19 Virtual Team Concepts in Projects: A Case Study - pmi.org · Workgroups Through Social Networks ... William Leban,PMP, DeVry University ... 19 Virtual Team Concepts in Projects:

Volume 40, Number 2

June 2009

7 Incorporating Employee Resourcing Requirements Into Deployment Decision MakingAndrew R. J. Dainty, Ani B. Raidén, and Richard H. Neale

19 Virtual Team Concepts in Projects: A Case StudyPalitha R. Kuruppuarachchi

34 Measuring Performance of Knowledge-IntensiveWorkgroups Through Social NetworksKon Shing Kenneth Chung and Liaquat Hossain

59 Contemporary Knowledge and Skill Requirements inProject ManagementHelgi Thor Ingason and Haukur Ingi Jónasson

70 Limits to Effective Leadership Style and Tactics inCritical Incident InterventionsDominic Thomas and Elliot Bendoly

81 Robust Project Network DesignDavid A. Ortiz de Orue, John E. Taylor, Arpamart Chanmeka, and Runi Weerasooriya

94 Availability-Impact Analysis of Project ManagementTrends: Perspectives From Allied DisciplinesYoung Hoon Kwak and Frank T. Anbari

cov1PMJ40-2.qxd 4/28/09 5:53 PM Page 1

EDITORIAL REVIEW BOARDVittal Anantatmula, PMP, Western Carolina University

Rick Bilbro, The Innova Group, Inc.

Tomas Blomquist, Umeå University, Sweden

Pierre Bonnal, CERN Accelerator and Beams Dept.

David Bonyuet, PMP, Delta Search Labs

Wanda Curlee, PMP, BearingPoint

Darren Dalcher, Middlesex University

Steven V. DelGrosso, PMP, IBM

J. Davidson Frame, University of Mgmt. & Technology

Ken Hartley, Parsons Brinckerhoff

Gary C. Humphreys, Humphreys & Associates

Peter Kapsales, Belcore

Young Hoon Kwak, The George Washington University

Alexander Laufer, Technion-Israel Inst. of Technology

William Leban, PMP, DeVry University

Ginger Levin, PMP, University of Wisconsin-Platteville

Stephen Leybourne, Plymouth Business School

Rolf A. Lundin, Jönköping University

Fred Manzer, PMP, Center for Systems Management

Peter Morris, University College, London

Ralf Müller, Umeå University, Sweden

Michael Okrent, Agilent Technologies Inc.

Blaize Horner Reich, Simon Fraser University

Paul B. Robinson, Market Strategy and Analytics Partners LLC

Richard L. Sinatra, PMP, Potomac, MD

James Snyder, Springfield, PA

Anders Soderholm, Umeå University

Paul Solomon, PMP, B-2 Earned Value ManagementSystems

Charles J. Teplitz, University of San Diego

Janice Thomas, Athabasca University

Zeljko M. Torbica, PMP, Cabi Developers, LLC

Derek H.T. Walker, RMIT University Australia

Terry Williams, University of Southampton

Graham M. Winch, The University of Manchester

EDITORIAL ADVISORY BOARDFrank R. Anbari, PMP, The George Washington University

Karlos Artto, Helsinki University of Technology (HUT),Finland

David Cleland, University of Pittsburgh

Lynn Crawford, University of Technology, Sydney

Brian Hobbs, Université du Québec à Montréal

Toshihiko Kinoshita, Waseda University

Mark E. Nissen, Naval Postgraduate School

Asbjørn Rolstadås, Norwegian University of Science andTechnology

Dennis P. Slevin, University of Pittsburgh

The Project Management Journal (Print ISSN 8756-9728); Online ISSN 1938-9507 at Wiley InterScience,www.interscience.wiley.com) is published quarterly byWiley Subscription Services, Inc., a Wiley Company, 111River Street, Hoboken, NJ 07030-5774.

Copyright © 2009 Project Management Institute, Inc.All rights reserved. No part of this publication may bereproduced in any form or by any means, except as permit-ted under Section 107 or 108 of the 1976 United StatesCopyright Act, without either the prior written permissionof the publisher, or authorization through the CopyrightClearance Center, 222 Rosewood Drive, Danvers, MA 01923;Tel: (978) 750-8400; Fax: (978) 646-8600.

The code and copyright notice appearing on the firstpage of an item in the journal indicates the copyrightholder's consent that copies may be made for personal orinternal use of specific clients, on the condition that thecopier pay for copying beyond that permitted by Sections107 and 108 of the U.S. Copyright Law. The per-copy fee isto be paid through the Copyright Clearance Center, Inc.This consent does not extend to other kinds of copying,such as copying for general distribution, for advertising orpromotional purposes, for creating new collective works, orfor resale. Such permission requests and other permissioninquiries should be addressed to the PermissionsDepartment, c/o John Wiley & Sons, Inc, 111 River Street,Hoboken, NJ 07030-5774; Tel. (201) 748-6011; please visithttp://www.wiley.com/go/permissions for more information.

NONMEMBER SUBSCRIPTION INFORMATIONPersonal rates: For print in the US, Canada, and Mexico,$125.00, rest of world, $149.00; electronic, all regions,$125.00; and for print and electronic, in US, Canada, andMexico, $138.00, rest of world, $162.00. Institutional rates:For print in the US, $299.00, in Canada and Mexico,$339.00, and rest of world, $373.00; electronic, all regions,$299.00; and for print and electronic, in the US, $329.00,Canada and Mexico, $369.00, and rest of world, $403.00.Claims for undelivered copies will be accepted only afterthe following issue has been received. Please enclose a copyof the mailing label or cite your subscriber reference num-ber in order to expedite handling. Missing copies will besupplied when losses have been sustained in transit andwhere reserve stock permits. Please address all subscriptioninquiries to Subscription Manager, Jossey-Bass, A WileyImprint, 989 Market Street, 5th Floor, San Francisco, CA 94103-1741; Tel. (888) 378-2537, (415) 433-1767 (International); E-mail: [email protected].

Postmaster: Periodical postage paid at NewtownSquare, PA 19073 USA and at additional mailing offices.Send address changes to Project Management JournalSubscription Distribution U.S., c/o John Wiley & Sons, Inc.,111 River Street, Hoboken, NJ 07030-5774.

Reprints: Reprint sales and inquiries should be direct-ed to the Customer Service Department, Gale Krouser, c/o John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ07030-5774; Tel: (201) 748-8789; Fax: (201) 748-6326; E-mail:[email protected].

MANUSCRIPTSAll manuscripts must be submitted electronically via the jour-nal’s Manuscript Central site (http://mc.manuscriptcentral.com/pmj). Questions regarding submission guidelines andmanuscript status should be sent to Natasha Pollard([email protected]).

All manuscripts submitted to the journal via ManuscriptCentral are assumed for publication and become the copy-right property of PMI if published.

EditorChristophe N. Bredillet, PhD, DSc, MBA,

Ingénieur EC Lille

■

PublisherDonn Greenberg;

[email protected]

Wiley Executive EditorIsabelle Cohen-DeAngelis;

[email protected]

Product EditorRoberta Storer; [email protected]

Project Management Journal AssociateNatasha Pollard;

[email protected]

Publications PlannerBarbara Walsh; [email protected]

Book Review EditorKenneth H. Rose, PMP

Contributing EditorsLisa M. Fisher

Vicki Hoenigke

■

The mission of the Journal is to provide information

advancing the state of the art of the knowledge of project

management. The Journal is devoted to both theory and

practice in the field of project management. Authors are

encouraged to submit original manuscripts that are

derived from research-oriented studies as well as practi-

tioner case studies. All articles in the Journal are the views

of the authors and are not necessarily those of PMI.

Subscription rate for members is $14 U.S. per year and is

included in the annual dues.

PMI is a nonprofit professional organization whose

mission is to serve the professional interests of its collec-

tive membership by: advancing the state of the art in the

leadership and practice of managing projects and pro-

grams; fostering professionalism in the management of

projects; and advocating acceptance of project manage-

ment as a profession and discipline. Membership in PMI is

open to all at an annual dues of $119 U.S. For information

on PMI programs and membership:

Project Management Institute Global Operations

Center, 14 Campus Blvd, Newtown Square, PA 19073-

3299 U.S.A.; tel: +1-610-356-4600; fax +1-610-356-4647;

E-mail: [email protected]; Web site: www.pmi.org

PMI Regional Service Centre: Europe - Middle East -

Africa (EMEA), 300, Avenue de Tervueren B-1150

Brussels, Belgium; tel. +32-2-743 15 73; fax +32-2-743 15

50; E-mail: [email protected]

PMI Asia Pacific Service Centre, #03-01, Rex House

73 Bukit Timah Road; Singapore 229832; tel. +65 6330

6733; fax. +65 6336 2263; E-mail: [email protected]

PMI Representative Offices, Beijing, China, and

Washington, D.C.

© 2009 Project Management Institute, Inc. All rights reserved.

“PMI” is a trade and service mark registred in the United States and other nations; “PMP” and the PMP logo are registered certification

marks in the United States and other nations; “PMBOK” is a trademark registered in the United States and other nations; and the PMI logo,

“PM Network”, “Project Management Journal”, “PMI Today”, and “Building professionalism in project management.” are trademarks of the

Project Management Institute, Inc.



Researchers are invited to submit research papers, poster and symposia proposals for presentation at the 2010 PMI Research Conference, the world’s premier project management research event.

We solicit proffered papers, posters and symposia proposals on any topic related to the theory and practice of project management.

Reviewers will include members of the global project management research community.All papers and posters will be subject to double-blind peer review. Students are also encouraged to apply.

Student posters will be entered in a separate student-level competition. PMI will award a prize to the best student poster.

New this year are panel and presenter symposia on any topic in project management. We invite submissions on the concept of global sustainability as a business imperative and the role of project management in this societal enterprise.

Symposia that draw participants from other disciplines (e.g. management, organizational psychology, adult education, linguistics, sociology, etc.) or from the practice community will be looked on favorably.

1 December 2009

For detailed information, please visit PMI.org/Researchsubmissions

PMI® Research Conference 2010

Making project management indispensable for business results.®

,SREPAP ROF LLACFACULTY POSTERS AND SYMPOSIA PROPOSALS

CALL FOR STUDENT POSTERS

CALL FOR SYMPOSIA PROPOSALS: NEW FOR 2010

submission Deadline

Submission REquirements

6th Biennial

Research Conference

Washington, DC USA

11-14 July 2010

Project Management Institute’s

© 2009 Project Management Institute, Inc. All rights reserved. “PMI”, the PMI logo, and “Making project management indispensable for business results” are registered marks of Project Management Institute, Inc.

For the full list of PMI’s legal marks, please contact PMI’s legal department.


Project Managem

ent Journal■

Volume 40, Num

ber 2■

June 2009

spinePMJ40-2.qxd 4/28/09 2:16 PM Page 1

June 2009 ■ Project Management Journal 1

T h e P r o f e s s i o n a l R e s e a r c h J o u r n a l o f t h e P r o j e c t M a n a g e m e n t I n s t i t u t e

2 From the EditorChristophe N. Bredillet, PhD, DSc, MBA, Ingénieur EC Lille

PAPERS

7 Incorporating Employee Resourcing Requirements Into Deployment Decision MakingAndrew R. J. Dainty, Ani B. Raidén, and Richard H. Neale

19 Virtual Team Concepts in Projects: A Case StudyPalitha R. Kuruppuarachchi

34 Measuring Performance of Knowledge-Intensive Workgroups Through Social NetworksKon Shing Kenneth Chung and Liaquat Hossain

59 Contemporary Knowledge and Skill Requirements in Project ManagementHelgi Thor Ingason and Haukur Ingi Jónasson

70 Limits to Effective Leadership Style and Tactics in Critical Incident InterventionsDominic Thomas and Elliot Bendoly

81 Robust Project Network DesignDavid A. Ortiz de Orue, John E. Taylor, Arpamart Chanmeka, and Runi Weerasooriya

94 Availability-Impact Analysis of Project Management Trends: Perspectives From Allied DisciplinesYoung Hoon Kwak and Frank T. Anbari

104 Cover to Cover—Book ReviewsKenneth H. Rose, PMP

JUNE 2009Volume 40, Number 2

PMJ40-2-toc.qxd 4/28/09 6:54 PM Page 1

2 June 2009 ■ Project Management Journal ■ DOI: 10.1002/pmj

IntroductionThis Letter From the Editor is the second part of a serieswhere I am presenting the results of ongoing researchundertaken in order to investigate the dynamic of evolu-tion of the field of project management.

I define here the theoretical foundation and themethod used in this research before introducing the research protocol and main findings in future LettersFrom the Editor.

Co-Word AnalysisTheoretical FoundationThe analysis of the dynamics of science has attractedmuch interest. A qualitative concern with scientificchange can be found in a range of disciplines like philos-ophy (Popper, 1959), social science (MacKenzie, 1978),history of science (Kuhn, 1970), and science policy(Weingart, 2005). Though these many writers have advo-cated a wide variety of theoretical perspectives, they allhave one thing in common; they do not make use ofquantitative indicators in order to handle aggregateddata. Quantitativists have worked in a quite differentway, using large databases to count publications, cita-tions, and patents (Garfield, Malin, & Small, 1978). Andthey share a common interest in the dynamics of science.However, it is necessary to build on this convergence.The reluctance of the qualitativists to use statisticalanalysis must be overcome and, on the other hand, thematerials collected and their application must con-tribute to a theoretically defensible concept of science(Callon, Law, & Rip, 1986).

Built on the actor-network theory, and as a conse-quence of interaction between actor-networks, resultingstructure of problems, and networks of problematization(Callon, Courtial, & Turner, 1986), the co-word analysistechnique was first proposed to map the dynamics of sci-ence. The most feasible way to understand the dynamicsof science is to take the force of science in present-daysocieties into account. “Actor network” is the theoreticalfoundation for co-word analysis to map the dynamics ofscience. Laboratories and literatures are considered astwo powerful tools for scientists to change the world.

They build complex worlds in laboratories and enforcethem on paper (Latour, 1987). This implies that scientistsattach particular importance to texts. They are usingtexts not only to publish their world built in the lab, butalso as a way to build a world and enroll others. Eventhough science cannot be reduced to texts alone, textsare still a prime source for studies on how worlds are cre-ated and transformed in the laboratory. Therefore,instead of following the actors to see how they changethe world, following the texts is another way to map thedynamics of science.

Based on the co-occurrence of pairs of words, co-word analysis seeks to extract the themes of science anddetect the linkages among these themes directly from thetexts’ subject content. It does not rely on any a priori def-inition of themes in science. This enables us to followactors objectively and detect the dynamics of sciencewithout reducing them to the extremes of either inter-nalism or externalism (Callon, Law, & Rip, 1986). Overall,co-word analysis considers the dynamics of science as aresult of actor strategies. Changes in the content of a sub-ject area are the combined effect of a large number ofindividual strategies. This technique should allow us inprinciple to identify the actors and explain the globaldynamics (Callon, Courtial, & Laville, 1991).

Co-word Analysis MethodCo-word analysis is a content analysis technique thatuses patterns of co-occurrence of pairs of items (i.e.,words or noun phrases) in a corpus of texts to identifythe relationships between ideas within the subject areasas presented in these texts. Indexes based on the co-occurrence frequency of items, such as an inclusionindex and a proximity index, are used to measure thestrength of relationships between items. Based on theseindexes, items are clustered into groups and displayed innetwork maps. For example, an inclusion map is used tohighlight the central themes in a domain, and a proxim-ity map is used to reveal the connections between minorareas hidden behind the central ones. Some other index-es, such as those based on density and centrality, areemployed to evaluate the shape of each map, showingthe degree to which each area is centrally structured and the extent to which each area is central to the others.By comparing the network maps for different time peri-ods, the scientific dynamic can be detected. The co-word

From the EditorChristophe N. Bredillet, PhD, DSc, MBA, Ingénieur EC Lille

Mapping the Dynamics of the Project ManagementField: Project Management in Action (Part 2)

Project Management Journal, Vol. 40, No. 2, 2–6© 2009 by the Project Management Institute Published online in WileyInterScience (www.interscience.wiley.com). DOI: 10.1002/pmj.20128

PMJ402_20128.qxd 4/28/09 10:32 AM Page 2

June 2009 ■ Project Management Journal ■ DOI: 10.1002/pmj 3

analysis technique was first developed in collaborationbetween the Centre de Sociologie de l’Innovation of theÉcole Nationale Superieure des Mines of Paris and the CNRS (Centre National de la Recherche Scientifique)of France during the 1980s, and their system was called“LEXIMAPPE.” For about 20 years, this technique hasbeen employed to map the dynamic development of sev-eral research fields.

Many examples (Callon, 1986; Coulter et al., 1993;Courtial & Law, 1989; Law & Whittaker, 1992; Turner &Callon, 1986) reveal that co-word analysis is a promisingmethod for discovering associations among researchareas in science and for revealing significant linkagesthat may otherwise be difficult to detect. It is a powerfultool that makes it possible to trace the structure and evo-lution of a sociocognitive network (Bauin, 1986). As such,it offers a significant approach to knowledge discovery.

In the following paragraphs, we are introducing themain metrics we are using in this study. A more detailedpresentation of the metrics used in co-word analysis canbe found in He (1999).

MetricsThe basis of a co-word analysis study is the calculation ofthe equivalence or link coefficient:

[E.ij] � ([C.ij]/[C.i]) [multiplied by] ([C.ij]/([C.j])

� [([C.ij]).2]/([C.i] [multiplied by] [C.j]) (1)

where:

[C.ij] is the co-occurrence frequency of the keyword

pair ([M.i] and [M.j]) in the set of articles;

[C.i] is the occurrence frequency of keyword [M.i] in the

set of articles;

[C.j] is the occurrence frequency of keyword [M.j] in the

set of articles; and

[E.ij] has a value between 0 and 1.

This coefficient is used to define the clusters anddraw the strategic diagram.

A cluster is a lexical structured set built from a co-occurrence analysis. It is constituted by a set of wordsstrongly associated or in strong co-occurrence in thedocuments. The co-occurrence analysis is conductedthrough the index (list of words to take into considera-tion) associated to the corpus of documents.

Density, Centrality, and Strategic DiagramA strategic diagram is used to illustrate the “local” and“global” contexts of themes. This diagram is created byputting the strength of global context on the X axis(called centrality) and putting the strength of local con-text on the Y axis (called density). This diagram is used inmany co-word studies. Two kinds of indexes (i.e., densityand centrality) are used to measure the strength of localcontext and global context, respectively.

DensityDensity is used to measure the strength of the links(equivalence or link coefficient) that tie together thewords making up the cluster. This is the internal strengthof a cluster and provides a good representation of thecluster’s capacity to maintain itself and to develop overthe course of time in the field under consideration(Callon et al., 1991). Ranking subject areas (clusters) interms of their internal coherence (density) is designed toprovide information for systematic discussion of a majorpolicy alternative. Further, sorting the keywords bydecreasing order of density can provide a precisedescription of the areas (Bauin et al., 1991). The value ofthe density of a given cluster can be measured in severalways. Generally, the index value for links between eachword pair is calculated first. Then, the density value canbe the average value (mean) of internal links (e.g.,Coulter, Monarch, & Kunda, 1998; Turner, Chartron,Laville, & Michelet, 1988), the median value of internallinks (e.g., Courtial, Callon, & Sigogneau, 1993), or thesum of the squares of the value of internal links (e.g.,Bauin, Michelet, Schweighoffer, & Vermeulin, 1991). Aninternal link means both of the words linked by it arewithin the cluster. Here we consider the average value(mean) of internal links.

CentralityCentrality is used to measure the strength of a subjectarea’s interaction with other subject areas. Ranking sub-ject areas (clusters) with respect to their centrality showsthe extent to which each area is central within a globalnetwork. The greater the number and strength of a sub-ject area’s connections with other subject areas, the morecentral this subject area will be in the network (Bauin et al., 1991). For a given cluster (area), its centrality canbe the sum of all external link values (e.g., Courtial et al.,1993; Turner et al., 1988) or the square root of the sum ofthe squares of all external link values (e.g., Coulter et al.,1998). More simply, it can be the mean of the values ofthe first six external links (e.g., Callon et al., 1991). Anexternal link is one that goes from a word belonging to acluster to a word external to the cluster. Here, we consid-er the mean of the values of the first six external links.

Strategic DiagramA strategic diagram that offers a global representation ofthe structure of any field or subfield can be created byplotting centrality and density into a two-dimensionaldiagram (Law, Bauin, Courtial, & Whittaker, 1988).Typically, the horizontal axis represents centrality, thevertical axis represents density, and the origin of thegraph is at the median of the respective axis values. Thismap locates each subject area within a two-dimensionalspace divided into four quadrants. The strategic diagramis used in many co-word analysis studies (e.g., Callon et al., 1991; Coulter et al., 1998; Courtial & Law, 1989;

PMJ402_20128.qxd 4/28/09 10:32 AM Page 3


From the Editor

Turner & Rojouan, 1991; Turner et al., 1988) and theanalysis based on it is similar among these studies.Generally, the subject areas in Quadrant 1 are both inter-nally coherent and central to the network in question.However, those areas in Quadrant 4 seem to be of onlymarginal interest to work in the global network.Coherent subject-specific areas always appear inQuadrant 3 of the diagram. These areas are internallywell structured and indicate that a constituted socialgroup is active inside them. However, they appear to berather peripheral to the work being carried out in theglobal network. Weakly structured areas are found inQuadrant 2. These subjects, individually, are linkedstrongly to specific interests throughout the network butare only weakly linked together. In other words, work inthese areas appears to be underdeveloped, but it couldpotentially be of considerable significance to the entirenetwork. All these characteristics of a strategic diagramcan be summarized in Table 1.

Dynamics of NetworksA striking feature of some strategic diagrams is the radicalchange in the configuration of the network at two peri-ods. This reflects the dynamics of science. Based on thestrategic diagram, we can analyze the stability of the net-works and foresee their changes in the future. This issueis addressed in many studies, and the methods used inthese studies fall into two categories: (1) the study ofstrategic diagrams and (2) the ratio of centrality to densi-ty. Here we use the former one. This method is used to study the stability of networks and is directly based on the strategic diagrams (e.g., Callon et al., 1991;

Turner & Rojouan, 1991). The findings can be summa-rized as showing that the probability for the content ofthemes situated in Quadrants 2 and 3 to change over timeis significantly higher than it is for themes situated inQuadrant 1. With a low density, the unstructured themesin Quadrant 2 tend to undergo an internal restructuringto improve their cohesiveness. With a low centrality, thescope of themes in Quadrant 3 is likely to be extended inorder to better articulate what is being done in the rest ofthe network. The reason, as well as the goal, for all ofthese changes is to place their work at the heart of theirfield (Quadrant 1). This can be done either by enlargingits scope or by improving its visibility through conceptu-al developments in the definition of the field.

Network ComparisonIn co-word analysis studies, several subnetworks (clus-ters) can be constructed concurrently even though eachnetwork changes over time. Detecting the differenceamong subnetworks simultaneously or subnetworks atdifferent times has been a long-standing research issue.The transformation of networks and their intersectionswith other networks across time periods providesinsights into theme emergence. The similarity of net-works in different time periods has also been studied byCoulter et al. (1998). In this study, the authors employ thesimilarity index (SI), which comes from Callon’s dissimi-larity (or transformation) index (Callon et al., 1991).

A transformation index (T), also called a dissimilari-ty index, is defined to measure the degree of dissimilari-ty between two given clusters. This index is defined as:

Density Centrality

Quadrant 1: Central and developed

Strategic heart of the field. Here we find the main themes.Clusters are very close from each other in terms of keywords(high density, strong association). Furthermore, as centralityis high, these themes are linked to several others.

Quadrant 2: Central and undevelopedClusters are linked by numerous keywords (high centrality)but are very different from each other (low density). Theseclusters represent central main themes but remain generic.

Here we find promising or past themes that are part of the dis-cipline or themes borrowed from other disciplines/contextthemes. This quadrant is important, as the themes are essen-tial for a good understanding of the field.

Quadrant 3: Peripheral and developedThe clusters are close to each other but they are specializedon one theme. We find specialized field themes here, eitherinternal themes constituting an autonomous subfield orexternal themes “imported” from other fields or disciplinesand having new development in the studied field.

Quadrant 4: Peripheral and undevelopedThese themes may evolve to the right, gaining centrality, andevolve upwards, gaining density. They might be at the originof new trends or development within the field.

Table 1: Strategic diagram.

PMJ402_20128.qxd 4/28/09 10:32 AM Page 4

T � ([W.i] � [W.j])/[W.ij] (2)

where:

[W.i] is the number of words in cluster [C.i];

[W.j] is the number of words in cluster [C.j]; and

[W.ij] is the number of words common to [C.i] and [C.j].

The similarity index is defined as follows:

S � 2 [multiplied by] ([Wij]/([W.i] � [W.j])) (3)

where:

[W.i] is the number of keywords (descriptors) in net-

work [N.i];

[W.j] is the number of descriptors in network [N.j]; and

[W.ij] is the number of descriptors common to [N.i]

and [N.j].

A constant 2 is multiplied to make the maximumvalue of SI to 1, which occurs when [N.i] and [N.j] haveidentical nodes (keywords or descriptors). SI is used tomeasure the point of descriptor intersection in two net-works and to examine the emergence of a network dur-ing a particular period.

Some Key Issues in Co-Word AnalysisThe maps obtained by co-word analysis are generallyconsidered very difficult to understand in isolation. Theyhave to be interpreted with caution. It is suggested thatthe interpretation must be active and based on the com-parison of maps (Callon, Courtial, & Turner, 1986). Giventhat the goal of co-word analysis is not just to photo-graph a field of knowledge but to reveal the strategies bywhich actors mutually define one another. Callon et al.(1991) suggest that the maps cannot be considered in sta-tistical isolation; they must be interpreted dynamically.

The choice of the words (keywords, descriptors) isanother issue to be carefully considered and has led tomany discussions (Leydesdorff, 1997; Whittaker, 1989). Aliterature review shows that the words used in co-wordanalysis are expanding from keywords in a lexicon towords in the full text (Bauin, 1986; Callon et al., 1991;Kostoff, Eberhart, Toothman, & Pellenbarg, 1997; Rotto &Morgan, 1997). The “normalization” of words must beconsidered as well (for example, many words for whichBritish and American spellings differ have been stan-dardized to the American spelling by the Institute forScientific Information when they are put into the cita-tion index databases). This has been addressed in sever-al studies (Courtial et al., 1993; Nederhof & van Wijk,1997; Turner et al., 1988).

Questions of meaning or change of word meaning atdifferent levels (during a period of time, from one authorto another, etc.) are also addressed (Leydesdorff, 1997,1998). Words are not used as linguistic items to meansomething in co-word analysis but are used as indicatorsof links between texts, whatever they mean. They are

chain indexes, allowing one to compute translation net-works. What is important for co-word analysis is not theexact meaning or definition of a word but the fact thatthis word is linked to word X in one case and word Y inanother case (Courtial, 1998).

The above-mentioned development will lead me topresent on meaningful foundations in the next “LetterFrom the Editor,” the research protocol I propose inmapping the dynamics of the project managementfield. ■

Ordo ab chaosChristophe N. Bredillet

ReferencesBauin, S. (1986). Aquaculture: A field by bureaucraticfiat. In M. Callon, J. Law, & A. Rip (Eds.), Mapping thedynamics of science and technology: Sociology of sciencein the real world (pp. 124–141). London: The MacmillanPress Ltd.

Bauin, S., Michelet, B., Schweighoffer, M. G., &Vermeulin, P. (1991). Using bibliometrics in strategicanalysis: “Understanding chemical reactions” at theCNRS. Scientometrics, 22, 113–137.

Callon, M. (1986). Pinpointing industrial invention: Anexploration of quantitative methods for the analysis ofpatents. In M. Callon, J. Law, & A. Rip (Eds.), Mappingthe dynamics of science and technology: Sociology of sci-ence in the real world (pp. 163–188). London: TheMacmillan Press Ltd.

Callon, M., Courtial, J.-P., & Laville, F. (1991). Co-wordanalysis as a tool for describing the network of interac-tions between basic and technological research: Thecase of polymer chemistry. Scientometrics, 22, 155–205.

Callon, M., Courtial, J.-P., & Turne, W. (1986). Futuredevelopments. In M. Callon, J. Law, & A. Rip (Eds.),Mapping the dynamics of science and technology:Sociology of science in the real world (pp. 211–217).London: The Macmillan Press Ltd.

Callon, M., Law, J., & Rip, A. (1986). How to study theforce of science. In M. Callon, J. Law, & A. Rip (Eds.),Mapping the dynamics of science and technology:Sociology of science in the real world (pp. 3–15). London:The Macmillan Press Ltd.

Coulter, N., Monarch, I., & Konda, S. (1998). Softwareengineering as seen through its research literature: Astudy in co-word analysis. Journal of the AmericanSociety for Information Science, 49(13), 1206–1223.

Courtial, J.-P. (1998). Comments on Leydesdorff’s arti-cle. Journal of the American Society for InformationScience, 49(1), 98.

Courtial, J.-P., Callon, M., & Sigogneau, A. (1993). Theuse of patent titles for identifying the topics of


PMJ402_20128.qxd 4/28/09 10:32 AM Page 5

invention and forecasting trends. Scientometrics, 26,231–242.

Courtial, J.-P., & Law, J. (1989). A co-word study of artifi-cial intelligence. Social Studies of Science (London), 19,301–311.

Garfield, E., Malin, M., & Small, H. (1978). Citation dataas science indicators. In Y. Elkana, J. Lederberg, R. K.Merton, A. Thackray, & H. Zuckerman (Eds.), Toward ametric of science: The advent of science indicators (Ch. 8,pp. 179–207). New York: Wiley.

He, Q. (1999). Knowledge discovery through co-wordanalysis. Library Trends, 48(1), 133–159.

Kostoff, R. N., Eberhart, H. J., Toothman, D. R., &Pellenbarg, R. (1997). Database Tomography for techni-cal intelligence: Comparative roadmaps of the researchimpact assessment literature and the Journal of theAmerican Chemical Society. Scientiometrics, 40, 103–138.

Kuhn, T. S. (1970). The structure of scientific revolutions(2nd ed.). Chicago: University of Chicago Press.

Latour, B. (1987). Science in action: How to follow scien-tists and engineers through society. Cambridge, MA:Harvard University Press.

Law, J., Bauin, S., Courtial, J.-P., & Whittaker, J. (1988).Policy and the mapping of scientific change: A co-wordanalysis of research into environmental acidification.Scientometrics, 14, 251–264.

Law, J., & Whittaker, J. (1992). Mapping acidificationresearch: A test of the co-word method. Scientometrics,23, 417–461.

Leydesdorff, L. (1997). Why words and co-words cannot map the development of the science. Journal ofthe American Society for Information Science, 48,418–427.

Leydesdorff, L. (1998). Reply about using co-words.Journal of the American Society for Information Science,49, 98–99.

MacKenzie, D. (1978). Statistical theory and social inter-ests: A case study. Social Studies of Science, 8, 35–83.

Nederhof, A. J., & van Wijk, E. (1997). Mapping thesocial and behavioral sciences worldwide: Use of mapsin portfolio analysis of national research efforts.Scientometrics, 40, 237–276.

Polanyi, M. (1958). Personal knowledge. Towards a post-critical philosophy. Chicago: University of ChicagoPress.

Popper, K. (1959). The logic of scientific discovery (trans-lation of Logik der Forschung). London: Hutchinson.

Rotto, E., & Morgan, R. P. (1997). An exploration ofexpert-based text analysis techniques for assessingindustrial relevance in U.S. engineering dissertationabstracts. Scientometrics, 40, 83–102.

Turner, W. A., & Callon, M. (1986). State intervention inacademic and industrial research: The case of macro-molecular chemistry in France. In M. Callon, J. Law, &A. Rip (Eds.), Mapping the dynamics of science and technology: Sociology of science in the real world (pp.142–162). London: The Macmillan Press Ltd.

Turner, W. A., Chartron, G., Laville, E., & Michelet, B.(1988). Packaging information for peer review: New co-word analysis techniques. In A. F. J. Van Raan(Ed.), Handbook of quantitative studies of science andtechnology (pp. 291–323). Amsterdam: Elsevier SciencePublishers.

Turner, W. A., & Rojouan, F. (1991). Evaluating input/output relationships in a regional research network using co-word analysis. Scientometrics, 22, 139–154.

Weingart, P. (2005). Impact of bibliometrics upon thescience system: Inadvertent consequences?Scientometrics, 62, 117–131.

Whittaker, J. (1989). Creativity and conformity in science: Titles, keywords and co-word analysis. SocialStudies of Science, 19, 473–496.


From the Editor

PMJ402_20128.qxd 4/28/09 10:32 AM Page 6

PA

PE

RS


INTRODUCTION ■

The Challenge of Resourcing in Project-Based SectorsEmployee resourcing is one of the key strategic human resource manage-ment (HRM) functions. It focuses on matching resources to the needs of theorganization (both strategic and operational) and ensuring the proper uti-lization of those resources (Armstrong & Baron, 2002). As such, it embodiescore HRM activities such as the recruitment, selection, and deployment ofemployees within the organization, which should be managed in such a wayas to support the strategic objectives of the firm (Taylor, 2008, p. 3). Despitethe vital importance of effective team selection and formation to the successof organizational performance, employee resourcing remains a reactive andlargely ad hoc function in many project-based firms (De Feis, 1987;Loosemore, Dainty, & Lingard, 2003, p. 83). A possible reason for this is thatallocating staff to projects demands that the (often competing) requirementsof the organization, the project, and individual employees be balanced.

Although short-term project and longer-range organizational resourcingrequirements are relatively easy to discern, they are not always mutuallycompatible, especially given the reactive decisions that characterize projectteam deployment (Turner & Müller, 2003). Moreover, capturing employeeneeds and integrating them into the decision process are even more difficult,given the tacit nature of individual aspirations, personal preferences, andprofessional contribution (Raidén, Dainty, & Neale, 2006). Eliciting suchknowledge requires detailed and sensitive involvement with each individual,a time-consuming activity in geographically distributed project-basedorganizations.

Although tacit knowledge is difficult to communicate or share with oth-ers (see Haldin-Herrgard, 2000; Smith, 2001), if it is excluded from thehuman resource allocation decisions, the result is likely to be dissatisfiedemployees, breached psychological contracts (see Guest & Conway, 2002),and, ultimately, attrition. However, despite the growth in project-basedworking, there is a paucity of research examining the process of humanresource allocation within such organizations. Indeed, there is a greater needfor structured workforce management strategies for the industry to informthe management of the workforce, as was recognized by Brandenburg, Haas,and Byrom (2006).

Whereas Brandenburg et al.’s work (2006) began to address the metricsand procedures necessary for a more strategic approach to workforce man-agement strategy, the emphasis in our article is on the proactive management

Incorporating Employee ResourcingRequirements Into DeploymentDecision MakingAndrew R. J. Dainty, Department of Civil and Building Engineering, LoughboroughUniversity, Leicestershire, United KingdomAni B. Raidén, Nottingham Business School, Nottingham Trent University, Nottingham,United KingdomRichard H. Neale, Research Office, University of Glamorgan, United Kingdom

ABSTRACT ■

Employee resourcing is the process of matchinghuman resource capabilities to the strategicand operational needs of the organization. Thisis exceptionally problematic in project-basedorganizations due to the competing priorities of the project, the individual employee, and thewider succession needs of the organization.This article presents the findings of researchexamining the human resource managementpractices that form the key components of theresourcing process. These included, inter alia,human resource planning, recruitment andselection, team deployment, performance man-agement, and human resource administration.Current practices were examined in seven lead-ing construction firms, all of which faceddynamic resourcing priorities. Within an induc-tive methodology, semistructured interviewswere carried out with senior executives, humanresource management (HRM) specialists, sen-ior operational managers, and project-basedstaff. Based on a synthesis of the promisingpractices extracted from the case-study organi-zations, an innovative approach to projectresourcing was developed that aims to balanceorganizational, project, and individual employeerequirements. Team deployment resides at the center of resourcing process for the project-based organization as it determines the suc-cess of the project, which in turn determines thecompetitiveness of the organization. Long-termplanning and employee involvement enableteam deployment to integrate with other ele-ments of HRM effectively and thus help to bal-ance the organizational strategic priorities,project requirements, and individual employeeneeds and preferences.

KEYWORDS: employee resourcing; decisionmaking; employment relationship

Project Management Journal, Vol. 40, No. 2, 7–18

© 2009 by the Project Management Institute

Published online in Wiley InterScience

(www.interscience.wiley.com)

DOI: 10.1002/pmj.20119

pmj402_20119.qxd 4/28/09 3:37 PM Page 7


Incorporating Employee Resourcing RequirementsP

AP

ER

S

of the resource allocation process. Thisresearch set out to develop a morerobust methodology for capturing andtaking account of employee require-ments in the resourcing process. Itexamines the practices of some leadingconstruction organizations, focusingon how the knowledge relating toemployee needs and abilities is cur-rently fed into the resourcing decision-making process. By combining effectiveelements of this practice, and supple-menting it with leading-edge thinkingin relation to the resourcing function,this research provides an improvedframework that could act as a point ofdeparture for incorporating such deci-sions into the resourcing process.

The Requirements andChallenges of EffectiveEmployee Resourcing inConstructionThe Resourcing ChallengeThe major components of employeeresourcing are staffing, performance, HRadministration, and change manage-ment (Taylor, 2008, pp. 3–4). These func-tions comprise several discreet butinterrelated management activities. Thestaffing and performance objectives aimto ensure that the right numbers ofemployees with the right skills and com-petencies are in the right place at theright time. This can be seen as a “balanc-ing act” in which managers are takinginto account the longer-term strategicconsiderations of human resource plan-ning while providing immediate solu-tions for the shorter-term operationalissues, such as recruitment and selec-tion, team deployment, dismissal, andredundancy (Beardwell & Holden, 1997;Rothwell, 1995). To be carried out effectively, management of staffing and performance should be concernedsimultaneously with managing perfor-mance and facilitating employees’ careerprogression and rewards (see Baruch,2003; Pilbeam & Corbridge, 2006, p. 179).HR administration focuses on the col-lection, storage, and use of employeedata supporting the monitoring and

analysis of HR information (CharteredInstitute of Personnel and Development,2000; Torrington, Hall, Haylor, & Myers,1991, p. 22). The change managementaspect of the function aims to facilitatethe continuous evolution of organiza-tional strategies and practices throughthe interrelated aspects of staffing, per-formance, and HR administration (seeTaylor, 2008). It is obvious that managingall of these functions concurrently repre-sents a complex and challenging task forany large organization resourcing manyprojects. However, for construction com-panies, the dynamic context of projectmobilization renders it even more prob-lematic.

Whereas in many industries theresourcing process represents the coor-dination of a set of fairly controllabletasks, the nomadic, transitory, and project-based nature of constructionoften results in employee needs beingsubordinated to project requirements(Raidén, Dainty, & Neale, 2009, p. 133).This is primarily because constructionprojects tend to have a short lead-intime that demands the rapid mobiliza-tion of teams. Moreover, resourcingpractices often rely on the assessmentsof line managers (see Druker, White,Hegewisch, & Mayne, 1996). A judgmenthas to be made of the skills and behav-ioral competencies of available staff,which are reconciled against require-ments (Dainty, Moore, & Cheng, 2005).However, given the likely organizationalpressures of successful project delivery,there is the potential for individualemployee requirements to be down-played in relation to immediate opera-tional requirements. This can, in turn,lead to high levels of staff turnover asemployees’ career aspirations are com-promised (Baruch, 2003), which is initself disruptive to the successful deliv-ery of projects. As such, poor resourcingpractices can lead to a recursive cycle ofinappropriate decision making and, ulti-mately, to organizational inefficiency.

Overcoming the extremely demand-ing resourcing context for both humanresource departments and line managers

requires a flexible approach to theemployee resourcing function in con-struction organizations. The differentcomponents of the employee resourc-ing function (i.e., staffing, performancemanagement, HR administration, andchange management) must be inte-grated to form an interconnected net-work of decision-making and supportprocesses (see Raidén et al., 2009, pp. 169–174).

Toward a More Effective ProjectResourcing ProcessEffective resourcing is clearly predicat-ed on having the requisite knowledgefor managing the process effectively. Aswas discussed previously, however,much of the data required for effectiveresourcing decisions will be tacit innature. For example, understanding anindividual employee’s aspirations andcareer development needs and reconcil-ing these with company performanceassessments is exceptionally difficult. A key factor in ensuring that the requi-site knowledge of employee needs iscaptured is the development of a cli-mate of trust in the employment relationship between employees andtheir managers. Trust is importantbecause employees will not be willingto share their work-related knowledgeunless they feel secure that it will beused for them rather than against them(see Ellis, 2001; Storey & Quintas, 2001).

In project-based sectors, it could beargued that the success of the resourc-ing process effectively depends uponidentifying: (1) the skills and experiencethat an employee brings to their role;(2) their limitations in relation to thetasks inherent in the project for whichthey are being considered; (3) learningopportunities in terms of further devel-oping the individual’s skills, competen-cies, and abilities; and (4) employees’personal needs and career aspirationsreconciled against the project opportu-nities.

Clearly, identifying and understand-ing the limitations of employees is prob-lematic given that most employees do

pmj402_20119.qxd 4/28/09 3:37 PM Page 8


not wish their weaknesses to be madeexplicit to their employer. Even theirpositive knowledge and competenciesmay be seen as a commodity that they may be reluctant to convey to theiremployer out of fear that thesestrengths might be exploited (e.g., causethe employees to be recruited to areas of the organization where they do notwish to work). Furthermore, there is alsothe issue that intrinsic, tacit knowledgeis what many people are employed for,and so exposing these explicitly maydiminish their worth to the organizationin terms of securing competitive advan-tage.

However, the greatest difficultyconcerns gaining insights on the per-sonal aspirations and developmentrequirements of individual employees.Because such needs may not accordwith organizational goals or opportu-nities, the individual may opt not todivulge them, and this may impede theefficient deployment of staff to meetproject requirements. Thus, findingways to incorporate individual careerneeds and aspirations into the decision-making process is fundamental to the successful management of theresourcing process. If managed suc-cessfully, appropriate opportunitiesfor employees to develop within theorganization can be identified and the possibility of harnessing and capi-talizing on employees’ contributionsenhanced.

The research presented here soughtto synthesize existing practice in man-aging this process among leading construction firms, in order to refine an improved resourcing process frame-work. The aim was to develop a knowledge-based resourcing model thatwill enable firms to make appropriateand rapid operational deploymentdecisions, while concurrently takingemployee needs and preferences intoaccount. In this way, project humanresource activities can be seen as inte-gral to defining the value of projectmanagement practice (see Thomas &Mullaly, 2007). Although effective imp-

rovement pathways, metrics, and pro-cedures have been developed for meas-uring and developing the effectivenessof people management practices (e.g.,Brandenburg et al., 2006; Curtis, Hefley,& Miller, 1995), these have not beendesigned to account for the particularchallenges of construction projectresourcing for management teams. Inthis article, we address this specificcomponent of HR strategy.

Research MethodA research approach was required forthis study that would enable the investi-gation and evaluation of the ways inwhich large construction firms capturedemployees’ knowledge, expectations,and aspirations so that they could betaken into account in the resourcingdecision-making process. Much of theprevious research on constructionemployee resourcing has focused onexamining particular aspects of the func-tion, such as recruitment and selection(Bresnen et al., 1985) and team building(Spatz, 2000). Rather than developingsolutions that seek to satisfy the organi-zational/project requirements andobjectives, in this research the primaryneed was to understand the environmentwithin which construction contractorsoperate and how challenges were over-come. Given that the researchers had noa priori knowledge of the likely approach-es adopted by construction companiesfrom which a hypothesis could be derived,an inductive methodology was chosenthat allowed for the examination of theresourcing process from a variety ofstakeholder perspectives.

The seven organizations that agreedto participate in the research were cho-sen as they operated throughout theUnited Kingdom, and their activitiesspanned every major sector of con-struction activity. All of the companieswere among the leading 20 U.K. con-tractors measured by annual turnover.Because they are some of the largestand most well-resourced constructioncompanies, it could be reasonablyexpected that their practices would be

likely to be more developed in compar-ison to other construction firms.

A total of 58 staff members wereinterviewed in a semistructured man-ner, chosen to reflect a cross-section ofviews from each organization. Theycomprised senior HRM staff, directors,and senior operational managers, all ofwhom have some responsibility forresourcing decision making. In addi-tion, a selection of project-basedemployees was interviewed in order toexplore their perceptions regarding theefficacy of the deployment approachesused. The age and experience profile ofthe respondents also varied. Some ofthe senior executives, senior managers,and more experienced project-basedemployees were approaching retire-ment age, having worked in the indus-try for several decades, but views werealso sought from new entrants andthose in their midcareer stage. The HRmanagers included those with overar-ching responsibility for HRM within theorganization as well as those with spe-cific responsibility for graduate recruit-ment and development, training, andperformance management. A summaryof the interviewees and their job roles isshown in Table 1.

