Content uploaded by Paul Chinowsky
Author content
All content in this area was uploaded by Paul Chinowsky on Jan 22, 2014
Content may be subject to copyright.
1
PROJECT ORGANIATIONS AS SOCIAL NETWORKS
Paul S. Chinowsky
1
, James Diekmann
2
, and John O’Brien
3
ABSTRACT
High performance teams achieve outcomes that exceed the expectations of the project and often
demonstrate unique or innovative approaches within a final solution. The foundation of this high
performance is the ability to focus on the success of the team over individual objectives.
However, the recognition of this emphasis is based on the establishment of professional trust and
strong communications between the team members. The Social Network Model of construction
introduced a dual-focused approach to enhancing these elements and creating high performance
project teams. The approach emphasizes balancing both a traditional project management
emphasis on efficiency of communications with a focus on the social factors that move the
project team from efficient to effective. In this paper the model is extended to present the results
of four studies of organizations that are full-service engineering companies that also provide
construction oversight services. The paper presents the results of these studies in terms of the
Social Network Model and the achievement of high performance in the project teams.
Analytical and graphical results are presented based on social network analysis techniques to
provide a multi-perspective analysis of the project teams.
KEYWORDS
Knowledge Management, Project Communications, High-Performance Teams
1
Assóciate Professor, Department of Civil, Environmental, and Arch. Engineering, University of Colorado,
Boulder, CO 80309-0428, paul.chinowsky@colorado.edu, (ph) 303-735-1063, (fax) 303-665-3697.
2
Professor, Department of Civil, Environmental, and Arch. Engineering, University of Colorado, Boulder, CO
80309-0428, james.diekmann@colorado.edu.
3
Research Assistant, Department of Civil, Environmental, and Arch. Engineering, University of Colorado, Boulder,
CO 80309-0428, john.obrien@colorado.edu.
2
INTRODUCTION
High performance teams focus on exceeding traditional measures rather than focusing on
meeting the benchmark accepted by previous project teams. This concept of high performance is
documented and routinely implemented in diverse industries including healthcare and
transportation (Poulton and West 1993). The Social Network Model of Construction introduced
a new perspective on achieving these project organizations within an engineering-construction
perspective (Chinowsky, Diekmann, and Gallotti 2008). Emphasizing the need to enhance the
free and open flow of knowledge as a basis for achieving high performance, the model
challenges organizations to reconsider the classic emphasis on communication as the driver for
project success. Specifically, the need to emphasize social relationships as an equal to project
communications is the core of the Social Network Model of Construction.
Historically, research in project success factors has provided organizations with specific
areas to emphasize in preparing for project success. The concept being that the identification of
specific impacts on project success will guide project managers to systematically engineer
critical issues prior to the start of the project (Ashley and Jaselskis 1991; Pinto and Covin 1989).
These efforts have succeeded in that repeated research demonstrates the value of project
planning and the underlying project success factors on which front-end planning is based
(Gibson et al 2006). However, models built on classic project success factors are limited by their
emphasis on project “efficiency” rather than project “effectiveness”. In contrast, an emphasis on
effectiveness changes the focus to the ability of team members to continuously exchange
knowledge and insights, in addition to project information, to enhance the collective group
output (Katzenbach and Smith 1993).
In this paper, the focus on project success moves from the historic efficiency factors to the
emphasis on effectiveness. Based on four studies of multi-office engineering organizations, the
paper introduces analysis findings that provide an initial demonstration of the need to introduce
the Social Network Model into project organization development. Building on a review of the
Social Network Model, this paper creates a case for a multi-dimensional approach to developing
high-performance organizations through an analysis of both leadership and team actions within a
project team framework.
Social Network Analysis
Social Network Analysis (SNA) has been an instrumental tool for researchers focusing on
the interactions of groups since the concept was introduced by Moreno in 1934 (Moreno 1960).
In the original concept formulation, sociograms were considered a formal representation of the
patterns of interpersonal relationships upon which larger social aggregates are created. This
sociology basis was extended into group dynamics with the concept that individuals or
organizations exchange information during the performance of any activity (Scott 1991;
Haythornthwaite 1996). Given the premise that any activity requires a transfer of information
and knowledge, the extension of this foundation is that these exchanges can be mapped within
sociograms where actors and information exchange become nodes and arcs within the graph
(Wasserman and Faust 1994). The translation of these social interactions to a mathematical basis
was the foundation of the strength and validity of the network approach to communication
analysis. Specifically, the ability to apply mathematical analysis to network information
3
exchange provides researchers with established measurements for analyzing the effectiveness
and weaknesses of the group being studied (Alba 1982).
