performance (see Hackman & Wageman, 2005). Moreover, in addition to specifying indicators of project quality (e.g., task efficiency), the present scale was expanded to include items pertaining to the construction industry, for example, change order requirements, cost reductions, and innovation creativity (Ibbs et al., 2007; Rojas & Kell, 2008). Items were prefaced with the following instructions to participants: “Compared to similar projects delivered under other methods (e.g., design-build, design-bid-build, construction management), in this project….”
Measurement Model
Before conducting the main analyses, the structural validity of the proposed measurement model was inspected using confirmatory factor analysis (CFA). This analysis was implemented to confirm the structural validity of the six-factor measurement model stipulated a priori (see Anderson & Gerbing, 1988; Hunter & Gerbing, 1982). Judgments regarding model fit were made by inspecting the residual matrix, which was produced using the lessR package in the R software environment (Gerbing, 2016; R Core Team, 2016). Items that produced large errors consistently were treated as invalid indicators of their respective latent factors (Anderson & Gerbing, 1988; Hunter & Gerbing, 1982). Additionally, the comparative fit index (CFI) and the standardized root mean square residual (SRMR) were estimated and implemented as additional indices of model fit; these fit indices were estimated using the lavaan package in the R software environment (R Core Team, 2016; Roseel, 2012). Given the small sample size (i.e., N = 191) and the degree of model complexity (i.e., 30 + items; six-factor solution), CFI values that approached .92 and SRMR values that were less than .09 were treated as indicators of adequate model fit (see Hair et al., 2007). Moreover, if model comparisons were required, the Akaike information criterion (AIC) was provided (see Singer & Willet, 2003).
Initial analyses of the six-factor measurement model produced marginal model fit, CFI = .90, SRMR = .08, AIC = 7631.65. Although the SRMR was modest, the CFI produced a value lower than desired. An inspection of the residual matrix revealed that numerous items produced large errors consistently. Consequently, these items were removed with the intent of improving the model's structural validity (Anderson & Gerbing, 1988; Hunter & Gerbing, 1982).
Reassessment of this abridged model produced better model fit, CFI = .93, SRMR = .07, AIC = 6672.05. Specifically, the values of all three fit indices improved. The difference between both model chi-squares was also substantial, Δχ2(Δdf = 155) = 292.42, p < .001. Hence, this abbreviated measurement model was preferred to the initial model. 3
Group-Level Constructs
Both antecedent and criterion variables were operationalized at the group level (i.e., the same level of analysis). Because each construct was assumed to represent a shared cluster construct, whereby measures are assumed to be isomorphic across team members and levels of analysis, median
Variable Means, SD, Correlations, and Coefficient Alphas
Note. Comm. is an abbreviation for communication. Coefficient alpha reliability coefficients are inserted in the diagonals. Listwise N = 21.
Results
To assess the accuracy of the hypotheses, ordinary least squares regression analyses were implemented. Given the small sample size due to responses aggregated at the team level, one-tailed p values were utilized when appropriate in the interest of increasing statistical power (Cohen et al., 2013). One-tailed analyses were also appropriate given the directionality of the hypotheses (Gravetter et al., 2020).
Main Effects
Goal Alignment
Analyses began by regressing goal alignment onto the project manager’s communication and team information sharing independent variables. This overall model was significant, F(2, 18) = 8.62, p < .01. R2 = .49, whereby both project manager communication and team information sharing had a positive impact on team members’ overall goal alignment. Specifically, increases in project manager communication were associated with increases in team members’ goal alignment (β = .46, t = 2.66, p = .01). Similarly, increases in team information sharing were associated with increases in team members’ goal alignment, (β = .45, t = 2.63, p = .01). Thus, as predicted, team members’ goals were more likely to be in sync when the project manager was involved in managing external and internal relationships and when team members shared information with one another.
Decision-Making Quality
Similar analyses were implemented when inspecting the effect of both independent variables on decision quality, F(2, 18) = 3.05, p = .07. R2 = .25. As expected, an increase in team information sharing was associated with increases in team decision-making quality (β = .48, t = 2.30, p = .02). Unexpectedly, however, upon controlling for the effects of team information sharing, project manager communication had a trivial effect on team decision-making quality (β = .09, t = 0.45, p = .33).