A relatively free-form researchinstrument was developed to guide thediscussions to cover all of the major top-ics within employee resourcing andHRM. The questions were worded insuch a way as to encourage informantsto answer from their own frame of ref-erence, consistent with an inductivemethodology. The interviews wererecorded, transcribed verbatim, and thencoded and analyzed using NVivo™qualitative analysis software. NVivo™proved invaluable in facilitating the codification and organization of theresearch material, and also in allowingfor exploratory searches to be carriedout on keywords, nodes, and/or attrib-utes (see Attride-Stirling, 2001). The datawere categorized according to whetherthe practices espoused by the participat-ing firms were likely to facilitate orimpede knowledge-based resourcing.

pmj402_20119.qxd 4/28/09 3:37 PM Page 9



AP

ER

S

ResultsAll of those with responsibilities for theprocess acknowledged the difficultiesinherent in managing resourcing with-in the construction project environmentand suggested that the development ofmore structured solutions had beenhampered by the unpredictability ofhuman resource needs. Many examplesof reactive resourcing practice thatundermined the strategic nature of thefunction emerged from the case stud-ies. Nevertheless, several promisingpractices emerged in terms of ascer-taining employee needs and takingthem into account within decision-making practices.

Examples of both positive and neg-ative practices emerging from the casesare summarized in Table 2 in terms ofwhether they facilitated or constrainedthe strategic resourcing function. Theassessment of whether a particularapproach was positive or negative hasbeen made on the basis of whether theinformants described it as such. In afew cases, there were divergent opin-ions as to whether a particular practice

or approach led to positive or negativeoutcomes in the resourcing process. Inthese cases, the majority view is shownin Table 2, with those generally deemedby the informants as having a positiveoutcome on the process denoted by theshaded cells. The practices identifiedare further elaborated below underheadings drawn from the table.

Human Resource PlanningWithin the case-study organizations,the human resource planning activitiesfound that were seen as positively sup-porting the resourcing process includedsuccession planning (Companies A andB), graduate and student recruitment(Companies A, D, and E), quarterlymeetings (Company C), and integrationof HR requirements with strategic busi-ness plans (Company G).

The succession planning involvedthe identification of people who showeddirector-level potential, who were sub-sequently placed on executive develop-ment programs. In Company B, HRMspecialists and departmental directorscollaborated more formally in identify-

ing suitable candidates for their inten-sive management development pro-gram. Their selection criteria weredrawn from business plans and currentorganizational capability charts.

Many organizations prioritizedgraduate development. Showing long-term commitment to developing grad-uates and offering them transparentprogression opportunities were seen askey factors to successfully retaining thebrightest candidates. This type of long-term approach to graduate recruitmentillustrates an effective integration of thestrategic human resource planning andrecruitment and selection activitiesintegrated with human resource devel-opment. In Company G, an overallstrategic plan was put forward by theboard of directors, which included HRinput with targets for each division onstaff development and retention. Seniordivisional managers reconciled the targets against their resourcing require-ments with a view of achieving appro-priately qualified and skilled staff available where needed. “What-if” sce-narios were run to notionally allocatestaff to projects for which the divisionhad bid. Numerical forecasting antici-pated volumes of staff required formeeting the objectives of the businessplans.

The only negative human resourceplanning activity was found withinCompany F, which operated weeklymeetings focusing on plugging skillsgaps through external recruitment and agency outsourcing, with no HRinvolvement. Clearly the short-termapproach contributed little to employ-ee resourcing within this organization.

Recruitment and SelectionCompanies A, C, D, and E demonstrat-ed positive and proactive recruitmentand selection practices. Following fromthe human resource planning activities,these included graduate and studentrecruitment (A, D, and E), active pro-motion of company culture (A), inter-nal activities (C), exemplary use of jobdescriptions (D), and a centralized

Table 1: Respondent profile.

Interviewee Category Interviewee Job Title No.

Senior executives (n � 7) Operational director 2HR director 3Departmental managing director 2

Senior management (n � 7) Contracts managers 4Chief estimator 1Chief surveyor 1Engineering manager 1

HRM staff (n � 11) HR managers 9Personnel/payroll administrators 2

Project-based employees (n � 33) Commercial manager 1Project managers 6Site managers/agents 8Design coordinators 2Estimators 2Quantity surveyors 7Engineers 4Foremen 3

Total 58

pmj402_20119.qxd 4/28/09 3:37 PM Page 10


Stra

tegi

c St

rate

gic

HRM

H

RM

Evid

ence

of P

ositi

ve a

nd N

egat

ive

Prac

tice

in E

ach

Case

Stu

dy O

rgan

izat

ion

Obje

ctiv

eAc

tivity

AB

CD

EF

G

Staf

fing

Hum

an

Succ

essi

on

Succ

essi

on

Quar

terly

St

uden

t St

uden

t W

eekl

y Di

visi

onal

reso

urce

pl

anni

ng;

plan

ning

plan

ning

pl

acem

ents

; pl

acem

ents

; m

eetin

gsta

rget

s fr

ompl

anni

nggr

adua

te

sche

dule

sgr

adua

te

recr

uitm

ent a

t st

rate

gic

plan

recr

uitm

ent

recr

uitm

ent

low

leve

ls

Recr

uitm

ent

Com

pany

Re

crui

tmen

t In

tern

al

Job

Cent

raliz

ed

Gene

ric

Agen

cy w

orke

rs;

and

sele

ctio

ncu

lture

and

sele

ctio

nre

crui

tmen

t de

scrip

tions

recr

uitm

ent

recr

uitm

ent

wor

d-of

-mou

thw

orks

hops

and

prom

otio

nsse

rvic

eW

eb s

itean

d he

ad-h

untin

g

Team

M

anag

emen

t Re

gion

al a

nd

Reso

urce

Ho

listic

Em

ploy

ee

Wee

kly

Wee

kly

mee

tings

; de

ploy

men

tde

velo

pmen

t na

tiona

l res

ourc

em

anag

emen

t ap

proa

chde

velo

pmen

tm

eetin

gsor

gani

zatio

nan

d em

ploy

ee

bank

sda

taba

se;

need

s pr

iorit

ized

invo

lvem

ent

avai

labi

lity

char

ts

Exit

Exit

“Hire

and

fire

” H

R co

nsul

ted

HR

cons

ulte

d Ex

it pr

oced

ure

No m

onito

ring/

Exit

inte

rvie

ws

ques

tionn

aire

man

agem

ent

[onl

y] o

n [o

nly]

on

anal

ysis

styl

e; e

xit

grie

vanc

e/

grie

vanc

e/

ques

tionn

aire

sdi

scip

line

disc

iplin

e m

atte

rsm

atte

rs

Perf

orm

ance

Perf

orm

ance

Re

activ

e H

R;

Seco

ndm

ent

Info

for

Fast

-tra

ckin

gAn

nual

St

ruct

ured

Pa

per-b

ased

m

anag

emen

tor

gani

zatio

n/

and

late

ral

staf

fing

and

appr

aisa

lsy

stem

annu

al a

ppra

isal

;in

dust

ry “d

rip

mov

esH

R ad

min

; pr

ojec

t rev

iew

feed

” cul

ture

proj

ect r

evie

w

Care

erEm

ploy

ee

Not d

iscu

ssed

Line

Ne

twor

k of

“L

eave

and

In

fo re

cord

ed

Fast

-tra

ckin

gm

anag

emen

tre

spon

sibi

lity

man

agem

ent

care

er p

aths

com

e ba

ck”

on p

aper

via

co

ntac

tcu

lture

appr

aisa

l

HR

adm

in.

Colle

ctio

n,

In-h

ouse

hum

anBa

sic

data

base

Reso

urce

In

tran

etIn

-hou

se

Basi

c M

ix o

f pap

er-

stor

age,

use

re

sour

ce

man

agem

ent

hum

an

com

mer

cial

ly

and

com

pute

r-of

em

ploy

ee

info

rmat

ion

data

base

reso

urce

av

aila

ble

base

d sy

stem

sda

tasy

stem

plan

ning

sy

stem

syst

em

Note

.Sha

ded

cells

indi

cate

pos

itive

pra

ctic

e fr

om a

reso

urci

ng p

ersp

ectiv

e.

Tabl

e 2:

A ty

polo

gy o

f dom

inan

t res

ourc

ing

prac

tices

use

d w

ithin

the

case

-stu

dy o

rgan

izat

ions

.

pmj402_20119.qxd 4/28/09 3:37 PM Page 11



AP

ER

S

recruitment service (E). The compatibil-ity of a company’s value system to thatof the potential candidate’s was cited asa major determinant of resourcing suc-cess. Thus, company A actively promot-ed the company culture within theirrecruitment process and advertise-ments in a bid to ensure that applicants’attitudes and abilities aligned with theirdesired values. By contrast, Company Cfocused on internal recruitment andpromotions. All vacancies were adver-tised on the company intranet and staffwere regularly encouraged to access theavailable information. Similarly, Com-pany E’s HR department operated acentralized recruitment and selectionservice that facilitated graduate andexecutive recruitment (managers andqualified professionals). This providedstaff and managers within the organiza-tion a visible point of contact. CompanyD’s strength was in job descriptions.These included job-specific roles andresponsibilities at entry and exit levels.The entry level outlined the minimumrequirements for the post. In reachingthe exit level, the post holder would beexpected to demonstrate a superiorquality performance at all aspects of thejob, at which stage he or she would gen-erally take on additional responsibilities(and be promoted) or be transferred to adifferent role.

The negative practices identifiedincluded company B’s recruitment andselection training/workshops, whichwere introduced as a reactive measureto combating increasing staff turnoverwithin newly recruited personnel.Company F’s generic recruitment Website failed to consider departmental/line management requirements in therecruitment and selection process,although it provided candidates with asingle route to submitting applications.Despite the extensive human resourceplanning process in Company G, theirrecruitment relied heavily on agencyworkers both short- and long-term.Word-of-mouth and “head-hunting”played a significant role in identifyingpotential new managers.

Team DeploymentPositive team deployment activitiesemerged within Companies A throughE. These firms relied upon regional and national resource banks that delin-eated the workforce according to theirneeds and preferences, and a resourcemanagement database that provideduseful availability charts. In Company A,the responsibility for team deploymentwas devolved to line management. Thishad resulted in some short-term, reac-tive practices, by which existing teamsmoved from project to project largelytogether, and any available people weredeployed to fill immediate staffing gaps.Their management development pro-gram had been used to address this byenabling employees to voice their pref-erences through their performancereviews.

Companies B and E had restruc-tured their operations to form separateregional and national businesses.Within the regional businesses, linemanagers took day-to-day responsibilityfor team deployment and other strate-gic HRM issues. HR personnel managedthe national business deployment via acentral resource bank, which membersof staff could voluntarily agree to join.This helped the company ensure thatonly personnel willing or keen to travelwere deployed to projects beyond theregional boundaries.

Company C utilized a resourcemanagement database to inform theirteam deployment activities. The systemcatalogued all employees’ job titles,their previous experience and projects,line manager, home address, and so on.Availability charts provided a basis fordecision making, which was focused onidentifying the required skills and com-petencies for the upcoming projectsand selecting appropriate personnel tofill the vacant posts. Secondary criteriaincluded appraisal records, careerdevelopment needs, location, salarypackage, client, etc. All of this informa-tion was held within the resource man-agement database and was thus easilyaccessible.

Company D operated a particularlyeffective knowledge-based approach tothe team deployment process. Fourmain sources of information weredrawn together to form a comprehen-sive picture of a potential team mem-bers’ suitability for a project—namely,technical competence reports (e.g.,from job descriptions, experience sum-mary sheets, chartered status records),personal development assessments(which included information on person-al aspirations, needs, and preferences),information on personal relationships(subjective knowledge on how theemployee works with other people/aspart of a team), and data on their avail-ability. A knowledge-based process wassaid to govern the dismantling andredeployment of the teams. Generally, a member of the team was moved to anew project with selected personnelalso being disbanded to other locations.This allowed for effective transfer ofknowledge from the project and itsteam members’ previous experience.Regional functional managers had beenintroduced to form a communicationlink between site and office staff. Theirrole had also been proven effective as analternative for direct-line managementcontact regarding appraisal interviews,career management discussions, andgrievance.

Again, the only potentially negativepractices that emerged were withinCompanies F and G, which relied onweekly meetings to agree on deploy-ment decisions within the organizationbased on the judgments of line man-agement. Company G’s criteria forteam selection prioritized the needs ofthe organization.

ExitCompanies A, E, and G had effective exitmonitoring and management practicesin place. Company A administered andanalyzed in-depth exit questionnairesof all leavers. Within Company E, exitswere managed in HR-line managementcollaboration via a more comprehen-sive procedure that was in place for

pmj402_20119.qxd 4/28/09 3:37 PM Page 12


dealing with voluntary turnover. Thisprocedure was carried out with anyoneindicating an intention to leave, includ-ing managers, professional staff, gradu-ates, and industrial placement students,among others. It was praised for givingthe company a chance to persuade aperson to stay and correct any issuesthat might exist within their workplace.Similarly, Company G operated a tieredexit interview protocol that was de-signed to encourage employee reten-tion and to help create and maintain anethos of “employer of choice.”

Companies B, C, D, and F did notmanage exits as effectively. Their prac-tices included an operations-focused“hire and fire” management style andrandom exit questionnaires (CompanyB), and HR consultation only on griev-ance/disciplinary matters (CompaniesC and D); Company F actually had nosystem in place for dealing with exitsnor any way of monitoring or analyz-ing relevant data. This area clearlypresents the weakest staffing activityoverall.

Performance ManagementOnly two positive performance man-agement activities were identified withinthe participating firms. In Company C,an information-gathering system feddirectly into the resource managementdatabase, thereby integrating the per-formance management and resourcingfunctions. The performance manage-ment system provided information onemployees’ skills, personal objectives,aspirations, and preferences, as well astheir training and career developmentneeds (in relation to both experienceand qualifications). It provided com-prehensive data that enabled potentialsuccession candidates to be identifiedand established how well their employ-ees shared the company’s values. Aswell as an annual appraisal, employeesalso had project appraisals that re-viewed employees’ performance on theproject in question, what they hadlearned, and what they would like to donext. Company F’s performance man-

agement system was highly structuredin order for it to be used as a tool forreviewing performance and rewardand for setting future goals. However,as Table 2 indicates, Company F didnot have equally robust supportingpractices to support the performancemanagement system. Companies A, B, D, E, and G operated reactiveapproaches to performance manage-ment. There, systems were said to belargely ignored by employees and linemanagers alike.

Career ManagementDespite its importance as a subactivityof performance management, onlyCompany D operated a knowledge-based approach to career managementthat included a total rewards packageand a network of career paths that wereguided by job descriptions (see recruit-ment and selection above). The totalrewards package included competitivesalary and benefits, as well as trainingand development options. The networkof career paths provided a map of trans-parent progression opportunities thatwere used to aid discussions on aspiredand realistic future moves.

Negative practices were apparent inCompanies A and E, where career man-agement was seen as an employeeresponsibility with no involvement/direction from HR or line managers. InCompany B, career management wasnot considered at all, this despite itsbeing identified on the exit question-naire as one of the main reasons thatpeople had left the company. AlthoughCompanies C and F recorded individu-als’ career ambitions on the paperappraisal forms, no systematic processexisted to enable managers to take thisinto account. Company G operated isolated pockets of fast-track develop-ment. These provided ambitious indi-viduals the opportunity to progresstheir careers at a fast pace and realizetheir aspirations far more quickly thanwould be otherwise possible. However,there were concerns as to the viabilityof such a practice in the future, as the

organization is unlikely to be able tofulfill employees’ raised expectationsindefinitely. This inability to sustainthis practice thus has a potentiallydetrimental effect on staff turnover. Inaddition, some of those on fast-trackcareer paths were pushed beyond theirlimits, resulting in stress-related illness-es. These individuals required extensiveperiods of time off work and carefulreintegration once they returned. Clearly,the effects were costly: sick pay andtemporary cover, reduced morale andproductivity, potential loss of valuablemembers of staff, and the subsequentcosts of recruitment and training ofreplacement personnel.

HR AdministrationCompanies A and E had developed theirown human resource information sys-tems, and Company C operated a com-prehensive resource management data-base. These held all basic personneldata such as information on employees’addresses, dates of birth, job titles, andabsence records. In Company E, the sys-tem had been developed to provide acentralized information source for thegroup as a whole and had been inte-grated with the main personnel data-base. In Companies A, E, and C, the HRdepartment had responsibility foradministering the system. This ensuredthat records were kept in a consistentmanner and ensured the comparabilityacross the database. Companies B and Fhad basic commercially available data-base systems in place, and Company Dhad no computerized human resourceinformation system in place at all. Thelack of retrievable information renderedtheir resourcing decision-making pro-cesses highly subjective and lacking inemployee input. In Company G, muchof the HR information was recorded andstored in manual paper files. Somecomputerized systems held trainingand development data, payroll details,and employee records in isolation fromeach other. Several members of HR andoperational staff were involved in main-taining and developing these systems.

pmj402_20119.qxd 4/28/09 3:37 PM Page 13



AP

ER

S

The most significant downside of such amix of approaches was the difficulty ingaining access to data—for example,when making resourcing-related deci-sions.

Toward a Framework forCapturing and IntegratingEmployee-Derived Data Withinthe Resourcing ProcessThe analysis of the practices used bythe seven case-study organizations hasrevealed the interrelatedness of theHRM aspects explored. None of theseHRM elements exists in isolation, butall mutually shape the other aspects ofthe resourcing process. The humanresource planning, recruitment, anddeployment functions form a set ofconjoined approaches for ascertainingfuture human resource requirementsbased on internal and external marketassessments, recruiting and selectingthe appropriate people, and utilizingthem in a way that supports the aimsof the business and projects. It istherefore essential that the outcomesof this process are fed back into futureplanning, recruitment, and deploy-ment decisions if the organization is todevelop effectively in the future. Thus,it is clear that improved resourcing isunlikely to be achieved through theapplication of an isolated element ofgood practice, but should be sensitive-ly designed to complement the otherelements of the resourcing process in a way that is compatible with the pre-vailing operational climate of the firm.

By extracting elements of each ofthe practices used by the firms dis-cussed above, a framework was devel-oped (see Table 3) that forms theunderlying process elements of astrategic employee resourcing model.The approaches advocated in Table 3represent a synthesis of the beneficialpractices identified within the case-study firms and how they overcome theinherent challenges posed by project-based working. The emphasis of theapproaches summarized is that they

should ensure that balanced decisionsare made that take into account theindividuals’ aspirations and capabilitieswithin the resourcing process.

Two key aspects underpin the effi-cacy of the framework. First, teamdeployment resides at the center of theresourcing process for the project-based organization as it determines thesuccess of the project, which in turndetermines the competitiveness of theorganization. Accordingly, the functionmust be managed with longer-termplanning in mind, incorporating a broadrange of variables into the decision-making process. It must recognize indi-vidual employee needs and preferencesand provide a route for balancing orga-nizational, project, and employee fac-tors. A second key factor is the role of employee involvement in enabling it to work effectively. This is an appro-ach to increasing organizational effec-tiveness through manager and employeecollaboration and sharing power andcontrol. The framework facilitatesemployee involvement through frequentand effective team briefing structuresand information provision, introduc-tion and maintenance of upward prob-lem solving and task participation, andthe integration of the team deploymentand other employee-resourcing func-tions (such as human resource planning,recruitment and selection, perform-ance, and exit). These seek to initiatemechanisms for employees to voicetheir needs and preferences in relationto their project allocation, develop-ment, careers, and other employee-resourcing and strategic HRM-relatedissues.

The performance and career man-agement sections of the frameworkseek to secure an integrated comprehen-sive strategy for maximizing individual,team, and organizational performancewhile facilitating employee careerdevelopment. This is a further tool formanaging the balance among organiza-tional, project, and individual employ-ee priorities, needs, and preferences. Italso promotes investment in the devel-

opment of people and alignment ofstrategic HRM practices with organiza-tional strategic decision making, as wellas encourages innovation. In terms ofcareer management, balance betweenemployee and manager responsibilitiesis important. The practical elementsinclude continuous review/feedbackmechanisms together with the appraisalprocess, personal development plans,job descriptions, and job profiles.Career management charts are alsoincluded to guide employees in thinkingabout their longer-term development.The HR administration component helpsbalance the organizational, project,and individual requirements for flex-ibility through the structured decision-support and comprehensive informationprovision. A human resource informa-tion system provides the underlyingknowledge-base for the framework andencourages employee involvement via“self-service” functionality.

Overall, the main benefits of theframework are derived from effectiveprovision of consistent informationwithin the organization, coherent man-agement practice throughout geo-graphical regions, operating divisions,and departments, and transparentorganizational culture. The model alsoprovides a communication channel forHR issues and could help to diversifylearning activities and promote cross-project and cross-function transfer ofknowledge. Flexibility is an inherentaspect of the framework. Effectiveknowledge management is the key con-tribution of the model beyond theemployee-resourcing specific func-tions. The model incorporates anunderlying human resource informa-tion mechanism, which correlates witha mechanistic approach to knowledgemanagement. Furthermore, the modelhelps in the transfer of knowledgeamong projects via efficient allocationof human resources and facilitateseffective sharing of company and otherinformation that supports an organic,people-centered approach to knowl-edge management.

pmj402_20119.qxd 4/28/09 3:37 PM Page 14


Mec

hani

sm fo

rPo

tent

ial B

enef

itsCh

alle

nges

of

Inco

rpor

atin

gfo

r the

Res

ourc

ing

Man

agin

g Pr

oces

sPr

omis

ing

Prac

tice

Empl

oyee

s Ne

eds

Proc

ess

for

Core

Res

ourc

ing

With

in a

Con

stru

ctio

nId

entif

ied

From

Into

the

Reso

urci

ngOr

gani

zatio

ns a

ndDi

men

sion

Proj

ect C

onte

xtCa

se-S

tudy

Firm

sPr

oces

sEm

ploy

ees

Hum

an re

sour

ce

Wor

kloa

d in

secu

rity

and

relia

nce

HR-

busi

ness

pla

nnin

gSe

lect

ion

crite

ria d

raw

n fr

omLo

ng-te

rm s

taff

ing

stra

tegy

plan

ning

on re

activ

e de

ploy

men

t lea

dsin

tegr

ated

(C a

nd G

)bu

sine

ss p

lans

and

cur

rent

inte

grat

ed w

ith tr

aini

ng a

ndto

sho

rt-ra

nge

plan

ning

and

Tran

spar

ent i

dent

ifica

tion

orga

niza

tiona

l cap

abili

tyde

velo

pmen

t to

optim

ize

reac

tive

deci

sion

mak

ing

of s

ucce

ssio

n cr

iteria

and

char

tsop

port

uniti

es fo

r em

ploy

ees

iden

tific

atio

n of

suc

cess

ion

Grad

uate

s of

fere

dCl

ear p

rogr

essi

on ro

utes

cand

idat

es (A

and

B)

tran

spar

ent p

rogr

essi

onan

d pr

omot

iona

l op

port

uniti

es u

nder

pinn

edop

port

uniti

es d

irect

ly

with

trai

ning

inte

rven

tions

linke

d to

em

ploy

ees’

ex

perie

nce

and

know

ledg

eba

se

Recr

uitm

ent a

nd s

elec

tion

Reac

tive

recr

uitm

ent d

riven

Or

gani

zatio

nal c

ultu

re

Inte

rvie

ws

prob

ed c

andi

date

s’

Recr

uitm

ent e

ffor

ts ta

rget

ed

by u

rgen

cy o

f res

ourc

ing

expl

icitl

y de

scrib

ed

com

patib

ility

with

eth

os a

nd

at s

uita

ble

popu

latio

nre

quire

men

ts le

ads

to

thro

ugho

ut th

e re

crui

tmen

tap

proa

ch o

f the

org

aniz

atio

nin

appr

opria

te re

crui

tmen

t pr

oces

s to

ens

ure

mut

ual

Cand

idat

es’ v

alue

sys

tem

deci

sion

sco

mpa

tibili

ty b

etw

een

mat

ches

that

of t

he

cand

idat

es a

nd

orga

niza

tion

orga

niza

tion

(A)

Low

er a

ttrit

ion

rate

s an

d gr

eate

r em

ploy

ee s

atis

fact

ion

Exit

Mob

ile w

orkf

orce

Exit

proc

edur

e (E

and

G)

Open

, tw

o-w

ay d

iscu

ssio

nSt

aff r

eten

tion

and

orga

niza

tiona

lle

arni

ng

Team

dep

loym

ent

Need

to ta

ke a

ccou

nt o

f Te

chni

cal c

ompe

tenc

e re

port

s,Em

ploy

ee n

eeds

cap

ture

d Be

tter-i

nfor

med

dec

isio

n pr

oces

ses

orga

niza

tiona

l suc

cess

ion

pers

onal

dev

elop

men

tth

roug

h th

e pe

rfor

man

ce

led

to h

ighe

r-qua

lity

wor

k, b

ette

r pl

anni

ng a

nd e

mpl

oyee

nee

ds

asse

ssm

ents

, inf

orm

atio

nm

anag

emen

t pro

cess

(see

st

aff m

oral

e, a

nd m

ore

effe

ctiv

e w

hile

sim

ulta

neou

sly

mee

ting

on p

erso

nal r

elat

ions

hips

, be

low

) and

fed

dire

ctly

into

kn

owle

dge

tran

sfer

am

ong

proj

ect

imm

edia

te p

roje

ct c

ompe

tenc

y an

d da

ta o

n th

eir a

vaila

bilit

yth

e de

ploy

men

t dec

isio

n st

aff

requ

irem

ents

com

bine

d to

sup

port

pr

oces

sde

cisi

on p

roce

ss (D

)Ca

ndid

ates

app

reci

ate

tran

spar

ent d

ecis

ion

mak

ing

and

that

thei

r nee

ds a

re ta

ken

into

acco

unt

(Con

tinue

s on

nex

t pag

e)

Tabl

e 3:

A fr

amew

ork

for i

mpr

ovin

g th

e re

sour

cing

pro

cess

.

pmj402_20119.qxd 4/28/09 3:37 PM Page 15



AP

ER

S

Mec

hani

sm fo

rPo

tent

ial B

enef

itsCh

alle

nges

of

Inco

rpor

atin

gfo

r the

Res

ourc

ing

Man

agin

g Pr

oces

sPr

omis

ing

Prac

tice

Empl

oyee

s Ne

eds

Proc

ess

for

Core

Res

ourc

ing

With

in a

Con

stru

ctio

nId

entif

ied

From

Into

the

Reso

urci

ngOr

gani

zatio

ns a

ndDi

men

sion

Proj

ect C

onte

xtCa

se-S

tudy

Firm

sPr

oces

sEm

ploy

ees

Perf

orm

ance

man

agem

ent

Oper

atio

nal p

ress

ures

lead

to a

St

ruct

ured

per

form

ance

Em

ploy

ees’

per

sona

l Em

ploy

ee c

omm

itmen

t(i

ncor

pora

ting

care

erre

lianc

e on

per

sona

l ass

essm

ents

man

agem

ent s

yste

ms

obje

ctiv

es, a

spira

tions

, Pr

ovid

es a

veh

icle

for

man

agem

ent)

of li

ne m

anag

ers

rath

er th

an

capt

ure

data

on

perfo

rman

ce,

pref

eren

ces,

and

trai

ning

em

ploy

ee v

oice

in d

eplo

y-st

ruct

ured

app

rais

als

train

ing

need

s, a

nd p

erso

nal

and

care

er d

evel

opm

ent

men

t dec

isio

ns th

at is

aspi

ratio

ns (C

and

F)

need

s ca

ptur

ed a

nd fe

d in

tore

gula

rly u

pdat

ed

reso

urci

ng p

roce

ss

Empl

oyee

s se

e ta

ngib

le im

pact

of p

erfo

rman

ce r

evie

w th

roug

hde

ploy

men

t pro

cess

HR

adm

inis

trat

ion

Low

-leve

l prio

rity

of th

e H

R Hu

man

reso

urce

pla

nnin

g Ce

ntra

lized

sys

tem

inco

rpo-

Com

preh

ensi

ve d

atab

ase

of

func

tion

lead

s to

low

take

-up

syst

em/h

uman

reso

urce

ra

ting

empl

oyee

sel

f-ser

vice

em

ploy

ee n

eeds

and

of

tech

nolo

gy th

at c

ould

in

form

atio

n sy

stem

use

d to

fu

nctio

nalit

y en

able

d re

cord

sre

quire

men

ts to

sup

port

en

hanc

e re

sour

cing

pro

cess

esfa

cilit

ate

and

expe

dite

the

to b

e re

gula

rly u

pdat

ed w

ithre

sour

cing

dec

isio

nsre

sour

cing

pro

cess

(A, C

, E)

little

adm

inis

trativ

e ov

erhe

adEm

ploy

ee re

assu

ranc

e th

at

pers

onal

requ

irem

ents

wer

e ta

ken

into

acc

ount

in H

Rde

cisi

ons

pmj402_20119.qxd 4/28/09 3:37 PM Page 16


Resourcing FrameworkValidationIn order to examine the practical appli-cation of the framework describedabove, it was tested in relation to theresourcing of a program for the con-struction of ten schools. Each of the tenprojects was allocated an individualproject manager, each of whom was inturn supported by a range of managerswho were responsible for the supervi-sory staff (foremen, engineers, trainees). A diverse range of staff had beendeployed to the projects, ranging fromnew employees recruited specifically forthe school projects to long-standingmanagers with over 30 years’ service.

This staffing strategy adopted by thecase-study organization before theframework was implemented (after oneof the projects had been resourced)resulted in multiple challenges that re-emerged throughout the program’sduration. The organization had no writ-ten policy or procedures for teamdeployment or recruitment. At the sitelevel, there was extreme pressure tomeet expected performance levels. Newrecruits were entrusted to undertakework, but the underdeveloped workingrelationships inevitably led to a reduc-tion in team synergy. This led to prob-lems in project delivery and to reactiveresourcing decisions having to be madeto address the problems that this hadcreated.

The resourcing framework wasimplemented in subsequent projectswithin the program to address issuesemerging from its early operation. Thesystem encouraged the production ofgood practice guides on employeeresourcing, training, and developmentand other elements of strategic HRM.This included recruitment and selec-tion procedures and team deploymentprocesses, which ensured the effectiveutilization of staff across the organiza-tion, its subdivisions, and its depart-ments, as well as across individual programs, as in this case. Thus, ineffi-ciencies in resourcing decisions amongthe projects were highlighted, and new

and experienced personnel were allo-cated to the projects in a more balancedmanner.

The HR administration element of theframework supported timely and accu-rate employee data storage and retrieval.This included not only explicit informa-tion about employee background andskills, but also tacit information thatrefers to their experience and expertise.This database was used as a directory of in-house expertise through whichresources could be deployed accordingto organizational and project require-ments. At the same time, employeeswere encouraged to input their needsand preferences in terms of work-lifebalance and career ambitions on thesystem, and thus they were incorporatedinto the managerial decision making. In this way, transparent and balanceddecision making took place, which was used to rapidly satisfy employee/clientinquiries. The user group suggested that,in the long term, the system should helpin planning employee progression andalso encourage repeat business. Thus,the framework could be considered tooffer a practicable method for aligningdeployment practices with organiza-tional needs.

ConclusionsA key aspect of strategic HRM functionis the deployment of human resourcesto operating teams and divisions. Inorder for it to be managed effectively, itrequires the systematic and concurrentcombination of dynamic organization-al, project, and employee data in orderthat appropriate resource allocationdecisions can be made. This article hasexplored how knowledge is currentlycaptured within leading U.K. construc-tion companies. Capturing individualemployee perspectives on their skills,competencies, and personal needs isthe cornerstone of any successful project resourcing strategy. Suchknowledge is difficult to take intoaccount, as it is tacit in nature and spe-cific to every individual employed bythe organization.

Although none of the firms partici-pating in this research had developed asolution that demonstrably resolvedthe resourcing challenge, a synthesis ofthe practices of those identified hasresulted in a more robust methodologyfor taking employee perspectives intoaccount in the resourcing process. Aframework for enabling the capture andeffective utilization of the tacit knowl-edge required for robust resourcing hasbeen developed, which has proved tobe practical through its application as aprototype in a construction program.However, its successful operation with-in other firms will demand a culture oftrust and openness, in which employ-ees will be willing to disclose their aspirations, abilities, and limitations.Although this may present a significantchallenge to construction organiza-tions, if both parties work to developsuch a culture, then the benefits shouldbe appropriate deployment of employ-ees that is sympathetic to both theirneeds and those of the organization. ■

ReferencesArmstrong, M., & Baron, A. (2002).Strategic HRM: The key to improvedbusiness performance. London:Chartered Institute of Personnel andDevelopment.

Attride-Stirling, J. (2001). Thematicnetwork analysis. Qualitative Research,1, 385–405.

Baruch, Y. (2003). Career systems intransition: A normative model fororganizational career practices.Personnel Review, 32, 231–251.

Beardwell, I., & Holden, L. (1997).Human resource management: A con-temporary perspective. London:Financial Times.

Brandenburg, S. G., Haas, C. T., &Byrom, K. (2006). Strategic manage-ment of human resources in construc-tion. ASCE Journal of Management inEngineering, 22(2), 89–96.

Bresnen, M., Wray, K., Bryman, A.,Beardsworth, A., Ford, J., & Keil, E.(1985). The flexibility of recruitment in

pmj402_20119.qxd 4/28/09 3:37 PM Page 17



AP

ER

S

the construction industry:Formalisation or recasualisation?Sociology, 19(1), 108–124.

Chartered Institute of Personnel andDevelopment. (2000). Retention of per-sonnel and other related records.London: Author.

Curtis, B., Hefley, W. E., & Miller, S.(1995). Overview of the people capabil-ity maturity MODELSM. Pittsburgh,PA: Software Engineering Institute,Carnegie Mellon University.

Dainty, A. R. J., Moore, D. R., & Cheng,M.-I. (2005). A comparison of thebehavioral competencies of client andproduction-focused project managersin the construction sector. ProjectManagement Journal, 36(2), 39–40.

De Feis, G. (1987). People: An invalu-able resource. Journal of Managementin Engineering, 3, 155–162.

Druker, J., White, G., Hegewisch, A., &Mayne, L. (1996). Between hard andsoft HRM: Human resource manage-ment in the construction industry.Construction Management andEconomics, 14, 405–416.

Ellis, K. (2001). Dare to share. Training,38(2), 74–80.

Guest, D. E., & Conway, N. (2002).Communicating the psychologicalcontract: An employer perspective.Human Resource Management Journal,12(2), 22–38.

Haldin-Herrgard, T. (2000). Difficultiesin diffusion of tacit knowledge inorganizations. Journal of IntellectualCapital, 1, 357–369.

Loosemore, M., Dainty, A. R. J., &Lingard, H. (2003). Human resourcemanagement in construction projects:

Strategic and operational approaches.London: Spon Press.

Pilbeam, S., & Corbridge, M. (2006).People resourcing: Contemporary HRMin practice (3rd ed.). London: FinancialTimes Prentice Hall.

Raidén, A. B., Dainty, A. R. J., & Neale,R. H. (2006). Balancing employeeneeds, project requirements, andorganisational priorities in teamdeployment. ConstructionManagement and Economics, 24,883–895.

Raidén, A. B., Dainty, A. R. J., & Neale,R. H. (2009). Employee resourcing inthe construction industry: Strategic con-siderations and operational practice.Abingdon, UK: Taylor & Francis.

Rothwell, S. (1995). Human resourceplanning. In J. Storey (Ed.), Humanresource management: A central text(pp. 167–202). London: Routledge.

Smith, E. (2001). The role of tacit andexplicit knowledge in the workplace.Journal of Knowledge Management, 5,311–321.

Spatz, D. (2000). Team building in con-struction. Practice Periodical onStructural Design and Construction,5(3), 93–105.

Storey, J., & Quintas, P. (2001). Knowledgemanagement and HRM. In J. Storey(Ed.), Human resource management:A critical text (2nd ed.; pp. 339–363).London: Thomson Learning.

Taylor, S. (2008). People resourcing(4th ed.). London: Chartered Instituteof Personnel and Development.

Thomas, J., & Mullaly, M. (2007).Understanding the value of projectmanagement: First steps on an

international investigation in search of value. Project Management Journal,38(3), 74–89.

Torrington, D., Hall, L., Haylor, I., &Myers, J. (1991). Employee resourcing.London: Pitman.

Turner, R. J., & Müller, R. (2003). Onthe nature of the project as a tempo-rary organization. InternationalJournal of Project Management,21(1), 1–8.

Andrew R. J. Dainty, PhD, is a professor of con-struction sociology and director of research atLoughborough University’s Department of Civiland Building Engineering. A renowned researcherin the field of human resource management inthe construction industry, he has publishedwidely in both academic and industry journalsand is a coauthor of six textbooks on the management of people within the industry.

Ani B. Raidén, PhD, is a senior lecturer in humanresource management at Nottingham BusinessSchool. Her career has developed from generalistHRM support in industry to employee resourcing–focused research in academia. Her work has beenpublished in both construction management andHRM journals and conferences.

Richard H. Neale, DSc, is a civil engineer who,after working for consulting engineers and con-tractors, was appointed lecturer atLoughborough University in 1974. He was sub-sequently appointed Professor of ConstructionManagement at the University of Glamorgan in1996, and held the posts of head of school anddirector of research. He retired in 2008 and is now professor emeritus. He has pub-lished eight books and more than 120 academic papers.

pmj402_20119.qxd 4/28/09 3:37 PM Page 18

PA

PE

RS


INTRODUCTION ■

The emergence of effective and relatively cheap information andcommunications technology (ICT), particularly web-based tech-niques, has led to the increased use of so-called “virtual teams.”While almost all the challenges associated with traditional teams

present for virtual teams, they also have additional challenges resultingfrom reduced face-to-face communication and lack of community amongparticipants (Danton, 2006). Virtual teams can be assembled on a “needbasis” to collaborate on projects (Andres, 2002; Jarvenpaa, Knoll, &Leidner, 1998), particularly utilizing the skills of the employees who aregeographically dispersed.

In July 2001, a virtual team structure was formed within the New SouthWales Police Force in Australia to undertake a project known as the CountryCapital Works Program (CCWP). This was a 3-year project costing more thanAUD $20.5 million. It was deployed in rural areas in the Australian state ofNew South Wales for the ongoing development and support of the NewSouth Wales Police Force radio systems. A virtual team structure was formedto implement the project because the distances involved in rural New SouthWales are vast, and this made it difficult for a colocated project team frompolice headquarters in metropolitan Sydney to implement the CCWP whilemaintaining effective liaison with, and involvement of, remote members ofthe project.

The author used the virtual team organization in implementing theCCWP project as a case study to demonstrate the virtual team concepts inprojects. Virtual teams can work collaboratively and effectively in a projectdespite the additional challenges, if a skilled project manager leads the pro-ject in overcoming the problems associated with the lack of direct person-to-person contact and immediate managerial oversight. In particular, difficul-ties can arise if it has been necessary for an organization to form a virtualteam from existing personnel who have had other responsibilities in theirday-to-day work in the organization. The author acknowledges that an earli-er version of the case study entitled “Managing Virtual Project Teams: How toMaximize Performance” (Kuruppuarachchi, 2006) was published in theHandbook of Business Strategy.

The remainder of the article is organized as follows. Following this intro-duction, the relevant literature is reviewed. In particular, the inherent prob-lems that exist in virtual organizations or virtual teams are identified. Theresearch question for the present study is then stated. A detailed case studyof the CCWP project is then presented. This includes an analysis of the func-tions undertaken in each functional area of project management. The les-sons learned from the case are then summarized. The article concludes witha brief discussion of the implications.

Virtual Team Concepts in Projects:A Case StudyPalitha R. Kuruppuarachchi, Radio Engineering Services, New South Wales Police Force,Sydney, Australia

ABSTRACT ■

This article uses a case study to demonstratethe application of virtual team concepts in avirtual project team formed from existing per-sonnel within an organization. The article is pre-sented as a literature review followed by a casestudy of a virtual team project entitled the“Country Capital Works Program” undertakenwithin the New South Wales Police Force,Australia. The case confirms the appropriate-ness of existing virtual team concepts in avirtual project team formed from existing per-sonnel for a specific purpose. The study has theinherent limitations of any case study in termsof the generalization of the findings.

KEYWORDS: project management; virtualproject teams; leadership





DOI: 10.1002/pmj.20110

pmj402_20110.qxd 4/28/09 5:10 PM Page 19


Virtual Team Concepts in Projects: A Case StudyP

AP

ER

S

Literature ReviewProjects and Project ManagementA Guide to the Project ManagementBody of Knowledge (PMBOK® Guide)(Project Management Institute [PMI],2004, p. 5) defines the term “project” as a“temporary endeavor undertaken to cre-ate a unique product, service, or result.”The PMBOK® Guide (2004, p. 8) alsodefines “project management” as the“application of knowledge, skills, tools,and techniques to project activities tomeet project requirements.”

The PMBOK® Guide (PMI, 2004) fur-ther categorizes the knowledge requiredfor project management into ninemajor areas: integration, scope, time,cost, quality, human resources, com-munication, risk, and procurementmanagement. This schema is followedin the case study in this article.

The PMBOK® Guide (PMI, 2004) alsoobserves that the use of virtual teamscreates new possibilities for acquiringproject team members. Finally, thePMBOK® Guide recognizes the signifi-cance of communication planning if avirtual team is to enjoy success; however,no details were provided.

Emergence of Virtual TeamsVirtual teams may exist across time,space, and cultural boundaries; theseteams, while sharing a common pur-pose, use technologies to communicateand collaborate effectively (Johnson,Heimann, & O’Neill, 2001; Lipnack &Stamps, 2000). A virtual team can for-mally be defined as a group of geo-graphically and/or organizationallydispersed coworkers assembled using acombination of information and com-munications technologies for accom-plishing an organizational task (Malhotra,Majchrzak, & Rosen, 2007; Townsend,DeMarie, & Hendrickson, 1998).

It is the ability of advances in infor-mation and communications technolo-gies (Jarvenpaa & Ives, 1994; Lipnack &Stamps, 2000), particularly web-basedtechniques (Danton, 2006), that facili-tates the development of electronicallyintegrated virtual organizations, offering

organizations global reach and collabo-ration capabilities. It is not a methodi-cally planned event (Bergiel, Bergiel, &Balsmeier, 2008). Moreover, relatively“flat” organizational structures arebecoming increasingly common in anattempt to reduce costs, improve quali-ty, compete globally, improve customerservice, and accelerate the product-development cycle (Akkirman & Harris,2005). This further supports the use ofvirtual organizations. However, flatorganization structures disperse employ-ees both geographically and organiza-tionally (Townsend et al., 1998). Anyteam of a virtual organization is often avirtual team. Another aspect is that acombination of “business pull” and“technology push” is likely to promotethe use of global virtual teams (Prasad &Akhilesh, 2002). The “business pull” isbeing produced by the expansion ofglobal operations, as firms attempt topenetrate new markets, access scientif-ic talent, and utilize the diverse human-resource capabilities of people fromacross the globe. The “technology push”is being produced by the advent of newelectronic technologies in communica-tions, ranging from simple e-mail tosophisticated groupware.

Drawbacks and Benefits of VirtualTeamsA number of potential drawbacks associ-ated with virtual teams have been report-ed (“Nortel and BP Succeed,” 2003).These include ineffective communica-tion in the absence of nonverbal compo-nents of messages, lack of leisure time forteam members because they tend to beoverloaded with work, resistance to theunstructured nature of the team, loss ofvision, security concerns in the onlineenvironment, lack of permanent records,too many members on some teams, andadded pressure due to overemphasis onspeed. Other reported obstacles that canhinder the performance of virtual teamsare multiple time zones, different lan-guages, and different approaches to con-flict resolution (Bergiel et al., 2008).Regarding conflict resolution, group

members have face-to-face opportuni-ties to immediately and directly discussconflicts and problems with each other(Andres, 2002), but virtual teams do notpossess such opportunities. Further-more, in virtual teams, lack of intimacyand the possibility of having site-specificcultures could lead to some conflicts(Hinds & Bailey, 2003).

Some of the downsides of virtualteams include the lack of knowledgeand/or expertise among the employ-ees about high-level applicationsrelated to virtual teaming, the possibil-ity that the team structure may not fitthe operational environment, and thatsome employees may be unfit for vir-tual teams psychologically (Bergiel etal., 2008). Virtual project team mem-bers rarely meet or sometimes nevermeet with each other or with the proj-ect manager (Das, Yaylacicegi, & Canel,2008). This could also create a specificset of problems: people managementis more difficult, there is an additionalcost for supporting different locations,and all team members may not beoperating on the same assumptions.Furthermore, quality management invirtual teams gives rise to a new set ofchallenges resulting from the practicalproblems of physically conductingthe audit at various locations, and thedifficulty of implementing a commonset of standards in different culturaland linguistic environments (Das et al.,2008).

A summary of the drawbacks ofvirtual teams previously discussed ispresented in Table 1.

However, a carefully designed andimplemented virtual team can offerbenefits.These benefits include improvedproductivity, reduced cost, increasedcompetitive advantage, and improvedcustomer service (Akkirman & Harris,2005); and improved business process,flexible working hours for employees,elimination of time-consuming travel to a central office, support of cross-functional and cross-divisional interac-tions, potential for expanding laborforce, flexibility in work scheduling,

pmj402_20110.qxd 4/28/09 5:10 PM Page 20


speedy dissemination of information,and enhanced knowledge sharing with-in organizations (Johnson et al., 2001).Stevenson and McGrath (2004) present-ed evidence to confirm that major com-panies in the United States—HewlettPackard, General Electric, IBM, and USWest—have benefited through substan-tial productivity increases by usingvirtual teams.