The value of Social Network Analysis is witnessed in a number of diverse applications
and domains. In popular culture, applications such as Facebook and MySpace are introducing
large-scale network formation to the general user (Boyd and Ellison 2007). In the context of
specific domains, social networks are being analyzed in areas including academic institutions
(Li, Xi, and Yao 2008), learning and innovation (Taylor and Levitt 2004), and political
connections (Krebs 2004) among a number of others. The connections within the networks
represent a significant number of differing relationships including knowledge transfer, learning,
trust, and communication. In an emerging area, the connections within molecules are being
visualized in 3-dimensional networks using the same concepts introduced in 2-dimensional
networks (MAGE 2008). In each of these cases, the network provides both a mathematical and
visual representation of the actors and their relationships. The advantage being that classic graph
theory analysis can be used together with visualization to establish measures of closeness,
centrality, and distance among others.
Recently, the network analysis approach is receiving attention within the engineering and
construction field where concepts such as trust and communication between project participants
is receiving significant attention (Morton et al 2006; Katsanis 2006). The understanding that
engineering projects are fundamentally unstable networks that get reinitiated for each project is
changing the focus on what constitutes a successful network team. Factors such as contract type,
project complexity, and litigation concerns are being analyzed in conjunction with the
relationships in a project network (Pryke 2004). This analysis is establishing a connection
between network stability, project success, and stakeholder familiarity.
THE SOCIAL NETWORK MODEL
The Social Network Model for Construction focuses on altering the emphasis of
construction project management from efficiency of projects to high performance projects.
Since this introduction, the model has been applied to project teams in a broader sense including
management teams. As discussed in the management literature, project-based organizations can
be both project specific and organization specific (Taylor and Levitt 2004). The Social Network
Model research incorporates this broader definition in the multiple scenarios in which it is being
fielded. In each scenario, the underlying premise remains on creating a team that has a greater
focus on the individuals within the team and their ability to collaborate to create a higher
standard of success for the entire team. In the Social Network Model, the underlying hypothesis
is that teams need to be managed as social collaborations to achieve results that exceed
traditional expectations. If projects can be viewed from a social collaboration perspective, then
an increased emphasis will be placed on developing teams that have shared values and trust
among the participants.
The Social Network Model incorporates a social network perspective while not
abandoning the positive aspects of the traditional information exchange model. The model
recognizes the need for both elements as complements in the overall achievement of high
performance. The social element is needed to recognize the importance of collaboration and
4
knowledge exchange, but the information element is still required to achieve the pragmatic
requirements of task completion. At the core of this relationship is the correlation between
knowledge exchange and trust. As detailed in the high performance literature, the key to high
performance is the recognition by the team that the success of the team is of primary importance
and that this success is based on the individuals openly exchanging knowledge for the benefit of
the solution (Losada, 1999). However, as further outlined in the research, the key to knowledge
exchange is a level of trust between the members of the team (Katzenbach and Smith, 1993).
This connection between trust and knowledge exchange provides the connection required for the
model to integrate traditional information-based perspectives with social dynamics. The
fundamental principle for the model being that the achievement of trust in a social network will
lead to a greater exchange of knowledge, thus resulting in enhanced solutions and high
performance results.
The Social Network Model contains two basic components, the Dynamics and the
Mechanics. The Dynamics focuses on the motivators for individuals to increase performance on
a project. The rationale behind this component is based on the research that high performance
teams require trust and shared values to achieve the knowledge sharing which results in
enhanced solutions (Kotter 1996). The second component in the Social Network Model, the
Mechanics, focuses on the information and knowledge that is exchanged during the completion
of the project. The overall concept behind these components is that the greater the level of
communication in the mechanics and the greater the move toward trust and shared values in the
dynamics, will ultimately lead to a greater focus on knowledge sharing and high performance.