Process Commitment
Regarding the effect of both independent variables on process commitment, the regression analysis failed to produce an overall model that was significant, F(2, 18) = 0.27, p = .77. R2 = .03. The effects of both project manager communication (β = −.11, t = −.45, p = .33) and team information sharing (β = .16, t = 0.66, p = .26) had trivial effects on process commitment.
Project Quality
Finally, regarding the effect of both independent variables on perceptions of project quality, a subsequent regression analysis produced an overall model that was significant, F(2, 18) = 5.72, p = .01. R2 = .39. Inspection of the main effects suggested strong effects for both the project manager communication (β = −.30, t = −1.61, p = .06) and team information-sharing variables (β = .61, t = 3.23, p < .01). As the direction of the effects indicate, team information sharing had the expected positive effect on perceptions of project quality, but project manager communication had a negative effect.
Summary of Main Effects
In summary, the general prediction that team information-sharing behaviors would facilitate positive team-level outcomes was generally supported (H1a–H1d). Increases in team information-sharing behaviors were associated with increases in goal alignment, better decision-making practices, and positive perceptions of project quality. In addition, the effect of team information sharing on process commitment was also positive, but this effect was smaller.
Regarding the effect of project manager communication on project-related outcomes (H2a–H2d), the results produced were decidedly more variable. Although project manager communication had a strong, positive impact on goal alignment among team members, it produced null or negative effects on other important team-level outcomes. The effect of project manager communication on decision quality was insubstantial albeit positive, and the effect on process commitment and project quality both trended negative.
Interaction Terms
To test the extent to which both independent variables combined nonadditively to affect participant responses, an interaction term was created by multiplying the values of both independent variables (Cohen et al., 2013). The criterion variables were then regressed onto the interaction term, controlling for the main effects of both independent variables. If the interaction variable produced an effect of substantial magnitude, then simple slopes were modeled at different levels of project manager communication (i.e., −1 SD, mean, and + 1 SD).
Goal Alignment
When inspecting the nonadditive effect of both independent variables on goal alignment, the overall model was significant, F(3, 17) = 7.32, p < .01, R2change = .07; moreover, the interaction variable produced an effect of notable magnitude, B = -1.79, t = -1.70, p = .05. Thus, this effect was graphed and explored further. Figure 2 shows the effect of team information sharing on goal alignment is strong and positive when project manager communication is low (B = 1.61), but becomes attenuated as project manager communication behavior increases (B = 0.07).

Team information sharing by project manager communication interaction.
Decision-Making Quality
When investigating the effects of the interaction term on decision-making quality, results were similar, F(3, 17) = 3.93, p = .03, R2change = .16; B = -1.59, t = -2.12, p = .03. Figure 3 shows the effect of team information sharing on perceptions of decision quality are strong and positive when project manager communication is low (B = 1.23), but becomes increasingly neutralized, if not somewhat negative, as project manager communication increases (B = −0.14).

Team information sharing by project manager communication interaction.
Process Commitment
Inspection of the effect of the interaction variable on process commitment produced a result consonant with the results reported previously, F(3, 17) = 1.78, p = .19, R2change = .21; B = -2.92, t = -2.16, p = .02. Specifically, as modeled in Figure 4, the effect of team information-sharing behaviors on process commitment is strong and positive when project manager communication is low (B = 1.53). Yet, as project manager communication behavior increases, the effect becomes increasingly negative (B = −0.98).

Team information sharing by project manager communication interaction.
Project Quality
Finally, when treating project quality as the main criterion variable of interest, the results suggested similar patterns in the data, F(3, 17) = 4.56, p = .02, R2change = .06; B = -1.19, t = -1.33, p = .10. As the simple slopes in Figure 5 suggest, high levels of project manager communication appear to attenuate the positive effect of information sharing in teams (B = 0.34). However, this positive effect is larger when project manager communication is low (B = 1.37).

Team information sharing by project manager communication interaction.