Other benefits (Bergiel et al., 2008)are the possibility to recruit talentedemployees, stimulate creativity andoriginality among team members, cre-ate equal opportunity in the workplace,and discourage age and race discrimi-nation. In particular, for softwareprojects, advantages of virtual teamsinclude the possibility of having a qual-ified labor force, speed in the productdevelopment cycle, having more flexi-bility in resource allocations, and takingadvantage of the availability of a pool ofexpertise regardless of location (Das et al., 2008; Lipnack & Stamps, 2000).The Human Resource ManagementInternational Digest (“Nortel and BP

Succeed,” 2003) presented a similar setof benefits—phenomenal cost savingsfrom not having face-to-face meetings,speedy dissemination of information,enhanced knowledge sharing within theagency, good customer relationships,flexibility on recruitment, talentedworkforce regardless of distance, andflexible scheduling.

A summary of the benefits of virtualteams discussed here is presented inTable 2.

Note that some of the benefits mayhave some negative effects, while someof the drawbacks may have some posi-tive effects. For example, too many vir-tual team members may be a drawbackfor communication, but at the same timeit could be a benefit to speed up productdevelopment. Similarly, diversity of theworkforce facilitates creativity, whilealso having a negative effect on commu-nication. Furthermore, this classifica-tion (advantages and disadvantages) issubjective. For example, “virtual teamscan create equal opportunity in theworkplace” is debatable, and “flexibility

in work schedules” may be dependentupon the situation.

Teams in GeneralDesign and support of teams are vital inleading any team (Hackman & Powell,2004). According to Hackman andPowell (2004), three key considerationsto decrease the adverse effects of teamsare (1) the team has to be a clearlybounded group of people with a sharedcollective responsibility for the out-come, (2) the team leader has to estab-lish basic norms of conduct and makethese explicit, and (3) the reward systems of the organization have to recognize collective performance of the team. Other considerations are that team members need to have therequired diversity of knowledge, skills,and experience; the team has to havethe right mix of personalities or behav-ioral styles; and the team size is limitedto less than 10. In a team, an individualwho is not fit for the team could hinderthe performance of the team.Furthermore, in contradiction to thenormal belief, harmonious relation-ships are often not a facilitator of teamperformance.

Some literature on self-managedteams (SMTs) are also worth mention-ing. Albert and Fetzer (2005) have con-sidered team theories and summarizedessential factors for team effectiveness:• skills, accountability, and commit-

ment;• vision, creation of clear mission,

development of goals, objectives, andaction plans;

• roles and goals, feedback, structure,problem solving, and relationships;and

• team environment, team design,teaming process, and work process.

Teams in highly innovative andtransformational environments oftenfind difficulties (Albert & Fetzer, 2005)in managing feedback, establishing agood structure, solving problems, andmanaging relationships. Furthermore,SMT often leads to disappointment

Drawbacks

• Ineffective communication in the absence of face-to-face communication

• Loss of vision—members may not know the goals and objectives clearly

• Structure may not fit the organization or operational environment

• Resistance to unstructured nature of teams

• Additional cost for setting up remote offices

• Too many members are possible on a team

• Lack of permanent reports or reports are not available centrally

• Lack of visibility of the work of the team members, including their workload and progress

• Conflicts are often invisible and complex—they could even be site-specific

• Quality control is difficult

• Some members may not be psychologically fit for virtual teams

• Supervision and monitoring and performance management are difficult

• Require managing multiple time zones, different cultures, and languages

• Require developing skills of employees on special virtual teaming supporting applications

• Require developing skills of individual members to work in virtual teams

Table 1: Drawbacks of virtual teams.

pmj402_20110.qxd 4/28/09 5:10 PM Page 21



AP

ER

S

because the SMT concept was poorlyimplemented by managers who did notunderstand the process of sharedmeaning (Flory, 2005). That is, man-agers often implemented the SMTwithout having much-needed dialoguewith the employees. These findingscould also be applicable to virtualteams; particularly that poorly imple-mented virtual teams do not providegood results.

Virtual Teams Versus Colocated TeamsMany of the elements relevant to suc-cessful colocated teams, particularlyhigh levels of trust, clear communica-tion, strong leadership, and the appro-priate level of technology, are also asso-ciated with successful virtual teams(Bergiel et al., 2008). Both types ofteams, colocated and virtual, “require aclear, well-founded and convincingpurpose”; however, virtual teams lackthe communication that “appeals to thehearts and minds of potential teammembers” (Morris, 2008, p. 34), asexplained next. Virtual teams are moredynamic than traditional teams, astheir members are dispersed geograph-ically in different locations with differ-ent functional roles.

In virtual teams, there is a lack of unplanned and informal socialexchanges (Putnam, 2001). That is, teammembers “missed the office atmosphereand the opportunities presented by strik-ing up a conversation in the cafeteria orhallway” (Oertig & Buergi, 2006, p. 25).Distant communication also preventscommunication through body language(Stough, Eom, & Buckenmyer, 2000). Inthe absence of rich face-to-face commu-nication, multiple means of communica-tion should be used in virtual teams tofacilitate information acquisition, shar-ing, and integration (Andres, 2002).Moreover, virtual teams rely on electron-ic communications technologies, andmany conversations can be asynchro-nous, such as those that rely on e-mail; incontrast, only a minority of conversa-tions are likely to be synchronous, such as those that use audio/videoconferencing (Prasad & Akhilesh, 2002).Asynchronous communication in virtualteams can preclude informal expressionsof appreciation for work that is well done(Lee-Kelley, Crossman, & Cannings,2004). As DeLuca and Valacich (2006)explained, media with low synchronicitysuch as e-mail may be appropriate forconveyance of information in newly

formed teams, while media with highsynchronicity such as face-to-face meet-ings and telephone may be more desir-able for convergence of shared meaning.

Furthermore, virtual teams are usu-ally more diverse than traditionalteams. Indeed, team membership cancross national boundaries and includepeople from a variety of culturalbackgrounds (Johnson et al., 2001).This heterogeneous workgroup canhamper team integration and commu-nication (Lau & Maurnighan, 1998;Oertig & Buergi, 2006). Thus, it isrequired to give careful attention to thepace of speech, slang, and differentaccents (Oertig & Buergi, 2006).

In addition, a lack of personalengagement in discussions limits thedevelopment of relationships amongteam members (Stough et al., 2000). Inparticular, if the team members areunknown to each other previously,there is unlikely to be trust amongmembers because they do not feel atease with each other (“Nortel and BPSucceed,” 2003). However, trust is a pre-requisite for virtual teams, becauseteam members rely on the trust, judg-ment, and self-motivation of talentedpeople working on a project, whiletheir structures often contradict estab-lishedcommand-and-controlstructures(Cascio, 2000; Shirley & Morton, 1998).Powell, Galvin, and Piccoli (2006) haveexplained this as virtual teams com-pared to colocated teams often show-ing strong relationships between workprocesses and trust, and between trustand effective commitment.

Human relationship breakdowns aswell as lack of trust could lead to virtualteam failures (“Nortel and BP Succeed,”2003; Pauleen, 2003). To avoid such fail-ures, face-to-face meetings and videoconferencing are essential componentsof virtual teams. In the absence of face-to-face interactions, managers shouldalso create alternative strategies fordeveloping mutual trust and reciprocalcommitments in such areas as thesupervision and coordination of projectstages, the clarification of questions,

Benefits

• Financial gains through improved productivity, reduced cost, reduced travel time, etc.

• Increased competitive advantages and improved customer satisfaction

• More flexibility on working hours for employees

• Improved business processes and cross-functional and cross-divisional interactions in the organizations

• Skilled, qualified, and talented workforce is possible regardless of the distance

• Availability of a pool of employees regardless of location, and possibility of easilyexpanding the workforce

• Enhanced information dissemination and knowledge sharing within the organization

• Stimulation of creativity and innovation most likely due to diversity of the workforce

• Creation of opportunities for employees in remote offices

• Flexibility in resource allocations and work scheduling

• Speed up product development and project management

Table 2: Benefits of virtual teams.

pmj402_20110.qxd 4/28/09 5:10 PM Page 22


and the conduct of performanceappraisals.

Lee-Kelley et al. (2004) used theoriesof social interaction to explain some ofthe relational issues created by the spa-tial and temporal separation character-istic of virtual teams. They groupedthese issues into two categories:• “Hard” (visible) issues: Operational

structures, terms and conditions,processes, systems, and communica-tion (methods, arrangements, andtechnology)

• “Soft” (invisible) issues: Trust andcommitment, motivation, communi-cation, role definition, and recognition

Piccoli, Powell, and Ives (2004)examined the effectiveness of virtualteams and advised to take considerablecare in transforming colocated teams tovirtual teams, because behavioral-relat-ed practices used in traditional teams—for example, motivation—cannot beused effectively in virtual teams, andcould even be counterproductive. Invirtual team dynamics, leadership isone of the most fundamental skills(Pauleen, 2003). According to Stevensonand McGrath (2004), managers are like-ly to incorporate intuitively the vari-ables effective team leadership, com-mitment, and regular personal contactas very important for virtual teams, whilethey are likely to overlook the variableseffective reporting procedures, solidwork structures, team hierarchy, strictassessment processes, and greater likeli-hood of communication breakdown.

Barczak, McDonough, andAthanassiou (2006) presented factorsessential for the formation of successfulvirtual teams as communication, trust,leadership, clear goals, and technology.Peters and Manz (2007) suggested thatdeveloped relationships, shared under-standing, and trust are the importantantecedents of virtual teams to collabo-rate effectively.

According to Peters and Manz(2007), the problems in virtual teamscompared with traditional teams aremainly due to a lack of opportunities for

team members to build relationshipsand trust, as well as to address issuescaused by heterogeneous membershipin terms of location and culture.Physical isolation and a lack of plannedand unplanned face-to-face interac-tions among team members can thushave adverse effects. Members mighthave different habits and methods ofworking, few opportunities for informalinformation exchange, and so on. Inaddition, team members might have to deal with mistrust, unequal (orunknown) expectations, and differentteam dynamics. Furthermore, the suit-ability of conventional managementstyles and techniques in dealing withthe issues encountered in virtual teamsare questionable (Lee-Kelley et al.,2004). Considering the lack of face-to-face contact within teams, differentskills are needed to interact in the glob-al community and succeed in a virtualworld. Part of the leadership functionssuch as monitoring team performance,implementation of solutions for prob-lems, development of team members,and so forth are to be accomplished byleadership substitutes and/or distribut-ing them to the team members them-selves (Hunsaker & Hunsaker, 2008).

Based on literature (including thepresented drawbacks and benefits), asummary comparison of virtual teamsand colocated teams is presented inTable 3. The comparison is presentedunder the headings “communication”and “team structure and leadership.”

Overcoming Inherent ProblemsThe initial effort required in designing,planning, and implementing virtualteams is enormous (Prasad & Akhilesh,2002); the dimensions to be consideredare team structure, strategic objectives,work characteristics, and situationalconstraints. It is also important to havea well-organized plan for success, par-ticularly considering managerial sup-port and establishing trust betweenmanagers and workers (Akkirman &Harris, 2005) are both dependent oneffective communication. A variety of

activities have been suggested to ame-liorate difficult issues in virtual teamsand to promote teamwork as follows.

Glacel (1998) stressed that a prereq-uisite for virtual teams is building a firmfoundation; thus, face-to-face relation-ship building is essential. Oertig andBuergi (2006) presented the same con-clusion: the first meeting should be aface-to-face meeting that facilitatesteam leaders to develop trust andrespect at the onset of a project for suc-cessful interaction of team members.People will report problems early, priorto them becoming critical, when thereis trust. It is also important to set clearroles for each team member by assign-ing formal individual responsibilities atthe outset to avoid ambiguity and mis-understanding (Lee-Kelley et al., 2004).

The much-needed team cohesionfor virtual teams can be encouraged ifface-to-face meetings are held at threepoints in the life of a virtual team (Lee-Kelley et al., 2004):• At commencement: When face-to-face

contact can create social relation-ships, build mutual trust, establishreciprocal commitment, and establisha shared set of business goals andobjectives

• At an intermediate stage: When work-shops can influence relationships,resolve misunderstandings, and clari-fy task issues

• At winding up: When gatherings canfinalize unresolved items, generatecommitment to output, and celebratesuccess

According to Lee-Kelley et al.(2004), the key requirements are thatregular face-to-face meetings should bescheduled (because there is evidencethat virtual teams with such face-to-face contacts perform better than thosewith less contact, or those that areentirely virtual), that a structured for-mal approach be used in recognizingteam members’ efforts (to motivatethem in the absence of immediate feed-back from peers and supervisors), andthat diversion of team members to

pmj402_20110.qxd 4/28/09 5:10 PM Page 23



AP

ER

S

other activities in the workplace berestricted (to reduce possible demoti-vating effects).

In the lack of face-to-face contactsin virtual teams, effectiveness of ongo-ing communication is very important.It is necessary to have a focus on sim-plicity and accuracy in communica-tion to keep members informed of thework in progress (Townsend, DeMarie,& Hendrickson, 1996). Effective dis-tance communication may include(Gould, 1997): having face-to-facetime, if at all possible; giving teammembers a sense of how the overallproject is going; establishing a code ofconduct to avoid delays, maintainingteam members’ calendars; augment-ing text-only communication; anddeveloping trust. Efficient monitoringmechanisms are also important(Melymuka, 1998) to detect early signs

of problems with the team or theschedules. The absence of contact orreduced contact between the managerand the team might create a situationwhereby the manager does not knowwhether people are actually engagedin their allotted tasks.

It is also important for virtual teammembers to possess a number of quali-ties (Johnson et al., 2001): good self-starters having self-discipline, a sense ofindividual accountability, and flexibilitywithin the virtual team. Furthermore,appropriate training for team membersmay assist virtual teams to improvegroup cohesiveness, perception of theteam process, and satisfaction with the team’s outcome (Beranek & Martz,2005). Brake (2006) argued that two keychallenges in leading virtual globalteams are isolation and confusion; iso-lation can be beaten through building

community, while confusion can bebeaten by promoting clarity. Ten practi-cal guidelines were presented: (1) beproactive (measures to be taken to meetthe most likely challenges); (2) applycultural intelligence (cultural differ-ences could be either assets or potentialliabilities); (3) build swift trust (connectfirst and then collaborate); (4) be aproblem solver (problem solving shouldbe pragmatic and not reactive); (5) stayperson-centric (connect with feelingsand not just with facts); (6) stay focused(everyone should understand the team’sgoals, objectives, strategies, and priori-ties the same way); (7) clarity of whoand what (team members should knowwhat is expected, by whom, and bywhen); (8) establish predictability (cre-ate common working grounds for keyactivities); (9) communicate context(provide all the contextual information

Communication Team Structures and Leadership

Colocated • A rich form of communication face-to-face • Well-developed traditional team leading is always available concepts can be used

• Regular personal interactions support building • Monitoring of work of the team members is not relationships and trust difficult

• Supervisor can motivate the members knowing their attitudes and requirements

• Easy to implement effective reporting procedures• Easy to implement a common set of standards

for various processes, including quality control

Virtual • A heavy reliance on electronic forms of communication, • Require skills to manage the diversity of the team often asynchronous media with several drawbacks membership, crossing national boundaries and

• Lacks opportunities for building relationships and trust different time zones• Requires careful planning for team integration and • Require skills to supervise without having direct

communication observation of team members’ work and work • Requires clear and precise communication, as progress

immediate feedback or clarification is difficult • Self-motivation and self-judgment are often • Problems can go unnoticed required from team members

• Different reward systems may be required, as it is difficult to admire well-done work and motivate members

• Establishment of standards for team processes, including quality control, is difficult

• Different locations may have different habits (may be site-specific) and different methods of working

• Different assessment systems may be requiredfor work, work structures, and work processes

Table 3: Comparison of virtual and colocated teams.

pmj402_20110.qxd 4/28/09 5:10 PM Page 24


to develop a shared mental model); and(10) drive for precision (be specific withlanguage and continuously probe forshared understandings).

Virtual team challenges may becomplicated by organizational policies;organizational human resource poli-cies, particularly concerning perform-ance incentives, may negatively impacton a team leader’s ability to manage ateam (Pauleen, 2003). According toJohnson et al. (2001), the following fac-tors lead to successful virtual teams:• human resource policies that recog-

nize, support, and reward virtual teammembers and leaders;

• an adaptable “flat” organizationalstructure, rather than a hierarchical,control-oriented organization;

• an organizational culture that valuescommunication, learning, teamwork,and the need for diversity;

• people who possess good verbal, listening, and writing skills; and

• a technologically advanced organiza-tion and people trained in usingtechnologies.

Hunsaker and Hunsaker (2008)provided detailed guides for project leaders who manage virtual projects, categorized into pre-project, project-initiation, midstream, and wrap-up. It seems that the suggested guidelines are not unique to virtualteams, but are also applicable to colo-cated teams:• Pre-project: Establish and communi-

cate project mission, priority andsuccess criteria, select team members,define roles, and determine technolo-gy requirements.

• Project initiation: Establish and man-age team boundaries, develop ashared mental model, create andmaintain awareness, and managecommunication processes.

• Midstream: Steps taken in initiationcontinue through managing teamboundaries and establishing workingconventions and norms.

• Wrap-up: Go over lessons learned andannotated successes.

Beal (2006) referred to the conclusionof the previously published article on thecase study presented in this article andhighlighted challenges for humanresource teams for managing virtual proj-ects. In that article, Kuruppuarachchi(2006) found the following:

Virtual project teams are cost effec-tive, but management of virtualteams requires skills over and abovethe management of co-located proj-ect teams . . . the organizationshould have high level project man-agement expertise and systemsprior to establishing virtual projectteams. In addition, the most appro-priate methods of managing virtualteams may not be fully compliedwithin the existing policies and pro-cedures of the organization. Teamsshould be empowered to own andcommit to the purpose, and shape itif it is required. It may demandorganizations to relinquish old stylecommand and control base man-agement and to provide freedom toteams. The team should work as areal virtual team rather than a dis-persed fragmented team. (p. 78)

Referring to an international cross-cultural virtual team in a matrix organi-zation, Oertig and Buergi (2006) outlinedmajor challenges reported by projectleaders. Factors identified as importantfor virtual teams are selecting creativeleaders with a collaborative leadershipstyle and excellent communication skills,top management support for continuingface-to-face communication and rela-tionship building, and ongoing invest-ment in language and intercultural com-munication training. Management ofproject tasks can be facilitated bydefining team operating guidelines, set-ting up simple and workable processes,explaining the rationale when collectinginformation, having agreement for writ-ten communication, doing follow-up,and keeping everyone well informed ofthe status (Oertig & Buergi, 2006).

Based on the literature, activities forovercoming the possible problems ofvirtual teams are suggested in Table 4.

Case StudyResearch QuestionThe research question addressed in thestudy can be expressed as follows:• To maximize performance, what spe-

cific undertakings are required in vir-tual project teams, compared withtraditional project teams?

This research question wasaddressed in this article by means of acase study of a 3-year police communi-cations project in rural New SouthWales (NSW), Australia, which wasknown as the Country Capital WorksProgram (CCWP). The CCWP com-menced in July 2001 and cost morethan AUD$20.5 million.

Background to Case StudyIn rural NSW, five communications centers—Tamworth, Wagga Wagga,Newcastle, Wollongong, and Penrith—provide radio dispatch services to policein various cities, townships, and ruralareas. Country radio communicationssystems in the rural NSW are linked tothese five centers. A center managercontrols each center with support frommaintenance centers, which are knownas “Radio Network Service” (RNS) units.Five of these RNS units are colocatedwith center managers; a further threeRNS units are located remotely.

NSW has an area of 801,600 squarekilometers. The distance from Sydney tothe remote city of Broken Hill is approxi-mately 1,200 kilometers. A virtual projectteam was formed to undertake theCCWP because the distance from Sydneyto the country centers and the distancesfrom country centers to remote radiosites made it difficult for a colocatedproject team in Sydney to carry out theproject. The project management struc-ture of the CCWP is shown in Figure 1.

Human Resources of Project TeamA project manager who was responsiblefor the day-to-day management of theproject was assigned to the CCWP on afull-time basis. Under this person’s man-agement, the CCWP was undertaken as

pmj402_20110.qxd 4/28/09 5:10 PM Page 25



AP

ER

S

a series of subprojects. The center man-agers and RNS units were responsiblefor the implementation of subprojects intheir respective areas. In each RNS unit,five to seven members actively partici-pated in the subprojects; all were full-time employees of the New South WalesPolice (NSWP) who worked on theCCWP on a part-time basis. In addition,six members from Sydney (also on a part-time basis) were included in the projectto provide technical and financial(accounting) support. In all, approxi-mately 45 members took part in theCCWP. It was not possible for the projectmanager to choose who should be on theteam.

Because all team members werefull-time employees of the NSWP, they

already knew each other when the proj-ect team was established. However, thisacquaintance across the organizationwas mostly limited to the familiarity ofnames of members and where theywere located rather than well-devel-oped relationships. Members at eachindividual RNS unit worked together ona daily basis, having autonomy withintheir areas of responsibility—that is,RNS units did not work together. Assuch, even though within each RNSunit, the team members were colocatedand had long-standing working rela-tionships with each other, when cross-ing the RNS unit boundaries, relation-ships, and attitudes toward each otherwere more problematical. The develop-ment of trusting relationships between

the project manager and individualRNS units was recognized as a key suc-cess factor for the project.

The project sponsor, who acted asthe “corporate champion,” chaired thesteering committee that was responsi-ble for providing overall project direc-tion and approval of key milestones. Aspreviously noted, the project managerwas responsible for the day-to-daymanagement of the project and thecountry teams were responsible for the delivery of assigned subprojects. The technical support teams reviewed specifications and requests for tenders/quotations, and were responsi-ble for recommending selection ofvendors. The financial support teamadministered other financial matters,

Area Suggested Undertakings

Communication • Have face-to-face meetings at commencement, intermediate stagesas required, and at close

• Use multiple channels of rich media for precise and effective ongoing communication to avoid miscommunication

• Provide feedback to members of the work progress, status of the project, and theoverall team environment

• Seek top management support to facilitate (to have additional) face-to-face meetings

• Use face-to-face meetings for problem solving whenever possible

Team Structures and Leading Teams • Establish team operational guidelines or code of conduct, particularly aiming to establish appropriate habitual routines, standard operating procedures, and regulation of team performance

• Set clear roles and responsibilities for individuals (i.e., clear engaging direction along with specific goals) as opportunities to resolve ambiguities and misunderstandings are greatly reduced in virtual teams

• At the onset of the project, develop individuals into a sound and well-integratedwork unit through the development of trust and respect, and cooperative relationships

• Use efficient structured formal approach to collect information in the absence of immediate feedback

• Establish efficient mechanisms for dissemination of information and collecting feedback, as well as effective monitoring mechanisms to detect problems early

• Maintain a snapshot of the members’ availability and work progress• Carefully select members (if possible) who have qualities for self-management

and motivation• Select skilled leaders with a collaborative leadership style and excellent

communication• Establish a supportive work climate, examine the existing policies and procedures

of the organization, and augment them if they are not supporting virtual teams• Provide the necessary training to team members on the use of technology and

project management

Table 4: Overcoming problems in virtual teams.

pmj402_20110.qxd 4/28/09 5:10 PM Page 26


such as processing orders and pay-ments.

Project Integration and ScopeManagementThe CCWP proceeded according to anapproved business case, whichdescribed a set of identified subprojectsfor each area. In accordance with thebusiness case, the “Project ExecutionPlan” (PEP) was developed to guide theplanning, execution, and control of the project.

Each subproject was planned to runfor a year, and the various subprojects ranconcurrently. In general, the implemen-tation of the subprojects was conductedindependently in various RNS areas;however, certain subprojects were com-bined across RNS units to gain value formoney and/or to execute certain subpro-jects efficiently.

At the beginning of the CCWP, aworkshop for center managers was con-ducted to ensure a shared understandingof the overall project, and to inform cen-ter managers of individual and groupresponsibilities. At the commencementof each year of the CCWP, remote teams(headed by center managers) were

empowered to refine the list of subpro-jects in consultation with users, withwhom center managers had relativelycloser contact than did central manage-ment. Consideration was then given tobudgetary allocations and time con-straints before the steering committee for-mally accepted the various subprojects. Informally accepting the subprojects, eachwas reviewed against corporate require-ments in consultation with countryteams.

Planning responsibility for subpro-jects, including the development of thework breakdown structure (WBS), wasassigned to remote teams. An analysisof the WBS at the subproject levelassisted team members in defining acluster of activities for each subproject.To ensure that they were formulated ina consistent manner across all RNSunits, guidelines and appropriate tem-plates were provided to team members.Thereafter, the project was presented asa whole; this included (1) project inte-gration activities, (2) refining subpro-jects, (3) combining appropriate sub-projects, (4) defining milestones, and(5) predicting cash flow. Milestones and cash flow established the basis for

project decisions and cost baselines(budgets) for measuring project per-formance. The progress of the projectwas continually monitored against tar-gets using periodical e-mail reports anddiscussions with team members.

During the course of the CCWP,changes occurred in several areas—including technology, the government’sradio strategy, and policing require-ments. These changes caused a few keycorrections to the scope of the overallproject. However, in many instances,virtual team members failed to followthe agreed procedures to initiatechanges in the scope of the allottedsubprojects.

A benefits-management processwas a formal part of the project. A“Benefit Management Plan” (BMP) pro-vided a structured process whereby the project team could demonstrate thebusiness benefits. As part of developingthe BMP, a series of workshops wasconducted, and business processeswere documented. This process of ben-efit management was undertaken atcountry centers, where the benefits ofthe project were measured, recorded,and verified according to the BMP.

WollongongCenter Manager

TamworthCenter Manager

Wagga WaggaCenter Manager

Project Manager Sydney

Penrith Center Manager

Penrith RNSWaggaWagga RNS

Tamworth RNS

NewcastleCenter Manager

Waratah RNS

Project Support TeamsSydney

Project SponsorSteering Committee

(Sydney)

Grafton RNSDubbo RNS

Warilla RNS

Cooma RNS

Figure 1: CCWP virtual team structure.

pmj402_20110.qxd 4/28/09 5:10 PM Page 27



AP

ER

S

Before commencing BMP measure-ments, team members needed skills ingathering data (or deriving necessarydata from existing information), vali-dating data, and documenting data asbaseline measures. They also needed toliaise with local business managers andother relevant stakeholders to deter-mine the realized benefits. Local staffmembers were very useful in thisprocess—an advantage of the use ofvirtual teams.

Time ManagementAll the constituent subprojects of theCCWP were carefully analyzed whenplanning project schedules and mile-stones. Schedules were specificallydesigned to facilitate concurrent sub-projects and empower country teams.The overall project management planwas presented in the PEP.

The scheduling of projects and set-ting of milestones were both found to bedifficult in the virtual team environ-ment. This was complicated by the largenumber of subprojects (more than 230)and certain engineering constraints,which required some subprojects to becompleted within certain critical timeperiods. In responding to the latterproblem, it was decided that time-criti-cal activities could be initiated: (1)before the completion of formal reviewsof preceding activities, and/or (2) with-out waiting for design approvals, and/or(3) without waiting for full details.

Although the overall CCWP was a3-year project, certain subprojects wereto be initiated and completed duringeach year. In the initial months of eachyear of the project, project milestoneswere used to monitor the commence-ments of critical activities, and towardthe end of each year the milestoneswere used to monitor the completion ofwork. Some activities also had interme-diate monitoring points at which theirprogress was reviewed. In addition, allsubprojects that had encounteredproblems were monitored continu-ously to detect and resolve difficultiespromptly.

The project schedules and mile-stones presented at the project levelguided the RNS units in their schedul-ing of subproject activities, and alsosynchronized project activities acrossRNS units. However, country teamstook ownership of the work andplanned and implemented their sub-projects with minimal control. Thisinvolvement of country teams acceler-ated the granting of required planningapprovals with regard to certain sitedevelopment activities because mem-bers of the country teams had goodrelationships with local governmentagencies in their areas.

Cost ManagementAs explained previously, the implemen-tation of the subprojects was conductedindependently in various RNS areas;however, certain subprojects were com-bined across RNS units to gain value formoney and/or to execute certain sub-projects efficiently. Furthermore, budg-etary allocations and time constraintswere considered before the steeringcommittee formally accepted the vari-ous subprojects. Milestones and cashflow established the basis for projectdecisions and cost baselines for meas-uring project performance.

The budget of the CCWP was con-trolled centrally, and this was found tobe an extremely efficient and usefulcontrol mechanism. The budget wasmonitored against time using simplemeans such as Excel worksheets andgraphs showing cumulative values.Rather than using special project man-agement tools, it was decided to useExcel worksheets, because all teammembers were familiar with informa-tion being provided in this format.

Quality ManagementQuality-control initiatives for the CCWPwere included in the PEP by identifyingapplicable quality standards and themeans of satisfying them. It was con-sciously decided not to employ central-ized resources for quality review and/orinspections at subproject levels,because it was difficult for quality

reviewers to visit diverse locationsthroughout rural NSW.

The quality-assurance activitiesthat were found to be efficient in theproject were as follows:• provision of necessary skills to coun-

try teams through appropriate train-ing programs;

• review of specifications, request fortender/quotations, and evaluationmethodologies developed by countryteams before their release;

• inclusion of product quality and rele-vant national and international stan-dards and accepted safety practices inthe PEP as references; and

• giving high priority to quality-man-agement experience of supplierswhen selecting contractors.

As it was difficult for quality review-ers to visit diverse locations throughoutrural NSW, the remote team memberswere trusted to deliver quality productswithout inspecting the intermediaryand final products by a nominated cen-tral team. As such, it was found that thequality assurance and quality control atthe product level was more difficult invirtual project teams.

Human Resource ManagementThe virtual team structure is presentedin Figure 1. As a mature organization,the NSWP has a well-developed humanresource management framework (sys-tems, structures, and guidelines),including policies and procedures.Virtual team concepts were used withinthe project structure, particularly forproject communication as described inthe “Communication Management”section, however without contradictingthe organizational human resourcemanagement framework. The centermanagers, as line managers, managedthe team members.

Communication ManagementThe NSWP already had systems, struc-tures, and guidelines in place to handleroutine information and the communi-cation needs of the CCWP, and all mem-bers had access to individual e-mail

pmj402_20110.qxd 4/28/09 5:10 PM Page 28


and telephone facilities, and sharedinformation repositories. Although theteam members of the CCWP knew eachother (by name and location) at thebeginning of the project in their capac-ities as full-time employees of theNSWP, this did not necessarily meanthat they had good working relation-ships. In general, team members inter-acted electronically; however, they hadoccasional face-to-face interactionswhen attending meetings and trainingprograms.

At the beginning of the project, for-mal procedures were developed forgathering information from countryteams. This was conducted on a needsbasis with the aim of ensuring thatrequired information was available forproject monitoring and control, as wellas for the provision of information tostakeholders. These formal procedureswere revised several times as the proj-ect proceeded. The progress of theproject was reported monthly; howev-er, it was necessary to collect informa-tion more frequently to ensure that asnapshot of the project was alwaysavailable. Formal reporting of the proj-ect took place in two broad forms: (1) project-specific reporting within theproject team (for information sharingand to assist in project execution) and(2) project status reporting (for provi-sion of selected information to variousstakeholders). For the latter, the exist-ing systems and structures of the NSWPwere used to provide formal reportsthrough the project manager. Withinthe project team, informal communica-tion was encouraged on the basis thatproject implementation in a virtualteam environment is dependent ontrust, cooperation, and teamwork.

There were formal reviews of theproject by the steering committee at thebeginning, midpoint, and end of eachyear of the project. These reviews werecarried out in face-to-face meetings withcountry members. In some cases, thecountry teams visited Sydney; in othercases, personnel from Sydney visitedcountry centers. In addition, video-con-

ferencing facilities were used asrequired—for example, in resolving criti-cal issues and/or formalizing procedures.

The monitoring and reportingmechanisms of the project supportedtrusting relationships. These mecha-nisms included the following:• Project activities, major milestones,

and other relevant information werereviewed and agreed to at the begin-ning of the project.

• Mechanisms, formats, and frequencyof reporting for monitoring of time-lines and budgets were agreed.

• Team members were consulted asrequired for progress reviews, andteam members were provided withfeedback when summary reports werepresented to stakeholders.

Three major problems were experi-enced that might be considered com-mon to all virtual teams:• At the beginning of the project, a con-

siderable effort was required for fullengagement of team members inaccepting the project managementframework and its monitoring andcontrolling mechanisms.

• To initiate appropriate correctiveaction, an efficient project-monitor-ing mechanism was required to iden-tify issues, problems, risks, scopecreep, and so on.

• As some team members might notappreciate the importance of docu-mentation, more attention wasrequired in maintaining comprehen-sive project documentation at thecentral office.

In addition, certain other problemswere experienced that might be attrib-uted to limitations of virtual teams, skilllimitations, or the culture of the partic-ular organization involved:• The project manager needed support

from the sponsor to implement someprocedures and to resolve a numberof problems.

• Some team members were deficientwith respect to experience in projectmanagement.

• Resource limitations were experi-enced in critical times because teammembers were also responsible forother activities.

• Decisions made at remote centers(and the results generated) were notstructurally captured and document-ed centrally in certain situations.

• Resolution of certain problems didnot follow the documented path (forexample, project manager to steeringcommittee to project sponsor).

Risk ManagementAt the time of defining subprojects,guidelines were provided to countryteam members to assist them in riskidentification, risk quantification, andresponse development for each sub-project. Thereafter, based on the infor-mation provided by country teams, anoverall risk management plan wasdeveloped for the project.

Preplanned risk responses workedeffectively in the project; for example:• The project was planned as a series

of locally managed subprojects,with only weak links existing amongthe majority of subprojects; thisallowed unsatisfactory subprojectsto be terminated in isolation if nec-essary.

• Certain activities (such as site-development works, installation oftowers, construction of huts, and soon) were not combined across sub-projects; this minimized the forma-tion of links among subprojects, eventhough it was possible to gain somevalue for money.

Similarly to quality assurance andquality control at the product level,responding to changes in risk over thecourse of the project and updating the risk management plans was foundto be difficult. There was often a timegap between an event happening andthe project manager noticing if therewas an associated risk. This happenedbecause the remote team memberswere either unable to identify riskyevents promptly or did not consider

pmj402_20110.qxd 4/28/09 5:10 PM Page 29



AP

ER

S

that it should be reported to the projectmanager.

Procurement ManagementProcurement activities of the CCWPwere planned carefully to satisfy two keyobjectives: (1) to gain value for moneyand (2) to adhere to NSW governmentpurchasing guidelines and policies.

Procurement guidelines wereincluded in the PEP to provide guidanceto country teams in all procurementundertakings—from decisions on whatto procure through to the completionand settlement of contracts. In addition,the project manager was readily avail-able to provide advice to team memberswhenever it was required.

An analysis of work structures at thesubproject level assisted team membersin defining a cluster of activities for eachsubproject. Thereafter, they identifiedthe project-delivery methods and docu-mented each procurement require-ment—such as obtaining off-the-shelfequipment, leasing arrangements withservice providers, use of contractors,and use of in-house resources. Theseactivities were included in the initialscope definition of each subproject.

The proposed procurement activi-ties of each subproject, as provided bycountry teams, were reviewed at thecentral office. Thereafter, an integratedprocurement plan was prepared tocover procurement activities acrossRNS units, including the monitoring ofprogress milestones. Certain procure-ment activities were combined to gainvalue for money; others were notcombined as part of a risk-reductionstrategy of minimizing linkages amongsubprojects. Thereafter, country teamswere allowed to initiate purchase requi-sitions, and these requisitions werereviewed at the central office as part ofthe approval process. High-level cen-tralized control on procurement assist-ed budgetary control and cash-flowmanagement of the project.

In addition, agreements with sup-pliers helped to procure equipment as needed for subprojects without

compromising quantity discounts. Thisenabled delivery of equipment in sever-al consignments, which assisted in sat-isfying procurement requirements on aneeds basis for subprojects.

Review of Project ClosureFollowing completion of the secondyear of the project, a structured studywas undertaken as part of the develop-ment of project closure report. Thisstudy included an appraisal of the viewsof participants in the project. Centermanagers, RNS staff, members of thesteering committee, and supportingstaff all participated in the developmentof the project closure report. A particu-lar focus was to propose improvementsto project management practices thathad been deployed in the project.

The process was guided. In provid-ing the information, participants wereto consider the learning requirements:what went right and why, what wentwrong and why, what could be donebetter with hindsight, and any ideasthat might help on another project. A table was provided, and participantswere asked to fill it in by providinginformation under nine major knowl-edge areas of project management andfour learning requirements. The lessonslearned from the case are presented inthe following section.

In addition, the internal AuditGroup of the NSWP conducted an audit review on the second year of theproject. The audit report made specificfindings related to some aspects ofproject planning, implementation, exe-cution, and finance.

Lessons Learned From the CaseCritical factors in maximizing the per-formance of virtual project teams werederived by combining informationfrom three sources: the project closurereport, the audit report from NSWP, andthe experience of the project manager.These factors have been published in aprevious paper (Kuruppuarachchi,2006). The factors can be categorizedinto three major areas:

• establishment of supportive systemsfor virtual teams;

• carefully planned launch of virtualprojects; and

• efficient ongoing monitoring andcontrolling.

The first of these, supportive systemsfor virtual teams, can be provided bytaking initiatives to improve the day-to-day activities of the organization.Possible initiatives include the following:• adherence to well-proven project-

management practices;• enhancement of project management

skills of staff members (without limit-ing their skills for day-to-day opera-tional activities);

• engagement of engineering teams (if available) to develop standardsand/or specifications for technicalproducts for forthcoming projectswithout waiting for project initiation;

• capture and storage of current infor-mation on user needs and availabletechnologies on a continuous basis;

• developing relationships amongmembers of the organization and withrelevant external agencies (govern-ment, power utilities, and carriers);

• minimization of delays in certain proj-ect activities—for example, utilizationof memoranda of understanding,agreements with land owners, and so on;

• establishment of knowledge manage-ment systems—especially withrespect to abstract information andvaluable documentary informationfrom experienced people (such astechnical problems of products, localconditions, and so on); and

• establishment of planning groups toplan for forthcoming projects—includinggroups to take action on time-critical project activities, even beforeinitiating a project (considering lengthyprocedures and associated long delays).

The second factor, a carefullyplanned launch, is vital for the successof any project. The following elementsneed attention in virtual teams:

pmj402_20110.qxd 4/28/09 5:10 PM Page 30


• identification at the central projectoffice of resource requirements formanaging the project—includingdetailed planning, handling procure-ment, budget control, and documen-tation;

• establishment of methodology andmechanisms for project control, riskmanagement, and quality control;

• agreement on the working responsi-bilities of the project office and thevirtual team, and agreement on pro-cedures for work coordination;

• holding of workshops to facilitate theestablishment of trust and achieve-ment of a common understandingregarding work responsibilities at thebeginning of a project;

• recognition of formal and informalcommunication needs, and agree-ment on formal communicationmechanisms for status reporting,monitoring, and controlling; and

• establishment of procedures forundertaking project activities (suchas capturing user requirements,undertaking procurements, and soon).

The third factor, ongoing manage-ment including monitoring and con-trolling of virtual teams, is more diffi-cult with virtual teams than with tradi-tional colocated teams. Attentionshould be given to the following ele-ments:• listening actively to virtual team

members and being sensitive to theirfeelings in the absence of face-to-facecommunication;

• using telephone, e-mail, video confer-encing, and so on (as appropriate for agiven situation);

• adherence to a code of conduct—especially with respect to requests forinformation, checking e-mail fre-quently, time limits, project scope,and work responsibilities;

• establishment of procedures to cap-ture useful information from informalcommunication and to documentdecision-making activities and resultsof decisions;

• having a snapshot of availability ofteam members (including absences,vacations, travel, or leave);

• recognition of differences in teamsand individuals, respect for diversity,facilitation of relationships, fosteringof motivation among team members,and so on;

• holding of regular meetings at two lev-els—(1) at the project level (convenedby the project manager), and (2) at thesubproject level (convened by remoteteam leaders);

• maintenance of trust and relation-ships;

• careful monitoring (including crosschecks) of accuracy of information;

• updating of schedules and provisionof feedback to team members using e-mail, organization’s intranet, and soon;

• recognition of any mismatch betweencorporate culture and team culture(and perhaps between policies andprocedures); and

• maintenance of documentation in a“lean” framework (but with cross-ref-erence to detailed information satisfy-ing corporate requirements).

ConclusionVirtual teams offer cost savings, flexibil-ity and many other benefits, but theyalso create various challenges, particu-larly associated with communicationand leadership. Lack of project visibili-ty, failure to see emotional aspects ofmembers, difficulty in contacts, tech-nology constraints, and so forth are all associated with communication.Problems resulting from miscommuni-cation should be avoided throughprecise and effective communication;this can be facilitated by more complexand interactive communications tech-nologies that are growing in popularity.It is also apparent that the managementof virtual teams requires skills that differ from those required for the man-agement of colocated project teams.That is, it is a challenge for leaders whoare more comfortable and familiar withtraditional face-to-face interactions to

manage virtual team projects. In thevirtual environment, it is difficult tomonitor performance of team mem-bers and implement solutions, as wellas develop team members throughmentoring and coaching. Furthermore,team members are remote from otherteam members and colocated supervi-sors. In this environment, trust, sharedunderstanding, and depth of relation-ships among team members serve as important antecedents for virtualcollaboration (Peters & Manz, 2007, p. 124).

The findings of the present study aregenerally in accordance with the aca-demic literature on the subject. Thestudy also shows that virtual teams forprojects can be established using theexisting personnel of an organization,even though project-specific activitiesoften require skills that differ from thoserequired of personnel in undertakingtheir ordinary day-to-day activities.However, the leadership of such teamsrequires high-level communication andcoordination skills if the work of teammembers is to be harmonized effective-ly. Furthermore, well-planned monitor-ing mechanisms are required to detectand resolve emerging problems that arenot immediately apparent as a result ofthe lack of direct managerial oversightof work undertaken in remote areas. It isalso apparent that quality managementand risk management are more difficultin the virtual environment.

The establishment of effective vir-tual teams is facilitated if employeeshave well-developed self-discipline anda clear understanding of project man-agement concepts from their other day-to-day work experiences—for example,they are well trained in quality reviewsand risk management. However, it isproblematic to apply existing central-ized systems for quality management,risk management, and performanceappraisal to virtual project teams in thesame organization. These mattersrequire further research.

Even though the case is not fully vir-tual (for example, the team did not exist

pmj402_20110.qxd 4/28/09 5:10 PM Page 31



AP

ER

S

across different time zones or use dif-ferent languages), the present casestudy confirms the applicability of vir-tual team concepts in projects. AsHunsaker and Hunsaker (2008) note,“While there may be pitfalls that virtualteams have relative to colocatedteams, these drawbacks can be addressedby developing effective virtual teamleadership” (p. 99).

AcknowledgmentsAn earlier version of the study was pub-lished under the title “Managing VirtualProject Teams: How to MaximizePerformance” in Handbook of BusinessStrategy, Volume 7, Issue 1, published byEmerald Group Publishing Limited. Theauthor is grateful to the NSW Police forpermission to publish the case study. ■

ReferencesAkkirman, A. D., & Harris, D. L. (2005).Organizational communication satis-faction in the virtual workplace.Journal of Management Development,24, 397–409.

Albert, S. R., & Fetzer, R. C. (2005).Smart community networks: Self-directed team effectiveness in action.Team Performance Management,11(5/6), 144–156.

Andres, H. P. (2002). A comparison offace-to-face and virtual software devel-opment teams. Team PerformanceManagement: An International Journal,8(1/2), 39–48.

Barczak, G., McDonough, E. F., &Athanassiou, N. (2006). So you want tobe a global project leader? ResearchTechnology Management, 49(3), 28–35.

Beal, B. (2006). Keep in touch, sayAustralian police: Virtual challenges forhuman-resource team. HumanResource Management InternationalDigest, 14(4), 26–28.

Beranek, P. M., & Martz, B. (2005).Making virtual teams more effective:Improving relational link. TeamPerformance Management, 11(5/6),200–213.

Bergiel, B. J., Bergiel, E. B., & Balsmeier,P. W. (2008). Nature of virtual teams:

A summary of their advantages anddisadvantages. Management ResearchNews, 31(2), 99–110.

Brake, T. (2006). Leading global virtualteams. Industrial and CommercialTraining, 38(3), 116–121.

Cascio, W. F. (2000). Managing a virtualworkplace. Academy of ManagementExecutive, 14, 81–90.

Danton, D. K. (2006). Using intranetsto make virtual teams effective. TeamPerformance Management, 12(7/8),253–257.

Das, S. R., Yaylacicegi, U., & Canel, C.(2008). Using ISO90003 for software“design and development” in large vir-tual teams. Industrial Management &Data Systems, 108, 775–793.

DeLuca, D., & Valacich, J. S. (2006).Virtual teams in and out of synchronic-ity. Information Technology & People,19, 323–344.

