Introduction
During the last decade, public and private organizations and institutions worldwide have undergone major changes including the use of information technology, distributed work patterns, social exchange, and speed of organizational processes. Information technology achievements, in particular, have propelled the development of a new set of “virtual” organizational identities that include virtual organizations, virtual teams, and virtual projects. In fact, 60 percent of projects are considered virtual (Guss 1998a). However, research on the effect of virtuality on project management processes and subsequent success is practically nonexistent. This paper focuses on two principal drivers of project performance in a virtual environment—project manager’s leadership and project teamwork.
Project manager’s competence always has been recognized as a critical project success factor (see Crawford 2000 for review), however today, when the majority of projects are run by project teams (Bounds 1998), the role of project manager is changing. Proper completion of classical project manager’s tasks, such as project planning and progress tracking, are not enough to deliver successful projects. The project manager should also facilitate the development of stimulating teamwork environment that fosters the team member involvement and commitment (Thamhain 2000). Such integration of task related and relational behaviors is referred to as project leadership (Bass 1990).
The virtuality of today’s projects causes even more challenges to the leadership and team process interaction. Characterized by the geographical distribution of team members and strong reliance on technology-mediated communications, “virtual projects” possess new social and task related dynamics within the project team (Kayworth and Leidner 2001/2002). These dynamics affect the impact of different leadership behaviors on project team processes and project outcomes. The discussion on the specifics of these interactions will begin with a review of relevant prior research followed by the development of the theoretical model. The subsequent section describes the research methodology used to empirically validate the model and presents preliminary empirical results.1 The paper is concludes with a discussion of contributions and future research directions.
Prior Research
Topics touched upon in this study have been previously addressed in project and team performance, leadership, and team virtuality research. This section outlines the aspects of these research streams most relevant to the model proposed in the paper.
Over the last few decades project management research has identified many factors that directly or indirectly lead to improved project performance (see overview by Morris 2000 and Kloppenborg and Opfer 2000) and the multidimensionality and complexity of those factors has been widely recognized (Murphy, Baker, and Fisher 1974; Pinto and Slevin 1988; Slevin and Pinto 1986; Pinto and Prescott 1988). Among those factors a project manager’s competence has been consistently identified as one of the most important factors (Jiang, Klein, and Margulis 1998). In addition, leadership constantly ranked as the most important among all the competencies (Crawford 2000). Because many of today’s projects are run by project teams (Bounds 1998), effective teamwork has been closely following leadership as an important contributor to project success (Crawford 2000; Kloppenborg and Opfer et al 2000). However, project management research has devoted little effort to investigating the causal relationship between leadership team processes and project performance (Jiang, Klein, and Chen 2001; Thamhain 2000). At the same time a considerable body of knowledge has been accumulated in the organizational behavior literature about the issues surrounding project teams (for review see Cohen and Bailey 1997) and leadership (Bass 1990). The following paragraphs provide a brief overview these two research streams as they apply to this paper.
Leadership Research
Leadership is one of the oldest management concepts, however there is still no consensus what constitutes leadership and how it is manifested (Bass 1990). Most of the leadership theories can be classified into one of three traditions: trait, behavioral, and contingency theories (Kayworth and Leidner 2001/2002; Horner 1997). Trait theories associate leader’s influence with some inherent characteristics of the person-leader. According to the trait view, leaders are “born, not made.” Behavioral theories focus on leader’s behaviors as a source of influence. Finally contingency theories argue that leadership effectiveness depends on the fit between particular leadership style used and specific circumstances (Kayworth and Leidner 2001/2002; Horner 1997). In the management context an overriding theme of the majority of these theories is that leaders (in one way or another) provide task structure and facilitate interpersonal interactions and positive working relations (Bass 1990). As such leaders directly contribute to their followers well being and efficiency. However, with the emergence of self-managed teams, this understanding of single leaders role should be reevaluated. Horner (1997) suggested that leadership in such teams often comes from within the team, where team members rotate or share leadership roles, thus the influence of single leader diminishes. In the virtual team context, the leadership behaviors become even more complex as the team members geographically distributed and are communicating using computer-mediated communication tools (Kayworth and Leidner 2001/2002).