Summary of Interaction Effects
In summary, both independent variables combined nonadditively to affect responses in a decidedly consistent manner. As all four figures show, project manager communication can provide project teams with a competitive advantage when team members regularly fail to share information with one another. Yet, the benefit of having an involved project manager is mitigated as teams take it upon themselves to share information diligently with one another. Thus, these results support H3b, but not H3a. Stated plainly, presuming that team members are already sharing information at adequate levels, communication activity by project managers appears to be either unnecessary or sometimes harmful.4, 5
Discussion
Results from this study suggest that team-level information sharing is beneficial in IPD teams. First, these findings are in line with past meta-analytic findings, where sharing information within teams is associated with increases in both decision and project quality (Lu et al., 2012; Marlow et al., 2018; Mesmer-Magnus & DeChurch, 2009). Indeed, information sharing in these project delivery teams facilitated the extent to which team member goals were in sync, produced decisions of a higher pedigree, and promoted the general effectiveness of IPD relative to other construction management delivery methods (Choi et al., 2019). As additional organizations adopt similar project delivery methods, facilitating and promoting effective team information-sharing behaviors will likely prove critical to fostering both team and organizational success. Prior research in the AEC industry suggests that team member information sharing is especially vital to discovering and correcting design errors and eliciting buy-in from members and their respective organizations (e.g., Baiden et al., 2006; Dossick & Neff, 2011; Mollaoglu-Korkmaz et al., 2013; Sun et al., 2015). The results presented herein buttress these initial works, and they contribute further by providing a direct quantitative assessment of the effects of team-level information sharing in IPD project teams.
Second, the results regarding project manager communication were less consistent or even contrary to expectations. Although project managers facilitated the alignment of goals in IPD teams (Senescu et al., 2012), their general involvement produced few benefits in terms of increasing decision-making quality and commitment to the project. Moreover, teams with communicatively involved project managers reported lower levels of project quality. Although project team managers are typically assigned to teams to promote information sharing within and between teams (Müller & Turner, 2010; Nofera et al., 2011), the findings here raise doubts regarding the wisdom of the general adoption of this approach (e.g., Iorio et al., 2012).
The interaction analyses, however, suggest that assigning a project manager to IPD teams can be quite beneficial under certain conditions. Project manager communication appeared beneficial when team members did not share adequate levels of information on their own. Thus, when team members do not actively engage in information sharing, having a project manager that encourages interaction among team members is decidedly beneficial. Ultimately, this positive effect likely stems from project managers’ ability to initiate interaction among project team members who do not communicate frequently (see Kotlarsky et al., 2015). Yet, project manager involvement was harmful when the IPD team was already active in sharing information. Under these latter circumstances, IPD project managers may clutter operations, thus complicating the decision-making process. Taken together, these results suggest that the decision to assign a project manager should be contingent on the extent to which project manager communication behaviors are required for fostering IPD team effectiveness. Of note, these results align with recent research by Hsu et al. (2017), which suggests that designated leaders can attenuate project quality when the team is already active in leading itself, but otherwise improve it when the team lacks such properties.
Uncovering the factors that either promote or preclude involvement from project managers in IPD teams will be decidedly critical to explaining this effect further. For example, the extent to which project managers can estimate the competence of their team members accurately might preclude the negative effects of unnecessary interference (Hoegl & Muethel, 2016). Future research would also benefit from uncovering the additional conditions under which project team management behavior and team dynamics combine to impact relevant team-level outcomes. The extent to which project managers have project-relevant knowledge might amplify the beneficial effects of productive team behavior, but otherwise attenuate it (Iorio et al., 2012). Beneficial effects of project managers’ involvement may also be contingent on the extent to which team members believe the project manager constitutes a credible source of information, or the extent to which they listen and adhere to the project manager's advice. Ultimately, continued investigation of these matters will be critical for practitioners as they attempt to understand the conditions under which it is either harmful or beneficial to assign a project manager to an IPD team. Such investigations could also be used to train project managers to become increasingly aware of these types of project dynamics.
The study's findings also raise several questions about team composition, training, and intervention (Kozlowski & Bell, 2013). In multiorganizational projects, there may be little input into the selection of organizational representatives. In some cases, individuals may have worked with each other before, have embedded feelings of trust or mistrust, or have stakeholders who interfere with trust building among members (Nofera et al., 2011). Although considerable attention is devoted to team processes or improving team members’ information sharing (e.g., AIA, 2007; Ebrahimi & Dowlatabadi, 2018), the extent to which organizations carefully weigh member selection or craft teams is uncertain. Uncertain also are the protocols or training in IPD (AIA, 2007) when project managers recognize that teams are sharing information at inadequate (Paik et al., 2017) or inappropriate levels (e.g., too much information; Hansen, 1999; 2002). The extent to which those overseeing project teams can recognize deficiencies and take remedial action is certain to enhance the likelihood of project team success.