Flory, M. (2005). Management fads:The case of the self-managed team.Benchmarking: An InternationalJournal, 12, 275–282.

Glacel, B. P. (1998). Virtual teamscreate new challenges. BusinessJournal, pp. 12–13.

Gould, D. (1997). Leading the virtualteam. Boeing Manager Magazine.Retrieved November 18, 2006, fromhttp://www.seanet/com/~daveg/ltv.htm

Hackman, J. R., & Powell, S. (2004).Viewpoint: Leading teams. TeamPerformance Management, 10(3/4),84–88.

Hinds, P. J., & Bailey, D. E. (2003). Outof sight, out of sync: Understandingconflict in distributed teams.Organization Science, 14, 615–632.

Hunsaker, P. L., & Hunsaker, J. S.(2008). Virtual teams: A leader’s guide.Team Performance Management,14(1/2), 86–101.

Jarvenpaa, S., & Ives, B. (1994). Theglobal network organization of thefuture. Journal of ManagementInformation Systems, 10(4), 25–58.

Jarvenpaa, S., Knoll, K., & Leidner, D.(1998). Is anybody out there?

Antecedents of trust in global virtualteams. Journal of ManagementInformation Systems, 14(4), 29–64.

Johnson, P., Heimann, V., & O’Neill, K.(2001). The “wonderland” of virtualteams. Journal of Workplace Learning,13(1), 24–29.

Kuruppuarachchi, P. (2006). Managingvirtual project teams: How to maxi-mize performance. Handbook ofBusiness Strategy, 7(1), 71–78.

Lau, D. C., & Maurnighan, J. K. (1998).Demographic diversity and faultlines:The compositional dynamics of orga-nizational groups. Academy ofManagement Review, 23, 325–340.

Lee-Kelley, L., Crossman, A., &Cannings, A. (2004). A social interac-tion approach to managing the “invisi-bles” of virtual teams. IndustrialManagement & Data Systems, 104,650–657.

Lipnack, J., & Stamps, J. (2000). Virtualteams: People working across bound-aries with technology (2nd ed.). NewYork: Wiley.

Malhotra, A., Majchrzak, A., & Rosen,B. (2007). Leading virtual teams.Academy of Management Perspective,21(1), 60–70.

Melymuka, K. (1998). What you heardis not what I said. Computerworld,32(28), 63.

Morris, S. (2008). How to get realresults from virtual teams. HumanResource Management InternationalDigest, 16(4), 33–35.

Nortel and BP succeed through virtualteamwork—But does the answer lie intechnology alone? (2003). HumanResource Management InternationalDigest, 11(4), 8–11.

Oertig, M., & Buergi, T. (2006). Thechallenges of managing cross-culturalvirtual project teams. TeamPerformance Management, 12(1/2),23–30.

Pauleen, D. J. (2003). Leadership in aglobal virtual team: An action learningapproach. Leadership & OrganizationDevelopment Journal, 24(3), 153–162.

pmj402_20110.qxd 4/28/09 5:10 PM Page 32


Peters, L. M., & Manz, C. C. (2007).Identifying antecedents of virtual teamcollaboration. Team PerformanceManagement, 13(3/4), 117–129.

Piccoli, G., Powell, A., & Ives, B. (2004).Virtual teams: Team control structure,work processes, and team effective-ness. Information Technology & People,17, 359–379.

Powell, A., Galvin, J., & Piccoli, G.(2006). Antecedents to team membercommitment from near and far: Acomparison between collocated andvirtual teams. Information Technology &People, 19(4), 299–322.

Prasad, K., & Akhilesh, K. B. (2002).Global virtual teams: What impactstheir design and performance? TeamPerformance Management: AnInternational Journal, 8 (5/6),102–112.

Project Management Institute (PMI).(2004). A guide to the project management

body of knowledge (PMBOK® Guide)—Third edition. Newtown Square, PA:Author.

Putnam, L. (2001). Distance teamwork:The realities of collaborating with vir-tual colleagues. Online, 25(2), 54–58.

Shirley, D., & Morton, D. (1998).Managing martians. New York:Broadway Books.

Stevenson, W., & McGrath, E. W. (2004).Differences between on-site and off-site teams: Manager perceptions. TeamPerformance Management, 10 (5/6),127–132.

Stough, S., Eom, S., & Buckenmyer, J.(2000). Virtual teaming: A strategy formoving your organization into the newmillennium. Industrial Management &Data Systems, 100(8), 370–378.

Townsend, A. M., DeMarie, S. M., &Hendrickson, A. R. (1996). Are youready for virtual teams. HR Magazine,41(9), 122–126.

Townsend, A. M., DeMarie, S. M., &Hendrickson, A. R. (1998). Virtualteams: Technology and the workplaceof the future. Academy of ManagementExecutive, 12(3), 17–29.

Palitha R. Kuruppuarachchi is the manager ofRadio Engineering Services in the New SouthWales Police Force, Sydney, Australia. He holds adoctor of technology, a master’s degree in busi-ness and technology, a master’s degree in com-puter science, and a bachelor’s degree in elec-tronic and telecommunication engineering. Hehas wide experience in formulating strategies,converting strategies to projects, and managingprojects, particularly information and communi-cations technology projects. He has publishedmany papers integrating the disciplines leader-ship, innovation, change management, andproject management.

pmj402_20110.qxd 4/28/09 5:10 PM Page 33

PA

PE

RS


INTRODUCTION ■

The performance of individuals in knowledge-intensive work in anyform of organization remains critical to the success of both individual-level and organization-level goals. Understanding the factors thatenhance and diminish the performance levels of individuals is there-

fore a necessity for monitoring and managing performance. Accordingly, agrowing body of research in management and organizational psychologyproposes to understand performance by decomposing its constructs basedon task level and contextual levels (Borman & Motowidlo, 1993). Theoriesfrom information systems (IS) research, for example, suggest that individualperformance can be understood by examining the “task-technology” fitwithin organizational human resources (Goodhue & Thompson, 1995). Inproject environments, much emphasis is on models of the personal attrib-utes of project personnel that relate to performance—for example, teamleadership effectiveness (Thamhain, 2004), management and leadershipstyles (Dvir, Sadeh, & Malach-Pines, 2006; Turner & Müller, 2005), and softskills such as motivation (Muzio, Fisher, Thomas, & Peters, 2007; Peterson,2007). However, these models do not account for the importance of socialprocesses that weave together a rich fabric of human or technology-enabledsocial and professional relationships that contribute largely toward per-formance. To this end, an emergent discipline of social networks theory andresearch takes as its central premise the embeddedness (Granovetter, 1985,p. 1065) of individuals in social networks. The novelty of this stream ofresearch lies in how it draws on the structural properties of individuals in asocial network to explain outcomes such as individual performance.

In this article, we examine the inherent relationship between profession-al network structure and individual performance by developing a theoreticalframework based on existing literature in the sociology, information science,and management science disciplines. By obtaining a pattern of network ofadvice-seeking interactions, we examine the fine-grained associationsbetween an individual’s network properties, measured by social networkstructure, position, and tie variables, and its relationship with performance,measured by the individual’s performance attitude about the various dimen-sions of task-level activities. The following section lays the conceptual foun-dation for the study, followed by the epistemological stance taken and theoretical justification of the hypotheses developed for the study. The latterparts of the article discuss the methodology for the study, including the sur-vey development procedure, followed by results and discussion.

Measuring Performance ofKnowledge-Intensive WorkgroupsThrough Social NetworksKon Shing Kenneth Chung, Project Management Graduate Programme, University ofSydney, AustraliaLiaquat Hossain, Project Management Graduate Programme, University of Sydney,Australia

ABSTRACT ■

In this article, we examine the effect of socialnetwork position, structure, and ties on the per-formance of knowledge-intensive workers indispersed occupational communities. Usingstructural holes and strength-of-tie theory, wedevelop a theoretical framework and a valid andreliable survey instrument. Second, we applynetwork and structural holes measures forunderstanding its association with perform-ance. Empirical results suggest that degreecentrality in a knowledge workers’ professionalnetwork positively influences performance use,whereas a highly constrained professional net-work is detrimental to performance. The find-ings show that social network structure andposition are important factors to consider forindividual performance.

KEYWORDS: social network; structure; ties;position; performance; knowledge-intensivework





DOI: 10.1002/pmj.20115

pmj402_20115.qxd 4/28/09 3:46 PM Page 34


BackgroundSocial NetworksBy social network, we mean a con-stituent of two or more actors (individ-uals) who are connected through oneor more relationships such as providingadvice, information, and so on. Socialnetwork studies have long been con-cerned with exploring structural and tieeffects, with a view toward illuminatingand explaining patterns of relation-ships in order to infer some outcome.For social network scholars, the raisond’être is that the structure of relation-ships among actors and the location ofindividual actors in the network haveimportant behavioral, perceptual, andattitudinal consequences both for theindividual units and for the system as awhole (Knoke & Kulinski, 1992). At theindividual level, the debate concen-trates on how the structural position ofan individual in the network impacts anoutcome, such as performance, of thatperson.

Social Networks and PerformanceNetwork effects on individuals’ abilityto perform better have been document-ed in studies on communications, soci-ology, and social psychology (Coleman,1988; Guetzkow & Dill, 1957; Leavitt,1951). Previous studies further demon-strate that actors with a dense socialnetwork perform better (Oh, Chung, &Labianca, 2004; Reagans & McEvily,2003). Furthermore, actors who are richin structural holes (i.e., having connec-tions to social clusters or groups whoare themselves not well connected) arebetter situated in their social networkto obtain, control, and broker informa-tion (Burt, 1992). Beginning with earlyliterature about communication pat-terns and the performance of groupsand individuals in project and organi-zational contexts, researchers demon-strated that, rather than being remote,impersonal, and rigid, knowledge-intensive work was actually communal,reflecting a strong interpersonal net-work of interconnected workers. Thestudies also suggested that informal

networks were equally or more impor-tant than formal networks in knowl-edge-intensive work, with the premisebeing that individual performance wasa function of network structure(Gabbay & Leenders, 2001). In fact,studies relating network structures toperformance have shown that in-degree centrality, “betweenness” central-ity, and density in network structuressuch as advice networks were related to coordination and project-relatedperformance (Hossain, Wu, & Chung,2006).

Problems Related to GeographicallyDistributed Knowledge-Intensive WorkThe quality of job performance inknowledge-intensive work is affectedby a variety of factors, such as experi-ence, education, keeping abreast ofwork-related and technologicalchanges, and so on. Holding such indi-vidual properties constant, perform-ance to a large extent is the product ofobtaining the right information toaccomplish the task at hand or to solvecomplex problems. For example, find-ing information and finding expertisefor handling the right information iscrucial for job performance. However,although knowledge and expertise arecritical resources, their mere presenceis insufficient to produce high-qualitywork. As Faraj and Sproull (2000) argue,expertise must be managed and coordi-nated in order to leverage its potential.This entails knowing where expertise islocated and where it is needed, andbringing needed expertise to bear. Thisproblem is accentuated when geo-graphical barriers are imposed. Grinter,Herbsleb, and Perry (1999) argue thatirrespective of the area of expertise,product structure, processes, and cus-tomized steps in organizational work,one of the most pertinent problems isthe location of expertise.

In distributed project environ-ments especially, Cross and Cummings(2004) claim that individuals who arenot aware of the location of expertiseelsewhere and who have fewer ties

spanning organizational and geo-graphical boundaries will have difficul-ty obtaining useful information forwork purposes. Furthermore, there isplenty of literature that emphasizes theimportance of social and professionalnetwork structure, position, and tiediversity. For instance, individuals whotend to be in closed networks tend tohave nondiverse ties, and the interac-tions are usually with the same individu-als. Such individuals are less successfulin adapting to a changing environ-ment and in receiving useful and novelinformation, and their work is thusmarked with low-quality performance(Ancona & Caldwell, 1992; Cummings,2004; Podolny & Baron, 1997; Reagans &Zuckerman, 2001).

Research QuestionsGiven the arguments above, the follow-ing questions motivate this research:• How can individual performance be

understood through the emergentpatterns of social processes that con-stitute performance?

• How can it be evaluated?• What is the role of social influence

and social networks (that create suchinfluence) in understanding individ-ual performance?

• Why is understanding social networkstructure and position important forunderstanding individual perform-ance?

• How does one account for social fac-tors, apart from personal and demo-graphic factors, that are important forenhancing individual performance inproject environments?

In order to shed light on the abovephilosophical questions, one needs toexplore possible answers by reviewingthe literature in the area of social net-works and performance. While there iscurrently a lack of literature that tiesthese constructs together coherently inproject contexts, it is important thatthese constructs are explored individual-ly, jointly, and holistically in a sequentialmanner. Figure 1 depicts a conceptual

pmj402_20115.qxd 4/28/09 3:46 PM Page 35


Measuring Performance of Knowledge-Intensive WorkgroupsP

AP

ER

S

framework, and the following sectionexplores some of the earliest works in thedomain of network structure and per-formance.

Classical Works of NetworkStructure and Performance Bavelas-Leavitt ExperimentOne of the earliest studies that relatedsociometric aspects of human commu-nication patterns to performance wasthat of the “Bavelas-Leavitt Experiment”(Bavelas, 1950; Leavitt, 1949), alsoknown as the MIT experiment. Drawingfrom the assumption that (1) success ofany classes of tasks depends upon aneffective flow of communication (hold-ing the nature and content of the com-munication constant) and (2) a fixedcommunication pattern affects taskperformance and individual outcome,the motivating question in the studyasked, “Under what principles may apattern of communication be deter-mined that will in fact be a fit one foreffective and efficient human effort?”The question sought to answer, througha laboratory-controlled experiment,how social network structure measuredin terms of patterns of communicationaffects the work and life of individualswithin groups.

The experiment consisted of fivepeople who had to communicate witheach other through enclosed cubicles

in order to solve a puzzle. Each subjectwas given five symbols from a set of six.All had unique symbols, but there wasone symbol from each group of five thatwas common to all of the groups. Thepuzzle was solved when a consensuswas reached as to what the commonsymbol was. The experiment was trialed15 times. None of the subjects kneweach other, nor did they know the num-ber of subjects in the study, or the configuration of the communicationstructure. The channels and flow ofcommunication was controlled by theexperimenter. The subjects could passas many messages as they wantedthrough the cubicle lines, depending onthe structure of the communicationchannels as shown in Figure 2.

The performance of network struc-tures was evaluated on the basis of pat-tern comparison and node-level analysis.Performance of the task-orientedgroups was measured in terms of thetime required to complete the puzzleand the number of errors made in theprocess of “guessing” the right answer.When the patterns of various structureswere compared, the completion time(i.e., time that it took to complete thepuzzle) for groups using the “star” and“Y” structures was on average relativelyshorter than for groups using the otherstructures (the “circle” and “line” struc-tures). The explanation offered byLeavitt (1951) was that centralizationwas key to influencing performance.Using centralization (measured as thesum of the internal distances of nodes xto y) as an operational construct, it wasfound that patterns that demonstratedhigher centralization performed better.When the subjects channeled all infor-mation through a central actor, theinformation was better coordinated andshared. Thus, groups using the “star”and the “Y” structures also used theleast number of messages comparedwith groups using the other structuresand also made the fewest errors. Whennode-level analysis was conducted, itwas discovered that structures that hadhigher centralization also tended tohave a leader emerge during the taskprocesses. The leaders emerged at posi-tions of highest centrality (measured by

Social Network

PersonalAttributes

Performance

Figure 1: Conceptual framework.

B C

A

D

E

B

C

A

D

E

B

C

A

D

E

B

C

A

D

E

Figure 2: The Y, star, circle, and line structures of communication.

pmj402_20115.qxd 4/28/09 3:46 PM Page 36


the degrees of communication activity).Thus, the “Y” and “star” structures hadnodes with extremely high-degree cen-trality compared with other nodes with-in the structure, which led to better performance.

Inevitably, a thought-provokingfinding that emerged from this studyback then was that centralized struc-tures, such as the “star” (or “hub-spokes”or “wheel”) network, were far more con-ducive to performance (solving the puz-zle faster), in contrast to decentralized orflatter structures, such as the “circle”network. The crux of the argument isthat information flow is inefficient indecentralized networks and, therefore,less conducive to performance.

However, later research byGuetzkow and Simon (1955) revealedthat decentralized structures actuallyworked better than centralized struc-tures when tasks become more com-plex. Complexity of tasks results inproblems and subtasks that cannot behandled by an individual alone. Thesame is true when central nodes areoverwhelmed with communications. Inthis context, the “circle” structureworked much better than the “star”structure. The performance of the struc-tures depended more on how well thechannels of communication were usedthan on the structure per se. In contrastto the “star” structure, the “all-channels”structure shown in Figure 3 (assumingthat there are no “ties-overloads”caused by having excessive ties pernode) provides a reconfigurable capaci-ty for task-relevant communication. Forthe actors, this allows an opportunity tonegotiate about the directions of com-munication, details regarding what thetask type is, and whether specific nodesare to be brokers of the communication.The resulting communication patternstend to be potentially more efficientthan if the network structure and flow ofinformation were rigid in form.

As Guetzkow and Simon (1955) andGuetzkow and Dill (1957) stipulate, theactors must solve two problems: (1) thatof developing an organization scheme

suitable for finding the common symbolwithin the constraints of their particularnetwork structures and (2) that of actu-ally identifying the common symbol.Ultimately, they would seek and find anorganizational structure that works andthen play with variations on it both tomaintain interest and to seek a “better”form.

The Bavelas and Leavitt experimentduring the early 1950s was a crucialmilestone, as it injected rejuvenatedvigor in the space of network structure-performance research. Their key findingwas that centralization leads toenhanced performance in simple tasksand that decentralization leads to effi-cient performance in complex tasks.Their conceptualization of the effectthat communication pattern in terms ofsocial structure and social position hason task performance opened up newresearch avenues. Consequently, itbegged new ideas and questions onunderstanding performance using asocial networks perspective.

Freeman’s Concept on Centrality

As described in the preceding section,the idea of centrality was applied to

human communication in the early1950s. Although the “structural central-ity and influence in group processes”hypothesis was proposed by Bavelas(1950) and reported in depth by Leavitt(1951), modifications and extension tothe original experiments provided con-tradictory and confusing results(Burgess, 1968). In the late 1970s,Freeman (1978) wrote a seminal articlethat clarified the conceptual founda-tions of centrality, which soon becamea core concept in social network stud-ies. His work laid the foundation forsocial network scholars to apply andextend the notion of structural central-ity at both the node and network level,conceptually and empirically.

Freeman (1978) reviewed literaturethat utilized the notion of centralityand showed that the concept wasapplied beyond “experimental groups’task-related” research such as in under-standing urban development, organi-zation of populated nations such asIndia, transportation and communica-tion networks in Russia, and patterns ofdiffusion of technological innovation inthe steel industry. Freeman (1978) thenreviewed the various measures and

E

B

C

D

A

Figure 3: The “all channels” structure.

pmj402_20115.qxd 4/28/09 3:46 PM Page 37



AP

ER

S

overlapping concepts of centrality byunifying the measures, while clarifyingcentrality’s range and limits of potentialfor application.

In particular, he defined centralityin terms of point, “betweenness,” andcloseness centrality, with each havingimportant implications on social out-come and processes. According toFreeman (1978), “these kinds of central-ity imply three competing ‘theories’ ofhow centrality might affect groupprocesses.” In-degree centrality can bemeasured in terms of degree (the num-ber of ties to and from an actor).Structurally, centrality is measured interms of closeness (the extent to whichan actor is close to all others in the net-work) and “betweenness” (the extent towhich an actor lies in the shortest pathto all others in the network). Each cen-trality concept has been related toimportant social occurrences: in-degree centrality being viewed as animportant indicator of an actor’s com-munication activity; “betweenness”centrality being viewed as an indicatorof the potential of an actor’s control ofcommunication; and closeness as anindex of minimum cost of time and effi-ciency for communicating with otheractors in the network.

In a subsequent study, Freeman,Roeder, and Mulholland (1979) revertedto the classic experiment by Bavelas(1950) to study the effects of structuralcentrality on human communication.Using 100 volunteers (university stu-dents) as subjects for the experiment,Freeman et al. (1979) analyzed theresults and demonstrated that centrali-ty is an important structural factorinfluencing leadership, satisfaction,and efficiency. In particular, out of thethree concepts of structural centrality,only two demonstrated interestingresults and significance in its effect onperformance—namely, the control-based measure of betweenness and theactivity-based measure of degree. Theindependence index based on close-ness was “vaguely” related to experi-mental results. Interestingly, another

structural factor, the overall density ofcommunication paths in the structuralform, also turned out to be relevant inunderstanding performance. Back atthe time of this study, these resultsbreathed new life into the MIT labora-tory experiments at a time when theresearch avenues from the experimentsseemed to have stagnated, as it raisednew questions as to what kinds of struc-ture (influenced by individuals orgroups) influenced differing types ofperformance (task or contextual).

In another classical work on theeffects of network structure on innova-tion diffusion, Coleman, Katz, andMenzel (1957) attempted to understandthe underlying social processes thataffected 125 doctors’ rate of adoption ofa new drug. Results suggested that doc-tors who were generally more integratedwith their peers—that is, with densernetworks—were quicker to adopt thenew drug compared with those whowere more isolated. The results fromColeman et al.’s (1957) study suggest thatthe larger number of ties an individualhas results in a higher likelihood to dif-fuse innovation faster. These individualsare quicker to capitalize on the noveltyof the information and are thus in a posi-tion to enhance individual outcomessuch as performance. These results resonated strongly with the similarfindings about the density concept byFreeman et al. (1979) described above.Since then, the notion of centrality,density, and centralization were one ofthe key network measures used forstudying effects on individual andgroup outcomes such as task efficiency,productivity, improved performance,and project coordination, as well asfavorable attitudes toward task-relatedwork (Ahuja, Galletta, & Carley, 2003;Bonacich, 1991; Brass, 1981, 1985; Cross &Cummings, 2004; Cummings & Cross,2003; Faust, 1997; Hossain et al., 2006;Mullen, Johnson, & Salas, 1991; Pfeffer,1980; Salancik & Pfeffer, 1978; Sparrowe,Liden, Wayne, & Kraimer, 2001).

The contribution of Freeman’s(1978) work was so influential that the

notion of centrality is now almostalways attributed to him. By expoundingon the intuitive notions of centrality—closeness, degree, and “betweenness”—Freeman (1978) not only unified thesemeasures mathematically, but also provided their respective theoreticaland practical implications, which were key contributions to the networkstructure and task-performance re-search. That is, “betweenness” centrality(the extent of communication con-trolled) and degree centrality (theextent of communication activity) have been shown to influence perfor-mance from a network structure per-spective, while closeness centrality (theextent of communication efficiency)does not.

Burt’s Structural Holes TheoryA major caveat in previous studies ofnetwork structure such as that ofColeman et al. (1957) is that theyassume that individuals are able tomaintain ties within their personal net-work consistently over time. They alsoassume that each tie is a channelprovider of unique information or com-munication. These drawbacks are theparadoxical reasons why an extremelydense network may paralyze an indi-vidual’s ability to perform better.

In the early 1990s, Burt (1992)made an influential contribution to thenetwork paradigm and phenomena ofstructural effects on individual out-come by shifting the focus from net-work structure and network relations tonetwork position. Burt’s (1992) theoryon structural holes offers a novel andsubtle but interesting perspective inexplaining why some individuals per-form better and others do not. Forexample, it takes Coleman et al.’s(1957) study and its assumptions a stepfurther by offering an explanation as towhy social processes such as innova-tion diffusion may occur faster from astructural positional point of viewrather than from a relational perspec-tive. The theory is linked to personalitytheory, suggesting that the personal

pmj402_20115.qxd 4/28/09 3:46 PM Page 38


attributes (such as locus of control,leadership skills, and ability to performwell) of an individual are associatedwith structural autonomy—an optimalsituation where an individual benefitsfrom nonredundant information bene-fits.

In contrast to network structureand relational ties, Burt (1992) arguesthat the structural configuration of anindividual’s social network that pro-vides optimized “brokerage” position iswhat dictates structural advantages,such as information novelty and con-trol. This argument is based on the factthat maximizing the number of ties(weak or strong) in an individual’s net-work does not necessarily provide suchbenefits. Instead, opportunity costscome into play, and the maintenance ofties becomes expensive in terms of timeand resources. Furthermore, as an indi-vidual’s personal network grows overtime, the extent of information comingfrom closely knit clusters tends tobecome redundant. Intuitively, an indi-vidual can maintain no more than 50 ormore close ties on a frequent basis. Thisfigure represents, at best, an individualstriving to maintain ties with his con-tacts. In particular, in a study byGurevitch (1961), where a diary ofacquaintances and friends was kept for100 days by 15 study participants, themaximum number of acquaintancesnoted was 658; the average was 500. Asone can imagine, to maintain ties withsuch numbers of contacts is sociallyexpensive and time-consuming. To thisend, Burt (1992) capitalizes on his theo-ry of structural holes by focusing on theimportance of structural position ratherthan structural properties such as den-sity or the size of the network.

The theory of structural holes issimple and intuitive yet empirically pro-found. “Holes” in the network refers tothe absence of ties that would otherwiseconnect unconnected clusters together.Individuals who bridge these holesattain an advantageous position thatyields information and control benefits.Therefore, structural holes theory is

based on the idea that actors are in a better position to benefit from inter-actions with others if the actors are con-nected to others who are not well connected themselves or are not wellorganized. In other words, the bridgingof connection to others provides oppor-tunities; the lack of connections amongthose others are the holes in the struc-ture (i.e., “structural holes”). Individualswho attain structural autonomy arethose who bridge all structural holeswhile the groups to whom the individ-ual is connected to are surrounded bystructural holes. Closer examination onthe crux of structural holes theoryreveals that it is based on the assump-tion of betweenness centrality: thatpower and influence accrue to thosewho broker connections betweenunconnected groups of people. It is thisconcept of betweenness centrality thatBurt (1992) capitalizes on and extendsto explain the role of “brokerage” as aform of obtaining structural autonomythat leads to improved performance,getting ahead, and obtaining goodideas. This theoretical contributionoffers a more insightful perspective onindividual performance, given thatGuetzkow and Simon (1955) note thatcentrality in itself is not always a keypredictor of individual performance.Instead, the theory offers insightfulexplanation beyond the concept of cen-trality and centralization in that an indi-vidual’s benefit accrues from the extentto which the individual’s network is effi-cient, effective, and constrained. Thefollowing section discusses networkefficiency and constraint in greaterdetail.

Network Efficiency and Effectiveness

In order to optimize a network by cap-italizing on structural holes, Burt(1992) claims that increasing networksize (number of direct contacts) with-out considering the diversity reachedby the contacts makes the networkinefficient in many ways. Therefore,the number of nonredundant contactsis important to the extent that redun-

dant contacts would lead to the samepeople and, hence, provide the sameinformation benefits. The term effec-tiveness is used to denote the averagenumber of people reached per primarycontact; while the term efficiency con-cerns the total number of people ofpeople reached with all primary con-tacts. Hence, effectiveness is about theyield per primary contact, while effi-ciency is about the yield of the entirenetwork. To illustrate, consider two net-work diagrams, first, an inefficient network (A) where “you” as the ego areconnected to the members of fourclusters of contacts who are all con-nected to each other within the clusterbut there are no intercluster connec-tions; and second, an efficient network(B) where “you” as the ego are con-nected to one member only from eachof the four clusters where all memberswithin the cluster are connected toeach other and there are no interclus-ter connections. In both networks,each cluster consists of four members.

Thus, in network A, the ego (“you”)maintains 16 ties with every contact inthe network. This represents a signifi-cant strain on the ego in terms of timeand opportunity cost that could beinvested in other contacts. Network B isfar more efficient, because the ego only needs to maintain ties with fourprimary contacts, thereby achievingefficiency at one-fourth of the costcompared with network A. Network B isfar more effective because the primarycontacts in this network are nonredun-dant in that they are connected to clus-ters that are not connected to eachother. An effective network thereforeregards the primary contacts as ports ofaccess to diverse clusters (because of nonredundancy), and thereforeachieves the yield of the entire network.

The term that Burt (1992) uses todenote effectiveness in networks iseffective size. In network A, the networksize is 16. The effective size, however, is4, because in effect the ego is able toobtain novel information and benefitsonly from the four clusters, which are

pmj402_20115.qxd 4/28/09 3:46 PM Page 39



AP

ER

S

not connected to each other exceptthrough “you.” The other three ties toeach of the clusters are redundantbecause they provide the same infor-mation that is available through thefourth. Efficiency in network A is there-fore 0.25 (measured as effective size[4]/network size [16]). In network B, thenetwork size is 4, and effective size is 4,resulting in perfect efficiency of 1 (4/4).Ideally, the number of nonredundantcontacts should increase with the num-ber of contacts to achieve optimal effi-ciency (i.e., 1). As one increases one’snumber of contacts and gradually startsto have a smaller number of nonredun-dant contacts, the individual’s networkefficiency decreases. Conversely, as thenumber of nonredundant contactsincreases relative to the lower numberof contacts, the individual’s networkefficiency increases.

Network Constraint

Constraint dictates the extent towhich an individual’s opportunitiesare limited by investing the bulk of hisor her network time and energy inrelationships that lead back to the sin-gle contact (Burt, 1992, p. 55). In otherwords, constraint measures thedegree to which an individual’s con-tacts are connected to each other andis therefore a proxy for redundancy ofcontacts. According to Hanneman(2001), constraint also measures theextent to which an ego is connected toothers who are connected to oneanother. So if the ego has many con-nections to others, who in turn havemany connections to still others, theego is quite constrained. At organiza-tional levels, individuals with highconstraint indices are unable to con-ceive novel ideas because of theredundant nature of information thatis sourced from a densely connectedgroup of individuals.

Previous research has consistentlydemonstrated that high-efficiency andlow-constraint indices are useful indi-cators of an individual’s ability to pro-duce good ideas (Burt, 2004), to “get

ahead” in terms of job performanceand promotion (Burt, 1992, 2005), andto enjoy greater career mobility(Podolny & Baron, 1997). In line withthese arguments, it is expected thatindividuals in knowledge-intensivework thrive on useful knowledge andinformation from peers. An individualin knowledge-intensive work with anefficient and low-constrained networkstructure is thus more likely to obtainuseful knowledge from diverse andnonredundant contacts, which hasbeen linked to improved performance.Therefore, the following hypotheses areformally derived:

H1a: Efficiency of an individual’s net-work position is positively associatedwith performance in knowledge-inten-sive work.

H1b: Constraint of an individual’s net-work position is negatively associatedwith performance in knowledge-inten-sive work.

Social Network Ties and PerformanceIt is obvious at this point, for social net-work scholars, that the raison d’être isthat the structure of relationshipsamong actors and the location of indi-vidual actors in the network haveimportant behavioral, perceptual, andattitudinal consequences both for theindividual units and for the system as awhole (Knoke & Kulinski, 1992). In thiscontext, individual or system-level out-comes are treated as a function of net-work structure and position (Gabbay &Leenders, 2001). However, at the indi-vidual level, the debate concentratesnot only on how structural positionimpacts individual performance, butalso concentrates on the relationalcomponents of an individual’s network.Evidence in the literature demonstratesthat just as structural position plays avital role in the effect of individual per-formance, tie strength has significanteffects as well (Borgatti, Jones, & Everett,1998; Hossain et al., 2006; Mehra, Kilduff,& Brass, 2001; Reagans & McEvily, 2003;Sparrowe et al., 2001).

Granovetter’s Theory on the Strength of Weak Ties

The most seminal work in social net-works with regards to the relationalcomponent of an individual’s socialnetwork almost always begins withGranovetter’s (1973) theory on thestrength of weak ties. Granovetter (1973)argues that individuals obtain new andnovel information from weak ties ratherthan from strong ties within the individ-ual’s group structure. The argument isbased on assumptions concerning thehomophilous nature of actors in a socialsystem, where strong ties tend to bondsimilar people to each other and wherethese similar people tend to clustertogether such that they all becomemutually connected. For this reason,information that originates and circu-lates at a high velocity among stronglytied cliques or clusters tends to becomeobsolete or redundant in a shortamount of time. Such network clustersor cliques of people bound together bystrong ties are therefore not conduciveto channels of innovation. That is, suchnetworks are closed networks and arenot highly receptive of new information.Granovetter (1973, p. 1373) suggests thatthe influx of new and novel informationmust therefore come from weak ties(hence, the theory of the strength ofweak ties), which serves as a bridge to adifferent cluster of people, from whichthe new information originates.

Although the strength of weak tiestheory has wide appeal, it has the draw-back of implying that maximizing thenumber of weak ties in one’s personalnetwork would yield novel informationbenefits that, in turn, allow one to per-form better. Furthermore, according toBurt (1992, p. 28), a weak tie is “thechasm spanned and the span itself. Bytitle and subsequent application, theweak tie argument is about the strengthof relationships that span the chasmbetween two social clusters” and there-fore not about information benefits. Bydefinition then, weak ties cannot pro-duce novel information benefits per se.Unlike structural holes, which provide

pmj402_20115.qxd 4/28/09 3:46 PM Page 40


information benefits, weak ties obscurethe control benefits that come alongwith information benefits.

Strength of Strong Ties

Subsequent to the seminal study on thestrength of weak ties, several researchershave examined the dichotomy of thestrength of ties and associated it withindividuals and group outcomes.Krackhardt (1992), for instance, notesthat the “affect” level of strong ties isimportant and cannot be ignored. In hisstudy (Krackhardt, 1992) of a SiliconValley firm where advice and friendshipnetworks of 36 employees were com-pared, he concluded that strong tieswere particularly important, especiallyin the generation of trust within propa-gators of major organizational change.In another study of 127 knowledge-intensive workers from a pharmaceuti-cal company, a bank, and an oil and gascompany, Levin and Cross (2004) report-ed that the relationship between strongties and receipt of useful knowledge wasmediated by benevolence-based trustand competence-based trust. They alsofound evidence that strong ties, more sothan weaker ties, led to the receipt ofuseful knowledge for improving per-formance in knowledge-intensive work.However, when their research modelcontrolled for the two dimensions oftrust, the structural benefit of weak tiesemerged, suggesting that it was theweaker ties that provided access tononredundant information. In this con-text, the results are consistent with priorresearch by Hansen (1999), who investi-gated the association between tiestrength, transfer of complex knowl-edge, and performance in terms of proj-ect completion times by 41 differentsubunits within an organization. Theconceptual model postulated by Hansen(1999) is that strong ties facilitate• low search benefits with moderate

transfer problems where knowledge isnoncodified or dependent,

• low search benefits with few transferproblems where knowledge is codifiedor independent, and

weak ties facilitate• search benefits with severe transfer

problems where knowledge is noncod-ified or dependent,

• search benefits with few transfer prob-lems where knowledge is codified orindependent.

According to Hansen (1999), weakties facilitate faster project completiontimes when the task is simple andenables faster search for useful knowl-edge among other organizational sub-units. However, it is strong ties ratherthan weak ties that foster complexknowledge transfer, as the transferprocess is slowed down when knowl-edge is highly complex, where the com-plexity of knowledge is determined by the degree to which it is tacit and bywhether an individual is dependent onanother for transfer and acquisition ofknowledge. Similar findings were alsoechoed in the study of Reagans andMcEvily (2003) of a sociocentric net-work of 104 highly skilled employeeswithin a contract research and develop-ment (R&D) firm; this study found apositive association between tiestrength and the ease of knowledgetransfer in performing knowledge-intensive task activities. Furthermore,the authors also found that diversity ofties (in terms of network range or ties todifferent knowledge pools) was posi-tively associated with ease of knowl-edge transfer. Therefore, where knowl-edge-intensive work is involved andwhere knowledge transfer and receiptof useful information is crucial for per-formance, strong ties rather than weakties facilitate complex knowledge trans-fer, especially to heterogeneous audi-ences (Reagans & Zuckerman, 2001). Inother words, for an individual engagedin knowledge-intensive work to per-form well, the importance of strong tiesof an individual cannot be discounted.Stated formally:

H2a: Strong ties will be positively associ-ated with individual performance inknowledge-intensive work.

Degree and Diversity of Ties

Apart from observing tie strength, vari-ous studies also examined the numberof ties as a significant predictor of indi-vidual performance. In this context, thenumber of direct ties that the individualhas to other individuals in his or hernetwork is also termed network size(Burt, 1992, p. 16) or degree centrality(Freeman, 1978). In a study that exam-ined tie correlates of the individual per-formance of 101 engineers and 125consultants, Cross and Cummings(2004) found significant support for thepositive association between an indi-vidual’s number of ties (both from andto departments outside his or herdepartment) and individual perfor-mance. Furthermore, they also foundthat an individual’s number of ties tohigher hierarchy levels and those thatspanned physical (geographical) barri-ers were positively related to perform-ance. These findings suggest that tiesthat span departmental boundaries andgeographical barriers, including thosethat reach senior personnel, enhancethe diversity of information flow, espe-cially where complex work with highintegration of specialized knowledge isinvolved. According to Cummings(2004), individuals in workgroups aremore likely to perform better if theyexchange knowledge externally withmembers in their professional networkwho are structurally diverse. Structuraldiversity here refers to individual differ-ences in geographic locations, function-al assignments, reporting managers,and business units. For example, pastresearch has shown that geographicallocation influences what individualsexperience and who they interact with,and therefore generates novel task-related information and knowledge-sharing opportunities (Monge, Rothman,Eisenberg, Miller, & Kirste, 1985). In par-ticular, although internal and externalknowledge sharing has direct implica-tions on performance, the latter is morevaluable when individuals in the work-group are more structurally diverse.Therefore, individuals with higher

pmj402_20115.qxd 4/28/09 3:46 PM Page 41



AP

ER

S

reach and diversity of information aremore likely to be exposed to unique andrelevant knowledge that is helpful insolving complex problems. More for-mally:

H2b: Degree centrality is positively asso-ciated with individual performance inknowledge-intensive work.

H2c: Geographical tie diversity is posi-tively associated with individual per-formance in knowledge-intensive work.

H2d: Functional tie diversity is positive-ly associated with individual perform-ance in knowledge-intensive work.

To summarize, the previous sec-tions of the review up to this pointhave critically analyzed key literatureconcerning the relationship betweensocial network and task-related per-formance. In terms of social networkstructure, Bavelas (1950) and Leavitt(1951) demonstrated that hierarchicalor centralized structures perform bet-ter when tasks are simple, but thatdecentralized structures are more con-ducive toward lesser errors, satisfac-tion, and speed of task completion incomplex tasks. Freeman (1978) laterexpanded on the notion of centraliza-tion to show that performance wasrelated to the individual’s property ofcentrality attributed by the structure of the network. In particular, he identi-fied that degree centrality connotesintensity of communication flow, thatbetweenness centrality indicated com-munication power and influence, andthat closeness centrality indicated efficiency of information flow. Burt’s(1992) notion of structural holes builtfurther upon the assumption ofbetweenness centrality that advocatedthe idea of a brokerage position as pro-viding information and control bene-fits. In fact, this shift from the focus onnetwork structure to network positionwas instrumental, and paved the wayfor further research in delineating therelationship of communication pat-terns and performance at the indi-vidual level. At the relational level of

network structure, the main theoryreviewed was the strength of weak tiestheory (Granovetter, 1973), which stip-ulates that weak ties provide novel anduseful information, as opposed tostrong ties within a densely knit clusterof people. However, later research inthe dichotomy of the strength of tieshas also led to claims that strong anddiverse ties are equally and respective-ly important for performance. Thisresearch, therefore, amalgamates theseconcepts together to propose that net-work structure, position, and relationsindividually and jointly impact individ-ual performance. In the next section,the implications of personal attributesand its relevance to the network per-formance model are introduced andincorporated.

Personal Attributes, ProfessionalSocial Networks, and PerformanceAlthough it has been ascertained thatthe structure of relations among actorsand their network positions haveimportant behavioral, perceptual, andattitudinal consequences for individu-als, one cannot discount the impor-tance of personal attributes that affectan individual’s network structure, rela-tional aspects, and network position. Inthis section, the importance of person-al demographic attributes such as orga-nizational affiliation, qualification,experience, and gender and theireffects on social network attributes arehighlighted.

First, gender has long been recog-nized as an important factor linked toindividual outcomes in the social sci-ences. Men and women often differ inthe nature of their interpersonalexchanges and in their opportunitiesfor social interaction (Dykstra, 1990).The differences in the interpersonalstyle of women compared with menhave stimulated a good deal of interestas a possible source of variation in theinterpersonal aspects of a variety ofdomains, including areas of knowl-edge-intensive and geographically dis-tributed work (Aries, 1996; Eagly &

Johnson, 1990). For example, womentend to reveal more information aboutthemselves in conversations (Dindia &Allen, 1992), are better facilitators andmoderators in communication thantheir male counterparts, and are moreintimate, warmer, and more engaged innonverbal communication than men(Hall, 1984). In the medical work prac-tice domain, for example, it is no differ-ent. Female physicians facilitate a moreopen and equal interaction withpatients as well as with colleagues, anda therapeutic milieu that is differentthan that of male physicians(Verbrugge & Steiner, 1981; Weisman &Teitelbaum, 1989). More recently, ameta-analytic review of 26 articles from1967 to 2001 accounted for physiciangender effects in medical communica-tion. It was concluded that female phy-sicians engage in significantly moreactive partnership behaviors, positivetalk, psychosocial counseling, psy-chosocial question-asking, and emo-tionally focused talk than male physi-cians (Roter, Hall, & Aoki, 2002). In lightof these arguments, it is anticipatedthat the network structure, position,and ties will be influenced by suchattributes in the process of professionalsocialization. Therefore, the followinghypotheses can be postulated:

H3a: There is a significant difference indegree centrality between gender-basedego networks in knowledge-intensivework.

H3b: There is a significant difference inconstraint indices between gender-based ego networks in knowledge-inten-sive work.

H3c: There is a significant difference inprofessional tie diversity (geographicallyand functionally) between gender-based ego networks in knowledge-intensive work.

Other valid demographic attributesthat are speculated to influence networkstructure and performance, irrespec-tive of gender, are education and pro-fessional qualifications of individuals

pmj402_20115.qxd 4/28/09 3:46 PM Page 42


in knowledge-intensive work. Inferringfrom previously discussed theoriesabout work structural diversity, ties,and performance (Cross & Cummings,2004; Cummings & Cross, 2003), andtechnological effects on performance(Aral, Brynjolfsson, & Alstyne, 2006), itis anticipated that individuals who areaccredited or affiliated with profession-al memberships related to their work orwho take on the responsibility of addi-tional functional tasks (such as hospitalappointments in primary care) willenjoy the advantage of greater diversityin professional network ties, a higherdegree of interaction, and a higher levelof performance as well. Therefore, thefollowing hypotheses are postulated:

H3d: There is a significant difference inego-network density between holdersand nonholders of professionalappointments in knowledge-intensivework.

H3e: There is a significant difference inego-network efficiency between holdersand nonholders of professionalappointments in knowledge-intensivework.

H3f: There is a significant difference indegree centrality between holders andnonholders of professional accredita-tions and memberships.

H3g: There is a significant difference infunctional tie diversity between holdersand nonholders of professional accredi-tations and memberships.

H3h: There is a significant difference inperformance attitude between holdersand nonholders of professional accredi-tations and memberships.

Toward a Social Networks–BasedModel for Performance Context of StudyIn order to validate the above hypothe-ses that form the conceptual and theoretical model for the study at anoperational level, the chosen context ofthe study is the rural general practi-tioners (GPs) of New South Wales,Australia. The Royal Australian College

of General Practitioners (RACGP) defines“general practice” as “the provision ofprimary continuing comprehensivewhole-patient medical care to individu-als, families, and their communities”(RACGP, 2004). Rural GPs are consid-ered knowledge-intensive workersbecause of the nature of their work,extensive medical expertise, highpatient-to-GP ratio, long work hours,usage of advanced medical technolo-gies, provision of diverse health careservices, and so on (Humphreys &Rolley, 1998). GPs working in ruralareas are geographically more occupa-tionally isolated from other practices.Furthermore, rural GPs often carry outprocedures in situations with limitedresources or personnel and are implic-itly required to adapt to the protocolsand codes of conduct of rural settings(Mellow, 2005). These problems andothers such as decreasing performanceas GPs age, lack of association withprofessional peers, obsolescence withmodern technology, and isolation fromcommunity not only hinder perfor-mance, but also make this study poten-tially interesting and practically useful(Choudhry, Fletcher, & Soumerai, 2005;Chung, Hossain, & Davis, 2005). It alsoprovides justification for the rural GPsas distributed knowledge-intensivework subjects for this study.

PerformancePerformance is essentially a multidi-mensional construct that varies in different contexts of work. Borman and Motowidlo (1993) conceptualizedperformance in terms of task-based performance and contextual-based per-formance. They defined task-based performance as “the proficiency withwhich job incumbents perform activi-ties that are formally recognized as partof their jobs, activities that contributeto the organization’s technical coreeither directly by implementing a partof its technological process or indirect-ly by providing it with needed materialsor services” (Borman & Motowidlo,1993, p. 75). Contextual performance,

by contrast, is defined as behavior thatsupports the broad organizational,social, and psychological environmentof the organization rather than theorganization’s technical core. In thisstudy, “performance” is about task-based activities that are core to a GP’spractice, such as the process of diagno-sis and prescription. In particular, itrelates to “the provision of primary con-tinuing comprehensive whole patientmedical care,” as per the definition ofgeneral practice.