SOCIAL NETWORK STUDIES
The initial testing of the Social Network Model occurred on a series of tenant
improvement projects for an international chain of coffee shops. The tests were designed to
demonstrate the capability of SNA to extract relationships within the projects and where these
positive or negative relationships might impact the project success. Using three identical
projects with only the subcontractors as a variable, the tests provided promising results that SNA
could identify areas that indicated underlying reasons for comparative levels of success in the
projects. Specifically, the SNA analysis revealed a relationship between the projects with better
outcomes in terms of client satisfaction and the ones that had a strong trust relationship within
the team. Although this relationship may not have a cause and effect relationship, it did indicate
a relationship that required further analysis. Using these results, the team refined the model and
applied it to the first large-scale test, the Solar Decathlon project at the University of Colorado
(Chinowsky, Diekmann, and Galotti 2008). Once again, the model illustrated that
communication and trust issues between key project personnel correlated to areas where the
project failed to meet its stated objectives. This test provided the preliminary results required to
field the model on several additional scenarios. Of particular interest were the questions of how
the social network model could evaluate the performance of project teams in two contexts; the
field and the project office. The following sections introduce the four studies that were
completed in relation to the latter question of how office performance can be analyzed using the
Social Network Model.
5
The Participants
The participants for the Social Network Model tests each volunteered to be a part of the
study based on presentations of the model to the American Council of Engineering Companies
(ACEC). Each of the companies met a specific set of criteria including: 1) multiple offices, 2)
multiple design disciplines, 3) actively participate in construction inspection and oversight, and
4) been in business for over 10 years. These criteria were set to meet two specific issues; 1) the
organizations needed to be in place for a sufficient amount of time to have potentially developed
a sense of collaboration, and 2) the offices were engaged in both the management of design and
construction to enhance the potential for collaboration and integration. Thus, in the context of
the studies, management coordination teams are the specific context in which project teams are
analyzed. In a parallel study, field teams are being analyzed by the authors to determine
relationships to the management coordination teams described here. The four multi-office
engineering design firm descriptions are as follows:
Company A has been in business over 60 years with six offices. The organization
focuses on custom engineering solutions. The organization has grown steadily and has
received numerous engineering awards for its work. The individual principals and
managers have been recognized for their experience and are well-respected for their
technical expertise. Company A had 31 participants in the study.
Company B is the largest of the firms with 16 offices nationally. The company has been
in business for over 70 years. The company works on all types of infrastructure projects
both domestically and internationally. The organization focuses on using its extensive
resources as a basis for innovation in its solutions. Company B had 100 participants in
the study.
Company C is a 100% employee-owned company established over 15 years ago and has
grown to 6 offices. The organization is focused on full-service design and construction
inspection services. The organization remains committed through its expansion to
continue focusing on client satisfaction as well as placing a significant emphasis on
community leadership and stewardship. Company C had 43 participants in the study.
Company D is a full-service engineering firm with over 40 years of history. The
organization has grown to 5 offices focusing on regional infrastructure services. The
organization places a strong focus on teamwork among its divisions and employees.
Additionally, the organization emphasizes the use of technology internally and for its
clients to support engineering solutions. Company D had 26 participants in the study.
Each of the four organizations meets the criteria set by the study. The organizations differ
slightly in size, but meet the fundamental requirements. Each meets the most important criterion
of delivering multiple services to support all phases of the design-construction process.
6
Methodology
The methodology for conducting each of the social network studies followed a three-step
process adapted from established social network protocols and put in place by the authors in the
original Social Network Model development phase to promote both repeatability and validity.
As a foundation for the process, the authors utilized commercial software that was developed
specifically for social network data collection and analysis. As detailed, the combination of
commercial software and specific analysis provided the basis for standardizing the process
among each of the testing scenarios.
Actor Identification
The first step in the study process was the identification of the actors who would be
included in the network. For this process, the Chief Executive Officer (CEO) of each
participating company identified individuals who have supervisory responsibilities based on the
organization chart. Using the CEO in combination with the organization chart to select the
participants provided the best opportunity to enhance consistency and reduce bias that may
emerge from either self-selection by individuals who want to be part of the study, or from a
single point of contact that may be biasing the sample due to selective inclusion of participants.
Finally, using this criterion provided the ability to establish comparable communities within the
organizations and establish defined populations for fielding the SNA study. From this selection
method, the participants were identified and established as actors within the social networks.
Each of the scenarios summarized in this paper represent results from these comparable networks
and are consistent in their use of comparable actors and positions in the organization charts and
thus provide the basis for comparing the results for each deployment.