Project Teams
A considerable body of team research has been devoted to the issues of teams working on projects (Cohen and Bailey 1997). This research includes cross-functional, new product development, and concurrent engineering teams. Majority of theories in this research stream draws on the three-domain framework of Context (Inputs), Process (Transformation), and Outcomes (Outputs) (Ancona 1990; Denison, Kart, and Hahn 1996). According to this framework, various team processes mediate the impact of contextual (environmental) factors on to team’s performance. One of the most important team processes identified is team cohesiveness, which is extent to which team members feel as part of the team and desire to remain in the team (Langfred 1998). As such team cohesiveness is a necessary prerequisite for team functioning and previous research suggests positive relationship between cohesiveness and performance (Evans and Dion 1991). However we have not found previous research that has investigated the relationship between team virtuality and team cohesiveness in project environment.
Virtual Teams
The concept of virtuality in the form of virtual teams, virtual organizations, and virtual enterprises was first introduced to the business literature during 1990s (Lipnack and Stamps 1997). Some have estimated that more than 60 percent of projects are considered virtual (Guss 1998a). However, there is no clear consensus on what comprises virtual teams. Different authors have identified somewhat different virtual characteristics or dimensions—time, space, and structure (Skryme 1998), people, purpose, and links (Lipnack and Stamps 1997), boundary crossing, resource sharing, geographic dispersion, participant equality, electronic communications (Jagers, Jansen, and Steenbakkers 1998), modularity, heterogeneity, time, and space distribution (Wassenaar 1999).
So far the main focus of virtual team research has been on the trust development within the team (Brown 2000; Jarvenpaa and Leidner 1998; Jarvenpaa, Knoll, and Leidner 1998) and the impact of electronic communication means on team processes (Ocker 2001; Warkentin, Sayeed, and Hightower 1997). The findings from these studies suggest that there are differences between virtual teams (defined as dispersed temporary teams that use electronically mediated communication tools) and conventional teams (defined as mainly using face-to-face meetings). Conventional teams are more satisfied with group process than virtual teams (Ocker 2001; Brown 2000; Warkentin, Sayeed, and Hightower 1997), and the net result is increased information exchange (Warkentin, Sayeed, and Hightower 1997), but this does not necessarily translate into better performance (Ocker 2001; Brown 2000).
Overall the virtual team research is still in an exploratory phase and is dominated by controlled experiments and/or case studies (Maznevski and Chudoba 2000) with very few examples of field research. Although the experimental results provide good extreme conditions for examining conventional teams versus virtual teams, there are several limitations. Student teams, by design, don’t have a team history or future prospects, however these factors have been identified as important to team performance (Delisle and Thomas 2000). Also project teams cannot be strictly divided into virtual versus nonvirtual, rather they differ based on the rate of virtuality (O’Leary and Cummings 2002). And of course there is the question of generalizability. Therefore the focus of this research is to extend virtual team concepts to real world setting.
Theoretical Model
The main focus of this study is the project and the team that manages it. The Project Management Institute (PMI®) defines a project as “a temporary endeavor undertaken to create a unique product or service” (Project Management Institute 2000, 4). In other words, a project is a set of tasks that all together lead to a unique product or service. The project is run by the project management team that is defined as the central management group, lead by the project manager, responsible for the management and successful outcome of the project (adopted from Wideman 1991). While both definitions match nicely, the distinction should be made between the project and the team in terms of their lifespan. In the simplest case (the way project teams are defined in a majority of virtual team experimental studies) the team lives or exists as long as the project is in the works. In practice, however, this is rarely the case. Usually team members know each other (e.g., have undergone some team development) before the project starts and might switch during the project execution. This study focuses on team processes during the project and uses project performance (not team performance) as a measure of the team accomplishments. Such an approach provides more meaningful conclusions, as is directly relates project success (the ultimate goal of project management) to the team’s processes (Exhibit 1).