Given that there is currently nogeneric performance index of GPs, aproxy to measure the effectiveness ofGPs in delivering medical care is toaccount for their behavior in actualpractice. Ideally, this would entailethnographic accounts of what wasactually observed in the consultationprocess and whether the practice canbe equated with components of “good-quality” patient care. However, thiswould most likely be an obtrusive andexpensive exercise time-wise andresource-wise. Therefore, an alternatemethod for studying behavior is to findout and empirically evaluate the self-perceived attitudes that the GP holdstoward the processes of medical care,as research in social psychology sug-gests a strong association between atti-tude and behavior.

Research in the area of social psy-chology suggests that the attitude that aperson holds toward any object is verylikely to be related to the overall patternof a person’s response to that object(Ajzen & Fishbein, 1980). Although therehave been previous studies showingonly weak correlations between meas-ures of attitudes and measures of behav-ior toward their objects (Breckler,Greenwald, & Pratkanis, 1989; Wicker,1969), Fishbein and Ajzen (1974) demon-strated that attitude and behavior arecorrelated given the following condi-tions: (1) when the observed behavior is judged to be relevant to the attitude,(2) when the attitude and behavior areobserved at comparable levels of speci-ficity, and (3) when mediation of the

pmj402_20115.qxd 4/28/09 3:46 PM Page 43



AP

ER

S

attitude-behavior relation by behavioralintentions is taken into account. Theirfindings were largely influential in thedevelopment of the theory of reasonedaction, or planned behavior, whichclaims that behavior is mediated bybehavioral intention, which in turn isinfluenced by attitude and social norms.Furthermore, Fazio (1986) demonstratedthat attitude and behavior are correlated(1) when the attitude is based on directexperience with the attitude object and(2) to the extent that the attitude is cog-nitively accessible.

Therefore, in work settings, atti-tudes about work practices are deter-mined by an individual’s past behaviorand how these behaviors come to beattributed to the environment or person.Using a social information-processingapproach to work attitudes and taskdesign, Salancik and Pfeffer (1978, p. 230) argue that “the process ofattributing attitudes from action is itselfaffected by (1) the individual’s commit-ment to the behavior, (2) the informa-tion about past behavior that is salientat the time the attitude is generated,and (3) social norms and expectationsthat affect what can be considered legit-imate or rational explanations for pastbehavior.” Plenty of evidence suggeststhat when individuals are committed toa situation, they tend to develop atti-tudes consistent with their commit-ment and their committing behavior(Igbaria & Tan, 1997; Kiesler, 1971).

In the case of general practice, GPs’practices are guided by physician char-ters, medical standards, and norms ofpractice. They are subject to norms ofdiagnosis and prescription, which arefairly standard in the consultationprocess. Moreover, the nature of theirtasks is also fairly specific to the extentthat their behavior affects the attitudeattribution process. The attitude for-mulated in turn is charged with emo-tion, which predisposes a class of medical actions toward consultationduring medical practice. Therefore,although behavior changes attitudes,attitudes are developed from justifying

past behavior. Given the above argu-ments, it is important to find out howGPs see their role in terms of theprocesses of care during the consulta-tion process; their perceptions are verylikely to influence their behavior in theconsultation, and they allow for com-parison with views held by other GPs(and patients, as well). Therefore, theargument acknowledges that althoughthe attitude of the GPs is not a sole orsufficient cause of behavior, it is a con-tributing cause. The theoretical modelfor the study is listed in Figure 4(Chung, 2008).

MethodologyA triangulation methodology was usedfor the study. First, qualitative interviewswere held with seven GPs, some ofwhom had dual job functions such asuniversity honorary associates and lec-turer. The qualitative interviews weresemistructured and were held for at least1 hour with each of the interviewees atappointed times over a 2-month period.The qualitative interviews were usefulfor exploring the validity and relevance of social network constructsused in the study. The interviews wereaudio-recorded, transcribed, and sum-marized as contact summary notes.Pattern-coding and memoing tech-niques were used to analyze the data.The analysis was used in conjunctionwith the existing literature to develop aconceptual model for the study.

An initial questionnaire survey wasdeveloped and pre-piloted among agroup of five students within theresearch laboratory. Ten copies of thesurvey were then sent out to rural GPs,with only three responding. With lowresponse rates, experts in the domain ofgeneral practice, including the formerpresident of a rural doctors’ associationin Australia, professor and head of dis-cipline of general practice in arenowned university, and rural GPs,were consulted about the survey instru-ment. Subsequently, the research designand theoretical constructs were furtherrefined. The experts also vetted the

instrument, which was then pretestedfor comprehension and ease of use. Thegeneral response from them was thatthe design of the network component inthe survey was visually complex andconfounding. As this not only detersresponse rates but also adds cognitiveload to survey completion, advice andsuggestions from the experts wereaccepted and the survey was modifiedaccordingly. The second version of thesurvey was designed for improved easeof comprehension and completion.Attribute items such as asking whetherthe GPs were trained overseas or locallywere included. Other items to deter-mine whether they were accredited withfellowships from the RACGP and theAustralian College of Rural and RemoteMedicine (ACRRM) were also includedin order to allow for cross-demographiccomparisons (although not shown inthis study). The survey was piloted to136 rural GPs practicing in two differentdivisions of rural general practice overthe period of December 2006 toFebruary 2007. Fifty-six GPs agreed to fillout the survey, thus achieving aresponse rate of about 41%. The surveywas mostly personally administered inorder to allow for capturing of surveyduration, respondent reaction, anderrors in the survey. Only one survey wasadministered over telephone, and fiveothers through postal mail.

After making cosmetic refinements,a final questionnaire was posted to1,488 GPs in the remaining 15 divisionsof rural general practice in New SouthWales. Telephone calls were made to thepractices 2 weeks in advance to gaugethe likelihood of survey response beforethe actual mail-out. Two weeks after themail-out, follow-up (reminder) callswere made to those who had not com-pleted the survey and who had previ-ously agreed to the survey. A total of 110rural GPs participated in the study,accounting for a response rate of 7.8%.

Egocentric Network MeasuresWe utilized the egocentric approach forcollecting network data because of

pmj402_20115.qxd 4/28/09 3:46 PM Page 44


its practicality and feasibility. In thisapproach, the actor of interest isreferred to as the “ego,” while the actorswhom the “ego” considers as being hisor her affiliates, advisors, friends, or rel-atives are known as “alters” (Scott,2000). Name generators are used inorder to elicit alters’ names. In ourstudy, we used the following name

generator to elicit names from a GP’sprofessional network:

By “professional network,” we meanprofessional people with whom youassociate, interact or work for theprovision of care to patients (e.g.,nurses, admin staff, specialists,pathologists, doctors). Looking backover the last 6 months, please iden-

tify people (up to a maximum of 15)who are important in providing youwith information or advice for pro-viding care to patients.

Name interpreter questions are alsocommonly asked to elicit some attrib-ute data about the alters and ties. In ourcase, we requested GPs to indicate theoccupational code (e.g., nurse, practice

Performance AttitudesDegreeCentrality

Efficiency

Constraint

Tie Strength

FunctionalDiversity

GeographicalDiversity

Structure

Position

Relations

H2bH1a

H1b

H2a

H2d

H2c

Aspects ofInterpersonal

Care

Aspects ofTechnical

Care

H3b

H3a

H3c

H3c

H3d

H3e

H3f

H3g

Gender

ProfessionalAppointments

ProfessionalAccreditations

Performance Attitudes

Performance Attitudes

Aspects ofInterpersonal

Care

Aspects ofTechnical

Care

DegreeCentrality

Efficiency

Constraint

Tie Strength

FunctionalDiversity

GeographicalDiversity

Structure

Position

Relations

Density

H3h

Figure 4: Theoretical model for the study (above, with personal attributes; below, without) (Chung, 2008).

pmj402_20115.qxd 4/28/09 3:46 PM Page 45



AP

ER

S

manager, specialist) as well as the geo-graphical location (e.g., same practice,other practice) of each alter. Thestrength of each tie, measured by “timethe person has been known,” “frequen-cy of interaction,” “type of relationship,”and “degree of closeness” (Marsden &Campbell, 1984), was also solicited.Attribute data about the frequency ofinteraction via e-mail, telephone, andvideo conferencing were also includedin the instrument to segregate face-to-face and technology media interactions.

To determine the relationshipbetween elicited alters in order to com-plete the network structure, we askedGPs to determine how the members oftheir professional network relate to eachother based on a 5-point degree of close-ness scale ranging from “especially close”to “do not know each other.” That is, foreach alter nominated, the GP woulddetermine a closeness scale for everyother alter. Although this approach hasbeen criticized in the past for its recallreliability and accuracy (Bernard,Killworth, Kronenfeld, & Sailer, 1985),later studies confirmed that people also remembered long-term or typicalpatterns of interaction with other peoplerather well (Freeman, Romney, &Freeman, 1987). Furthermore, the freerecall method elicits a richer data on the social networks of people, whereasthe fixed-choice method influences people to elicit accurate information onthe most important relationships (i.e.,strong ties; Hammer, 1984).

Efficiency and Constraint

Effective size is a measure of the num-ber of alters minus the average degreeof alters within the ego network, notcounting ties to the ego (Burt, 1992, p. 55).The effective size of an actor’s (ego) network is thus:

where i is the ego, actor j is a primarycontact, and actor q is also a primarycontact who has strong ties with theego i (represented by strong tie piq) and

ajc1� a

qpiq mjq d ,̊ q � i, j

actor j (represented by marginal tiemjq). Efficiency is measured by dividingthe effective size by the number ofalters in the ego’s network.

Ego constraint measures the oppor-tunities held back by the extent towhich the ego has invested time andenergy in relations with alters that leadback to a single contact (Burt, 1992, p. 55). In other words, it measures theextent to which the ego’s connectionsare to others who are connected to oneanother. Constraint on an actor’s net-work is defined as:

where i is the ego, actor j is a primarycontact, and actor q is also a primarycontact who has strong ties with the egoi (represented by piq) and actor j (repre-sented by pqj).

Degree Centrality

Degree centrality (CD) is measured asthe count of the ties (a) to the ego (pk)(Freeman, 1978). In graph theoreticalterms:

where a(pi, pk) � 1 if and only if they areconnected by a line, and 0 otherwise.

Tie Diversity

Diversity is measured using an entropy-based diversity index developed byTeachman (1980):

where if there are N possible states inwhich the system can be, Pi is the prob-ability that the system will be found instate i, the only exception being whenthe state is not represented, in whichcase the value is 0 (Ancona & Caldwell,1992). Tie diversity in this study ismeasured in terms of functional (occu-pational) and geographical diversity.Therefore, if a GP has ties to profession-

H � � as

i�1

pi 1lnPi 2

CD1pk 2 � an

i�1

a1pi, pk 2

apij � aq

piqpqj b2

,̊ q � i, j

als from diverse occupations, his or herfunctional diversity would be high.Similarly, if a GP has ties to profession-als who are from differing cities andtowns, his or her occupational diversitywould be considered high.

Performance MeasuresA validated and reliable questionnaire(Cockburn, Killer, Campbell, & Sanson-Fisher, 1987) for assessing GPs’ per-formance attitudes to medical care toascertain their perceived effectivenessof clinical and interpersonal care wasused in this study. In order to validatethe item sets to see if they could beclustered into few dimensions, a factoranalysis was conducted, first perform-ing the principal components analysiswithout rotation on the 17 items.Principal components analysis revealedthe presence of five factors with eigen-values greater than one. These factorsaccounted for 60.6% of the variance,which was considered adequate in pre-vious research (Hair, 1995). However,the first two factors captured muchmore of the variance (37.79%) than theother three factors. This was also clear-ly evidenced from the scree plot shownin Figure 5.

Using Catell’s (1966) scree test, itwas decided that two factors should beretained (note the clear break after thesecond component) for further inves-tigation. The second step involvedperforming the exact same steps butrotating the two factors to simpleorthogonal structure using directoblimin with Kaiser normalization. Thetwo-component solution explained atotal of 37.8% of the variance, with thefirst factor contributing 23.3% and thesecond factor contributing 14.5%.Finally, items with a high factor loadingof at least 0.40 on only one factor in thecomponent matrix were retained as list-ed in Table 1. One item (g15) that loadedrelatively low on both factors (lesserthan 0.4) was discarded. Factor 1 can bedescribed as aspects of interpersonalcare, and factor 2 as aspects of technicalcare; this is consistent with literature in

pmj402_20115.qxd 4/28/09 3:46 PM Page 46


quality of care in general practice(Brook, McGlynn, & Shekelle, 2000;Campbell, Roland, & Buetow, 2000).

A reliability analysis demonstratedgood internal validity and consistencyfor both factors 1 (interpersonal care)and 2 (technical care), with Cronbach’sa � 0.771 and 0.720, respectively. Thescores were summed up, forming acomposite score for technical andinterpersonal care.

Demographic ItemsDemographic details were also solicit-ed from GPs. GPs were asked to statethe number of years they have beenpracticing as GPs and the number ofyears they have practiced in the cur-rent practice, in order to ascertain theiryears of experience. This is an impor-tant variable, as research has shownthat age and clinical experience arecovariates of performance. In particular,the systematic review of the literatureon GP experience and performance by Choudhry et al. (2005) shows thatquality of care delivered declines asGPs age. Other items asked in the sur-vey were: year of graduation, medicalcollege, and country from which the

GP graduated, gender, number of GPsin the practice, whether they had hos-pital appointments, and if they weremembers of the Fellowship of theAustralian College of Rural and RemoteMedicine and the Fellowship of theRoyal Australian College of GeneralPractitioners. Inclusion of these itemsets and others in the instrument wereconfirmed and validated with GPs fromNew South Wales, including those whoheld senior positions (e.g., president ofthe Rural Doctor’s Network and pre-sident of the RACGP) in the field ofgeneral practice.

ResultsResults from the respondents surveyedindicate that the typical rural GP hasbeen in rural practice for 20.24 years,with 13.63 years in the current practice.Although there are solo practices (a one-doctor-only practice), the typical ruralGP, as indicated by the results, workswith at least four other GPs in the same practice. Furthermore, 90 (or81.8%) are male and 20 (or 18.2%) arefemale, and 85 (or 77.3%) have hospitalappointments. Table 2 also lists the

descriptive statistics for the variables:network structure, position, ties, andperformance.

Given that there were no clear out-liers in the data distribution of all vari-ables and that the histograms of thedependent variables were fairly normal,we examined Pearson’s product-momentcorrelations indices. In summary, thecorrelations are suggestive of the fol-lowing:1. Demographics: Years in rural practice

and years in current practice (whichaccount for experience) are stronglycorrelated positively (r � 0.673, p �

0.000). In particular, the former vari-able is negatively associated withgeographical diversity (r � �0.231, p � 0.05), while the latter is negativelycorrelated with functional diversity (r � �0.230, p � 0.05).

2. Network Structure: As expected,there is a strong negative correlationbetween ego-network density andefficiency (r � �0.979, p � 0.000) anda strong positive correlation betweenego-network density and constraint(r � 0.455, p � 0.000). There is a smallnegative correlation between densityand average tie strength as well (r �

0.199, p � 0.05). Interestingly, degreecentrality is positively associated withgeographical diversity (r � 0.207, p �

0.05), functional diversity (r � 0.301, p � 0.01), and interpersonal care (r � 0.258, p � 0.01).

3. Network Position: Constraint is nega-tively associated with geographicaldiversity (r � �0.286, p � 0.01), func-tional diversity (r � �0.285, p � 0.01),and interpersonal care (r � �0.217, p � 0.05).

4. Network Relation: There exists amedium positive correlation betweengeographical diversity and functionaldiversity (r � 0.422, p � 0.000).

5. Performance: While there exists apositive correlation between inter-personal care and technical care (r �

0.230, p � 0.05), there are no othervariables that bear any significantcorrelations on technical care. Theother variables that bear significant

4

2

Eig

enva

lue

Component Number

0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Figure 5: Scree plot of factor eigenvalues for performance attitudes items.

pmj402_20115.qxd 4/28/09 3:46 PM Page 47



AP

ER

S

correlations with interpersonal careare degree centrality (r � 0.258, p � 0.01) and constraint (r � �0.217,p � 0.05).

Hypothesis Testing

Efficiency and Performance

To test the first hypothesis (H1a), theindependent samples t-test was adopt-ed to test for the significant differencebetween two independent groups(high- and low-efficiency actors) on acontinuous measure of performanceattitude scores. A t-test allows us to findout if more efficient actors score higheron their attitudes to performance thanthose with lower efficiency. If the differ-ence between high- and low-efficiencyactors is statistically significant, then itis evidence that efficiency is associated

Table 1: Rotated pattern matrix for performance attitudes.

Item Component

No. Item Description 1 2

g8 Partnership with patient 0.813

g4 Counseling patients about personal problems helps them cope better 0.674

g9 Emotional support for patients 0.643

g6 Identifying modifiable risk factors is very important 0.633

g10 Be frank and open with patients 0.621

g3 Listening to patients’ worries is an important part of my role 0.561

g2 Ensuring that explanation of prescribed treatment is understood 0.504

g11 GPs are influential for lifestyle change 0.465

g16 Don’t help with psychological problems 0.723

g12 Treating physical disease is most important 0.676

g14 Patient involvement in treatment decisions 0.604

g13 Patients are likely to follow GP advice 0.598

g17 GP is responsible for medical problems 0.565

g5 My medical expertise is wasted because I see so many healthy patients 0.486

g7 Patients make convenience of me 0.444

g1 It is a waste of time trying to persuade patients to give up smoking 0.415

Note. Extraction Method: Principal Component Analysis. Rotation Method: Oblimin with Kaiser Normalization. Rotation converged in 11 iterations.

Table 2: Descriptive statistics.

Mean Median SD Min Max

DemographicsYears in rural practice 20.24 – 10.44 1 50Years in current practice 13.63 – 10.37 1 43No. of GPs in current practice 4.54 – 4.11 0 25

Network StructureDegree 8.77 8.0 3.94 1.00 15.0

Network PositionConstraint 0.402 0.35 0.197 0.078 1.00Efficiency 0.475 0.40 0.297 0.067 1.00

Relations (Ties)Geographic diversity 0.700 0.67 0.367 .000 1.71Functional diversity 0.799 0.86 0.447 .000 1.58

Performance (attitudes)Interpersonal care 47.45 48.00 6.25 13.00 56.00Technical care 39.07 40.00 7.05 19.00 53.00

pmj402_20115.qxd 4/28/09 3:46 PM Page 48


with performance attitudes in knowl-edge-intensive work. Whether the asso-ciation is positively or negatively relatedcan be evidenced in the direction of thedifference (i.e., which group is higher).

The division or cut-point of thehigh- and low-efficiency groups wasmade first by sorting the data based onthe efficiency index in ascending order.This is equivalent to ranking the casesof data based on the efficiency index.The median of the index was then chosen as the cut-point that allowed fordivision of the groups based on effi-ciency. The median for the efficiencyscore was 0.406 in this study. Therefore,GPs who have an efficiency score ofgreater than or equal to the median aregrouped as the “high-efficiency group,”and those GPs with efficiency scoreslower than the median are termed the“low-efficiency group.”

For interpersonal care, the two effi-ciency groups (high-efficiency and low-efficiency) have no statistically signifi-cant difference in the interpersonalattitude care scores for the high (M �

47.85, SD � 5.67, n� 54) and low (M �

47.14, SD � 6.83, n � 55) efficiencygroups, t (107) � �0.587, p � 0.559 (two-tailed). The magnitude of the differ-ences in the means (mean difference ��0.70640, 95% CI: �3.093 to 1.680) isquite small (eta squared � 0.003).1

Therefore, the null hypothesis statingefficiency of an individual’s networkposition is not associated with attitudesto interpersonal care in knowledge-intensive work cannot be rejected. Inother words, there is no associationbetween efficiency and attitudes to per-formance with respect to interpersonalcare. With respect to attitudes to tech-nical care aspects, the two efficiencygroups also show no significant differ-ence in the technical care attitude scoresfor the high- (M � 38.68, SD � 7.23, n � 47) and low- (M � 39.52, SD � 6.86,n � 62) efficiency groups, t(107) � 0.933,

p � 0.508 (two-tailed). The magnitudeof the differences in the means (meandifference � 0.89630, 95% CI: �1.779 to 3.572) is very small (eta squared �

0.004). Therefore, the null hypothesisthat the efficiency of an individual’snetwork position is not associated withtechnical care attitudes in knowledge-intensive work cannot be rejected.Consequently, there is no support forthe alternative hypothesis (H1a) thatefficiency is positively associated withattitudes to technical care.

Constraint and Performance

For H1b, the same principle using t-testwas applied as in testing for H1a. In thiscase, the median constraint score was0.355. GPs with constraint scoresgreater than or equal to this cut-pointwere termed a “high-constraint group”and those lesser than the median scorewere termed a “low-constraint group.”

In terms of attitudes to interperson-al care, the t-test reveals a significantdifference in the interpersonal care atti-tude scores of high-constraint groups(M � 45.69, SD � 7.18, n� 53) and low-constraint groups (M � 49.19, SD �

4.72, n � 56); t (107) � 3.021, p � 0.003(two-tailed). The magnitude of the differ-ence in the means (mean difference �3.49, 95% CI: 1.2024 to 5.7942) is mod-erate (eta squared � 0.079). The direc-tion of the difference shows that thelower-constraint group has a higherinterpersonal care mean of 49.1964 andthat the higher-constraint group has alower interpersonal care mean of45.6981 in attitudes to interpersonalcare. Further investigation from thecorrelation results in Table 3 shows asignificant negative correlation (r �

�0.217; p � 0.05) between constraintscores and attitudes to interpersonalcare. Consequently, there is no evi-dence to support the null hypothesisthat constraint in an individual’s net-work position is positively or not asso-ciated with attitudes to interpersonalcare. Therefore, there is sufficient evi-dence to support the hypothesis (H1b)stated in terms of attitudes to interper-

sonal care. For technical care attitudes,there is no significant difference in thescores of high-constraint groups (M �

38.02, SD � 6.97, n � 53) and low-constraint groups (M � 40.23, SD �

6.97, n� 56); t(107) � 1.657, p � 0.101(two-tailed). The magnitude (effectsize) of the differences in the means(mean difference � 2.2132, CI: �0.435to 4.86) is small (eta squared � 0.02).Therefore, the null hypothesis that con-straint in the individual’s network posi-tion is positively or not associated withtechnical care attitudes cannot be reject-ed. In other words, there is no supportfor the alternative H1b that constraint isnegatively associated with technical careattitudes and is thus rejected.

Strong Ties and Performance

Hypothesis 2a tests the positive associa-tion of strong ties with attitudes to inter-personal and technical care. As thedependent variables and the averagestrength of ties variable is reasonablynormally distributed, parametric tests tocompare groups (t-tests) may be con-ducted. In terms of hypothesis testing,the t-test tests the probability that thetwo sets of scores (strong tie and weaktie) came from the same population.This allows for the t-test to revealwhether there is a statistically significantdifference in the mean score for thestrong-tie group and the weak-tie group.

As stated earlier, the operationaliza-tion of tie strength is determined fromthe average of frequency of contact andcloseness measures of the “ego” GP to allother GPs within his or her network. Themaximum value for frequency of contacton a scale of “daily” (coded as 1 butreverse-coded for analysis as 5) to “lessoften” (coded as 5 but reverse-coded foranalysis as 1) is 5, which indicates fre-quent interaction and, intuitively, a veryclose tie. The maximum value for degreeof closeness on a scale of “especiallyclose” (coded as 1, but reverse-coded inanalysis as 4) to “distant” (coded as 4 butreverse-coded in analysis as 1) is 4,which indicates a stronger relationshipand a stronger tie. Therefore, in this

1Effect-size calculations are not shown. They are calculated

as .t2

t2 � 1N1 � N2 � 22

pmj402_20115.qxd 4/28/09 3:46 PM Page 49



AP

ER

S

Tabl

e 3:

Pear

son'

s pr

oduc

t mom

ent c

orre

latio

n.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

Year

s in

rura

l pra

ctic

e (1

)1

0.67

3 (*

*)�

0.04

4�

0.05

7�

0.16

40.

040

0.08

1�

0.09

4�

0.23

1 (*

)�

0.16

60.

003

�0.

140

Year

s in

cur

rent

0.

673(

**)

1�

0.03

80.

181

�0.

091

0.18

5�

0.14

6�

0.02

4�

0.13

7�

0.23

0 (*

)�

0.04

6�

0.05

7pr

actic

e (2

)

No o

f GPs

in c

urre

nt

�0.

044

�0.

038

10.

183

0.05

60.

130

�0.

145

0.00

1�

0.00

3�

0.04

00.

162

0.02

0pr

actic

e (3

)

Ego

dens

ity (4

)�

0.05

70.

181

0.18

31

0.03

60.

455

(**)

�0.

979

(**)

0.19

9 (*

)�

0.10

3�

0.01

6�

0.04

10.

064

Degr

ee (5

)�

0.16

4�

0.09

10.

056

0.03

61

�0.

720

(**)

�0.

180

�0.

196

(*)

0.20

7 (*

)0.

301

(**)

0.25

8 (*

*)0.

176

Cons

trai

nt (6

)0.

040

0.18

50.

130

0.45

5 (*

*)�

0.72

0 (*

*)1

�0.

296

(**)

0.26

5 (*

*)�

0.28

6 (*

*)�

0.28

5 (*

*)�

0.21

7 (*

)�

0.02

8

Effic

ienc

y (7

)0.

081

�0.

146

�0.

145

�0.

979

(**)

�0.

180

�0.

296

(**)

1�

0.16

00.

053

�0.

044

0.00

2�

0.08

3

Aver

age

tie s

tren

gth

(8)

�0.

094

�0.

024

0.00

10.

199

(*)

�0.

196

(*)

0.26

5 (*

*)�

0.16

01

�0.

184

0.13

9�

0.07

4�

0.11

8

Geog

raph

ical

div

ersi

ty (9

)�

0.23

1 (*

)�

0.13

7�

0.00

3�

0.10

30.

207

(*)

�0.

286

(**)

0.05

3�

0.18

41

0.42

2 (*

*)0.

026

�0.

080

Func

tiona

l div

ersi

ty (1

0)�

0.16

6�

0.23

0 (*

)�

0.04

0�

0.01

60.

301

(**)

�0.

285

(**)

�0.

044

0.13

90.

422

(**)

10.

126

0.06

3

Inte

rper

sona

l car

e (1

1)0.

003

�0.

046

0.16

2�

0.04

10.

258

(**)

�0.

217

(*)

0.00

2�

0.07

40.

026

0.12

61

0.23

0 (*

)

Tech

nica

l car

e (1

2)�

0.14

0�

0.05

70.

020

0.06

40.

176

�0.

028

�0.

083

�0.

118

�0.

080

0.06

30.

230

(*)

1

N11

011

011

010

910

910

910

911

010

910

911

011

0

*Cor

rela

tion

is s

igni

fican

t at t

he 0

.05

leve

l (tw

o-ta

iled)

.

**Co

rrel

atio

n is

sig

nific

ant a

t the

0.0

1 le

vel (

two-

taile

d).

pmj402_20115.qxd 4/28/09 3:46 PM Page 50


study, the theoretical maximum valuethat the “average tie strength” can haveis 4.5 ((5 � 4)/2). In order to distinguisha strong tie from a weak tie, a cut-pointneeds to be determined from this theo-retical maximum value by dividing it by2. Therefore, the cut-point is 4.5/2 �

2.25. Consequently, GPs who have anaverage strength of tie score that isgreater than or equal to 2.25 are grouped“strong ties,” and those lesser than 2.25are termed “weak ties.”

With respect to interpersonal care,there is no significant difference in theattitudes scores to interpersonal care forGPs with strong ties (M � 47.26, SD �

7.19, n � 54) and GPs with weak ties (M � 47.64, SD � 5.25, n � 56); t (108) �.320, p � 0.749 (two-tailed). The magni-tude of the differences in the means(mean difference � 0.38, 95% CI: �1.99to 2.76) is very small (eta squared �

0.0009). With regard to technical care,there is also no significant difference inthe attitude scores to technical care forGPs with strong ties (M � 19.45, SD �

5.81) and GPs with weak ties (M � 18.52,SD � 5.59); t(108) � �1.030, p � 0.305(two-tailed). The magnitude of the differ-ences in the means (mean difference ��1.3849, 95% CI: �4.049 to 1.279) is verysmall (eta squared � 0.009). Therefore,given the results, it is clear that there isnot sufficient evidence to reject the nullhypothesis that strong ties will not beassociated with individual perform-ance in knowledge-intensive work. Forthis reason, the alternative hypothesisH2a cannot be supported.

Degree Centrality and Performance

Again, the t-test was deployed to assessif there is a statistically significant dif-ference between the means of the per-formance attitude scores of those withhigh-degree centrality and those withlow-degree centrality. In order todichotomize groups into high- andlow-degree centrality, the mediandegree was chosen as the cut-point. Inthis study, the median degree centralityis 8. Therefore, those with a degree cen-trality lower than the median were

categorized as being in the “low-cen-trality group,” and those with degreecentrality equal to or greater than themedian were categorized as being inthe “high-centrality group.”

In terms of interpersonal care, thereis a significant difference in the atti-tudes scores to interpersonal care forGPs with high-degree centrality (M �

48.58, SD � 4.59, n � 64) and GPs withlow-degree centrality (M � 45.95, SD �

7.87, n � 45); t(107) � �2.189, p � 0.031(two-tailed). The magnitude of the differ-ences in the means (mean difference ��2.62, 95% CI: �4.99 to �0.247) is small(eta squared�0.04). Furthermore, inspe-cting the correlations, there exists avery significant positive correlation (r �

0.258; p � 0.01) between degree cen-trality scores and attitudes to interper-sonal care. Therefore, there is sufficientevidence to support the alternativehypothesis H2b in terms of attitudes to interpersonal care. With respect totechnical care, there is no significantdifference in the attitudes scores totechnical care for GPs with high-degreecentrality (M � 39.55, SD � 7.08, n � 64)and GPs with low-degree centrality (M �

38.6, SD � 6.99, n � 45); t(107) � �0.691,p � 0.491 (two-tailed). The magnitudeof the differences in the means (meandifference � �1.37, 95% CI: �3.66 to1.77) is very small (eta squared � 0.004).As such, the null hypothesis that thereis no association between degree cen-trality and interpersonal care cannot berejected. Therefore, there is sufficientevidence to reject the alternativehypothesis H2b in terms of attitudes totechnical care.

Geographical Tie Diversity andPerformance

Following the t-test method in the previ-ous tests, results revealed no significantdifference in the interpersonal care atti-tude scores of GPs with high geographi-cal diversity (M � 47.44, SD � 4.77, n �54) and GPs with low geographicaldiversity (M � 47.46, SD � 7.44, n � 56);t(108) � 0.017, p � 0.987 (two-tailed). The magnitude of the differences in the

means (mean difference � 0.01984, 95%CI: �2.354 to 2.394) was very small (etasquared � 0.00). Therefore, the nullhypothesis that geographical diversity isnot associated with attitudes to inter-personal care cannot be rejected. Assuch, there is no evidence to supportthe alternative hypothesis H2c, and itshould thus be rejected. Similarly, thereis also no significant difference in thetechnical care attitude scores of GPswith high geographical diversity (M �

38.35, SD � 6.54, n � 54) and GPs withlow geographical diversity (M � 39.77,SD � 7.5, n � 56); t(108) � 1.054, p �0.355. The magnitude of the differencesin the means (mean difference � 1.41,95% CI: �1.247 to 4.079) was very small(eta squared � 0.01). Therefore, the nullhypothesis cannot be rejected, and thealternative hypothesis that geographi-cal diversity is positively associated withattitudes to technical care should thusbe rejected. In conclusion, there is nosupport for H2c in terms of both inter-personal and technical care.

Functional Tie Diversity andPerformance

As discussed earlier, to test thishypothesis, the median (0.8676) waschosen as the cut-point. GPs, withfunctional diversity indices greaterthan equal to the median categorizedas a “high-functional diversity group”and those with indices lower than themedian value categorized as a “lowfunctional diversity group.” The t-testshows that for attitudes to interperson-al care, the two functional diversitygroups (high functional diversity andlow functional diversity) have no statis-tically significant difference in theinterpersonal attitude scores for thehigh functional diversity (M � 48.05,SD � 4.33, n � 54) and low functionaldiversity (M � 46.87, SD � 7.66, n � 56)groups. The probability value of the test(equal variances not assumed) is notless than or equal to 0.05, so the resultis not significant, t(108) � �0.999, p �0.320 (two-tailed), with the magnitudeof the means (mean difference � �1.18,

pmj402_20115.qxd 4/28/09 3:46 PM Page 51



AP

ER

S

95% CI: �3.54 to 1.18) being very small(eta squared � 0.009). In other words,there is no association between func-tional diversity and attitudes to per-formance with respect to interpersonalcare. In terms of attitudes to technicalcare, the two functional diversitygroups also show no significant differ-ence in the technical care attitudescores for the high (M � 39.05, SD �

7.03, n � 54) and low (M � 39.08, SD �

7.13, n � 56) groups; t(108) � 0.025, p� 0.980 (two-tailed). The size of the dif-ference (mean difference � 0.03373,95% CI: �2.643 to 2.71) is extremelysmall (eta squared � 0.000). In conclu-sion, there is no support for the alter-native hypothesis (H2d) that function-al diversity is positively associated withattitudes to technical care.

Gender Effects

The independent samples t-test wasmost suited to test hypotheses H3a toH3c, as it tells us whether there is a sta-tistically significant difference in themean scores for the two groups (i.e., ofmale and female GPs). Statistically, it istesting the probability that the two setsof scores (for males and females) camefrom the same population.

For H3a, t-test results suggest thatthere is a statistical difference in thedegree centrality scores for males (M �

8.40, SD � 3.88) and females (M � 10.4,SD � 3.88); t(107) � 2.075, p � 0.04(two-tailed). The magnitude of the dif-ferences in the means (mean difference� 1.99, 95% CI: 0.088 to 3.9) was small(eta squared � 0.03). Furthermore, it isevident that the mean degree centralityscores of females are higher than thoseof males. Therefore, while there is sup-port for H3a, one can also infer thatfemales have higher-degree centralityin their ego networks than males inknowledge-intensive work.

For H3b, there is a statistical differ-ence in the constraint scores for males(M � 0.4192, SD � 0.206) and females (M � 0.3283, SD � 0.1317); t(43.005) ��2.475, p � 0.017 (two-tailed). The mag-nitude of the differences in the means

(mean difference � �0.09086, 95% CI:�0.16489 to �0.01683) was near moder-ate (eta squared � 0.05). Furthermore,it is evident that the mean constraintscores of males are higher than those offemales. Therefore, although there issupport for H3b, one can also infer thatmales have higher constraint in theirego networks than females in knowl-edge-intensive work.

Regarding H3c, in terms of geo-graphical diversity, the results suggestthat there is a statistical difference in thegeographical diversity scores for males(M � 0.6613, SD � 0.376) and females (M� 0.8767, SD � 0.266); t(38.194) � 3.002,p � 0.005 (two-tailed). The magnitudeof the differences in the means (meandifference � 0.21538, 95% CI: 0.07019 to0.36058) was fairly moderate (etasquared � 0.07). Also, it is evident thatthe mean geographical diversity scoresof females are higher than those ofmales.

In terms of functional tie diversity, itappears that there also is a statisticaldifference in the functional diversityscores for males (M � 0.7347, SD � 0.446)and females (M � 1.0884, SD � 0.329);t(107) � 3.342, p � 0.001 (two-tailed).The magnitude of the differences in themeans (mean difference � 0.35375,95% CI: 0.14394 to 0.56357) was fairlymoderate (eta squared � 0.09). Further-more, it is evident that the mean func-tional diversity scores of females arehigher than those of males. Therefore,while there is strong support for H3c,one can also infer that females havehigher tie diversity in their ego net-works than their male counterparts inknowledge-intensive work.

Regarding H3d, it is clear that thereis a significant difference in the egonetwork density indices for holders ofhospital appointment (M �0.6557, SD �

0.328) and nonholders of hospital ap-pointment (M � 0.4674, SD � 0.377);t(107) � �2.426, p � 0.017 (two-tailed).The magnitude of the differences in themeans (mean difference � �0.188, 95%CI: �0.3420 to �0.0344) was small tonear moderate (eta squared � 0.05). In

particular, it also appears that thosewho hold professional appointmentsalso have a denser network than thosewho do not (by inspecting the means).Therefore, there is sufficient evidence toreject the null hypothesis in favor of H3d.

In terms of ego-network efficiency,it is clear that there is a significant dif-ference in the ego-network efficiencyindices for holders of hospital appoint-ment (M � 0.4327, SD � 0.283) andnonholders of hospital appointment (M � 0.6198, SD � 0.302); t(107) � 2.85,p � 0.005 (two-tailed). The magnitudeof the differences in the means (meandifference � �188, 95% CI: 0.0569 to0.3172) was moderate (eta squared �

0.07). In particular, it also appears thatthose who hold professional appoint-ments also have a less efficient networkthan those who do not (by inspectingthe means). Therefore, there is suffi-cient evidence to reject the nullhypothesis in favor of H3e.

With regard to H3f, fellowship accre-ditations are of two forms: Fellowship ofthe Royal Australian College of GeneralPractitioners (FRACGP) and Fellowshipof the Australian College of Rural andRemote Medicine (FACRRM). In terms ofthe FACRRM accreditation, there is a sig-nificant difference in the degree central-ity indices for holders of FACRRM (M �

10.55, SD � 4.13) and nonholders ofFACRRM (M � 8.125, SD � 3.69); t(107) ��2.787, p � 0.004 (two-tailed). The mag-nitude of the differences in the means(mean difference � �2.42, 95% CI:�4.065 to �0.78818) was near moderate(eta squared � 0.05). In particular, italso appears that those who hold theFACRRM accreditation also have higher-degree centrality in their ego networksthan those who do not (by inspectingthe means). Therefore, there is suffi-cient evidence to reject the nullhypothesis in favor of H3f in terms ofFACRRM.

In terms of H3g, there is a signifi-cant difference in the functional diver-sity indices for holders of FACRRM (M �

0.944, SD � 0.31342) and nonholders ofFACRRM (M � 0.7472, SD � 0.47791);

pmj402_20115.qxd 4/28/09 3:46 PM Page 52


t(107) � �2.49, p � 0.015 (two-tailed).The magnitude of the differences in themeans (mean difference � �0.19672,95% CI: �0.35408 to �0.03936) was nearmoderate (eta squared � 0.05). In par-ticular, it also appears that those whohold the FACRRM accreditation alsoexperience higher functional tie diver-sity in their professional networks thanthose who do not (by inspecting themeans). Therefore, there is sufficientevidence to reject the null hypothesis infavor of H3h in terms of FACRRM.

With respect to H3h: first, in termsof performance attitudes to interper-sonal care for FRACGP-accredited GPs,there is a significant difference in theinterpersonal care attitudes scores forholders of FRACGP (M � 49.068, SD �

4.353) and nonholders of FRACGP (M �

46.38, SD � 7.08); t(108) � �2.251, p � 0.026 (two-tailed). The magnitudeof the differences in the means (meandifference � �2.6893, 95% CI: �5.057to �321) was small (eta squared � 0.04).Second, in terms of performance atti-tudes to technical care for FRACGP-accredited GPs, there is a significantdifference in the technical care atti-tudes scores for holders of FRACGP (M � 41.34, SD � 6.80) and nonholdersof FRACGP (M � 37.56, SD � 6.85);t(106.331) � �3.469, p � 0.001 (two-tailed). The magnitude of the differ-ences in the means (mean difference ��6.6893, 95% CI: �10.512 to �2.866)was small (eta squared � 0.04).

Therefore, in terms of the associa-tion between FRACGP and perform-ance attitudes (for both technical andinterpersonal care), the evidence is suf-ficient to reject the null hypothesis andlend support to H4h. It is also clear thatthose who are professionally accreditedwith the FRACGP are more likely tohave a higher performance attitudescore than those who are not (byinspecting the means for both interper-sonal care and technical care attitudes).

ConclusionThe findings from the study providerich insights into the social network

model for explaining individual per-formance in geographically distributedknowledge-intensive work environ-ments. Although ego-network efficien-cy is claimed to be an influential predic-tor of performance (Aral et al., 2006),the results in this study indicate no sup-port for this particular variable in theresearch model. The correlation coeffi-cients, for instance, are indicative of thefact that there is no significant associa-tion between efficiency and perform-ance. This can be attributed to the factthat individuals (such as rural GPs)involved in knowledge-intensive workwithin dispersed communities andgroups are satisfied or are performingrelatively well with sources of informa-tion and advice based on their profes-sional network of peers and colleagues.It does not seem crucial for such knowl-edge workers to be efficient in terms ofobtaining information for the deliveryof quality care at the technical or interpersonal level. Unlike real estateprojects or corporate projects wherethe competition for information gainsis likely to provide competitive advan-tage so as to boost sales incentives,bonuses, and salary promotions (Burt,2007; Burt, Hogarth, & Michaud, 2000;Crowston, Sawyer, & Wigand, 2001),GPs in rural areas have no such motiva-tions, and the nature of their occupa-tion is noncompetitive.

Having said this, the study’s find-ings shift the focus toward the impor-tance of ego-network constraint. Anindividual’s professional network ishighly constrained to the extent thatthe individual seeks advice and infor-mation from peers and colleagues thatlead back to the same person. In thecontext of rural GPs, results show thatconstraint has a marginal detrimentaleffect on performance, not in terms oftechnical care but in terms of attitudesto interpersonal care. That is, the higherthe constraint for the individual GP, thelower the score for attitudes to interper-sonal care (r � �0.217, p � 0.05). Ahighly constrained professional net-work for a GP means that the GP seeks

advice and information from within hisor her professional network that leadsback to the same contact. This con-strains the GP from learning novelideas, interacting with a new anddiverse range of personnel in the relat-ed profession, who would contribute toa better level of interpersonal care.

Figure 6 illustrates a professionalnetwork of a rural GP. The rural GP isthe ego in the sociogram. The density ofthe overall network is 0.924 (whichindicates high cohesion and possiblyinformation redundancy). Despite thehigh density level, the ego’s constraintscore is 0.249 and is relatively uncon-strained compared with others in thenetwork. The size of the nodes alsoshows the relative constraint index (i.e.,larger nodes have a higher constraintindex). The ego GP has quite a largenumber of strong ties with others (col-ors of node depict occupation) in thenetwork, demonstrated by the thick-ness of lines (1 � distant, 5 � veryclose). Thicker lines denote strongerties. The low constraint score of the egoGP indicates his or her ability to seekadvice and information from nonre-dundant contacts.

With respect to performance interms of attitudes to interpersonal care,the finding of ego-network constraintbeing negatively associated with per-formance conforms with literature(Aral et al., 2006; Rosenthal, 1997).However, it was surprising and interest-ing to note that efficiency did not dis-play its hypothesized correlation withperformance in individual knowledge-intensive work. It can thus be speculat-ed that although the measures of ego-network position such as efficiencywere developed based on theories ofsocial structure and competition, itsdocumented effects toward perform-ance may not be clearly reflected orapplicable in occupational communi-ties of knowledge-intensive work whereindividuals are geographically dis-tributed in isolated settings. This natu-rally translates into an avenue for fur-ther research with respect to the effects

pmj402_20115.qxd 4/28/09 3:46 PM Page 53



AP

ER

S

of ego-network efficiency and perfor-mance in noncompetitive domains.

In terms of ego-network density,there was no association (negative orpositive) with performance. Degreecentrality, on the other hand, demon-strated significant positive associationbetween performance with respect toattitudes to interpersonal care, the cor-relation coefficient for the associationbeing r � 0.258, p � 0.01. Furthermore,the t-tests also demonstrated signifi-cant difference in the scores of attitudesto interpersonal care between twoindependent centrality groups (high-and low-centrality groups) of GPs. Inparticular, GPs with higher-degree cen-trality scored higher than those withlower-degree centrality in terms of atti-tude to interpersonal care. These find-

ings resonate with studies by Cross andCummings (2004), where significantsupport was found between degreecentrality (measured as an individual’sties to and from his or her department)and individual performance ratings inknowledge-intensive projects. Consensuscan, therefore, be reached regardingdegree centrality in that it representsthe extent of information activity,which in turn is conducive to perfor-mance (Freeman, 1978).