On-Line Survey
The second component in the study analysis was to obtain input from each of the actors
in the designated network. SNA researchers focus on surveys to obtain this data and this
research follows this precedent. The basis for the current survey is the Mechanics and Dynamics
levels developed for the Social Network Model. Each level within the model required a
corresponding question for the data input process. Once again, the reader is referred to the
original paper introducing the Social Network Model for full definitions of each model
component (Chinowsky, Diekmann, and Galotti, 2008). Working in conjunction with a
sociologist specializing in social networks, the team developed a survey that would elicit the
responses required to create a network representation of the organization (Appendix A). The
survey questions focus on understanding relationships between each of the actors in the network
for each component of the network model. The network represents a point-in-time reflection for
each of the model components.
As illustrated in the survey, the questions focus on frequency of communications and
knowledge transfer for the Mechanics segment and on levels of reliance, trust, and values for the
Dynamics section. The focus on frequency for the Mechanics section is intended to provide an
indication of the levels of interaction that are occurring within the organization. This focus
returns to the original research question of how to increase interactions to achieve knowledge
transfer and thus high performance. Each of the questions in the Mechanics section builds on the
previous by first establishing a communications basis, and then using this group as a subset to
determine knowledge transfer. Similarly, the Dynamics section uses levels of reliance or trust to
7
determine which individuals have an above average level of reliance and subsequently trust
between each other. In each section, the questions provide a continual refinement to the
population to focus the responses from the participants. The questions provide the opportunity
for the actors to provide input on their relationships with each of the other actors in the network.
The combination of the communications frequency, the knowledge transfer frequency, and the
trust levels provides the basis for determining the degree of collaboration currently within the
organizations and ultimately the potential for achieving high performance.
The delivery of the survey was completed using Network Genie, an on-line survey
system designed specifically for managing social network analysis (Hansen et al 2008). The
survey was distributed to the predefined list of actors in each network. The actors were
individually notified to complete the survey with documentation outlining the goals of the study
and the need to get 100% participation within the network. The deployment process was
successful in each case, with a 100% response rate achieved in each scenario.
It should be noted at this point that the focus of this survey was to obtain the perceptions
of each actor in the organization. The research did not focus on the quality of communications
being exchanged, nor did it focus on the mode in which communications were conducted.
Addressing the quality of communications and determining how information technologies such
as e-mail may affect the networks was outside the scope of this research effort. Rather, the intent
of this study is to focus on the relationships between frequency of communications, levels of
trust, frequency of knowledge exchange and achieving high performance. There are limitations
associated with this approach such as different types of communications can require different
frequencies and different types of knowledge may require multiple contacts. These limitations
are noted and are put forth as future steps in the research effort.
Network Analysis
Finally, the results collected from the surveys were analyzed using the UCINET Social
Network Analysis software (Borgatti, Everett, and Freeman 2002). The UCINET software
provides the mathematical measurements as well as the graphical representations required to
conduct a Social Network Analysis. A separate analysis was completed on each of the survey
questions to acquire the relationships outlined in the Social Network Model. The survey
responses collected from the individual questions was used to create a corresponding matrix for
each question. The matrix was subsequently used to create both a graphical representation of the
network as well as a set of mathematical measurements. In terms of the graphical
representations, each actor in the survey was represented by a node in the network. The
relationships between each actor were represented by lines in the graph at the appropriate
magnitude. In this manner, a relationship such as frequency of communications specific to
organization issues could be isolated for each of its possible values, or groups of values, from
less than once a month to weekly. Similarly, the matrix relationships were used by the UCINET
system to analyze the network from a series of graph theory perspectives. Although multiple
measurements can be obtained, the current research effort focuses on the following based on
results from the initial model development.
Network density – A measure to indicate the amount of interaction that exists between the
network members. Density reflects the number of actual links that exist between members in
8
comparison to the number of potential links that exist if all members were connected through
relationship links. The larger the density number that is calculated, the greater the number of
relationships that actually exist in the network.
Centrality – A key measure that reflects the distribution of relationships through the network.
In a highly centralized network, a small percentage of the members will have a high
percentage of relationships with other members in the network. In contrast, a network with
low centrality will have relatively equal distribution of relationships through the network.
An example of a highly centralized network is one where an individual such as the project
manager serves as a filter for a high percentage of communications rather than
communications being distributed throughout the network.