The competent project manager or successful project champion always has been recognized as a prerequisite for project success (Crawford 2000; Pinto and Slevin 1989). These people posses various characteristics and exhibit certain behaviors that altogether facilitate “getting the job done” and many ways they can be considered the art and science of project management. While this works well for projects where there are distinct project champions, the truth is that there are not enough people with the appropriate skill set to serve the increasing number of projects in our society. It is, therefore, important to identify the underlying behaviors that can be transferred (taught) to other people fulfilling the project manager’s role. One of such characteristics is leadership, which is defined as “the ability to identify work that has to be done and to select the people who are best able to tackle it. Leadership is about setting goals and objectives and generating enthusiasm and motivation amongst project team members and stakeholders to work towards these objectives” (CRMP 1999). This conceptualization of leadership incorporates the two main themes identified in leadership literature that is initiating structure and facilitating interpersonal relationships (Bass 1990). In project context it also incorporates project planning and team development that have been ranked among the top project manager competence factors (Crawford 2000). This understanding of leadership follows the behavioral tradition of the leadership theories (Kayworth and Leidner 2001/2002), suggesting that a leader’s influence is achieved through his behaviors, not personality characteristics.
In the project team context leadership works in two main directions (Exhibit 1). In one direction, the structualization and goal setting contributes towards project success. In the other direction, the facilitation of interpersonal relationships creates the team atmosphere among project team members thus contributing to the team cohesiveness (Bass 1990). However in practices it impossible to distinguish the two sub dimensions and the project leader should exercise both types of behaviors in order to be successful (Thamhain 2000).
Exhibit 1. Team Virtuality Impact on Efficiency of Project Manager’s Leadership and Team Cohesiveness
Cohesiveness denotes the extent to which team members feel as part of the team and their desire to remain in the team (Langfred 1998; Seashore 1954, 11). As such cohesiveness is a team specific concept and is entirely based on the team member attitudes towards each other irrespective of the task team is set to accomplish (Langfred 1998). On one side team cohesiveness provides the coherence necessary for effective teamwork. On the other side, a team with exceptionally high cohesiveness could isolate itself from the environment, thus failing to address the environmental constraints during the project execution (Ancona and Caldwell 1992b). However in today’s changing project environments it is not believed that team cohesiveness would reach the levels that would seriously diminish team’s performance. Therefore it is expected to have positive relationship between team cohesiveness and project success.
Virtuality refers to something whose existence is inferred from indirect evidence (Merriam Webster Dictionary). Following this conceptualization a virtual team is a team whose existence can only be detected indirectly (e.g., through the job it accomplishes). Two factors play an especially important role in creating team virtuality—physical (geographical) distribution and reliance on computer mediated communications (Lipnack and Stamps 1997).
On the dimension of physical distribution, a conventional team is closely collocated while a virtual team is geographically dispersed. In current practice, however, the extremes are highly unlikely and project teams differ on their level of physical distribution (O’Leary and Cummings 2002). Physical distribution has two main effects. First, physical proximity is strongly related to face-to-face communications; beyond a distance of thirty yards the total amount of communications becomes asymptotically low (Zahn 1991; Allen 1977). Second, physical proximity leads to team member membership in strong organizational groups with key informants within the thirty-yard range (Keller and Holland 1983). Physical distribution will have three effects on leadership—team cohesiveness interaction. First, it will have negative effect on team member cohesiveness, since limited relational information exchange between team members will prevent establishment of interpersonal contacts that are at the core of team cohesiveness. Second, leader also facilitates interpersonal relationships that are heavily based on social information exchange. As the social information exchange decreases the efficiency of leadership declines. Third, the increase of team virtuality (decrease of information exchange) will decrease overall efficiency of leadership in influencing project performance.
An unavoidable consequence of physical distribution of project team is a reliance on technologically mediated (e.g., videoconferencing, telephone, email, database sharing, web-conferencing) communication (as opposed to face-to-face). Different communication means differ on their ability to transmit (see Media Richness Theory [Daft and Lengel 1984]; Social Presence Theory [Short, Williams, and Christie 1976]) and/or rate of transmission (see Social Information Processing Theory [Walther 1992, 1993]) of different types (e.g., task related, social) of information. According to these theories face-to-face communications can transmit all types of information at the fastest rate, thereby contributing to effectiveness of leadership and team cohesion. Technologically mediated communications are slower and perceived to be less capable to transmit social information, however they still contribute to the impression building of one person about the other (Walther 1992, 1993). Technologically mediated communications will diminish the efficiency leadership, however they can still ensure the exchange of information needed for task completion and enough social information for team coherence. This reasoning provides support to the common practice of holding initial project meeting face-to-face to introduce team members (Guss 1998b; Delisle 2000; Maznevski and Chudoba 2000).