The ego-network ties hypotheseswere examined from the perspective ofstrength of relational ties and of theirdiversity. First, this research hypothe-sized that strong ties in an individual’sego network will be positively associat-ed with performance in knowledge-intensive work. Beginning with the

theory on the strength of ties (Granovetter,1973), arguments to formulate thehypothesis were put forward regardinghow strong ties connect individualswho work frequently with each other,how these individuals have greatermotivation to be of assistance, and howstrong ties are typically more availablethan weak ties (Granovetter, 1983).Furthermore, it was claimed that strongties generate trust that allows influx ofuseful knowledge (Levin & Cross, 2004;Reagans & McEvily, 2003) and that their“affect” level is conducive to perfor-mance (Krackhardt, 1992). Second, tiediversity in terms of geographical andfunctional (or occupational) diversitywas postulated to be positively associ-ated with performance. This hypothesiswas derived from previous research

Network Density � 0.924

GPPractice Manager

Legend

SpecialistNurseReceptionistEgo

0.249 0.256

0.2490.249

0.249

0.249

0.255

0.250

0.2520.249

0.250

0.252

0.252

0.252

0.249

0.249

Ego

Figure 6: A GP's professional network.

pmj402_20115.qxd 4/28/09 3:46 PM Page 54


primarily by Cummings (2004) andMonge et al. (1985), who found thatstructural diversity played an impor-tant role in obtaining new and novelinformation for performance. However,contrary to inferences from literature,findings from this study show no statis-tical support for the hypotheses thatstrong ties or tie diversity are positivelyassociated with individual perform-ance in knowledge-intensive work. Theresults are surprising, as they contra-dict past research, thus questioning thegeneralizability of the strong-tie and tiediversity hypotheses to the extent ofindividuals within geographically dis-tributed knowledge-intensive work. Inshort, this study argues that in ego networks of individuals in knowledge-intensive and geographically distributedwork, although demographic proper-ties (e.g., professional accreditation)are important, network structure (i.e.degree centrality) and network position(e.g., constraint) are equally potent pre-dictors of performance.

To summarize, empirical resultssuggest that network structure, posi-tion, and ties of knowledge workers playa crucial role in individual performance.In particular, degree centrality wasfound to be positively associated withperformance, while ego-network con-straint was found to be negatively corre-lated with performance. The resultsfrom this study resonate with findingsfrom past literature and extend tradi-tional theory of social networks andperformance within the micro-level to include geographically dispersedindividuals involved in knowledge-intensive work. For individuals in suchnoncompetitive settings, traditionalnetwork theories such as structural holestheory still apply. However, a key findingis that network structure is a muchmore potent predictor of performance,although network position is important.

Methodologically, the study con-tributes toward a triangulation approachthat utilizes both qualitative and quanti-tative methods for operationalizing thestudy. The quantitative method includes

a nontraditional “networks” method ofdata collection and analysis to serve as afine complement to traditional researchmethods in behavioral studies. The out-come is a valid and reliable survey instru-ment that allows the collection of bothindividual attributes and social networkdata. The instrument is theoreticallydriven, practically feasible to implement,time-efficient, and easily replicable forother similar studies.

At the domain level, key findingsfrom this study contradict previous lit-erature that suggests that professionalsin projects and occupational commu-nities such as GPs decline in perform-ance as they age. In fact, findings fromthis study suggest that age and experi-ence do not affect for performance.Furthermore, degree centrality is alsopositively associated with professionalaccreditations, such as the Fellowshipof the Royal Australian College ofGeneral Practitioners, which is con-ducive to performance in terms of atti-tudes toward interpersonal and techni-cal care. The contextual implicationfrom the quantitative and qualitativeevidence of this study is that whenstrategies for enhancing performance,such as improving attitudes to qualityof care at the technical and interper-sonal level, are contemplated, theimportance of social structure, posi-tion, and relationships in the profes-sional network needs to be consideredcarefully as part of the overall individ-ual-level, project-level, and organiza-tion-level goals. ■

ReferencesAhuja, M. K., Galletta, D. F., & Carley,K. M. (2003). Individual centrality andperformance in R&D groups: Anempirical study. Management Science,49(1), 21–38.

Ajzen, I., & Fishbein, M. (1980).Understanding attitudes and predictingsocial behavior. Englewood Cliffs, NJ:Prentice Hall.

Ancona, D. G., & Caldwell, D. F. (1992).Demography and design: Predictors of

new product team performance.Organization Science, 3, 321–341.

Aral, S., Brynjolfsson, E., & Alstyne,M. V. (2006, December 10–13).Information, technology and informa-tion worker productivity: Task levelevidence. Paper presented at theInternational Conference onInformation Systems, Milwaukee,Wisconsin, USA.

Aries, E. (1996). Men and women ininteraction: Reconsidering the differ-ences. New York: Oxford UniversityPress.

Bavelas, A. (1950). Communicationpatterns in task-oriented groups.Journal of Acoustical Society ofAmerica, 22, 725–730.

Bernard, H. R., Killworth, P. D.,Kronenfeld, D., & Sailer, L. (1985).On the validity of retrospective data.Annual Review of Anthropology, 13,495–517.

Bonacich, P. (1991). Simultaneousgroup and individual centralities.Social Networks, 13(2), 155–168.

Borgatti, S. (1996). Communicationstructure and its effects on task per-formance. Retrieved September 13,2006, from http://www.analytictech.com/mb021/commstruc.htm

Borgatti, S. P., Jones, C., & Everett, M.(1998). Network measures of socialcapital. Connections, 21(2), 27–36.

Borman, W. C., & Motowidlo, S. J.(1993). Expanding the criteriondomain to include elements of contex-tual performance. In N. Schmitt & W. C. Borman (Eds.), Personnel selec-tion in organizations (pp. 71–98). San Francisco: Jossey Bass.

Brass, D. J. (1981). Structural relation-ships, job characteristics and workersatisfaction and performance. Admini-strative Science Quarterly, 26, 331–348.

Brass, D. J. (1985). Technology and thestructuring of jobs: Employee satisfac-tion, performance and influence.Organizational Behaviour and HumanDecision Making, 35, 216–240.

Breckler, S. J., Greenwald, A. G., &Pratkanis, A. R. (1989). Attitude structure

pmj402_20115.qxd 4/28/09 3:46 PM Page 55



AP

ER

S

and function. Hillsdale, NJ: Erlbaum.

Brook, R. H., McGlynn, E. A., &Shekelle, P. G. (2000). Defining andmeasuring quality of care: A perspec-tive from US researchers. InternationalJournal for Quality in Health Care, 12,281–285.

Burgess, R. L. (1968). Communicationnetworks and behavioral conse-quences. Human Relations, 22,137–159.

Burt, R. S. (1992). Structural holes: Thesocial structure of competition. Boston:Harvard University Press.

Burt, R. S. (2004). Structural holes andgood ideas. American Journal ofSociology, 110, 349–399.

Burt, R. S. (2005). Brokerage and clo-sure: An introduction to social capital.New York: Oxford University Press.

Burt, R. S. (2007). Secondhand broker-age: Evidence on the importance oflocal structure for managers, bankers,and analysts. Academy of ManagementJournal, 50(1), 119–148.

Burt, R. S., Hogarth, R. M., & Michaud,C. (2000). The social capital of Frenchand American managers. OrganizationScience, 11(2), 123–147.

Campbell, S. M., Roland, M. O., &Buetow, S. A. (2000). Defining qualityof care. Social Science & Medicine, 51,1611–1625.

Catell, R. B. (1966). The screen test fornumber of factors. MultivariateBehavioral Research, 1, 245–276.

Choudhry, N. K., Fletcher, R. H., &Soumerai, S. B. (2005). Systematicreview: The relationship between clini-cal experience and quality of healthcare. Annals of Internal Medicine, 142,260–283.

Chung, K. S. K. (2008). Understandingattitudes towards performance inknowledge-intensive work: the influ-ence of social networks and ICT use.Unpublished doctoral dissertation,University of Sydney, Sydney, Australia.

Chung, K. S. K., Hossain, L., & Davis, J.(2005, October 20–21). Social networks

and ICT correlates to individual per-formance. Paper presented at theApplications of Social Network Analysis2005, University of Zurich, Switzerland.

Cockburn, J., Killer, D., Campbell, E., &Sanson-Fisher, R. W. (1987). Measuringgeneral practitioners’ attitudes towardsmedical care. Family Practice, 4(3),192–199.

Coleman, J. S. (1988). Social capital inthe creation of human capital.American Journal of Sociology, 94,S95–S120.

Coleman, J. S., Katz, E., & Menzel, H.(1957). The diffusion of an innovationamong physicians. Sociometry, 20,253–270.

Cross, R., & Cummings, J. N. (2004).Tie and network correlates of individualperformance in knowledge-intensivework. Academy of ManagementJournal, 47, 928–937.

Crowston, K., Sawyer, S., & Wigand, R.(2001). Investigating the interplaybetween structure and informationand communications technology inthe real estate industry. InformationTechnology & People, 14(2), 163–183.

Cummings, J. N. (2004). Work groups,structural diversity, and knowledgesharing in a global organization.Management Science, 50, 352–364.

Cummings, J. N., & Cross, R. (2003).Structural properties of work groupsand their consequences for perform-ance. Social Networks, 25(3), 197–210.

Dindia, K., & Allen, M. (1992). Sex differences in self-disclosure: A metaanalysis. Psychology Bulletin, 112,106–124.

Dvir, D., Sadeh, A., & Malach-Pines, A.(2006). Project and project managers:The relationship between project man-agers’ personality, project types, andproject success. Project ManagementJournal, 37(5), 36–48.

Dykstra, P. A. (1990). Disentanglingdirect and indirect gender effects onthe supportive network. In C. P. Kees,M. Knipscheer, & T. C. Antonucci(Eds.), Social network research

(pp. 55–65). Amsterdam: Swets andZeitlinger.

Eagly, A. H., & Johnson, B. T. (1990).Gender and leadership style: A meta-analysis. Psychology Bulletin, 108,233–256.

Faraj, S., & Sproull, L. (2000).Coordinating expertise in softwaredevelopment teams. ManagementScience, 46, 1554–1568.

Faust, K. (1997). Centrality in affilia-tion networks. Social Networks, 19(2),157–191.

Fazio, R. H. (1986). How do attitudesguide behavior. In R. M. Sorrentino &E. T. Higgins (Eds.), The handbook ofmotivation and cognition: Foundationsof social behavior (pp. 204–243). NewYork: Guilford Press.

Fishbein, M., & Ajzen, I. (1974).Attitudes toward objects as predictorsof single and multiple behavioral crite-ria. Psychological Review, 81, 59–74.

Freeman, L. C. (1978). Centrality insocial networks: Conceptual clarifica-tion. Social Networks, 1, 215–239.

Freeman, L. C., Roeder, D., &Mulholland, R. R. (1979). Centrality insocial networks: ii. Experimentalresults. Social Networks, 2, 119–141.

Freeman, L. C., Romney, A. K., &Freeman, S. C. (1987). Cognitive struc-ture and informant accuracy. AmericanAnthropologist, 89, 311–325.

Gabbay, S. M., & Leenders, R. T. A. J.(2001). Social capital of organizations.Amsterdam and London: JAI.

Goodhue, D. L., & Thompson, R. L.(1995). Task-technology fit and indi-vidual performance. MIS Quarterly, 19,213–236.

Granovetter, M. (1973). The strength ofweak ties. American Journal ofSociology, 78, 1360–1380.

Granovetter, M. (1983). The strength ofweak ties: A network theory revisited.Sociological Theory, 1, 201–233.

Granovetter, M. (1985). Economicaction and social structure: The prob-lem of embeddedness. AmericanJournal of Sociology, 91, 481–510.

pmj402_20115.qxd 4/28/09 3:46 PM Page 56


Grinter, R. E., Herbsleb, J. D., & Perry,D. E. (1999). The geography of coordi-nation: Dealing with distance in R&Dwork. Paper presented at the Group 99Conference, Phoenix, Arizona.

Guetzkow, H., & Dill, W. R. (1957).Factors in the organizational develop-ment of task oriented groups.Sociometry, 20, 175–204.

Guetzkow, H., & Simon, H. (1955). Theimpact of certain communication netsupon organization and performance intask-oriented groups. ManagementScience, 1(3/4), 233–250.

Gurevitch, M. (1961). The social struc-ture of acquaintanceship networks.Boston: Massachusetts Institute ofTechnology Press.

Hair, J. F. (1995). Multivariate dataanalysis with readings (4th ed.).Englewood Cliffs, NJ: Prentice Hall.

Hall, J. A. (1984). Nonverbal sex differ-ences: Communication accuracy andexpressive style. Baltimore, MD: JohnsHopkins University Press.

Hammer, M. (1984). Explorations into the meaning of social networkinterview data. Social Networks, 6,341–371.

Hanneman, R. A. (2001). Introductionto social network methods. RetrievedAugust 8, 2004, from http://faculty.ucr.edu/~hanneman/SOC157/NET-TEXT.PDF

Hansen, M. T. (1999). The search-transfer problem: The role of weak tiesin sharing knowledge across organiza-tion subunits. Administrative ScienceQuarterly, 44(1), 82–111.

Hossain, L., Wu, A., & Chung, K. S. K.(2006). Actor centrality correlates toproject based coordination. Paper presented at the 2006 20th Anniver-sary Conference on ComputerSupported Cooperative Work, Banff,Alberta, Canada.

Humphreys, J. S., & Rolley, F. (1998). Amodified framework for rural generalpractice: The importance of recruit-ment and retention. Social ScienceMedicine, 46, 939–949.

Igbaria, M., & Tan, M. (1997). The con-sequences of information technologyacceptance on subsequent individualperformance. Information &Management, 32(3), 113–121.

Kiesler, C. A. (1971). The psychology ofcommitment: Experiments linkingbehaviour to belief. New York:Academic Press.

Knoke, S., & Kulinski, J. H. (1992).Network analysis. Newbury Park, CA:Sage.

Krackhardt, D. (1992). The strength ofstrong ties: The importance of philos inorganizations. In N. Nohria & R. Eccles (Eds.), Networks and organi-zations: Structures, form and action(pp. 216–239). Boston: HarvardBusiness School Press.

Leavitt, H. (1949). Some effects of cer-tain communication patterns on groupperformance. Boston: MassachusettsInstitute of Technology Press.

Leavitt, H. (1951). Some effects of cer-tain communication patterns on groupperformance. Journal of Abnormal andSocial Psychology, 46, 38–50.

Levin, D. Z., & Cross, R. (2004). Thestrength of weak ties you can trust: Themediating role of trust in effectiveknowledge transfer. ManagementScience, 50, 1477–1494.

Marsden, P., & Campbell, K. E. (1984).Measuring tie strength. Social Forces,63, 482–501.

Mehra, A., Kilduff, M., & Brass,D. (2001). The social networks of highand low-self monitors: Implications for workplace performance. Admini-strative Science Quarterly, 46, 121–146.

Mellow, M. (2005). The work of ruralprofessionals: Doing theGemeinschaft-Gesellschaft Gavotte.Journal of Rural Sociology, 70(1), 50–69.

Monge, P., Rothman, L., Eisenberg, E.,Miller, K., & Kirste, K. (1985). Thedynamics of organizational proximity.Management Science, 31, 1129–1141.

Mullen, B., Johnson, C., & Salas, E.(1991). Effects of communication net-work structure: Components of posi-

tional centrality. Social Networks, 13(2),169–185.

Muzio, E., Fisher, D. J., Thomas, E. R., &Peters, V. (2007). Soft skills quantifica-tion (SSQ) for project manager compe-tencies. Project Management Journal,38(2), 30–38.

Oh, H., Chung, M.-H., & Labianca, G.(2004). Group social capital and groupeffectiveness: The role of informalsocializing ties. Academy ofManagement Journal, 47, 860–875.

Peterson, T. M. (2007). Motivation:How to increase project team perform-ance. Project Management Journal,38(4), 60–69.

Pfeffer, J. (1980). A partial test of thesocial information processing model ofjob attitudes. Human Relations, 33,457–476.

Podolny, J., & Baron, J. (1997).Resources and relationships: Socialnetworks and mobility in the work-place. American Sociological Review,62, 673–693.

Reagans, R., & McEvily, B. (2003).Network structure and knowledgetransfer: The effects of cohesion andrange. Administrative ScienceQuarterly, 48, 240–267.

Reagans, R., & Zuckerman, E. W.(2001). Networks, diversity, and pro-ductivity: The social capital of corpo-rate R&D teams. Organization Science,12, 502–517.

Rosenthal, E. (1997). Social networksand team performance. TeamPerformance Management, 3, 288–294.

Roter, D. L., Hall, J. A., & Aoki, Y. (2002).Physician gender effects in medicalcommunication: A meta-analyticreview. JAMA, 288, 756–764.

Royal Australian College of GeneralPractitioners (RACGP). (2004). Ruralgeneral practice—Definition. RetrievedNovember 14, 2004, from http://www.racgp.org.au/document.asp?id�15301

Salancik, G. R., & Pfeffer, J. (1978). Asocial information processingapproach to job attitudes and task

pmj402_20115.qxd 4/28/09 3:46 PM Page 57



AP

ER

S

design. Administrative ScienceQuarterly, 23, 224–253.

Scott, J. (2000). Social network analy-sis: A handbook. London: SAGE.

Sparrowe, R. T., Liden, R. C., Wayne,S. J., & Kraimer, M. L. (2001). Socialnetworks and the performance of individuals and groups. Academy ofManagement Journal, 44, 316–326.

Teachman, J. D. (1980). Analysis ofpopulation diversity: Measures of qual-itative variation. Sociological MethodsResearch, 8, 341–362.

Thamhain, H. J. (2004). Team leader-ship effectiveness in technology-based project environments. ProjectManagement Journal, 36(1), 49–61.

Turner, J. R., & Müller, R. (2005). Theproject manager’s leadership styles asa success factor on projects: A litera-ture review. Project ManagementJournal, 36(1), 49–61.

Verbrugge, L. M., & Steiner, R. P.(1981). Physician treatment of men and women patients: Sex bias or appropriate care? Medical Care, 19,609–632.

Weisman, C. S., & Teitelbaum, M. A.(1989). Women and health care com-munication. Patient Education andCounseling, 13, 183–199.

Wicker, A. W. (1969). Attitudes versusactions: The relation of verbal andovert behavioral reponses to attitudeobjects. Journal of Social Issues, 25,41–78.

Kon Shing Kenneth Chung is a lecturer at theProject Management Graduate Programme atthe University of Sydney. He holds a PhD fromthe School of Information Technologies atSydney University and holds a BS with first-class honors in information systems and abachelor’s of commerce degree majoring inaccounting and information systems. His PhDresearch received the RailCorp Prize for“Addressing Society’s Needs in Research ProjectWork” at the 2005 Research Conversazionesheld at the University of Sydney. Prior to com-mencing his doctoral study, he was a systemsintegrator at Protocom Technologies and theTelstra Messaging (Text & SMS) Group. His workspanned several large-scale information tech-nology projects that involved development ofintegration and testing processes for mobileSMS text software from development to rollout in production environments. Prior toTelstra, he was online ad campaign coordinatorfor DoubleClick Australia (now acquired byGoogle). He currently enjoys teaching projectmanagement and research in the area of social networks and its influence on social out-comes such as coordination, performance, andinnovation. On a broader level, his currentresearch interest lies in the triangulation ofinterdisciplinary theories and methods fromsocial networks, sociology, information sys-tems, and management science. His recentwork has been published in top-tier conferences such as Computer-Supported Cooperative Work, European Conference onInformation Systems, Human-ComputerInteraction, and SIG-Management of Information Systems.

Liaquat Hossain is the director of the ProjectManagement Graduate Programme at theUniversity of Sydney. He was a postdoctoral candi-date at the Massachusetts Institute of TechnologyCenter for Technology, Policy, and IndustrialDevelopment during 1997. He completed his PhD ininformation and communications technology fromthe University of Wollongong during 1997. He com-pleted his MSc in computer and engineering man-agement in 1995 and bachelor’s of businessadministration in 1993 from Assumption Univer-sity. His work aims to explore the effects of differenttypes of social network structures and patterns ofinformation technology use on coordination in adynamic and complex environment. The primaryfocus of his research is in the area of networkanalysis of organizational and social systems. Heapproaches this using social networks theory andanalytical methods and applies theories and meth-ods from sociology and social anthropology tostudy coordination problem in a dynamic, complex,and distributed environment. He further appliesnetwork-based theories and methods to explorethe phenomenon of globally distributed work-groups (referred to as outsourcing in businessmanagement literature) and its management chal-lenges. Overall, he is interested in exploring theeffects of different types of social network struc-tures on coordination and organizational perform-ance from a theoretical and an applied perspective.In his research, he uses methods and analyticaltechniques from mathematical sociology (i.e.,social networks analysis), social anthropology(i.e., interview and field studies), and computer sci-ence (i.e., information visualization, graph theoret-ic approaches and data-mining techniques such asclustering) to explore coordination problems in adynamic, distributed, and complex setting.

pmj402_20115.qxd 4/28/09 3:46 PM Page 58

PA

PE

RS

June 2009 ■ Project Management Journal ■ DOI: 10.1002/pmj 59■ Project Management Journal ■ DOI: 10.1002/pmj 59

INTRODUCTION ■

What knowledge and skills are essential for a competent projectmanager in the present-day context? To address this question,one can start by looking at two well-recognized definitions ofproject management. On the one hand, the Project Management

Institute (PMI) defines project management as “the application of knowl-edge, skills, tools, and techniques to project activities to meet projectrequirements” (PMI, 2004). On the other hand, the International ProjectManagement Association (IPMA) states: “Project Management (PM) is theplanning, organising, monitoring, and controlling of all aspects of a projectand the management and leadership of all involved to achieve the proj-ect objectives safely and within agreed criteria for time, cost, scope, and per-formance/quality. It is the totality of coordination and leadership tasks,organisation, techniques, and measures for a project. It is crucial to optimisethe parameters of time, cost and risk with other requirements and to organ-ise the project accordingly” (IPMA, 2006). Although these are useful defini-tions, they are rather wide and therefore do not give a comprehensiveanswer to the question of what knowledge and skills must be mastered bythe project manager in the future in order to excel—an issue that is of impor-tance not only for project managers, project sponsors, clients, customers,contractors, and other project stakeholders, but especially for everyone whois training future project managers.

The project management profession has in the past strongly emphasizedtechnically supported methods of planning and execution as a core compe-tence, and continues to do so today. However, while project managementtoday remains firmly focused on this traditional “objective” or “hard” per-spective, there seems now to be an increasing focus on the more “subjective”and “soft” factors—leadership, motivation, group dynamics, interpersonalcommunication, culture, and ethics—that could be regarded as essential toall professional endeavors.

In light of this, we see that the PMI and IPMA definitions of project man-agement do not provide a comprehensive or all-inclusive description ofproject management as such, and that they do not take into considerationthe development that project management has undergone and the extent towhich it has changed since it was first formally defined in the middle of thetwentieth century. This development has, however, been discussed in anumber of textbooks and papers.

Kerzner (2006) gives an overview of project management from 1945 to 2006and highlights the development, over three different periods, of non-project-driven enterprises into hybrid enterprises that are primarily production-driven but also include numerous projects. The present-day situation andthe future will, according to Kerzner, focus on project-driven enterprises,where the project manager has profit-and-loss responsibilities and where

Contemporary Knowledge and SkillRequirements in Project ManagementHelgi Thor Ingason, University of Iceland, Reykjavik, IcelandHaukur Ingi Jónasson, University of Iceland, Reykjavik, Iceland

ABSTRACT ■

This article provides an overview of some recentdevelopments within the field of project manage-ment and presents metrics used to assess current trends within the discipline in order toidentify its future direction. Graphical tools areintroduced to measure the span and areas ofemphasis within project management, such asradar diagrams and the recently introducedInternational Project Management Association(IPMA) “Eye of Competence.” The researchshows that the discipline seems to focusincreasingly on interpersonal competences,relationship management, resource manage-ment, and strategic alignment. The results areevaluated and discussed, and the article con-cludes with some ideas about the future devel-opment of project management.

KEYWORDS: metrics; eye of competence;categories of project management





DOI: 10.1002/pmj.20122

pmj402_20122.qxd 4/28/09 5:16 PM Page 59


Contemporary Knowledge and Skill Requirements in Project ManagementP

AP

ER

S

project management is a recognizedprofession.

Milton D. Rosenau Jr.’s textbook onproject management, Successful ProjectManagement: A Step-by-Step Approach(Githens & Rosenau, 2005), is a well-established work in its field. In connec-tion with the release of the fourth editionof the book, an interview with theauthor was published in C2M Journalunder the title “What’s New in ProjectManagement?” (Scanlan, 2006). Askedabout what changes had actually takenplace since the release of the third edi-tion in 1998, Rosenau replied that little,in fact, had changed in project manage-ment as such. He observed, however,that project management as a profes-sion has undergone tremendousgrowth in the past few years, as can beseen, for example, in the increasednumber of project management profes-sional (PMP®) credential holders certi-fied by PMI, rising from 10,000 in 1998to 100,000 in 2004.

Jugdev and Müller (2005) discussthe way our perception of project suc-cess has changed over the past 40 years,and they retrospectively identify threedistinct periods: The first (1960s–1980s)was the “project implementation andhandover period”; the second (1980s–1990s) was the “project critical successfactor (CSF) list period”; and the third(1990s–present) is the “CSF frameworkperiod.” They suggest that the next onewill be the “strategic project manage-ment period.”

The applications of project man-agement methodology seem, therefore,to be expanding. It is increasingly beingused as a general management tool, as suggested by Kerzner (2006) and assupported in the findings of Jugdev andMüller (2005), and is being used by anincreasing number of people in all sec-tors of business, as pointed out byRosenau (Scanlan, 2006). This means,though, that project management, if itis to be viewed as a general toolbox forthe manager, must be well equippedenough to be applicable in all businesssectors and must include a wider range

of dynamic management tools than itdid in the earlier days.

A recent publication, Competenciesin Project Management, NationalCompetence Baseline for Scandinavia(Fangel, 2005), points in this direction;according to its author, the competen-cies of modern project managers arenow to be based on the intertwining ofleadership behavior, project manage-ment experience, and competence intraditional project management meth-ods. This new way of thinking is sup-ported by the IPMA CompetenceBaseline v. 3 (2006), issued in June 2006,which illustrates the necessary compe-tences of the modern project managerby the innovative “Eye of Competence”and where contextual, behavioral, andtechnical competences are all repre-sented. This diagram has been an eye-opener for the field.

Other standards, such as A Guide tothe Project Management Body ofKnowledge (PMBOK ® Guide—Thirdedition) (PMI, 2004) and ISO10006:2003(International Standards Organization,2003), do certainly define project man-agement in terms of a coherent body ofknowledge; however, they seem toreflect somewhat outdated viewpoints.This is undesirable, as project manage-ment as a discipline is, and should be,dynamic and in a state of incessantdevelopment. For this reason also, thequestion of what knowledge and skillswill be needed in the future for the proj-ect manager to excel can probablynever be fully answered. The history ofproject management points to an ongo-ing maturation process, requiring ouranswers to this important question tobe modified from one period to thenext. The objective of the present arti-cle, however, is to answer this questionfor the current period.

Previous Research and ThisStudyMany researchers have attempted todefine measurements to identify themost crucial elements of project man-agement and have done so by analyzing

various sources of information. Zobeland Wearne (2000) analyzed the relativefrequency of attention to different proj-ect management body of knowledgetopics in some of the papers presentedat PMI and IPMA seminars, symposia,and congresses during the period1996–1998. A total of 44 important topics were identified and used as a ref-erence guide. Urli and Urli (2000) useda “scientometric method” of text inorder to identify tendencies in the asso-ciation of keywords in the electronicdatabase ABI-INFORM in all paperspublished between 1987 and 1996 thatwere considered relevant to projectmanagement. The method, which wasnot based on any previous classifica-tion, was meant to illustrate what the presenters themselves deemed themost significant elements.

In a paper by Crawford, Pollack, andEngland (2006), the highlighted trendswithin project management literaturein the period 1994–2003 were identifiedbased on an analysis of articles in theInternational Journal of Project Manage-ment and in the Project ManagementJournal. Their objective was to providean overview of how the discipline mightbe changing by using as a basis foranalysis a list of 48 project managementtopics covering 18 main topic cate-gories. These categories and topics areshown in Table 1.

Following in the footsteps of thesestudies, our main purpose here, howev-er, is to explore the use of a new metricsthat can be used to assess the currentstatus of the discipline of project man-agement. We have, therefore, developedsome graphical representational toolsfor this purpose. We believe that thesetools can be applied to gauge differentinformation sources in order to providea more holistic view of project manage-ment as a discipline in both the presentand the future. One reason for the study(other than to contribute something ofinterest for the Project ManagementJournal and our interested colleagues)was to create a comprehensive over-view of information in order to be

pmj402_20122.qxd 4/28/09 5:16 PM Page 60


better equipped to continue our devel-opment of a new graduate program inproject management (Master of ProjectManagement) at the Faculty of Engi-neering at the University of Iceland(further information can be obtained atwww.mpm.is). The graphical tools thatwe developed are principally used toassess the current status of the disci-pline of project management in termsof education and demands for compe-tence in the year 2008, by analyzing thefollowing three sources of information:• selected textbooks on the subject

published in or after 2003;• papers published in the Project Mana-

gement Journal and the InternationalJournal of Project Management in theperiod 2003–2008, and

• 17 Master of Project Management(MPM) programs, in the United States,Australia, and Europe.

Research MethodFor this purpose, two different measure-ments were used. The first metric isbased on the 18 categories suggested byCrawford, Pollack, and England (2006).In addition to the 48 topics listed inTable 1, however, more references wereneeded in order to map the data into the18 categories, and in order to obtain thisobjective, Kerzner (2006), PMI (2004),PMI (2005), and Wideman (2006) wereused. These categories were then trans-formed into a graphical diagram basedon two main dimensions used to sug-gest the further arrangement of the cat-egories: (a) the horizontal leadership–craftsmanship dimension reflects thebalance between leading the projectand the project work itself, and (b) thevertical strategy-execution dimensionreflects the balance between settingstrategy and executing it.

In order to locate the different cate-gories within the dimensions, we calledon a panel of 24 project managementexperts, all of whom had at least a master’s degree in the field and wereactively working as project managers inbusinesses or in the public sector. Byusing a data sampling through a simple

Category Topics

Cost management Cost management

Cross-unit outcomes EstimatingIntegration managementProject context/environmentProject life cycle/phasing

Finalization Project closeout/finalizationTesting, commission, handover, and acceptance

Interpersonal Conflict managementLeadershipNegotiationProblem solvingTeamwork

Legal issues Legal issuesRegulationsSafety, health, and environment

Marketing Marketing

Product functionality Configuration managementDesign managementRequirements managementTechnology managementValue management

Program management Program management

Project evaluation and improvement Organizational learningPerformance managementProject evaluation and review

Project planning and control Change controlProject monitoring and controlProject planning

Project start-up Goals, objectives, and strategiesProject initiation/start-upSuccess (criteria and factors)

Quality management Quality management

Relationship management Benefits managementDocument managementInformation and communication managementReportingTeam building and development

Resource management Personnel/human resource managementProcurementProject organizationResource management

Risk management Risk management

Scope management Scope management

Strategic alignment Business caseFinancial managementProject appraisalStrategic alignment

Time management Time management

Note: From Crawford, L., Pollack, J., & England, D. (2006). Uncovering the trends in project management:Journal emphases over the last 10 years. International Journal of Project Management, 24(2). Reprinted [orAdapted] with permission.

Table 1: Project management categories and corresponding topics.

pmj402_20122.qxd 4/28/09 5:16 PM Page 61



AP

ER

S

Web-based survey, the participants wereasked to place each category withineach of the two dimensions. The graph-ical instrument used is the 360� radardiagram and is shown with the survey’sresults in Figure 1. Only the 18 main cat-egories are shown, but the remaining 48topics, mentioned before, were used asa reference guide for arranging the datafrom each information source in theapposite place within the diagram.

The data was then mirrored into the radar diagram in such a way that themost important category for each of the three sources of information scored100% (full slice), and the other cate-gories scored relatively on this basis(i.e., their relative importance is repre-sented on the basis of their area withinthe slice). In this way, the radar diagramrepresents the internal relative impor-

tance of all the categories for eachsource of information used in the study:journal articles, master’s programs, andrecent textbooks.

The second metric is based on the “Eye of Competence” (EOC) intro-duced in 2006 by the IPMA. The EOC isshown in Figure 2.

The subjects that fall under each ofthe three competences are as follows:• Behavioral competences: Leadership,

engagement and motivation, self-control, assertiveness, relaxation,openness, creativity, results of orien-tation, efficiency, consultation, nego-tiation, conflict and crisis, reliability,value appreciation, and ethics.

• Contextual competences: Project ori-entation, program orientation, port-folio orientation, project program andportfolio implementation, permanentorganization, business, systems prod-ucts and technology, personnel man-agement, health, security, safety andenvironment, finance, and legal.

• Technical competences: Project man-agement success, interested parties,project requirements and objectives,risk and opportunity, quality, projectorganization, teamwork, problem res-olution, project structures, scope anddeliverables, time and project phases,resources, cost and finance, procure-ment and contracts, changes, controland reports, information and docu-mentation, communication, start-up,and closeout.

Based on these three main areas ofcompetences and their corresponding

Cra

ftsm

ansh

ip

Lead

ersh

ip

Execution

Strategy

Prog

ram

man

agem

ent

Proj

ect p

lann

ing

and

cont

rol

Reso

urce

man

agem

ent

Relationship managementInterpersonal

Project start-up

Proj. evaluation & improvement

Strategic alignment

Scope managem

ent Cro

ss-u

nit o

utco

mes

Risk

man

agem

ent

Mar

ketin

g

Product functio

nality

Legal issues

Cost managementQuality management

Time m

anagement

Finalization

Figure 1: Project management categories and dimensions.

Contextual competences

Behavioral competences

Technical competences

Figure 2: The Eye of Competence (EOC)—IPMA version, June 2006.

pmj402_20122.qxd 4/28/09 5:16 PM Page 62


subjects, the data (i.e., the definedcategories identified in the surveydescribed above) were then furthermapped into the proper competencesand subjects of the EOC. The relativebalance between the competenceswas calculated and represented in apie chart, which revealed our findingsin a specific corresponding version of the EOC for each source of infor-mation.

ResultsPapers in the Project ManagementJournal and International Journal ofProject ManagementAll papers published in the ProjectManagement Journal and the Interna-tional Journal of Project Managementduring the period June 2003–February2008 were registered, including titles,authors, and keywords. A total of 2,338keywords were found in 484 papers.Keywords that were too general or toobroad were omitted—for example,“project management” (81 occur-rences), “studies” (47 occurrences),and “managing projects” (56 occur-rences)—as were keywords indicatingcountries and specific project types. Inall, 687 keywords were omitted, andthe analysis was based on the remain-ing 1,651 keywords. The mapping isshown in Figure 3.

The mapping of the article key-words into the 18 categories shows that three categories were dominant:interpersonal competences, resourcemanagement, and strategic alignment,followed by relationship management.These seem, according to our findings,to be the areas in which most activeresearch is currently occurring withinproject management. This focus inresearch will probably influence thedirection that project management willtake during the coming years, and itsimpact can already be seen in the focusof some cutting-edge master’s-degreeprograms in project management. Wediscuss this focus in research in thenext section.

Graduate Programs in ProjectManagementThe MPM is far from being as estab-lished as an MBA degree. Increasingnumbers of universities are, however,offering specific graduate programs thatfocus on project management. In orderto obtain some indication of the educa-tional focus of these programs, 17 pro-grams were selected for evaluation. Theselection was based on an extensiveWeb exploration in which the selectioncriteria were that the program was on amaster’s-degree level, that it was offeredby a recognized university, and that suf-ficient information on the program wasavailable online for the analysis. The fol-lowing programs were included:• Euro Master of Project Management.

University of Applied SciencesDortmund (Germany), University ofthe Basque Country (Spain), Norwegian

University of Science and Technology(Norway), University of Zaragoza(Spain), ESC (France), and Universityof Maribor (Slovenia) (EuroMPM,2008);

• Master of Project Management,Pennsylvania State University (UnitedStates) (Pennsylvania State University,2008);

• Master of Project Management, Uni-versity of Iceland (Iceland) (Universityof Iceland, 2008);

• Master of Project Management, BondUniversity (Australia) (Bond University,2008);

• Master of Project Management, EdithCowan University (Australia) (EdithCowan University, 2008);

• Master of Project Management, KellerGraduate School of Management(United States) (Keller GraduateSchool of Management, 2008);

Cra

ftsm

ansh

ip

Lead

ersh

ip

Execution

Strategy

Prog

ram

man

agem

ent

Proj

ect p

lann

ing

and

cont

rol

Reso

urce

man

agem

ent


Project start-up


Strategic alignment

Scope managem

ent Cro

ss-u

nit o

utco

mes

Risk

man

agem

ent

Mar

ketin

g

Product functio

nality

Legal issues


Time m

anagement

Finalization

Figure 3: Focus in journals in 2003–2008.

pmj402_20122.qxd 4/28/09 5:16 PM Page 63



AP

ER

S

• Master of Project Management,Queensland University of Technology(Australia) (Queensland University ofTechnology, 2008);

• Master of Project Management,University of Queensland (Australia)(University of Queensland, 2008);

• Master of Project Management, Uni-versity of South Australia (Australia)(University of South Australia, 2008);

• Master of Science in Project Man-agement, City University of Seattle(United States) (University of Seattle,2008);

• Master of Science in Project Man-agement, George Washington Univer-sity (United States) (George WashingtonUniversity, 2008);

• Master of Science in ProjectManagement, Royal Institute ofTechnology (Sweden) (Royal Instituteof Technology, 2008);

• Master of Science in Project Man-agement, University of Alaska (UnitedStates) (University of Alaska, 2008);

• Master of Science in Project Man-agement, University of San Francisco(United States) (University of SanFrancisco, 2008);

• Master of Project Management,Syddansk Universitet (Denmark)(Syddansk Universitet, 2008);

• Stanford Advanced Project Man-agement (United States) (StanfordUniversity, 2008); and

• Master of Science in Strategic ProjectManagement, Heriot-Watt University(Scotland), Politecnico di Milano(Italy), Umeå School of Business andEconomics (Sweden) (Anonymous,2008).

A total of 24 universities in Europe,Australia, and the United States areinvolved, in one way or another, in run-ning the 17 graduate programs listed.The analysis of the programs wasbased on both course descriptions andthe number of credits given for thecompletion of each course, which thencombined to yield the weight of the dif-ferent categories in each program. Allof the programs have an equal weight

in the outcome. Their coalesced educa-tional focus in terms of our 18 cate-gories is represented in Figure 4.

The radar diagram shows that threecategories are dominant: interpersonal,strategic alignment, and project plan-ning and control, followed closely byrelationship management. There is,interestingly, a low score in the categoryfinalization. These results are quiteconsistent with those demonstrated inFigure 3, which illustrated the focus ofresearch topics in journal articles. It isnot unreasonable to expect that in thefuture these master’s-degree programswill increase their focus on resourcemanagement and product functionalityand that the research being conductedin these areas will be disseminated inthose programs.

Textbooks on Project ManagementThe last source of information that weexamined was professional books onproject management. A large number ofworks on project management are avail-able and many of them are universitytextbooks. For the purpose of this study,we selected a number of these text-books based on a search on thewww.amazon. co.uk Web site on May 1,2008. The search term was “projectmanagement,” and the search was lim-ited to books published in 2003 or later.Books that were too specific or too spe-cialized were omitted from the list, suchas books on specific project manage-ment software, and only books thatwere over 200 pages in length wereincluded. The final list consisted of thefollowing 21 books:

Cra

ftsm

ansh

ip

Lead

ersh

ip

Execution

Strategy

Prog

ram

man

agem

ent

Proj

ect p

lann

ing

and

cont

rol

Reso

urce

man

agem

ent


Project start-up


Strategic alignment

Scope managem

ent Cro

ss-u

nit o

utco

mes

Risk

man

agem

ent

Mar

ketin

g

Product functio

nality

Legal issues


Time m

anagement

Finalization

Figure 4: Focus in graduate project management programs in 2008.

pmj402_20122.qxd 4/28/09 5:16 PM Page 64


• A Guide to the Project ManagementBody of Knowledge (PMBOK ® Guide;PMI, 2004),

• Project Management for Dummies(Portny, 2007),

• Project Management (Maylor, 2005),• Software Project Management for

Dummies (Luckey & Phillips, 2006),• Information Technology Project

Management (Schwalbe, 2005),• Guide to Project Management (Roberts,

2007),• IT Project Management: On Track

from Start to Finish (Phillips, 2004),• Project Management (Lock, 2007),• Project Management in Construction

(Walker, 2007),• Project Manager: Mastering the Art of

Delivery in Project Management(Newton, 2005),

• Reinventing Project Management: TheDiamond Approach to SuccessfulGrowth and Innovation (Shenhar &Dvir, 2007),

• Software Project Management (Hughes& Cotterell, 2006),

• The Definitive Guide to Project Man-agement: The Fast Track to Getting theJob Done on Time and on Budget(Nokes, 2003),

• The Handbook of Project Man-agement: A Practical Guide to EffectivePolicies, Techniques and Processes(Young, 2007),

• Project Management: A ManagerialApproach (Meredith & Mantel, 2006),

• Project Management—A SystemsApproach to Planning, Scheduling,and Controlling (Kerzner, 2006),

• Project Management—Planning andControl Techniques (Burke, 2003),

• The Fast Forward MBA in ProjectManagement (Verzuh, 2005),

• PRINCE2 Revealed: Including How ToUse PRINCE2 for Smaller Projects(Bentley, 2006),

• Project Management: The ManagerialProcess (Gray & Larson, 2006), and

• Project Management: A StrategicPlanning Approach (Gardiner, 2005).

Once the list was defined, the con-tent of each of the 21 books was

mapped into the 18 categories on thebasis of chapter and subchapter titlesand their lengths. Each book was giventhe same weight in the final representa-tion, and their focus in terms of contentmirrored in our 18 categories is shownin Figure 5.

This diagram indicates that theselected textbooks put the strongestemphasis on project planning and con-trol, relationship management, andresource management. There is littlefocus on program management, mar-keting, and legal issues. These findingsalso fit well with the previous findings.The strong emphasis on project plan-ning and control in project manage-ment textbooks is not unexpected, and,based on what seems to be the trend inresearch, we might expect a strongerfocus on strategic alignment and interpersonal competences in project

management textbooks in the comingyears.

Holistic ViewThe information presented thus far canfurther be consolidated in such a way asto provide us with an overview of theentire current situation. By drawing asingle radar diagram that includes all ofour findings as illustrated in the previousdiagrams, we get the representationdemonstrated by Figure 6, which showsthe entire picture covering all of our threesources mirrored in the 18 categories.

The variation in emphasis on the 18 different categories is striking: interper-sonal, relationship management, resourcemanagement, project planning and con-trol, and strategic alignment are all areasof strong focus, whereas there is muchless focus on marketing, legal issues, final-ization, and program management.

Cra

ftsm

ansh

ip

Lead

ersh

ip

Execution

Strategy

Prog

ram

man

agem

ent

Proj

ect p

lann

ing

and

cont

rol

Reso

urce

man

agem

ent


Project start-up


Strategic alignment

Scope managem

ent Cro

ss-u

nit o

utco

mes

Risk

man

agem

ent

Mar

ketin

g

Product functio

nality

Legal issues


Time m

anagement

Finalization

Figure 5: Focus in project management textbooks after 2003.

pmj402_20122.qxd 4/28/09 5:16 PM Page 65



AP

ER

S

Another way to represent our find-ings is to produce a still new perspec-tive by reflecting the information fromthe three sources into the three compe-tences of the newly introduced EOC.The results are shown in Table 2.

Combining the three diagrams thusobtained yields the representationgiven in Figure 7.

In the picture, the shaded areas rep-resent an overlap between the different

sources. In general, however, this viewis quite consistent with the relativeweight of competences as exhibited inthe original IPMA EOC (Figure 2).Technical competence is still the mostimportant feature of professional proj-ect management, followed closely bybehavioral competences and then con-textual competence—which is interest-ing in view of the fact that the originalweighting of competences in the IPMA’s

EOC was based on the intuitive impres-sion of its authors (according to ourconversation with one of the authors,Gerrit C. L. Koch, at the IPMA Congressin Shanghai in October 2006).

In general, the IPMA EOC can beused as a tool to provide a holisticimpression of project management. Forthose, however, who like to play withmore specific categorization, the 18-category view, based on the work ofCrawford and her coworkers, might givea somewhat more detailed notion. Themeasurement tool provided in ourstudy, however, might benefit fromsome improvements with respect notonly to mirroring our sources in thenumbers and description of Crawford’scategories, but also to the Leadership–Craftsmanship and the Vertical Strategy–Execution dimensions that were used toplace the categories within the radardiagram in the first place.

ConclusionThe combined view of project manage-ment in Figure 6 indicates that there isa stronger focus on “leadership” than“craftsmanship” and a somewhatstronger focus on “execution” than“strategy.” The “leadership–execution”section of the radar diagram is, there-fore, strong. This may support the general impression that many organi-zations do better in defining theirobjectives and strategy than in actuallyexecuting them in a proficient andeffective way. A renewed focus on tradi-tional project management themes inthe “craftsmanship” section could assistin adjusting this balance.