Power – The power variable works in conjunction with centrality. Whereas centrality
measures the total number of relationships that an individual may have, power reflects the
influence of an individual in the network. Individuals who are giving information to others
in the network, who are in turn passing along that information to others, has a high degree of
influence or power according to SNA research. Individuals, who are mainly on the receiving
end of communications may be central in the network, but have little power as they do not
influence other actions.
Betweenness - This variable measures the amount of information that is routed through an
individual during team discussions. This rating indicates which individuals are involved in
discussions that are occurring within the network. The rating reflects the total number of
loops within the network which an individual is included. The greater number of loops that
are included, the greater the level that the individual is participating in discussions.
Banding Considerations
A variable that must be considered in the analysis of the social networks is the size of the
network. In contrast to SNA studies that focus entirely on mapping and determining
relationships, the current study is focusing on evaluating the level of these relationships in
relation to achieving high performance. In making this evaluation, the size of the network is a
consideration since a 20 person network will have a different level of network density than a 100
person network. Incorporating this difference into the analysis is based on research in human
psychology. The research in psychology finds that humans operate effectively in bands of
networks with a preferred number of no more than 30 members (Dunbar 1993). Given this
banding concept, the total frequency of communications and knowledge exchange is modified
based on the size of the network and the associated number of bands. Specifically, the greater
the number of bands that exist in a network, the greater the analysis looks at the distribution of
communication frequencies. For example, in a network of one band (30 or fewer members), the
analysis focuses exclusively on the preferred threshold of weekly communications due to the fact
that each actor has the capacity to communicate with every other member on a regular basis. For
two bands this analysis is increased to proportionately take into account monthly
communications as the individuals do not have the capacity to communicate with every
individual on a weekly basis. Similarly, networks with three or more bands are proportionately
analyzed to take the same capacity effect into consideration. In this manner, networks of
different sizes can be compared using similar evaluation criteria.
9
The Network Studies
The management studies in the current research effort were undertaken to examine the
relevance of the Social Network Model to organization management teams that support projects
as well as to traditional project field management. The underlying premise is that significant
similarities exist in both instances in terms of achieving high performance. Specifically, both
scenarios require a diverse group of individuals to collaborate in an effort to achieve high
performance in a project context. This collaboration is based on trust and communication, the
same requirement as field-based teams.
The focus of the social network studies on the four organizations emphasized the
collaboration between the team members to establish a high performance environment. The
study looked at achieving high performance from two perspectives; leadership and collaboration.
In terms of the former, the study utilized the network measurements of centrality, power, and
betweenness to determine if the organization leadership was involved with client issues and
providing guidance for collaboration. In the second perspective, the organizations were analyzed
based on three items of interest for achieving high performance based on the Social Network
Model; 1) the professional trust within the team, 2) the amount of communications the team had
in terms of client projects, and 3) the amount of knowledge exchange that was occurring within
the team. Each of these focal points was measured with the density function obtained by the
network analysis.
The Leadership Analysis
Leadership can significantly assist teams to achieve a state of collaboration. This
leadership may be predetermined according to an organizational chart or develop organically as a
team responds to its conditions and environment. In either case, leadership can develop a
guiding force that promotes collaboration. In the leadership analysis segment of the study, the
leaders of the organizations according to the supplied organization charts were analyzed based
on their communications in client issues. The leadership was analyzed for three levels of the
chart, the top executive, the director of engineering or operations, and the office or discipline
principals. The three levels were selected to analyze if the organizations were placing an
emphasis on collaboration beyond the top levels to the individual project or discipline levels.
If the organizations were operating according to the anticipated collaboration model, the
numerical ratings should indicate a greater level of integration at the discipline manager level as
the organization places these individuals in the position of implementing the collaboration
requirement. If the numbers did not reflect this tendency, then it would be possible that either
the top executives were micromanaging the client solution process, or the offices were focusing
more on individual concerns than collaborative issues. In either case, this component of the
study focused on determining if the individuals tasked with guiding collaboration were in fact
performing this function in terms of client project involvement.
Of particular interest in this study was the group of discipline managers. This group is
important as it serves as a boundary spanner between senior management and staff. Specifically,
it is the responsibility of these mid-level managers to ensure that the strategic vision developed
by the senior executives is implemented by the project teams. Thus, in this study, the authors
were interested if the discipline managers were providing leadership in implementing the
10
collaboration objectives. One key measure of this leadership is how central the discipline
mangers are within the client-focused communications. Using the centrality measure, the
authors examine the managers according to their relative position in the organization in terms of
centrality. The higher the level of centrality in communications, the higher the manager is
located in percentile on the graph. Each of the four organizations demonstrates a level of
inconsistency in terms of discipline manager involvement when examined with this measure.