Two general sets of limitations are considered here. First, the survey asked respondents to answer broad questions about project coordinator behaviors. As noted earlier, project team manager and project coordinator are often interchangeable in terms and personnel. To promote a more inclusive response, the survey used the terminology of project coordinator. Yet, the extent to which team members were responding about one project coordinator, and one project coordinator only, remains unclear. The severity of this limitation is mitigated by the fact that substantial within-group agreement was established for the project coordinator variable (i.e., team members provided similar responses). The severity of this limitation is mitigated further by the fact that it is typical for only one project coordinator or manager to be assigned per IPD team (Mollaoglu-Korkmaz et al., 2014). Nevertheless, future research endeavors can rectify this potential limitation by specifying the project coordinator or manager of each team when soliciting responses. Future scholarship could also explore whether informal facilitators of team information sharing emerge as the project progresses (Iorio et al., 2012).
A second set of limitations involve the nature of the sample. Only 21 teams were sampled in this study, which can be problematic because small sample sizes have more sampling errors. This, in turn, makes it more difficult to reject the null hypothesis and thus substantiate findings. It is worth mentioning, however, that many of our findings were able to attain conventional levels of statistical significance, and that the patterns of results were generally consistent with one another.
In addition to the size of the sample, sample composition also merits consideration. Sample demographics indicated that men represented an overwhelming majority of the sample (90.8%), which raises questions about the generalizability of our results to other samples with greater diversity in biological sex. Nevertheless, the disproportionate percentage of males in the sample does not invalidate the reported findings for two reasons. First, the construction management arena is largely comprised of men (Hickey & Qui, 2020), suggesting that these findings do in fact apply to a large portion of the construction industry. Second, a recent meta-analysis by van Dijk et al. (2012) found that gender diversity had virtually no effect on task performance. Nevertheless, we acknowledge that the extent to which gender diversity impacts other relevant outcomes (e.g., team goal alignment) remains unclear. Presuming the effects were negative, it would be important to mitigate interpersonal conflicts that generally arise due to surface-level differences (e.g., Manata, 2016, 2019; see also van Knippenberg & Mell, 2016).
Third, most of our project teams were reporting on projects that were either in the construction phase or had already completed (82.2%). It is reasonable to suggest that any negative effects from project management involvement would be less severe in earlier stages of project development given IPD's emphasis on sharing information in its initial phases (see also Lu et al., 2012). Conversely, any negative effects from project manager involvement may be even greater during the earlier stages, especially if team members were sufficiently competent to perform the task on their own. High-performing teams, in particular, may avoid discussion conflicts in the early stages of a project (Jehn & Mannix, 2001). We encourage future investigations to explore the dynamics of project manager involvement and team-level information sharing over the entire life cycle of IPD construction projects.
Finally, we acknowledge further that these results may not generalize to non-IPD construction delivery methods (e.g., design-build, construction management at risk). Yet, given the general applicability of the project manager and team effectiveness literatures, there is little reason to believe that they should not be considered when evaluating other project team or construction contexts (see also Kozlowski & Bell, 2019). Additional research will also be required to investigate the extent to which our results generalize to other relevant outcomes (e.g., intentions to quit; motivation).
Conclusion
In conclusion, given the rarity with which IPD projects are investigated in a quantitative manner, these results make a marked set of substantive contributions to the IPD literature. These results provide empirical evidence to indicate that information sharing in IPD teams is perceived as beneficial for numerous team- and project-level outcomes, whereas the effects of assigning a project manager are more nuanced. More specifically, these results indicate that assigning a project manager should be contingent on the extent to which the team is already operating at optimal information-sharing levels. Overall, the integrative model proposed and analyzed herein is of marked value because it contributes to our understanding of the specific team-level conditions that either promote or attenuate IPD project quality. Research in IPD and other team settings (e.g., McGrath, 1984) in other construction contexts that seek to replicate these findings will contribute substantially to our understanding of team information sharing and project managers’ communication behaviors (see also McEwan et al., 2018).
Footnotes
Acknowledgments
Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. We would also like to thank the reviewers and editor for providing helpful comments.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Science Foundation through Grant SES-123-1206.