There is an overall consistencybetween the essential focal views ofproject management represented inFigures 3 through 5, and this can clearlybe seen in the combined illustrationshown in Figure 6. Journals have thestrongest focus on resource manage-ment, interpersonal competence, andstrategic alignment, and MPM pro-grams show a very similar focus,although with the three strongest cate-gories being interpersonal competence,

Cra

ftsm

ansh

ip

Lead

ersh

ip

Execution

Strategy

Mar

ketin

g

Legal issues

Finalization

Pro

t start-up

alu

gem

it o

y ma

me m

ana

P

t

alu

maana

ge

u

roj

manost

Quality m

Time

m

m

ment

oss-

un

nag

m

ro

n

y m

me

ag

m

i

Reso

Re

Reso

ur

Relat

Reso

Rel

I

p

aluation & im uct func

cuct

St

men

t

Inter

Strateg ana

ojec

t pla

nnin

g an

d co

ntro

l

ourc

e man

agem

ent

lationship managementrpersonal

& improvement

tegic alignment

Cro Ri

sk m

aProduc

Co

Qur

ateg ma

In

ourc

lat

& im uctuct

Str

oje

mRe

unc

Co

Qu

up

aluo

na

Figure 6: Combined view of project management.

Competence Journal MPMAreas Papers Programs Textbooks

Technical 48% 49% 57%

Behavioral 26% 24% 21%

Contextual 25% 26% 22%

Table 2: Relative focus of journal papers, MPM programs, and textbooks in the competence areasof the IPMA 3.0 competence baseline.

pmj402_20122.qxd 4/28/09 5:16 PM Page 66


project planning, and control andstrategic alignment. The two strongestcategories of textbooks are project planning and control and relationshipmanagement, and the third mostfocused-on category is resource man-agement. It is reasonable to assume thatthere is a certain time lag between thethree sources considered in this study.The journals represent active researchin the discipline that is often connectedwith universities and disseminated tobusiness through the Master of ProjectManagement programs. The time lapsebefore research results are incorporatedinto project management textbooks islikely to be longer. Naturally, traditionalproject management themes will con-tinue to be strongly emphasized inMPM programs and textbooks, regard-less of research and development onnew frontiers. But given the strong focusin both journals and MPM programs onstrategic alignment and interpersonalcompetence, these will very likely bemore visible themes in the project man-agement textbooks in the future.

In view of the strong trends identi-fied in the data studied, we concludethat the traditional baseline used todescribe the discipline of project man-agement could be improved. TheIPMA’s recent EOC and the Compet-encies in Project Management, NationalCompetence Baseline for Scandinavia

(Fangel, 2005) represent a step in thisdirection. Further development of the360 � radar diagram presented in thisarticle may also be of use. Such devel-opments, however, would necessarilyinvolve not only a review of the cate-gories used, but also a more sophisti-cated analysis—both of the dimensionsthat were defined for this study and oftheir links to the categories within theradar diagram. This is an interestingsubject for further research aimed atgetting a still-clearer picture of the cur-rent eye-opening undertakings of theproject management field.

AcknowledgmentsWe thank Thorbjörg Sæmundsdóttir,MSc, for her help in the processing ofdata for this study. We also thankProfessor Hans J. Thamain, PhD, for hisvaluable comments. ■

ReferencesAnonymous. (2008). Master of sciencein strategic project management.Retrieved April 1, 2008, from www.polimi.it/english/academics/english_courses/master_mspme.php?id_nav�

-289&apri�-300

Bentley, C. (2006). PRINCE2 revealed:Including how to use PRINCE2 forsmaller projects. Oxford, UK: Elsevier.

Bond University. (2008). Master ofproject management. Retrieved April 1,

2008, from www.bond.edu.au/study/courses/btsd/pg/m-p-m-structure.html

Burke, R. (2003). Project management—Planning and control techniques(4th ed.). Chichester, UK: Wiley.

Crawford, L., Pollack, J., & England, D.(2006). Uncovering the trends in proj-ect management: Journal emphasesover the last 10 years. InternationalJournal of Project Management, 24,175–184.

Edith Cowan University. (2008). Masterof project management. Retrieved April1, 2008, from http://www.aipm.com.au/html/0802-aetp-edithcowan.cfm

Euro MPM. (2008). Retrieved April 1,2008, from http://www.eurompm.net

Fangel, M. (2005). Competencies inproject management—National com-petence baseline for Scandinavia.Hilleroed, Denmark: Associations ofProject Management in Denmark,Norway and Sweden.

Gardiner, P. D. (2005). Project manage-ment: A strategic planning approach.London: Palgrave McMillan.

George Washington University. (2008).Retrieved April 1, 2008, fromhttp://business.gwu.edu/grad/mspm/index.html

Githens, G. D., & Rosenau, M. D.(2005). Successful project management:A step-by-step approach with practicalexamples (4th ed.). Hoboken, NJ: Wiley.

Gray, C. F., & Larson, E. W. (2006).Project management: The managerialprocess. New York: McGraw-Hill Irwin.

Hughes, B., & Cotterell, M. (2006).Software project management.Maidenhead, UK: McGraw-HillEducation.

International Project ManagementAssociation. (2006). ICB IPMA compe-tence baseline version 3.0. Nijkerk, TheNetherlands: Author.

International Standards Organization(ISO). (2003). 10006:2003(E) qualitymanagement systems—Guidelines forquality management in projects.Geneva: ISO Copyright Office.

Contextual competence

Behavioral competence

Technical competence

Figure 7: Overview of project management, based on the Eye of Competence.

pmj402_20122.qxd 4/28/09 5:16 PM Page 67



AP

ER

S

Jugdev, K., & Müller, R. (2005).Retrospective look at our evolvingunderstanding of project success.Project Management Journal, 36(4),19–31.

Keller Graduate School ofManagement. (2008). Master of projectmanagement. Retrieved April 1, 2008,from http://www.devry.edu/keller/programs/m_project_management/courses.jsp

Kerzner, H. (2006). Project manage-ment—A systems approach to plan-ning, scheduling and controlling (9thed.). Hoboken, NJ: Wiley.

Lock, D. (2007). Project management(9th ed.). Burlington, VT: Gower.

Luckey, T., & Phillips, J. (2006).Software project management for dum-mies. Hoboken, NJ: Wiley.

Maylor, H. (2005). Project management(3rd ed.). Edinburgh: PearsonEducation.

Meredith, J., & Mantel, S. J. (2006).Project management—A managerialapproach (6th ed.). Hoboken, NJ:Wiley.

Newton, R. (2005). Project manager:Mastering the art of delivery in projectmanagement. Edinburgh: PearsonEducation.

Nokes, S. (2003). The definitive guideto project management: The fast trackto getting the job done on time and onbudget (2nd ed.). Edinburgh: PearsonEducation.

Pennsylvania State University. (2008).Master of project management.Retrieved April 1, 2008, from www.worldcampus.psu.edu/MasterinProjectManagement_Degree_Map.shtml

Phillips, J. (2004). IT project manage-ment: On track from start to finish.Emeryville, CA: McGraw-Hill Education.

Portny, S. (2007). Project managementfor dummies (2nd ed.). Indianapolis,IN: Wiley.

Project Management Institute. (2004).A guide to the project management bodyof knowledge (PMBOK® Guide) —Thirdedition. Newtown Square, PA: Author.

Project Management Institute. (2005).Organizational project managementmaturity model. Newtown Square, PA:Author.

Queensland University of Technology.(2008). Master of project management.Retrieved April 1, 2008, from http://www.courses.qut.edu.au/cgi-bin/WebObjects/Courses.woa/wa/selectMajorFromMain?structureID�16258&courseID�4935#16258

Roberts, P. (2007). Guide to projectmanagement: Achieving lasting benefitthrough effective change. London:Profile Books.

Royal Institute of Technology. (2008).Retrieved April 1, 2008, from www.kth.se/studies/master/programmes/sth/Project%2BManagement%2Band%2BOperational%2BDevelopment?l�en

Scanlan, J. (2006). What’s new in proj-ect management—Janice Scanlaninterviews Milton D. Rosenau, Jr.Consulting to Management (C2M)Journal, 19(2), 35–36.

Schwalbe, K. (2005). Introduction toproject management (4th ed.). Boston:Thomson Course Technology.

Shenhar, A. J., & Dvir, D. (2007).Reinventing project management: Thediamond approach to successful growth& innovation. Boston: HarvardBusiness School Press.

Stanford University. (2008). RetrievedApril 1, 2008, from http://apm.stanford.edu/courses.htm

Syddansk Universitet. (2008). Master iprojektledelse. Retrieved April 1, 2008,from http://www.sdu.dk/Uddannelse/Uddannelsesoversigt/Master/Master_i_projektledelse.aspx

University of Alaska. (2008). Master ofscience in project management.Retrieved April 1, 2008, from http://soe.uaa.alaska.edu/espm/introduction/programs/project_management/index.html

University of Iceland. (2008). RetrievedApril 1, 2008, from http://www.hi.is/page/mpm

University of Queensland. (2008).Master of project management.Retrieved April 1, 2008, fromhttp://www.uq.edu.au/study/pro-gram.html?acad_prog�5289

University of San Francisco. (2008).Master of science in project manage-ment. Retrieved April 1, 2008, fromhttp://www.usfca.edu/catalog/cps_grad_pmgt.html

University of Seattle. (2008). Master ofscience in project management.Retrieved April 1, 2008, from http://www.cityu.edu/som_gd_ms_projmgmt.htm

University of South Australia. (2008).Master of project management. Re-trieved April 1, 2008, from http://www.unisanet.unisa.edu.au/programs/program.asp?Program�IMPJ&Year�

2008

Urli, B., & Urli, D. (2000). Project man-agement in North America, stability ofthe concepts. Project ManagementJournal, 31(3), 33–43.

Verzuh, E. (2005). The fast forwardMBA in project management (2nd ed.).Hoboken, NJ: Wiley.

Walker, A. (2007). Project managementin construction. Oxford, UK: Blackwell.

Wideman, M. (2006). Web page of MaxWideman. Retrieved November 15,2006, from http://www.maxwideman.com

Young, T. (2007). The handbook ofproject management: A practical guideto effective policies, techniques andprocesses. London: Kogan PageLimited.

Zobel, A. M., & Wearne, S. H. (2000).Project management topic coverage inrecent conferences. ProjectManagement Journal, 31(2), 32–37.

Helgi Thor Ingason holds a PhD in process metallurgy from the Norwegian Institute ofTechnology and an MSc in mechanical andindustrial engineering from the University ofIceland. He is an associate professor of industri-al engineering at the University of Iceland and

pmj402_20122.qxd 4/28/09 5:16 PM Page 68


lectures in project management, quality man-agement, and facilities planning. He is the headof the Master of Project Management program atthe university (www.mpm.is). He is a memberof the Nordica Consulting Group in Iceland andcofounder and chairman of Alur, alvinnsla hf, arecycling company that works with the alu-minum industry in Iceland. He is the coauthor ofthree books on quality management, productdevelopment, and project management. Hisresearch fields range from quality and projectmanagement to renewable energy, production,

transport, and utilization. He is an IPMA-certifiedsenior project manager.

Haukur Ingi Jónasson holds a PhD in psychiatryand religion from Union Theological Seminary inNew York. He is an assistant professor in man-agement at the School of Engineering andNatural Sciences and a research fellow at CooperUnion for the Advancement of Science and Arts inNew York. He is an associate director of theMaster of Project Management program at the

university, a partner with the Nordica ConsultingGroup ehf, and an associate with HLP-HirzelLeder & Partner consultants in Frankfurt,Germany. His research interests include organi-zational behavior and development, manage-ment and project management, industrial andorganizational psychology, communication,negotiation, conflict management, group dynam-ics, human growth and development, coaching,higher-order thinking (critical/creative/ethical),psychoanalytical and depth psychological theo-ries, religion, culture, and project ethics.

pmj402_20122.qxd 4/28/09 5:16 PM Page 69

PA

PE

RS


INTRODUCTION ■

The success of projects relies critically on the effectiveness of theirleadership. The growth in project management training and certifica-tion programs, along with an increased shift toward recognizing proj-ect-based work in the United States and abroad, indicates a trend in

practitioner understanding of this need. A 2006 State of the Profession studyby the Project Management Institute found that instituting project manage-ment offices, which enable project leadership to be better organized andtrained, leads to better budget, scope, and completion outcomes (all find-ings statistically significant at p � 0.05; Andrews, 2007). Other recent reportsindicate that project work has been increasing as a proportion of total orga-nizational work and now constitutes one-fifth of the world’s gross domesticproduct (GDP), or approximately U.S. $7.2 trillion annually (Bredillet, 2007).

The academic literature has certainly not ignored the importance ofleadership or project management. Recent research on teams and their pro-ductivity in project work reinforces the critical role leadership must play ingetting productivity from teams with heterogeneous specialization. Thisresearch also underscores the need for further study of how leaders mayaffect these outcomes (Shin & Zhou, 2007). Another recent study calls forfocus on work design within and between organizations and points to thetrend toward more complex work, often involving new arrangements char-acterized by multiorganizational team projects (Sinha & Van de Ven, 2005).Few studies have focused on the role of the team leader (whom we refer tointerchangeably as the project manager) and how he or she may manage and improve team productivity in emergent forms of multiorganizational anddistributed team projects. In one exception, researchers examined formal-ization (e.g., explicitly outlining how certain interactions should take place)and discretion (e.g., favoring and promoting certain interactions over oth-ers) given to team members and the resulting impact on productivity(Naveh, 2007). Yet while research has outlined virtues of various contributorsto project success, such as balancing formality and discretion, there remainsa clear need to understand the dynamics of their effectiveness. In particular,there is a need to further specify how a project manager may apply leader-ship styles and tactics when managing actual projects, what synergistic free-dom these styles and tactics afford, and of what limitations in these practicesmanagers should be cognizant.

In this study, we specifically consider the synergies and limitations of twofactors concerning project leader action: breadth of leadership style, whichencompasses both Theory X– and Theory Y–oriented leadership, and scopeof leadership tactics, which relates to the general extent of intervention at a

Limits to Effective Leadership Styleand Tactics in Critical IncidentInterventionsDominic Thomas, Goizueta Business School, Emory University, Atlanta, GA, USAElliot Bendoly, Goizueta Business School, Emory University, Atlanta, GA, USA

ABSTRACT ■

Leadership effectiveness takes many forms. Itsimpact on performance can be critical, but byno means does it stand alone, and hence can-not truly be evaluated in isolation. Nor can it beassumed that more of any one effective style isnecessarily a good thing. In project manage-ment contexts, these issues are particularlysalient. Here we study the impact of multipleleadership styles on project performance, aswell as the joint role that intervention tactics ingeneral play. Our empirical study demonstratesthe benefit of moderate mixes of leadershipstyle and the declining returns on active inter-ventions in these settings.

KEYWORDS: leadership; critical incidents;interventions; tactics





DOI: 10.1002/pmj.20118

pmj402_20118.qxd 4/28/09 4:20 PM Page 70


given point in a project. In order to pro-vide contextual focus, in our investiga-tion we specifically consider each ofthese issues as they relate to the effec-tiveness of leaders in managing com-munication and worker interactionswithin projects. We furthermore focuson the use of these styles and tactics,specifically with regard to dealing withcritical incidents of projects or points atwhich work breaks down due in part tointeraction failures. Encountering atleast a single incident of breakdown isnot uncommon among projects,although these events themselves areoften ephemeral and difficult to ade-quately capture in research. Our deci-sion to pursue this focus is aimed atproviding meaningful and practicalextension of existing thought on inter-related project manager behaviors. It isalso designed to allow for a re-examina-tion of standing theoretical assump-tions regarding the applicability ofleadership styles and tactics within atypical context of knowledge workerprojects.

Research and HypothesesWe begin our consideration of project

leadership engagement by acceptingthe premise that project managersmust manage levels of formalizationand discretion in any and all projectsthat involve creativity among groups ofknowledge workers (Naveh, 2007).Today’s project managers face a funda-mental problem when it comes to themanagement of formalization and dis-cretion levels. They must act to manageteam interaction among workers whoare often linked across organizations ina supply chain and who probably arenot collocated (i.e., so-called “virtualteams”) (Lipnack & Stamps, 2000). Thisproblem is complicated by the verynature of project manager roles as wellas their own personal professionalbackgrounds. Project managers tend tobe leaders who come from a single dis-ciplinary area within a field relevant toa project (Project ManagementInstitute [PMI], 2004), and they will

often focus only on the management of team interaction at points wheninteraction breaks down (Hackman,2002). In computer-mediated team-work, researchers have identified animportant and common version ofthese breakdowns related to the inabilityto interact through information and communications technologies. Theseoccasions have been termed criticalincidents (Thomas & Bostrom, 2008) ordiscrepant events (Majchrzak, Rice,Malhotra, King, & Ba, 2000). Successfulleader interventions during these inci-dents enable better project outcomesand have the potential to avert projectcollapse (Thomas & Bostrom, 2008).This being the case, one might ask howproject leaders can effectively operatethrough the use of various leadershipstyles and interventions during criticalincidents. To answer this, we turn to theleadership literature and its discussionsof both leadership style and interven-tion.

Leveraging Leadership StyleExisting leadership literature highlightsthe notion that project leaders’assumptions about their project work-ers influence the nature and methods oftheir resulting leadership style regard-ing formalization and discretion(McGregor, 2006). According to this lit-erature, “Theory X” leaders representthose who assume their subordinates tobe less-than-ideally self-motivated (i.e.,“lazy” by certain interpretations) and inneed of direct task assignment. In con-trast, “Theory Y” leaders assume work-ers to be independently motivated andin need of general direction andencouragement (McGregor, 1960). Thenotions of Theories X and Y continue topermeate discussions of leadershipstyle and action and relate to the con-cepts of formalization and discretion.

A Theory Y–oriented leader wouldbe more likely to take actions thatenable team members to have morediscretion, whereas a leader takingmore of a Theory X–based approachwould take actions to formalize rules

and tasks. In keeping with theseassumptions, discussions of leadershipstyle in practitioner sources and even insome research indicate that a TheoryX–oriented leader may be considered interms of being a micromanager or anindividual likely to intervene moreoften, even to the point of disruptingcreative processes (Stuart & Tax, 2004).In the management of professionalworkers, it has been suggested thatleaders leaning toward a Theory Yapproach will tend to achieve higherperformance (Mintzberg, 1998; Stuart &Tax, 2004). Although a critical perspec-tive might suggest that such results arelargely due to the professional and self-motivated nature of the workers thatallow such leaders to engage in aTheory Y mode in the first place(Fiedler, 1967), empirical evidencefrom virtual projects has supported thisfinding, noting that leadership effortsto directly control the behavior of high-ly qualified workers often have negativeconsequences (Piccoli & Ives, 2003).Based on this research, we can expectthat project managers working withqualified professionals (e.g., those withprofessional certifications) in particu-lar would achieve higher success byemploying a predominantly Theory Yapproach. We have no reason to antici-pate such a dynamic to be restricted tostrictly “virtual” settings, but expectthat it might be relevant in all projectmanagement settings (where commu-nication of some sort is fundamentallycritical). In a later section of this article,we discuss our method to isolate theactions associated with a Theory X orTheory Y approach, but for the timebeing, we present hypotheses referringto Theory Y actions or Theory X actions.

H1: Leadership actions characterizedpredominantly by Theory Y–orientedactions will exhibit higher levels of per-formance than those characteristic ofTheory X–oriented actions.

Of course, not all project team mem-bers can be characterized as knowl-edgeable and self-guiding, or even as

pmj402_20118.qxd 4/28/09 4:20 PM Page 71


Limits to Effective Leadership Style and TacticsP

AP

ER

S

fully cognizant of individual taskdynamics. Recent advances in the theo-ry of leadership in cognitively andbehaviorally complex settings suggestthat, under conditions such as thosefaced by most project managers, partic-ularly in projects involving critical inci-dents and, hence, in need of well-ledredirection, Theory Y actions may beuseful but not sufficient to handle thepossible combinations of variables thatmight occur (Bennis, Heil, & Stephens,2000). Actions much more akin to aTheory X style may be needed (Bobic &Davis, 2003). Thus, leaders in highlychallenging project settings, such asthose involving critical incidents,would likely benefit from the use ofboth Theory X and Theory Y approach-es in tandem (Denison, Hooijberg, &Quinn, 1995; Hooijberg, Hunt, &Dodge, 1997). We could expect to seesynergies in the joint use of Theory Xand Theory Y actions, particularly insuch contexts.

H2a: The interaction effect betweenTheory X and Theory Y actions will be asignificant positive predictor of per-formance in projects involving criticalincidents.

Although we formally pose H2 as adynamic to guide our research andanalysis, we do not ignore the fact thatother competing lines of thought exist.One such line of thought in particularprovides for a very different predictionregarding the mixed use of manage-ment styles and resulting productivitylosses due to team conflict. Projectteams suffer substantial productivitylosses when they devolve into conflict(Behfar, Peterson, Mannix, & Trochim,2008). While it is possible that, overtime, conflict management in teams willbe aided by carefully designed work sys-tems embedded in communicationstechnologies (Chidambaram, Bostrom,& Wynne, 1991), team technologies inuse today still lack such features. In themeantime, a key moderator of the illeffects of conflict on team productivityis shared identity and context (Hinds &

Mortensen, 2005). Shared team identityhelps unify expectations about eachother’s behavior during times of stress,such as a critical incident, and sharedunderstandings about how to refocusand intervene to improve outcomes(Cannon-Bowers, Salas, & Converse,1993; Mathieu, Goodwin, Heffner, Salas,& Cannon-Bowers, 2000). This being thecase, the team members may have exist-ing expectations about their leader andhow he or she leads, and if the leaderwere to use a mix of actions drawn fromboth Theory X and Y approaches, thismay confuse and impede shared actiontoward improvement during a criticalincident. According to such a perspec-tive, actions consistent with either aTheory X or Theory Y approach might beeffective on their own, but when used inalternation, or when excessive fusionsof the two styles are attempted, theresult may be considerably disruptive,and perhaps even crippling. We hencesimultaneously pose a competinghypothesis to that outline by H2a:

H2b: The interaction effect betweenTheory X and Theory Y actions will be asignificant negative predictor of per-formance in projects involving criticalincidents.

Tempered Managerial InterventionAside from sufficient conditions forguiding a project team through the res-olution of project complications, whatsort of structural interventions by proj-ect managers are necessary during suchincidents? To what extent can they beeffectively deployed? Although theseare nontrivial questions, one criticalpoint of fact is that project work today islargely computer-mediated, even amongcolleagues within the same building(Davenport, 2005). Teams whose mem-bers are also distributed, as in virtualproject teams, will be both computer-mediated and drawn from multiple or-ganizations (Lipnack & Stamps, 2000).In such computer-mediated knowledgework, teams experience “repair events”when computer-mediated interactionbreaks down and something must be

done to get everyone working togeth-er again (Massey & Montoya-Weiss, 2006).

Leader involvement in theseevents would consist of taking actionsto manage communications technolo-gies, team processes, and team mem-bers, thereby aiding successful repairby helping the team effectively coordi-nate themselves to get back on track.There could be an endless number ofpossible intervention actions thatcould be taken only by a project leaderby virtue simply of their level ofauthorization, regardless of whetherthe leader tends to be Theory Y– orTheory X–oriented. The Chinese havean ancient proverb explaining that inorder to improve one’s life, one “mustmove 27 things.” How many and whatsort of actions form this sweet spot forleaders intervening to improve projectinteraction? Clearly, insufficient inter-ventions may offer an insufficientimpact on project work during criticalperiods. On the other hand, if projectmanagers allow themselves to becomeoverly engaged with changing the useof communications technologies,roles, and tasks, teams might not settleinto normal productive work. This tiesin Fiedler’s contingency model of lead-ership in that the creation of poorground scenarios can easily derail anyleadership effectiveness otherwiseanticipated (Fiedler, 1967). Thus, theextent to which the team is redirectedby leader intervention actions during aproject is likely to show something ofan inverted-U relationship with overallproject performance. Such relation-ships have become increasingly citedin the operations management litera-ture (Bendoly & Hur, 2007; Linderman,Schroeder, & Choo, 2006) and are like-ly to have application here, as well.Because we expect such effects to befundamentally critical to the contextsin which leadership style is applied, wehypothesize that such a dynamic willtranscend and, hence, be observableregardless of the influences of leader-ship style.

pmj402_20118.qxd 4/28/09 4:20 PM Page 72


H3: The extent to which managementinterventions are applied in projectsinvolving critical incidents will exhibitan inverted-U relationship with overallproject performance, observable aboveand beyond the effects exhibited byleadership style.

Method and DataIn order to formally test the proposedhypotheses, and in so doing shed lighton discussions regarding the effective-ness of specific leadership approachesduring critical incidents of projects, wepursued an empirical investigation intothe joint use and extent of project man-ager action in such contexts. Becauseour intention here was not to explorenew means by which to measure suchactivity but rather the impact of suchactivity as observable through existingscales, we draw heavily on the extantliterature in our overall measurementapproach. Further, our own profes-sional experience within computer-mediated teamwork contexts and indications in prior literature guidedour approach. In that the critical inci-dents needed to be sampled from actual project work in distributed, com-puter-mediated contexts to properlyidentify the pressures that lead to inter-action breakdown over the long life ofactual projects, it was obvious that wewould: (1) face the possibility of record-ing a considerable quantity of uselessdata with no guarantee of capturing anincident in any specific projectobserved and (2) be forced to compre-hensively monitor a wide variety (e.g.,12�) of information and communica-tions technologies being used simulta-neously in project work interactions(Thomas, Bostrom, & Gouge, 2007).

Our own experience and priorresearch also made it evident that criti-cal incidents involving relationshipbreakdowns or other emotionallycharged conflict would likely be accom-panied by a switch in media from moreexplicit, archival, and, therefore,observable media, such as e-mail, tomore synchronous and rich media,such as phone or face-to-face meetings

(Barki & Hartwick, 2001; Dennis &Valacich, 1999). As a result, attemptingto capture logs and directly observeinteraction would be challenging andperhaps misleading. We could not guar-antee that we would see a critical inci-dent, due to the potential motivationfor individuals to resolve it discreetlythrough no one observable interaction.As a result, we decided to use as our pri-mary approach the collection of anony-mous, retrospective data as a bestmeans for motivating full disclosure,ensuring usable data from each partici-pant sample, and capturing all types ofbreakdown.

Operationalization of Key FactorsTo operationalize and understandmethods of enacting formalization orenabling discretion in terms of leaderactions, we draw on prior literatureidentifying forcing and linking actionsby project managers in distributedinformation systems developmentprojects (Thomas & Bostrom, 2008).Forcing actions include occasionswhen leaders confront employees,define roles, or block employees fromtaking action. These actions associatewith formalization in that they enforceformal structure in a project, add for-mal structure, or define the nature ofteam, task, and technology and howthe team must behave. If team mem-bers are interested in having cleardirection, there is some evidence thatthese actions would lead to improvedteam performance (Bobic & Davis,2003; McGregor, 1960). Enforcingactions indicate project managementconsistent with a Theory X approach(McGregor, 1960, 2006). Following thelead of Thomas and Bostrom (2008),we asked leaders whether they enactedeach of several commonly exhibitedenforcing actions. Items in our con-struct mimic those used in past studiesand included “Enforced existingexplicit rules through penalties for vio-lation” and “Mandated adherence tonew or existing explicit roles for inter-acting.”

In contrast, linking actions involvetraining employees, encouraging themto interact, developing consensus, initi-ating/modeling desired behavior, andsuggesting solutions for specific prob-lems. Such actions have been associ-ated with leaders as coaches and havebeen found effective for achieving im-proved team performance (Hackman,2002). These actions associate withbuilding discretion in that they empha-size interaction among team members,individual team member selection ofhow to act in given situations based onmodeling or suggestion, and consensus-based problem-solving. Encouragingactions indicate a leader acting in amanner consistent with Theory Y. Theytend to be associated with a moretransformational leader perspective,with focus on relationship building inwhich providing resources, connec-tions, and general direction is expectedto achieve improved outcomes (Bass &Avolio, 1994; McGregor, 2006). Again,following the lead of Thomas andBostrom (2008), we operationalizedthis construct through items such as“Explained why it is important to shareinformation in a timely fashion” and“Reminded team members of exist-ing inexplicit rules/norms for interac-tion.”

Because we also had a particularinterest in how leaders may interveneto improve team interaction, it was crit-ical to follow our conceptualization andoperationalize such actions as orthogo-nal to our construction of encouragingand enforcing. The leadership literatureidentifies initiating structure, behaviorssuch as setting up modes of interactionand providing means for interaction, asa core intervention action that leaderstake within groups (Hanson, 1996).These are behaviors we believe inde-pendent of either a Theory X or Yapproach and indicative of interveningin a group process. In that “effectiveleadership is characterized by highinitiation of structure” (Halpin, 1966, p. 127), we expect that, to achieveimproved project outcomes, some

pmj402_20118.qxd 4/28/09 4:20 PM Page 73



AP

ER

S

degree of intervening will be requiredin addition to enforcing or encouragingactions.

Project performance from a projectmanager perspective is typically char-acterized by effectively achieving acomposite goal known as “the tripleconstraint” (budget-time-scope). Thisgoal consists of bringing in a project onor under budget, on or before its dead-line (time), and fully achieving thedesired scope. Empirical evidence sup-ports the need to approach project performance from a composite per-spective for effective management(Dimitroff, Schmidt, & Bond, 2005) andsuggests that measures of strategicvalue beyond the triple constraint alsoneed to be considered to maximize per-formance (Norrie & Walker, 2004). Ourmeasure of improved project outcomestakes such a composite perspective andincludes indicators for time, budget,and scope in addition to factors associ-ated with team and leader perfor-mance: organizational learning (Robey& Sahay, 1996), improvement of teammembers’ skills and abilities (Bass &Avolio, 1994), client satisfaction, andoverall ability to complete the work(PMI, 2004). As project managers haveinherent bias in their reporting of suc-cess factors, in particular to overweighttime as a factor in failure (Dilts & Pence,2006), our composite measure helpsensure reduced bias in reporting per-formance. It is worth noting, of course,that any retrospective report of resultsmay also be subject to recollectionerror. Although this is common to mostempirical survey methods, in our casethe merger of subjective assessmentswith rich case reporting amelioratedthe risk of recall failure (Bendoly & Hur,2007). Respondents scored each meas-ure (on a 5-point Likert-type scale)according to the degree to which theybelieved their actions led to improve-ments or whether their actions weredetrimental. We aggregated theseresponses as a composite index forimproved project outcomes attributa-ble to the leaders’ actions.

Standard controls attempting toaccount for common issues such asproject size and scope, measured byitems such as the number of groupmembers involved, project duration,and average education level of groupmembers, were each taken into accountas controls in the analysis. Given ourspecific interest in the impact of projectmanager interventions in project inter-action, we also decided to include inour controls an index representing theextent to which various critical forms ofcommunication were available acrossthe project duration. This index wasessentially a sum product of projectmanager ratings of the criticality of avariety of communication channels(e.g., phone, fax, IM) and ratings of theavailability of these channels. The useof composite indices is not uncommonin the management literature and canbe particularly useful in serving as gen-eral controls for multiple relatedthough not synonymous externalities(Bharadwaj, Bharadwaj, & Bendoly,2007; Schroeder, Bates, & Junttila,2002). Our hope in including this infra-structural control, which we will refer togenerally as “infrastructure” in ouranalysis, was to mitigate the potentialmisrepresentation of our highly proac-tive interpersonal independent vari-ables with the effects of passive techni-cal communication infrastructuresinherent to the project setting.

Data CollectionOur data collection took place in multi-ple phases over the course of a 5-hourworkshop in which our subject poolvoluntarily participated. Our targetpopulation was practicing members ofthe Project Management Institute inattendance at the annual PMI GlobalCongress. The subjects were fully awarethat they would be providing input ontheir on-the-job activities prior to theworkshop and, hence, had the opportu-nity to opt out of the workshop alto-gether or to leave the workshop, effec-tively dropping out of our sample set, atany point. Regardless of these options,

our final sample represented a fairlylarge and representative sample of PMIproject managers with project effortsfocusing on IT infrastructure (22%) andorganizational systems improvement(16%), as well as construction (13%),research and development (R&D; 9%),and manufacturing improvement (6%).

Of 115 participating project man-agers, 77 provided complete data andconsent. As a further test of nonre-sponse bias, we compared the demo-graphics of respondent and firm attrib-utes (e.g., number of employees andsale volume) to those of the generalPMI constituency (Baruch, 1999). Nostatistical differences were detected,suggesting sufficient representation.Projects reported had an average of 3.4 organizations, a budget of U.S.$500,000, and a duration of 6 months toa year. Fifty-seven percent of projectteams were dispersed to the degree thatthey could not meet face-to-face regu-larly, and, characteristic of currenttrends in project management andmatrixed organizations, managersreported that, on average, few of theproject team members were undertheir direct supervisory authority(Hackman, 2002). Project teams con-tained at least seven members on aver-age, the point at which the proactivemanagement of team interaction andvalue of information technologiesappears to become much greater(Fjermestad & Hiltz, 1999). The respon-dents also reported that they were high-ly experienced professionals with anaverage of 9 to 10 years’ experience as a project manager. The team membersmeet our expectation for managingknowledge work in that, on average,managers reported that all members oftheir teams had a college education oradvanced training.

Although individual managerswere providing both dependent andindependent variable data, our designfor collection drew heavily on the sug-gestions outlined by past research in anattempt to ameliorate issues associatedwith a mono-respondent bias. In line

pmj402_20118.qxd 4/28/09 4:20 PM Page 74


with the suggestions of past authors(e.g., Miller & Roth, 1994; Podsakoff,MacKenzie, Lee, & Podsakoff, 2003), weprovided a clear separation betweenthe collection of dependent and inde-pendent variables by interminglingthese collection phases of our studywith actual nonsurvey activities thatthe participants in our study engagedin. This iteration served as an excellentmechanism by which to both provide aclear temporal separation betweenreports on factor items of interest, aswell as instill a psychological separa-tion to reduce the risk of additionalmental model development and factorassociations. Anonymity was ensuredthrough the use of a coding system, andthe notion that there were no “right orwrong” responses was reiteratedthroughout the collection period in anattempt to avoid norm-seeking behav-ior in responses (Das & Joshi, 2007;Podsakoff & Organ, 1986). The use offactor-item interspersing and reverse-coding of specific items was also appliedin an attempt to reduce halo-type

effects that might bias our data andanalysis.

Analysis and ResultsTo assess the extent to which the itemsreported on could be used to representeach of the three key independent fac-tors of interest to our study, factoranalysis was conducted. The results ofthe analysis following a principal com-ponents approach with varimax rota-tion are provided in Table 1.

As reported, all item loadings wereas predicted by their design and factorconceptualization, and above the rec-ommended levels established in the lit-erature on convergent validity (e.g.,Fornell & Larcker, 1981). Cronbach’salpha values for all three factors wereabove the 0.7 threshold recommendedas well (Nunnally & Bernstein, 1994).

In part, the distinctiveness of thesefactors along the items of interest initself provided support that commonmethod variance was not a major con-cern, at least not among the independ-ent variables. An additional unrotated

factor analysis inclusive of the primarydependent variable, as well, in accor-dance with Harman’s single-factor test,demonstrated that a total of four factors—the three independent factorsand the main independent variable—could be clearly distinguished(Podsakoff & Organ, 1986). This test hasbeen used increasingly in similar worksin this field (cf. Gowen, McFadden,Hoobler, & Tallon, 2006; Kaynak &Hartley, 2006; Klein, Rai, & Straub,2007). Overall, 66% of constituent vari-ance was accounted for across the fourfactors derived in this test, with the firstderived factor accounting for 17%,hence further reducing concern oversingle-source biases. As a final checkagainst such biases, we followed therecommendation of Bardhan (Bardhan,Krishnan, & Lin, 2007) in conducting amethods factor test (Williams, Cote, &Buckley, 1989), along with a directproduct model check (Bagozzi & Yi,1990).

Factor reliability on the dependentvariable (mean � 3.91, SD � 1.02) was

Factor 1: Factor 2: Factor 3: Enforcing Encouraging Intervening

Created new, explicit rules for interacting 0.734 0.120 0.104

Enforced existing explicit rules through penalties for violation 0.660 0.044 0.026

Enforced existing explicit rules through visibility 0.703 0.141 0.212

Mandated adherence to new or existing explicit rules for interacting 0.532 0.080 –0.034

Reminded team members of existing inexplicit rules/norms for interaction 0.162 0.706 –0.044

Encouraged team to share information with each other 0.192 0.798 0.057

Explained why it is important to share information in a timely fashion 0.073 0.874 0.055

Explained why it is important to share information that is accurate –0.056 0.856 –0.046

Manipulated team information or communications technologies 0.117 –0.072 0.892

Blocked usage of a communication technology 0.031 –0.086 0.719

Set up and made available a new additional communication technology –0.151 0.057 0.701

Reconstructed or reconfigured an existing communication technology 0.275 0.158 0.673

Cronbach’s alpha for item sets including only heavily loading 0.775 0.813 0.723(italicized score) items

Table 1: Factor loadings and alphas on selected item sets.

pmj402_20118.qxd 4/28/09 4:20 PM Page 75



AP

ER

S

also established, with a Cronbach’salpha on the measure of outcomeimprovement of 0.865. A last check ondiscriminant validity among the inde-pendent factors was provided throughan examination of the correlationsamong these factors. As shown in Table 2, while each had moderately sig-nificant correlations with the outcomemeasure, no correlations among theindependent factors appeared signifi-cant.

Regardless of seemingly low corre-lations among the key independentvariables, as a necessary step for testingthe interaction hypotheses of interest,we followed the standard guidelines formoderated regression analysis suggest-ed in the literature (Cohen, Cohen,West, & Aiken, 2003; Jaccard, Turrisi, &Wan, 1990). To minimize the risk ofmulticollinearity among interactionterms and their component main-effect terms, we first mean-centered allof the independent variables involvedin creating the interaction terms. Thevariance inflation factors in all caseswere substantially below the criticalvalue of 10, indicating that multi-collinearity was not an issue (Neter,Wasserman, & Kutner, 1990).

In further accordance with recentwork on interaction effects in projectmanagement settings (cf. Bendoly &Swink, 2007; Tatikonda & Rosenthal,2000), we adopted a hierarchical multi-variate regression approach to analysis.The stepwise increases in independentfactor variance accountability and

significance of the factors and interac-tions of interest are depicted in Table 3.

Demonstrated by the significantincreases in variance accountability inthe second step, the key factors indica-tive of leadership style activity (enforc-ing and encouraging) as well as struc-tural change (intervening) do seem tohave direct relevance to outcomeimprovement in projects involving criti-cal incidents. However, a test for differ-ences in the coefficients of enforcingand encouraging showed nonsignifi-cance. This was true for both the secondstep as well as the full model. As a result,the notion that Theory Y actions domi-

nate those of Theory X with respect toproject benefits (H1) could not be sup-ported.

The third step, or full model,demonstrates the particularly signifi-cant presence of nonlinear effects onoutcome improvement both in theform of leadership-style interactionand the quadratic intervention term(see Table 3). The coefficients on theleadership-style interaction (enforc-ing � encouraging) were negative andcomparable in magnitude to directeffects of either leadership style. Hence,contrary to some elements of populartheory, the hypothesis regarding poten-tial synergies drawn from the use ofmultiple styles (H2a) was not well sup-ported. When high levels of bothTheory Y– and Theory X–orientedactions are pursued, rather than pro-viding positive synergies, negativeeffects manifest. Our arguments lead-ing to our alternative hypothesis (H2b)certainly could shed some light on thiseffect; however, this alternative hypo-thesis cannot be viewed as entirely supported, either. As suggested by themagnitudes of the coefficients, it appearsthat a very modest level of activity

Improved Outcome Enforcing Encouraging Intervening

Improved Outcome 1.000

Enforcing 0.165* 1.000

Encouraging 0.251** 0.121 1.000

Intervening 0.163* 0.148 0.102 1.000

* p � 0.1. ** p � 0.05.*** p � 0.01.

Table 2: Pearson product moment correlations.

Step 1 Step 2 Step 3

Intercept –0.260 –0.517 –0.091

Infrastructure 0.461* 0.456* 0.464*

Enforcing 0.266* 0.475**

Encouraging 0.252* 0.517***

Enforcing � Encouraging –0.402**

Intervening –0.192* 0.805***

Intervening2 –0.799***

F for the step 10.948*** 5.907* 4.868*

F for the regression 10.948*** 6.601*** 5.873***

Adjusted R2 0.163 0.286 0.395

Note. Aside from the intercept, the main table contains standardized coefficient betas. No other controlsaside from our index of infrastructure were significant at any step.

* p � 0.05.** p � 0.01.*** p � 0.001.

Table 3: Hierarchical regression with outcome improvement as dependent variable.

pmj402_20118.qxd 4/28/09 4:20 PM Page 76


indicative of an opposing leadershipstyle may be leveraged to successfullyaugment a more dominantly appliedform.

To test further for the potential ofsuch dynamics in our data, since thereis no guarantee that a quadratic termfully or appropriately captures thehypothesized nonlinearities, we con-ducted a simple multivariate analysis ofvariance (MANOVA) comparison ofoutcome improvement among projectinterventions at alternating quartiles ofenforcing and encouraging. The signifi-cantly highest performance wasobserved among interventions simulta-neously positioned at both the secondlowest quartiles of one style and thehighest quartile of the other (p � 0.05).This result is again consistent with thesaddle dynamic on leadership stylesuggested by our hierarchical modelresults.

The other nonlinear dynamic indi-cated by our regression model relates tothe role of structural intervention activ-ities. In support of H3, such actions doappear to be well represented by aninverted-U dynamic (positive beta onmain effect, negative beta on quadrat-ic), more so than a simple lineardynamic. Furthermore, this effect isdistinct from and significant above andbeyond effects of leadership-styleactions, as indicated by the incrementalR-square levels, beta significance lev-els, and the convergence and discrimi-nant analysis presented earlier.

These results suggest that the use ofstructured interventions is necessarywhen leaders encounter critical inci-dents of interaction breakdown. Anadditional test for interactions betweenintervening and the styles failed toshow additional significance. This find-ing suggests that interventions employ-ing either dominant style are equally aseffective regardless of the amount ofintervention, and casts doubt on priorpractitioner expectations that leaderswho employ a Theory X style may bemicromanagers who intervene morethan Theory Y–oriented leaders, at least

within the context of responding to crit-ical incidents.

DiscussionIn this study, we expand the currentbody of knowledge on project manage-ment by empirically examining howleaders may be more effective interven-ing to improve project outcomes. Thisarea has been neglected in existingresearch and requires an examinationof leader actions during their interven-tions, particularly during critical inci-dents, somewhat ephemeral and elu-sive occasions, which are very difficultto observe and sample in practice dueto their existence in computer-mediat-ed and often distributed team contexts.With a focus on these contexts, we havere-examined the notion that the stylesimplied by Theory X/Y are exclusive.Through our analysis, we demonstratethe potential synergy of the styles,counter to traditional interpretations ofTheory X/Y. Leaders may employ someenforcing or encouraging actions inaddition to their actions from the otherapproach during interventions and inso doing enjoy particularly positive out-comes. At the same time, however, thecounterargument posed by the litera-ture on leadership under behaviorcomplexity (i.e., that leaders need to beultimately flexible and responsivelyselecting their behaviors to fit the con-texts in which they act) does not seemto be entirely applicable here. In a criti-cal incident intervention scenario, proj-ect managers need to communicate acoherent or at least dominant identityto their teams. Ultimately, the keyappears to be one of focusing on a sin-gle approach predominantly character-ized by either enforcing or encouragingactions, and then augmenting such anapproach with modest and selectiveuse of a contrasting one. This findinghas implications for how we train futureleaders. As a follow-up, when we infor-mally polled 40 MBA students, havingexplained the option of a leader inter-vening in a critical incident with eithera predominantly Theory X or Y or mixed

approach, two responded that Theory Xwould be best, seven responded thatTheory Y would be best, and 31responded that a mix would be best.Our evidence suggests that the basicnotions they now retain will not servethem well for interventions in contextsthat are increasingly common to every-day business. It may be that for day-to-day leadership, a mixed approach isbest on a generalized basis. Within spe-cific interventions, this may not be true.Albeit limited to a sample size of 77complete records, the analysis of ourdata indicates that it is not true for crit-ical incident interventions.

Based on existing literature on lead-ership style and effectiveness, we mightwonder whether leaders’ assumptionsabout workers impact their perfor-mance as leaders. Some research overthe past several decades on the topic ofleadership style and approach supportsthis notion. Is the same true aboutworkers? Do worker assumptions abouttheir leaders impact their performance?

We found no studies exploring thisnotion, but it stands to reason that theinverse of leaders’ assumptions aboutworkers impacting their behavior may be true, especially the more werecognize the ability of workers toautonomously think and judge. Thus,among professional, educated workerstrained to a degree of specialization intheir own fields, we can expect that theywould make assumptions about theirleaders and act as followers accordingto these assumptions. To the degreethat this is true, our findings that lead-ers may be equally effective takingeither a Theory X or Y approach, butfocusing on one approach and onlymodestly drawing on the other for aug-mentation, does seem to make sense.Team members need to be able tocoherently understand the expecta-tions of behavior being changed byleaders during interventions. It appearslikely that, in agreement with the theo-ries supporting H2b, team membersdraw on their understanding of project-manager identity in interpreting leaders’

pmj402_20118.qxd 4/28/09 4:20 PM Page 77



AP

ER

S

actions. This enables them to follow andadapt more effectively during criticalincident interventions aimed at improv-ing team interaction and productivity.What is needed is a comprehensiveempirical examination of both manage-rial actions and work reactions duringinterventions to validate this notion.