Company C is the most consistent with only one manager falling below the 70
th
percentile in
terms of centrality. Company D demonstrates the most inconsistency in the group with
managers ranging from the 70
th
percentile down to the 20
th
percentile. Ideally, the managers in
each of these organizations would be at the highest percentile levels if they were successfully
performing their boundary spanner responsibilities.
The Collaboration Analysis
The collaboration analysis centers on the Social Network Model where a combination of
professional trust and client-specific communications are necessary to establish the environment
for knowledge exchange. To determine these values for the four organizations, a network
representation was created for each of the variables. The density of each network was then
calculated to provide an indication of the network interactions occurring within the specific team.
Table 1 provides the density measurements obtained for each of the variables for each of the
specific organizations.
An open question for this research effort is to establish a preferred density level
representing frequencies and levels for each of the variables. Given that a higher density
indicates a higher relationship, the current effort establishes that a larger density is a preferred
number. However, an exact percentage as a preferred number remains an open research
question. This relationship is illustrated with the first element of collaboration, professional
trust, where the study examines the level of trust that the team members have within the network.
A positive trust relationship occurs when an actor indicates an above-average level of trust with
another actor. As illustrated, Company D has the highest levels of professional trust at both a
monthly and weekly level. Although Company A and Company C are close to these levels, the
analysis indicates that more work is required to build professional trust between the team
members. Company B has a density level almost as high as Company D when the banding
Organization No.
Mgrs.
Professional
Trust Density
Client Specific
Communication
Density -
Weekly
Client Specific
Communication
Density - Monthly
Knowledge
Exchange
Density –
Weekly
Knowledge
Exchange
Density -
Monthly
Company A 31 45% 6% 13% 6% 11%
Company B 100 51% 10% 20% 6% 13%
Company C 43 41% 13% 25% 7% 9%
Company D 26 58% 12% 30% 7% 19%
Table 1: The results of the density analysis for the three variables impacting high performance
on the Integration Matrix.
11
concept is incorporated into the calculations. Similar relationships exist between the companies
for Client Specific Communication where the networks analyze how often the members
communicate about specific client issues. As illustrated, each of the companies has a weekly
density below 15% and a monthly density that is predominantly in the 20-30% range.
Finally, the third variable focused on the amount of knowledge being exchanged within
the network. This measure indicates that the members are moving beyond the efficiency of
information transfer to the effectiveness of knowledge transfer. Once again, the frequency level
from the number of bands is considered as a threshold for a positive response. As illustrated,
each of the organizations has a knowledge exchange density that is lower than the
communications density. However, Company D stands out as achieving a noticeably higher
level of interchange within its network. In general, the organizations display a weakness in this
variable indicating a greater focus on individual silos than collaboration throughout the network.
It should be noted that the Social Network Model proposes a correlation between the trust,
communication, and knowledge exchange numbers. However, additional data is required before
a statistical certainty can be determined for the relationship.
Collaboration Network Interpretation
The data collected for the four scenarios indicates from a numeric perspective a similarity
in the circumstances for each organization. Each organization demonstrates a propensity toward
isolation versus collaboration. Although there are variances in the results for each factor, the
overall tendencies remain similar. However, when these results are analyzed graphically in a
social network context, differences emerge for the root of each scenario. Figure 1 illustrates the
social network for the knowledge exchange variable for each organization. In these four
diagrams, the first band threshold is used to illustrate where the primary knowledge exchange
relationships exist for each network.
Figure 1-a illustrates the knowledge exchange network for Company A. As illustrated,
the organization is split into distinct groups. In this organization, the groups reflect the
geographic distribution of the offices. Collaboration has been hindered due to a greater focus on
office independence and individual profit centers. The primary gates for exchanging knowledge
are the senior principals in each office. The challenge for organization leaders is how to retain
their desire for office independence while enhancing collaboration between the locations.
Figure 1-b illustrates the knowledge exchange network for Company B. As illustrated,
there is a difference from Company A in that Company B, while still isolated in distinct groups,
relies on fragile connectors to link the groups together. The geographic boundaries still
influence the groupings, but the organization is attempting to establish links between key
individuals. Unfortunately, these connections are fragile in that they rely on single individuals
with little or no redundancy to the links. The challenge for the organization is to encourage
greater links without placing an undue burden on a large organization.