Research Methodology and Measures
To test the described model this study is using a web/paper2 survey to collect data. The participants for the study were recruited through the organizations represented in the local chapter of PMI. After initial contact, organizations were asked to identify project team members, project manager, and project sponsor for the projects to participating in the study. The objective was to use multiple informants at different levels in order to capture the possible variation in project performance assessment between different stakeholders as reported in previous studies (Ancona 1990).
The web/paper format was chosen because it facilitates efficient distribution and collection of data from virtual (geographically dispersed) project team members. At the same time this format provides response validity, response rate, and data quality compatible with paper-based (mail) surveys (Zhang 2000; Weibl and Wallace 2000; Sheehan and McMillan 1999).
The study design is similar to several previous multi-informant project team studies (e.g., Pinto and Pinto 1991; Pinto, Pinto, and Prescott 1993; Ancona and Caldwell 1992a, b). We expect to have 200–250 respondents representing 30–40 projects. Data analysis used in the study will follow the methods used in the aforementioned studies. This will involve the evaluation of psychometric properties of the constructs, testing of inter-rater reliability and agreement to warrant aggregation of responses to the project level (Klein et al 2000), and use of path analysis to make inferences regarding hypothesized relationships and evaluate overall model fit.
To extend the previous findings and ensure the reliability of the constructs, the measures that have exhibited strong psychometric properties in previous studies were selected for this study. Project success is measured with adopted scale from the previous project success studies (Pinto and Prescott 1988; Pinto and Slevin 1988). Team cohesiveness is measured using the scale that based on the work of Seashore (1954), and later used in several project team studies (c.f. Ancona and Caldwell 1992b; Keller 2001). The leadership scale was adopted from project manager’s performance scale used to investigate leadership in information system projects (Jiang, Klein, and Chen 2001; Jiang, Klein, and Shepherd 2001). All of these constructs were measured using five-point Likert’s scale with anchor points of “strongly agree” and “strongly disagree.”
To calculate the physical distribution index (O’Leary and Cummings 2002) the location of each team member’s office (in terms of building and floor) is obtained from each organization. The use of different communication means is measured by asking the respondents to indicate the frequency of use of different communication tools using five-point Likert’s scale with anchor points of “never” and “very frequently.”
Preliminary Results
This paper reports preliminary results of an ongoing field research of the impact of project team virtuality on leadership, project team processes, and project performance. The following results are based on the responses from seventy-eight project team members representing twenty-five projects in education, information system development, waste management construction, and financial sector. The findings are based on correlation and regression analysis of individual level data.
The results of the preliminary analysis indicate strong psychometric properties of the measures used in the study (Exhibit 2). The data provides support for the impact of leadership and team cohesiveness on project success (statistically significant correlation coefficients of 0.357 and 0.285). However the effect of team cohesiveness disappears in the presence of leadership. Further analysis support the assertion that leadership and face-to-face communications are strong predictors of team cohesiveness.
Exhibit 2. Preliminary Regression Results
Conclusion
Rapid globalization of company operations during the last decade has triggered the development of new management practices, which have influenced nearly all areas of project management. For example, we are witnessing a rapid increase in geographically distributed project teams that are reliant on modern information and communication technologies. Contemporary organizational literature refers to such teams as “virtual teams.” The characteristics of these teams are significantly different from those of “conventional teams,” where close proximity and frequent face-to-face contacts were considered prerequisites for good team performance. As the team becomes more virtual, that is, geographically dispersed and technology-mediated-communication reliant, the team dynamics change. The main contribution of this research is the integration of project management, project team, and virtual team research streams. More specifically this study investigates how the team virtuality affects the efficiency of project leadership in building team cohesiveness and ensuring project success.
Unlike previous research using controlled experiments, this study investigates the project team virtuality in real project setting. That allows investigating virtuality at various levels, thus providing a refined understanding of its impact on project leadership and team process.
Finally, on the practical side, this research will provide organizations with a model to assess the advantages and disadvantages of the use of virtual teams in project management. That will help them to develop an infrastructure that supports virtual teams and maximizes their efficiency.
Notes
1. The empirical testing of the model is supported by the Buffalo Chapter of the Project Management Institute.
2. Web/paper format means that the survey is primarily distributed thought the Internet, however respondents have an option to print and mail back the survey.