Figure 1-c illustrates the knowledge exchange network for Company C. In this network
we see a deviation from the previous two networks. Specifically, this network displays the focus
on a small group of individuals who are the centerpiece of the knowledge network. The
12
organization is relying on a centralized approach where a key group of individuals are
responsible for ensuring knowledge is distributed through the organization. The challenge in this
case is for the organization to release control of knowledge exchange and encourage individuals
to leverage the expertise of others directly rather than through key individuals.
13
Finally, Figure 1-d illustrates the knowledge exchange network for Company D. As the
smallest of the organizations, Company D has the greatest opportunity to meet the desired
knowledge exchange threshold. However, as illustrated, the organization is hindered by the
separation seen in the previous examples. Specifically, Company D is split into two primary
exchange sub-networks that are based on internal discipline and geographic offices. While the
organization has the fewest number of individuals with whom to introduce a change in process,
the distinct sub-networks could prove to be the most difficult to integrate.
In summary, the four scenarios present similar numerical evaluations, but the actual
knowledge exchange networks exhibit subtle variances that only become apparent when
analyzed from a graphical network perspective.
Figure 1 a-d (clockwise from upper left): The primary band knowledge exchange networks
for the four companies illustrating the differences in the challenges facing the organizations
to enhance knowledge exchange.
14
CONCLUSION
The Social Network Model for construction introduced an innovative and transformative
approach to enhancing project team performance. The current paper extends the introduction of
the model by presenting data from four organizations that are actively engaged in project
management and are focusing on enhancing collaboration within their organizations. The data
collected from these example scenarios provides an indication of the relationship between trust,
communication, and knowledge transfer. In each scenario, the organizations have placed
internal barriers to collaboration based on decisions to divide the organization along geographic
or discipline boundaries in exclusion of collaborative emphasis. The result being that knowledge
networks have been established around the isolated silos rather than throughout the organization.
In conclusion, the current research is demonstrating the need for project organizations to
begin expanding their focus away from efficiency to a broader emphasis on effectiveness. As
outlined in the Social Network Model, it is a combination of dynamics and mechanics which lead
to high performance. An over-emphasis on either side will result in lack of performance for the
project team. In the context of the Integration Matrix, project organizations that desire to move
to a state of collaboration need to focus on improving communication to achieve high
performance.
REFERENCES
Alba, Richard D. (1982). Taking stock of network analysis: A decade’s results. Research in the
Sociology of Organizations, 1, 39-74.
Ashley, David, and Jaselskis, Edward (1991). Optimal Allocation of Project Management
Resources for Achieving Success, Journal of Construction Engineering and Management,
ASCE, Vol. 117, No. 2.
Borgatti, S.P., Everett, M.G. and Freeman, L.C. 2002. Ucinet for Windows: Software for Social
Network Analysis. Harvard, MA: Analytic Technologies.
Boyd, Danah M. and Ellison, Nicole B. (2007). “Social Network Sites: Definition, History, and
Scholarship,” Journal of Computer-Mediated Communication, 13(1), 210-230.
Chinowsky, Paul S. Diekmann, James, and Galotti, Victor (2008). “The Social Network Model
of Construction,” Journal of Construction Engineering and Management, 134(10), 804-
810.
Dunbar, R. I. M. (1993). Coevolution of neocortical size, group size and language in humans.
Behavioral and Brain Sciences 16 (4): 681-735.
Gibson, G. Edward, Jr., Wang, Yu-Ren, Cho, Chung-Suk, Pappas, Michael P. (2006). “What is
Preplanning, Anyway?” Journal of Management in Engineering, ASCE, 22(1), 35-42.
Hansen, William B., Reese, Eric, Bryant, Kelvin S., Bishop, Dana, Wyrick, Cheryl H., and
Dyreng, Douglas I. (2008). Network Genie User’s Manual, Tanglewood Research, Inc.
Haythornwaite, C. (1996). Social network analysis: an approach and technique for the study of
information exchange. Library and Information Science Research, 18(4), 323-342.
Katsanis, C.J. (2006). Network Organizations: Structural and Strategic Implications, Proceedings
of the 2nd Specialty Conference on Leadership and Management in Construction, May
2006, Louisville, CO: PM Publishing, 108-115.
Katzenbach, Jon and Smith, Douglas K 1993, The Wisdom of Teams, MA: Harvard Business
School Press
15
Kotter, J.P. (1996). Leading Change. MA: Harvard Business School Press.
Krebs, Valdis E. (2004). Managing the Connected Organization, Orgnet.com, www.orgnet.com.
Li, Pengxiang, Xi, Youmin, and Yao, Xiaotao (2008). “Where Does Help Come From: A Case
Study of Network Analysis in an Academic Group,” Connections, International network
for Social Network Analysis, 28(1), 73-87.
Losada, M. (1999). The Complex Dynamics of High Performance Teams. Mathematical and
Computer Modelling 30: 179-192.
MAGE Page (2008). http://kinemage.biochem.duke.edu/kinemage/magepage.php, last viewed
December 2008.
Moreno, Jacob L. (1960). The Sociometry Reader, Glencoe, IL: The Free Press.
Morton, S.C., Dainty, A.R., Burns, N.D., Brookes, N.J., Backhouse, C.J. (2006). Managing
Relationships to Improve Performance: A Case Study in the Global Aerospace Industry,
International Journal of Production Research, 44(16).
Pinto, J.K. and J.G. Covin, (1989), "Critical factors in project implementation: a comparison of
construction and R&D projects," Technovation, 9 (1), 49 – 62.
Poulton, B.C. and West, M.A. (1993) Effective multidisciplinary teamwork in primary health
care. Journal of Advanced Nursing, 18, 918-925.
Pryke, S.D. (2004) “Analysing construction project coalitions: exploring the application of social
network analysis,” Construction Management and Economics, 22(8), 787-797.
Scott, John (1991). Social network Analysis: A Handbook, Sage: London.
Taylor, J., & Levitt, R. 2004. “Understanding and managing systemic innovation in project-
based industries.” In D. Slevin, D. Cleland & J. Pinto (Eds.), Innovations: Project
management research. Newton Square, Pennsylvania: Project Management Institute.
Wasserman, Stanley, & Faust, Katherine. (1994). Social network analysis. Cambridge, MA:
Cambridge University Press.
16
Appendix – Survey Questions
The section on Client Project questions has been deleted due to similarity to the organization-
focused questions.
1. What individuals do you communicate with on any topic, work related or not, at least once every
three months?
2. Which of the following individuals do you discuss issues that are SPECIFIC to the organization at
least once every three months?
3. How often do you have specific communications with {name} about organization issues?
a) Rarely - Less than once a month
b) Infrequently - Less than once a week
c) Weekly - 1-2 times per week
d) Frequently - Several times per week
e) Daily - At least once per day
4. How often do you exchange KNOWLEDGE with {name} to develop new solutions or approaches to
organization issues?
a) Never
b) Rarely - Less than once a month
c) Infrequently - Less than once a week
d) Weekly - 1-2 times per week
e) Frequently - Several times per week
f) Daily - At least once per day
5. Who do you RECEIVE information from that is necessary for you to perform/complete your
organization-related initiatives? Select individuals you receive information from.
6. Who do you GIVE information to that is necessary for them to perform/complete their
organization-related initiatives? Select individuals that you give information to
12.
The following set of questions move to a focus on individual relationships. The set of potential
picks returns to the master list. Which of the following individuals have you had experience
working with on organization initiatives or Client Projects in the past 12 months?
13.
Rate the amount of RELIANCE you have on {name} to complete their tasks so that you can
perform/complete yours?
a) No Reliance
b) Little Reliance
c) Moderate Reliance
d) Above Average
e) Strong Reliance
14.
How much do you TRUST {name} to take actions that are mutually beneficial to BOTH you and
they based on your SOCIAL interactions with this person rather than professional project
interaction?
a) No Trust
b) Little Trust
c) Moderate Trust
17
d) Above Average
e) Strong Trust
15.
How much do you TRUST {name} to take actions that are mutually beneficial to BOTH you and
they based on your PROFESSIONAL interactions with this person rather than social interaction?
a) No Trust
b) Little Trust
c) Moderate Trust
d) Above Average Trust
e) Strong Trust
16.
Which individuals do you believe share similar values as you do in relation to the organization; For
example: openness, integrity, respect, flexibility, teamwork, and excellence