Above and beyond interpersonalactions as indicative of leadership style,we have also contributed to the literatureon project management by demonstrat-ing the distinctive role of interveningstructural actions that may benefit out-come improvements in critical incidentscenarios. The nonlinear nature of therelationship observed between interven-ing actions and outcome suggests a cru-cial trade-off between the importance ofbeing an active manager involved in the operations of team interaction and theimportance of not overly disrupting thecontext into which future work will tran-sition. Too little intervention has littlechance of making a positive impact,while too much intervention may lead tostill worse performance.

However, aside from these findings,one might ask what are the warningsigns of too much intervention in teaminteraction design? Some indicationscome from the cases in which we seesuch activity to be high. Although ourgoal here was not to investigate the sub-causes of critical incidents, in the caseswhere several types of actions weretried, we consistently see that incidentswere characterized by trust and relation-ship problems involving two or moregroups within a team not cooperatingwith each other. These sorts of incidentsinvolving unwillingness to cooperateappear particularly challenging forinterventions and may require the mosttypes of action by leaders. Our mathe-matical analysis suggests that some-where between three and six types ofaction overall (project managers indicat-ed the types of actions they took—wedid not directly observe and count theexact number of actions taken due todesign considerations discussed earlier)characterize the sweet spot for interven-

tion. The aforementioned incidentsinvolving cooperation breakdown werecharacterized by at least 16 types ofactions each.

ConclusionOur research demonstrates the impor-tance of examining the behavior ofproject managers and, in particular, theimportance of specific choices in lead-ership styles and interventions aimedat successful outcomes. Contrary tosome indications in literature about thevalue of an overly customized approachto leading and concerning the impor-tance of a facilitative approach forknowledge workers, our data reveal thatleaders need to clearly demonstrate adominant approach during critical inci-dent interventions. That approach maybe either Theory X– or Y–dominated butmay gain from a relatively minor degreeof approach modification. Moreover, assuggested by situational theories suchas that of Fiedler (1967), effective lead-ership cannot be expected to flourish inthe presence of excessive interventions.Managers should be forewarned totemper their tactics, as well as remainconscious of the dominant leadershipstyle they chose to apply during criticalincidents. ■

ReferencesAndrews, E. (2007, August). Pulse ofthe profession. Presented at the PMIGlobal Congress, Atlanta, GA.

Bagozzi, R. P., & Yi, Y. (1990). Assessingmethod variance in multitrait-multi-method matrices: The case of self-reported affect and perceptions atwork. Journal of Applied Psychology, 75,547–560.

Bardhan, I. R., Krishnan, V., & Lin, S.(2007). Project performance and theenabling role of information technolo-gy: An exploratory study on the role ofalignment. Manufacturing & ServiceOperations Management, 9, 579–595.

Barki, H., & Hartwick, J. (2001).Interpersonal conflict and its manage-ment in information systems develop-ment. MIS Quarterly, 25, 195–228.

Baruch, Y. (1999). Response rate in aca-demic studies: A comparative analysis.Human Relations, 52, 421–438.

Bass, B. M., & Avolio, B. J. (1994).Improving organizational effectivenessthrough transformational leadership.Thousand Oaks, CA: Sage.

Behfar, K. J., Peterson, R. S., Mannix,E. A., & Trochim, W. M. K. (2008). Thecritical role of conflict resolution inteams: A close look at the links betweenconflict type, conflict managementstrategies, and team outcomes. Journalof Applied Psychology, 93(1), 170–188.

Bendoly, E., & Hur, D. (2007).Bipolarity in reactions to operational“constraints”: OM bugs under an OBlens. Journal of OperationsManagement, 25(1), 1–13.

Bendoly, E., & Swink, M. (2007).Moderating effects of informationaccess on project management behav-ior performance and perceptions.Journal of Operations Management,25, 604–622.

Bennis, W., Heil, G., & Stephens, D.(2000). Douglas McGregor, revisited:Managing the human side of the enter-prise. New York: Wiley.

Bharadwaj, S., Bharadwaj, A. S., &Bendoly, E. (2007). The performanceeffects of complementarities betweeninformation systems, marketing, man-ufacturing, and supply chain process-es. Information Systems Research, 18,437–455.

Bobic, M. P., & Davis, W. E. (2003). Akind word for Theory X: Or why somany newfangled management tech-niques quickly fail. Journal of PublicAdministration Research and Theory,13, 239–264.

Bredillet, C. (2007). From the editor.Project Management Journal, 38(2),3–4.

Cannon-Bowers, J. A., Salas, E., &Converse, S. A. (1993). Shared mentalmodels in expert team decision mak-ing. In N. J. J. Castellan (Ed.), Currentissues in individual and group decisionmaking (pp. 221–246). Hillsdale, NJ:Erlbaum.

pmj402_20118.qxd 4/28/09 4:20 PM Page 78


Chidambaram, L., Bostrom, R. P., &Wynne, B. E. (1991). A longitudinalstudy of the impact of group decisionsupport systems on group develop-ment. Journal of ManagementInformation Systems, 7(3), 7–28.

Cohen, J., Cohen, P., West, S. G., &Aiken, L. S. (2003). Applied multipleregression/correlation analysis for thebehavioral sciences. Philadelphia:Erlbaum.

Das, S. R., & Joshi, M. P. (2007). Processinnovativeness in technology servicesorganizations: Roles of differentiationstrategy, operational autonomy andrisk-taking propensity. Journal ofOperations Management, 25, 643–660.

Davenport, T. H. (2005). Let them allbe power users. The HBR list: Break-through ideas for 2005. HarvardBusiness Review, 83(2), 41–42.

Denison, D. R., Hooijberg, R., & Quinn,R. E. (1995). Paradox and performance—Toward a theory of behavioral com-plexity in managerial leadership.Organization Science, 6, 524–540.

Dennis, A. R., & Valacich, J. S. (1999).Rethinking media richness: Towards atheory of media synchronicity.Proceedings of the 32nd HawaiiInternational Conference on SystemSciences, Hawaii, USA (pp. 1–10).

Dilts, D. M., & Pence, K. R. (2006).Impact of role in the decision to fail:An exploratory study of terminatedprojects. Journal of OperationsManagement, 24, 378–396.

Dimitroff, R. D., Schmidt, L. A., &Bond, T. D. (2005). Organizationalbehavior and disaster: A study of con-flict at NASA. Project ManagementJournal, 36(2), 28–38.

Fiedler, F. E. (1967). A theory of leadership effectiveness. New York:McGraw-Hill.

Fjermestad, J., & Hiltz, S. R. (1999). Anassessment of group support systemsexperimental research: Methodologyand results. Journal of ManagementInformation Systems, 15(3), 7–149.

Fornell, C. R., & Larcker, D. F. (1981).Evaluating structural equation models

with unobservable variables and meas-urement error. Journal of MarketingResearch, 18(1), 39–50.

Gowen, C. R., McFadden, K. L.,Hoobler, J. M., & Tallon, W. J. (2006).Exploring the efficacy of healthcarequality practices, employee commit-ment, and employee control. Journal ofOperations Management, 24, 765–778.

Hackman, J. R. (2002). Leading teams:Setting the stage for great performances.Boston: Harvard Business School Press.

Halpin, A. W. (1966). Theory andresearch in administration. New York:Macmillan.

Hanson, E. M. (1996). Educationaladministration and organizationalbehavior. Boston: Allyn and Bacon.

Hinds, P. J., & Mortensen, M. (2005).Understanding conflict in geographi-cally distributed teams: The moderat-ing effects of shared identity, sharedcontext, and spontaneous communica-tion. Organization Science, 16, 290–307.

Hooijberg, R., Hunt, J. G., & Dodge,G. E. (1997). Leadership complexityand development of the Leaderplexmodel. Journal of Management, 23,375–408.

Jaccard, J., Turrisi, R., & Wan, C. K.(1990). Interaction effects in multipleregression. Newbury Park, CA: Sage.

Kaynak, H., & Hartley, J. L. (2006).Using replication research for just-in-time purchasing construct develop-ment. Journal of OperationsManagement, 24, 868–892.

Klein, R., Rai, A., & Straub, D. (2007).Competitive and cooperative position-ing in supply chain logistics relation-ships. Decision Sciences, 38, 611–646.

Linderman, K., Schroeder, R. G., &Choo, A. S. (2006). Six Sigma: The roleof goals in improvement teams.Journal of Operations Management, 24,779–790.

Lipnack, J., & Stamps, J. (2000). Virtualteams: People working across bound-aries with technology. New York: Wiley.

Majchrzak, A., Rice, R. E., Malhotra, A.,King, N., & Ba, S. (2000). Technology

adaptation: The case of a computer-supported inter-organizational virtualteam. MIS Quarterly, 24, 569–600.

Massey, A. P., & Montoya-Weiss, M. M.(2006). Unraveling the temporal fabricof knowledge conversion: A model ofmedia selection and use. MISQuarterly, 30, 1–22.

Mathieu, J. E., Goodwin, G. F., Heffner,T. S., Salas, E., & Cannon-Bowers, J. A.(2000). The influence of shared mentalmodels on team process and perfor-mance. Journal of Applied Psychology,85, 273–283.

McGregor, D. (1960). The human sideof the enterprise. New York: McGraw-Hill.

McGregor, D. (2006). The human sideof management (Annotated ed.). NewYork: McGraw-Hill.

Miller, J. G., & Roth, A. V. (1994). A tax-onomy of manufacturing strategies.Management Science, 40, 285–304.

Mintzberg, H. (1998). Covert leader-ship: Notes on managing profession-als. Harvard Business Review, 76(6),140–147.

Naveh, E. (2007). Formality and discre-tion in successful R&D projects.Journal of Operations Management,25(1), 110–125.

Neter, J., Wasserman, W., & Kutner,M. H. (1990). Applied linear statisticalmodels: Regression, analysis of variance,and experimental design. Boston: Irwin.

Norrie, J., & Walker, D. H. T. (2004). Abalanced scorecard approach to proj-ect management leadership. ProjectManagement Journal, 35(4), 47–56.

Nunnally, J. C., & Bernstein, I. H.(1994). Psychometric theory. New York:McGraw-Hill.

Piccoli, G., & Ives, B. (2003). Trust andthe unintended effects of behaviorcontrol in virtual teams. MIS Quarterly,27, 365–395.

Podsakoff, P. M., MacKenzie, S. B., Lee,J. Y., & Podsakoff, N. P. (2003).Common method biases in behavioralresearch: A critical review of the literature and recommended remedies.

pmj402_20118.qxd 4/28/09 4:20 PM Page 79



AP

ER

S

Journal of Applied Psychology, 88,879–903.

Podsakoff, P. M., & Organ, D. W. (1986).Self-reports in organizational research:Problems and prospects. Journal ofManagement, 12, 531–544.

Project Management Institute (2004).A guide to the project managementbody of knowledge (PMBOK ® guide)—Third edition. Newtown Square, PA:Author.

Robey, D., & Sahay, S. (1996). Trans-forming work through informationtechnology: A comparative case studyof geographic information systems incounty government. InformationSystems Research, 7(1), 93–110.

Schroeder, R. G., Bates, K. A., &Junttila, M. A. (2002). A resource-basedview of manufacturing strategy and therelationship to manufacturing per-formance. Strategic ManagementJournal, 23(2), 105–117.

Shin, S. J., & Zhou, J. (2007). When iseducational specialization heterogene-ity related to creativity in research anddevelopment teams? Transformationalleadership as a moderator. Journal ofApplied Psychology, 92, 1709–1721.

Sinha, K. K., & Van de Ven, A. H. (2005).Designing work within and betweenorganizations. Organization Science,16, 389–408.

Stuart, F. I., & Tax, S. (2004). Toward anintegrative approach to designing serv-ice experiences: Lessons learned fromthe theatre. Journal of OperationsManagement, 22, 609–627.

Tatikonda, M. V., & Rosenthal, S. R.(2000). Successful execution of prod-uct development projects: Balancingfirmness and flexibility in the innova-tion process. Journal of OperationsManagement, 18, 401–425.

Thomas, D. M., & Bostrom, R. P. (2008).Building trust and cooperationthrough technology adaptation in vir-tual teams: Empirical field evidence.Information Systems Management,25(1), 45–56.

Thomas, D. M., Bostrom, R. P., &Gouge, M. (2007). Making knowledgework successful in virtual teams viatechnology facilitation. Communica-tions of the ACM, 50(11), 85–90.

Williams, L. J., Cote, J. A., & Buckley,M. R. (1989). Lack of method variancein self-reported affect and perceptionsat work: Reality or artifact? Journal ofApplied Psychology, 74, 462–468.

Dominic Thomas joined the Goizueta BusinessSchool in 2005 after receiving his PhD in man-agement information systems from theUniversity of Georgia. Prior to pursuing his PhD,

he worked in education, education systemreform, and international development in theUnited States, Russia, Japan, and Nepal. He haspersonally managed large, multinational proj-ects in Nepal and Japan. He has served as anadvisor for two large-scale information systemsprojects, one involving implementation of anenterprise resource planning system in a globalenergy company and one involving re-engineer-ing the core processes and systems with anobject-oriented architecture at a governmentagency. His research is in press at the Journal ofthe ACM, Information Systems Research, and theJournal of Management Information Systems.

Elliot Bendoly is an associate professor atEmory University’s Goizueta Business School.He holds a PhD from Indiana University in opera-tions management and decision sciences, withan information systems specialization. Heserves on the editorial boards of the Journal ofOperations Management and Decision Sciences.His research has been published in numerouselite academic journals, including Production andOperations Management, Information SystemsManagement, MIS Quarterly, the Journal ofApplied Psychology, the International Journal of Operations and Production Management,Information & Management, Journal of OperationsManagement, Decision Sciences Journal, andDecision Support Systems. He is also coeditor of Strategic ERP Extension and Use and author ofExcel Basics to Blackbelt.

pmj402_20118.qxd 4/28/09 4:20 PM Page 80

PA

PE

RS


INTRODUCTION ■

Project managers have traditionally relied on planning tools andexperience to predict the performance of construction projects (Jin &Levitt, 1996). Popular project planning tools, such as the critical pathmethod (CPM) and the program evaluation and review technique

(PERT), have not been successful in planning projects that are subject todynamic uncertainties (Vaziri, Carr, & Nozick, 2007). A CPM schedule canalert the project manager to the activities most impacted by unexpectedchanges; however, it does not provide project managers with a means forplanning for that uncertainty. PERT considers the impact of uncertaintybased on previous project experience; however, uncertainties vary widelyfrom project to project and over the life cycle of a project. Project managersneed better tools and methods to plan projects reliably, given the highdegree of uncertainty endemic to the project execution environment.

Uncertainties on construction projects are generated from both internaland external sources stemming from the project organization and the build-ing environment (Kolltveit, Karlsen, & Gronhaug, 2004). Ben-Haim andLaufer (1998) argued that uncertainty has become one of the major factorsinfluencing project performance and successful project delivery. In this arti-cle, we develop an integrated approach that project managers can use todesign project networks that perform reliably in dynamically uncertain proj-ect environments. To achieve this, we integrate a project organizational designcomputational simulation model (Virtual Design Team) and robustdesign experimental method to develop a robust project network designapproach. We apply this approach in the context of a specific test case proj-ect to examine three interrelated research questions:1. Can we integrate a validated project organization simulation tool and a

validated robust design methodology?2. Can we employ the integrated approach to design a robust combination of

interventions for a case project?3. Can the robust project network design framework described in this article

produce results that extend the capabilities of project managers to designproject networks in dynamically uncertain project environments?

BackgroundProject Organization DesignOrganization design research focuses on the conception and develop-ment of organizations whose targets have been established in advance(Baligh, Burton, & Obel, 1996). Lawrence and Lorsch (1967) first arguedthat organizational designs are contingent on the business environment.They found that an organization’s fit with its environment determined its

Robust Project Network DesignDavid A. Ortiz de Orue, The Shaw Group Inc., Baton Rouge, LA, USAJohn E. Taylor, School of Engineering and Applied Science, Columbia University,New York, NY, USAArpamart Chanmeka, University of Texas at Austin, TX, USARuni Weerasooriya, Spire Consulting Group, LLC, Austin, TX, USA

ABSTRACT ■

Traditional project planning tools and method-ologies fail to address the dynamic uncertain-ties that can arise during project execution. Inthis article, we develop an approach to enableproject managers to design project networkscapable of performing predictably under uncer-tainty. We integrate a validated project organi-zation design simulation tool (Virtual DesignTeam) with a robust design experimentalmethod to enable robust project networkdesign. We test the framework on a 13-month,U.S. $11.5M public renovation project. We findthat: (1) it is possible to integrate the projectorganization design tool with the robust designexperimental method for the test case project;(2) the integrated approach enables projectmanagers to identify and simulate the combina-tion of interventions that will lead to the mostrobust project performance; and (3) the inte-grated approach can extend the capabilities ofproject managers to design project networks indynamically uncertain project environments.The findings highlight the need to explore newways to achieve predictable project perform-ance in dynamically uncertain project execu-tion environments.

KEYWORDS: project networks; robustdesign; simulation; uncertainty





DOI: 10.1002/pmj.20109

pmj402_20109.qxd 4/28/09 5:51 PM Page 81


Robust Project Network DesignP

AP

ER

S

performance. Thompson (1967) extend-ed organizational design theory byconsidering the interactions betweenactivities in the task environment. Theextent of activity interdependenceimpacts where organizational bound-aries are structured to achieve optimalperformance. Galbraith (1973) addresseduncertainty in organizations by examin-ing how organizations make appropriateplanning decisions and how informa-tion is processed within organizations.According to Galbraith, uncertaintiesare proportional to the amount of infor-mation processed by project partici-pants. He concluded that organizationaldesigns must consider the massiveamount of information organizationsmust process along with the low infor-mation-processing capacity of mostorganizations.

Organization design researcherslaid the foundation for project organi-zation design research. The projectorganization design approach focuseson assessing the work capacity, coordi-nation, and communication of eachproject participant and their ability todeal with a variety of managerial chal-lenges. These challenges include theactivity interdependencies and infor-mation-processing issues described byearly organization design researchers.Project organization design researchwas reified into a computational simu-lation model by Levitt and colleaguescalled the Virtual Design Team (Jin &Levitt, 1996).

Virtual Design Team SimulationResearchThe Virtual Design Team (VDT) is anobject-oriented, discrete event-drivensimulation system that enables users tocombine models of project organiza-tions and project work plans by allowingspecific project tasks to be assigned tospecific project participants (Jin & Levitt,1996). VDT was originally developedto quantify the additional coordina-tion load and rework required whenproject schedules are compressed (Kunz,Christensen, Cohen, Jin, & Levitt, 1998).

When the project organization andwork plan are combined in the VDTmodel, the project participants’ microlev-el actions are simulated and aggregatedto produce simulated output measuresof project progress (e.g., quality, cost,and time). The additional coordinationrequired by schedule compression isthen represented by, for example,extensive project participant work back-log, missed meetings or deadlines, orincreased quality risk of project activi-ties executed by teams with significantwork backlog.

When VDT is used to model a con-struction project, project managers ini-tially develop a baseline schedule for aproject based on traditional planningmethods. Unlike traditional planningmethods, however, VDT predicts coor-dination requirements and rework thatarise during concurrent work. Toachieve this, users input activity modelvariables (e.g., activity durations, failuredependencies between activities, activ-ity precedence, and required skill level),actor model variables (e.g., number ofteam members, skill level, responsibili-ty assignments, role in organizationalhierarchy, and task experience), andproject model variables (e.g., degree ofcentralization of decision making,degree of formalization of communica-tions, communication requirements,and team experience working togeth-er). VDT then predicts organizationalbreakdowns and generates alternativesto remodel the project organizationdesign (Levitt et al., 1999). AlthoughVDT predicts project breakdowns andassesses the performance of projectparticipants, it does not address thedynamic nature of project uncertain-ties. Project managers can revisit theschedule as uncertainties that impactproject execution arise. However, VDTis limited in the extent to which it canpredict the combination of interven-tions that will minimize sensitivity todynamic uncertainties that arise duringproject execution.

VDT has been used by research-ers as a simulation tool to extend

organizational theories and examine arange of factors that impact projectperformance. Levitt and colleagues(1999) applied VDT to examine thebridge between psychological micro-organization theory and socioeconom-ic macro-organizational theory forproject teams. This research was under-taken through the VDT modeling of acomplex satellite launch vehicle.Thomsen, Levitt, and Nass (2004) uti-lized VDT to examine the execution ofless routine tasks that have consider-able flexibility in how they are per-formed. In doing so, they extendedGalbraith’s (1977) information process-ing model. Horii, Jin, and Levitt (2005)used VDT to examine the impact ofvarying national culture on projectorganization performance, comparinginternational joint ventures composedof Japanese and U.S. constructionteams. Using VDT, they were able todraw a distinction between practice-based differences and value-baseddifferences in terms of project perfor-mance. Khosraviani, Levitt, and Koza(2004) were able to combine geneticprogramming with VDT in order to findnear-optimal solutions to optimizeproject performance considering arange of static uncertainty variables.Finally, Ibrahim and Nissen (2007) usedVDT to examine the interrelationbetween knowledge flow theory andorganizational design in an affordablehousing project case to extend organi-zational theories to address discontinu-ous organizations. In this article, wefollow in the tradition of using VDT as avalidated simulation tool upon whichto base a research investigation. Wecombine the use of VDT with an actualbuilding construction project to exam-ine the question of how robust designexperimental methods can be used todesign project organizations that per-form reliably in dynamically uncertainproject execution environments.

Robust Design Experimental MethodTaguchi developed the robust designmethod, which has been hailed as one

pmj402_20109.qxd 4/28/09 5:51 PM Page 82


of the most notable contributions toquality management (Antony & Antony,2001; Bendell, Disney, & Pridmore,1989; Peace, 1993). Robust design, alsoknown as the Taguchi method, hasbeen used to improve quality in thefabrication and production of numer-ous manufactured products (Antony &Antony, 2001; Bendell et al., 1989). Thegoal of robust design is to arrive at atargeted level of quality by developingstable products that perform consis-tently under varying degrees of uncer-tainty (Antony & Antony, 2001). Robustdesign methods reveal the optimalcombination of controllable factor lev-els that leads to the design of a productor process that is least sensitive to theinfluence of uncontrollable factors(Kwon & Lin, 2005; Taguchi & Clausing,1990). Robust design involves (1) iden-tification of the most relevant charac-teristics and quality drivers of theproduct or process; (2) identification ofthe controllable and uncontrollablefactors; (3) selection of the controllablefactor-setting levels along with possi-ble combinations to diminish the effectsof noise; and (4) assessment of eachcombination’s response at differentsetting levels to determine which oneyields the most robust product or processdesign.

The manufacturing industry hasleveraged Taguchi’s robust designmethod to enhance product and processquality. However, this is not the case inmany industries that adopt project net-work forms of organization, which havein many cases been slow to adopt man-ufacturing best practices to improveproject performance. Some best prac-tices aimed at improving quality per-formance include lean production andSix Sigma. Lean production, which hasbeen adopted by the constructionindustry as lean construction, is aprocess that aims to improve perform-ance by eliminating waste and unnec-essary work (Beary & Abdelhamid, 2005;Salem, Solomon, Genaidy, & Minkarah,2006). Motorola developed the Six Sigmaprocess as a means for detecting and

eliminating product performance vari-ability by using statistical methods. SixSigma methods have been used in theconstruction industry to improve proj-ect efficiency and quality (Beary &Abdelhamid, 2005). However, neither ofthese manufacturing practices thathave been implemented in architec-ture, engineering, and constructionprescriptively minimizes the effect ofdynamic uncertainties that can ariseduring project execution.

By implementing robust design inthe architecture, engineering, and con-struction industry and other industriesthat adopt project network forms oforganization, project networks can bedesigned to perform more reliably indynamically uncertain project environ-ments. The selection of the best combi-nation of factor-level settings for arobust design is based on the signal-to-noise ratio (S/N) measure of perfor-mance. In general, the higher the S/Nvalue, the lower the variation from thetarget and the better the performanceof the project. Taguchi developed anumber of equations to determine theS/N value associated with each combi-nation of controllable factors. EachS/N equation is tailored to the mainproduct or process goals. The mainobjective of a robust design would beto yield a project network with minimalvariance while maximizing perfor-mance. To accomplish this, the control-lable factors must be adjusted until themean of the project performancevalues (time, cost, and quality) isimproved with minimal variance. Thecharacteristic value of the S/N ratio isdetermined by applying the followingset of “smaller is better” Taguchi equa-tions (Roy, 1990):

(1)

(2) Sm �(�n

i�1 yi)

2

n

Sm � an

i�1

y 2i

n

� 10 #

Log10.1

n#

Sm � Ve

Ve

S�N � �10 * Log10(Sm)S

N

where:yi � objective function value per com-

bination,n � the number of data points,Sm � sum of the squares of the mean,

andS/N � signal-to-noise ratio.

The objective function (yi) is a valuethat measures the performance of aproject by considering its most relevantcharacteristics, such as cost, time, andquality. An objective function is deter-mined for each combination of control-lable factors. The objective function forproject networks is specified in moredetail for the case project examined inthis article in the following sections.

Project NetworksThe utilization of project networks as ameans for organizing industrial pro-duction is growing across industries(Kanter, 1991). In project networks,groups of two or more firms worktogether in the interdependent produc-tion of goods or services (Powell, 1990).Researchers have begun to examinearchitecture, engineering, and con-struction (A/E/C) project networksfrom various perspectives. Project net-work researchers examining A/E/Cproject networks have investigated theimpact of this form of organization oninnovation (Harty, 2005; Taylor & Levitt,2004, 2007), cross-organizational tech-nology implementation (Taylor, 2007),information dependency (Pekericli,Akinci, & Karaesmen, 2003), learning(Taylor, Levitt, & Villarroel, 2006), andleadership (Chinowsky & Taylor, 2007).However, we lack research on how todesign a robust project network. If theuse of project networks is growingacross industries and uncertainty is akey issue in designing effective projectnetworks, then we need to develop newapproaches to improve project networkdesign under conditions of uncertainty.In this article, we combine projectorganization design and robust design

Ve �(�n

i y2i ) � Sm

n � 1

pmj402_20109.qxd 4/28/09 5:51 PM Page 83



AP

ER

S

methods to develop a robust projectnetwork design approach that enablesproject managers to design projectnetworks that perform reliably indynamically uncertain project execu-tion environments.

Research MethodologyOur first research question examineswhether integrating validated projectorganization design and robust designapproaches is possible. Previousresearchers have docked VDT withother methodological approaches toexamine questions of whether suchintegration is possible. Horii et al. (2005)succeeded in integrating Hofstede’s(1991) cultural values framework withVDT and used the simulated output toenable project managers to examine theimpact of cross-cultural factors on proj-ect performance. Khosraviani et al.(2004) were able to integrate geneticprogramming algorithms, which theythen proved outperformed users of thesimulation at identifying the most effi-cient project organization design basedon a set of static uncertainties. We fol-low this tradition by attempting to dockVDT with the robust design method toascertain whether such a combinedframework can enable project man-agers to identify the most robust projectnetwork design in the presence ofdynamic uncertainty.

To develop the framework, we beganby modeling a building renovationproject as a test case in VDT. Workingwith the project manager, we estab-lished the activity model, the actormodel, the project model, and the base-line levels of uncertainty on the project.The development of the VDT model isdescribed in detail in the following sec-tion. The activity, actor, and projectmodels were then combined to deter-mine the baseline project networkmodel to be simulated. Once the base-line VDT model was established, weadded in the identified static uncer-tainty level and identified the criticalbottlenecks on the project. As will bediscussed in the following section, we

found that the mechanical and drywallsubcontractors were predicted to havesignificant work backlog and, therefore,should be the focus of interventions inthe robust project network designexperiment. We then worked with theproject manager to determine an objec-tive function for the project.The objectivefunction is the weighted and normal-ized combination of factors (cost, dura-tion, quality) that correspond to theowner’s expectations for the projectperformance. This output value wasthen used to compare project perfor-mance in a robust design experiment.

In order to conduct the robustdesign experiment with the VDT simu-lated output, we needed to identify a setof interventions that project managerswould feasibly employ in a project toaddress the delays in the drywall andmechanical contractor’s work. We inter-viewed project managers to design a setof possible interventions to improveperformance on the case project reli-ably. Interviewees were given the proj-ect schedule and the objective functionand were asked to suggest interven-tions that would best enable achievingthe project objectives while perform-ing reliably when unforeseen changesoccurred in the project executionenvironment. Interventions wereapplied to enhance project perfor-mance by altering the baseline model’sproject network, resource allocation,and work process. We then ran eachintervention across a range of uncon-trollable factor levels to arrive at themost robust combination of control-lable factors. It is the range of uncer-tainty levels that we test in the robustdesign experiment that extends VDTto consider dynamic uncertaintiesthat are prevalent in the building proj-ect execution environment.

We next sought to understandwhether our robust project networkdesign framework extends the capabili-ties of project managers. To answer thisquestion, we interviewed 20 projectmanagers. We asked them to select thecombination of controllable factors

(at this point, these were limited tochanges in crew size and experiencelevel in the drywall and mechanicalsubcontractor organizations) thatwould lead to the most robust projectperformance in the presence of dynam-ic uncertainties for the case project. If asignificant proportion of the projectmanagers failed to identify the mostrobust combination of controllable fac-tor interventions produced by thesimulation, then there is evidence tosupport the assertion that the robustproject network design approach canextend the capabilities of project man-agers to design robust project networkorganizations.

Integrating ProjectOrganizational Designand Robust DesignCase StudyIn order to test whether an integratedrobust project network design approachis possible, we modeled a specific testcase project. We chose a building reno-vation project and used VDT to simu-late the renovation of the top two levelsof a U.S. $11.5M four-story public reno-vation project with a scheduled projectduration totaling 13 months. Thepurpose of modeling this building ren-ovation project was not to test VDT’scapabilities. VDT is a well-validatedproject organization simulation toolthat has been used extensively inresearch and by industry. Rather, thebuilding renovation project provided uswith a test case from which we coulduse the simulated outputs relating toschedule, cost, and quality to producean objective function to be used in arobust design experiment. Using a realproject as a test case also enabled us to:(1) discuss the case with project man-agers to identify appropriate interven-tion strategies for the robust designexperiment and (2) share the case withother project managers with relevantexperience to ascertain through inter-views whether the robust project net-work design approach extends thecapabilities of project managers.

pmj402_20109.qxd 4/28/09 5:51 PM Page 84


Project Manager A (PM-A) facilitat-ed our study of the building renova-tion project by providing schedule,cost, labor, material, equipment, andmanagerial data. PM-A led theresearch team on several site visits,during which he identified the mostcritical elements of the project.Many considerations involved updat-ing mechanical work, establishingAmericans With Disabilities Act (ADA)compliance, and preserving the build-ing’s historical value. The project wasdesigned to be performed concurrent-ly for all four levels of the building.From the project schedule provided byPM-A, we modeled the third- andfourth-floor baseline cases with 23 and25 construction activities, respectively.Each floor had two absolute, interimmilestones (initial demolition com-plete and cover-up), and meetings wereassigned in the VDT model as specifiedby PM-A. The budgeted amount forthese two floors was U.S. $3.6M, andthe total project durations for the thirdand fourth floor were 184 and 179days, respectively. The structure ofthe project was such that PM-A was inthe position of highest authority, fol-lowed by the project engineer and twosite superintendents. In addition tothe general contractor’s personnel,the project network was composed ofeight subcontractor organizations,simultaneously working on the toptwo floors. The subcontractors on thejob directly reported to the projectengineer.

With the assistance of PM-A, taskswere assigned to subcontractors, allcrew members were set at the mediumexperience level, and activity durationswere assigned. Project cost was alsomodeled using the cost data providedby the general contractor. Once the ini-tial architecture of the baseline modelfor the case was developed and the datawere modeled, we ran the simulation.The results reflected the project planningand resource allocation choices origi-nally made by the project manager. Infact, the simulation results predicted

the same total project duration as PM-Ahad estimated using the critical pathmethod. However, neither communica-tion and rework issues nor uncontrol-lable factors had yet been taken intoaccount.

Communication and Rework LinksA refined project network model wasdeveloped by adding communicationand rework links to the baseline modelas illustrated in Figure 1. Activitiesrelated to MEP systems associated bycommunication links presume thatmechanical, electrical, and plumbingsubcontractors must communicatewith each other before or during theexecution of MEP-related activities(e.g., plan revision meetings to avoidmaterial interferences). Activities con-nected by rework links are interdepen-dent. If one activity fails to achieve itsperformance expectations, the nextactivity will be affected, and thus itsperformance will likewise be impacted.Consequently, the modeling of reworklinks is vital to predicting more accurateproject duration and cost. Love, Mandal,and Li (1999) found that rework hasbecome a negative and inherent char-acteristic of construction projects,increasing the probability of scheduleand cost overrun. To simulate theimpact of communication and rework,we added links to connect all interde-pendent activities and to connect allactivities that require communicationacross organizational boundaries in thenetwork. These are represented asdashed lines connecting activity boxesin Figure 1.

Setting Activity PropertiesDuring preconstruction planning, it isessential to consider the technicalskills required to complete differentproject tasks. Not all of the activities ina construction project require the samelevel of skill and experience. VDT takesthese variances into account by permit-ting the user to assign different activityproperties, such as a task’s priority, worktype, skills, requirement complexity,solution complexity, and uncertainty.

We tested the accuracy of our model byreviewing each activity and assignmentwith PM-A.

Uncontrollable FactorsDaily information exchanges in con-struction projects, such as telephonecalls, impromptu meetings, and activityrework, affect the quality of construc-tion projects and their ultimate suc-cess. VDT considers the influence ofthese unpredictable conditions andinterprets them as project probabilityrates. In our robust project networkdesign approach, the project probabili-ty rates are defined as uncontrollablefactors that represent uncertainties inthe project execution environment.Project probability rates are dividedinto four distinct uncontrollable fac-tors. Every time an organization is per-forming an activity, VDT considers thepotential for functional errors, projecterrors, information exchange lapses,and noise probabilities by taking intoaccount the number of activities linkedby communication and rework links,the interdependencies among activi-ties, the nature and number of activitiesassigned to each organization, and thework complexity of each activity. If aproject network is designed to performreliably under different levels of unfore-seen conditions or uncertainties, thelikelihood of determining the correctallocation of resources, work-processsequence, and project-organizationstructure is higher.

We set probabilities for informationexchange, noise, functional errors, andproject errors according to the rangesof values suggested in the SimVisionUser Guide (2005). SimVision is thecommercially available version of VDTdistributed by ePM, LLC. We chosevalues within the ranges consideringmultiple factors, including the generalcontractor’s familiarity with renovationprojects, the sophistication of theowner and subcontractors, and the pro-ject’s scope and size. We set probabilitiesfor information exchange and noiseprobabilities to 0.4 and 0.1, respectively.

pmj402_20109.qxd 4/28/09 5:51 PM Page 85



AP

ER

S

Given the familiarity with this projecttype, we set the functional and projecterror probabilities each to 0.075, whichis 25% lower than the recommendedvalue. Achieving robust project networkdesign requires that the controllablefactors perform with minimal variance

across a range of uncontrollable factorlevels. We set three levels for eachuncertainty factor. We increased anddecreased the factor-level settings foreach uncontrollable factor by 50%.These factor-level values are summa-rized in Table 1.

Robust Project Network DesignInterventions and AnalysesProject Performance AssessmentThe first step in designing a robustproject network is to analyze the proj-ect performance by measuring themean performance and variation fortime, cost, and quality. Once the com-munication and rework links, activitysettings, and project probability rateswere added to the baseline model, weran the simulation model. The resultswere significantly different from thoseobtained by running the initial baselinemodel. The simulation forecasted thatthe completion of the fourth floor wouldbe delayed. On the other hand, althoughthe completion of the third floor wasstill on schedule, the completion of the

G.C., Owner, UserMeeting

Project Participantsand Stake Holders

Project Engineer

Site Inspection

Project Manager

Site Superintendent

Mechanical

Start

RemovingExisting

AcousticalCeiling Tile

Install ElectricalRough-In

InstallMechanicalRough-In

InstallPlumbingRough-In

InstallMech Piping

Rough-In

Install ConcreteMech PadsRough-In

Fire SpinkleRough-In

InstallFraming Install

CMU

InstallElevator

ShaftWalls

Install MepEquipment

Install NewDormerFraming

InstallDrywall

InstallCeramic

Tile

InstallTape, Float,

Paint

InstallMep

Finishes

Finish

InstallPlumbingFeatures

InstallToilet

Partitions

InstallAcoustical

Rid

InstallAcoustics

Tile

InstallFloor

Covering

4th Floor

Cover Up

RemoveConcrete

Slab@MechShaft

RemoveConcreteSlab@Elev

Shaft

RemoveExisting

Mep Systems

PerformAsbestos

Abatement

InitialDemolitionCompleted

Electrical Plumbing Sprinkler Concrete Finding Drywall/Framing/Ceiling

Demon Asbestos

Figure 1: Fourth-floor project organization simulation model.

Table 1: Project uncontrollable factor-level settings.

Information Functional ProjectExchange Noise Error Error

Probability Probability Probability Probability

Level 1 0.2000 0.1500 0.0375 0.1125

Level 2 0.4000 0.1000 0.0750 0.0750

Level 3 0.6000 0.0500 0.1125 0.0375

pmj402_20109.qxd 4/28/09 5:51 PM Page 86


two interim milestones was alsodelayed. The results indicated that thedelays were due to the mechanical anddrywall subcontractors. These twocrews were responsible for excessiveposition and person backlogs, amongother issues, such as high project reworkand communication volume. Backlog isdefined as the amount of work that aperson or crew has to work beyond theplanned duration time and is given inunits of time. Delays on this projectwould incur hefty penalties for liqui-dated damages. Excluding the potentialliquidated damages, the simulationforecasted a cost overrun of U.S. $65,800for both floors.

Quality was assessed by analyzingthe project quality risk calculated byVDT across seven risk factors: function-al risk index (FRI), project risk index(PRI), product quality, process quality,communication, meeting, and coordi-nation risk. FRI refers to the probabilityof a project containing functionalexceptions (i.e., those that affect onlyspecific tasks as opposed to propagat-ing rework across other activities). PRIrefers to project-level exceptions. Thisis the probability that a problem ordefect will arise in one task that mayimpact other activities linked to thatactivity. Both FRI and PRI are used toestimate the probability of quality riskson a project. The higher the risk value,the higher the project’s quality risk. Thesimulation predicted a project FRI andPRI of 49.2% and 35.5%, respectively.

Identification of the ControllableFactorsIn order to design a robust project net-work, it is critical that the factors affect-ing project performance be defined. Inthis way, the selection of the optimalcombination of controllable factorsleads to the least project performancevariance from target. We applied inter-ventions to the baseline case by adjust-ing the levels of selected controllablefactors in order to gauge their effects onproject performance. The controllablefactors proposed by project managers

surveyed during this research projectwere evaluated and compared with thecontrollable factors that are built intoVDT. Two key controllable factorsemerged as important to achieving proj-ect objectives and as available control-lable factors within VDT: (1) increasingthe skill level of the workers and(2) increasing the number of workers.We isolated the effects of each control-lable factor by modeling each combina-tion and analyzing the project resultsafter running the simulation model.Altering the levels of the controllablefactors resulted in varying impacts onproject performance.

Mechanical and drywall teamswere predicted by VDT to be responsi-ble for the aforementioned delays. We

adjusted the controllable factors of themechanical and drywall crew to threedifferent levels. In the case of labor, wechanged the number of workers oneach crew by �50%. For labor experi-ence, we similarly raised and loweredthe labor experience for each crew,considering that the medium numberof laborers and the medium laborerexperience was the default settingmodeled in the refined baseline model.Three variations in the number ofworkers and three variations in thelevel of laborer experience were com-bined to produce nine possible inter-ventions. Tables 2 and 3 summarizethese interventions for the drywallsubcontractor and mechanical sub-contractor, respectively.

Number of Labor Experience, Mechanical SubcontractorWorkers,

Mechanical Low Medium HighSubcontractor (EL) (EM) (EH)

Low WL EL WL EM WL EH(WL) � 5 (Case M5)

Medium WM EL WM EM Simulated WM EH(WM) � 10 baseline combination (Case M4)

High WH EL WH EM WH EH(WH) � 15 (Case M1) (Case M2) (Case M3)

Table 3: Mechanical subcontractor interventions.

Number of Labor Experience, Drywall SubcontractorWorkers,Drywall Low Medium High

Subcontractor (EL) (EM) (EH)

Low WL EL WL EM WL EH(WL) � 5 (Case D5)

Medium WM EL WM EM Simulated WM EH(WM) � 10 baseline combination (Case D4)

High WH EL WH EM WH EH(WH) � 15 (Case D1) (Case D2) (Case D3)

Table 2: Drywall subcontractor interventions.

pmj402_20109.qxd 4/28/09 5:51 PM Page 87



AP

ER

S

The experimental matrix is devel-oped by calculating the number of pos-sible combinations of drywall subcon-tractor and mechanical subcontractorcrew size and skill level. Nine possibleinterventions for the drywall andmechanical crews yielded 81 possibleresponses from project managers.However, no project manager chose anintervention below the default assign-ments of medium numbers of workersand medium labor experience. Theyfocused their answer on five main com-binations, which are identified as casesin Tables 2 and 3. Thus, the five possibleinterventions for each crew resulted in25 combinations, which we will test in therobust design experiment. These are out-lined in Table 4.

Defining the Objective FunctionThe objective function is used to assessthe fit between simulation results andproject objectives. From the case studypresented earlier, the objective func-tion was defined by the project dura-tion, cost, and quality risks factors thatmake up the FRI and the PRI. The strin-gent project delivery date set up by theowner placed the project duration asthe most heavily weighted value. Weworked with the project manager toassign weights to each project metricbased on the project’s performancelevel that would satisfy the stakehold-ers’ requirements, including the deliv-ery date, owner’s budget, qualityrequirements, and regulations compli-ance, such as the preservation of thehistorical character of the building and

ADA requirements. The project manag-er suggested weighting all factors equally,with the exception of project duration,which he suggested weighting 50% morethan the other three. This resulted in anobjective function value for each com-bination of interventions as follows:

yi � 0.34 PDi � 0.22 PCi � 0.22 FRIi

� 0.22 PRIi (3)

where:yi � objective function value per com-

bination,PDi � project duration per combination,PCi � project cost per combination,FRIi � functional risk index per combi-

nation, andPRIi � project risk index per combination.

To determine the objective function(yi) per intervention, it is first necessaryto normalize the project performancevalues (duration, cost, FRI, and PRI) bytransforming them into the same scale.The normalization of each project per-formance value at the three noise levelsis calculated by determining the ratiobetween each single factor and themean value. The objective function isderived from the VDT model output ofthe building renovation project sim-ulation. Each combination of interven-tions is simulated for each of threenoise levels. These objective functionvalues for each combination of inter-ventions are then combined in therobust design experiment to identify themost robust project organization. Hence,the objective function is the critical out-put value of 75 simulated combinations

of interventions and varying levels ofuncontrollable factor levels.

Calculating the Signal-to-NoiseRatio (S/N)Equation 1 determines the S/N ratio’svalue as a function of the mean projectperformance values. The mean of theproject performance values is the meanof the objective function’s valuesassigned in each combination of inter-ventions across the three noise levels.Table 5 includes the mean and signal-to-noise ratio values for each combina-tion of factor-level settings resultingfrom the integrated project organiza-tion design and robust design approachdeveloped in this article.

Robust Project NetworkDesign ResultsThe influences of each controllable fac-tor (experience and number of workersin drywall and mechanical subcontrac-tors) on the mean and signal-to-noiseratio for the project outcome were calcu-lated and plotted as shown in Figures 2,3, and 4, respectively. In terms of theeffect of controllable factors on the proj-ect outcome (Figure 2), we seek the low-est value, since the objective function isto minimize the project duration, thecost, and the quality risk in order toimprove the project’s performance.Figure 2 indicates that D2 and M3 wouldbe the combination of controllable fac-tors resulting in the minimum meanproject duration, cost, and quality risk.Increasing the number of workers of thedrywall crew while hiring more workerswith higher experience for the mechan-ical crews would most enhance theproject performance. However, thiscombination may not be the best proj-ect network design if performancevaries significantly with uncontrollablefactor levels.

Identifying the most robust combi-nation of controllable factors involvesmore than calculating the mean projectobjective outcome value. The signal-to-noise (S/N) ratio indicates which combi-nation of interventions will provide themost robust project network design

Table 4: Experimental matrix of intervention combinations.

Drywall Subcontractor (D)

Case D1 D2 D3 D4 D5

M1 D1M1 D2M1 D3M1 D4M1 D5M1

MechanicalM2 D1M2 D2M2 D3M2 D4M2 D5M2

Subcontractor (M)M3 D1M3 D2M3 D3M3 D4M3 D5M3



pmj402_20109.qxd 4/28/09 5:51 PM Page 88


performance by taking into accountvariance. The larger the S/N ratio, themore robust the project network per-formance is to uncertainty. Figure 3shows that from the 25 possible inter-ventions that PM-A can invoke, the S/Nratio in the D2 case is the most robustsolution for the drywall contractor.Figure 4 shows that the most robust

solution for the mechanical contractor isthe M3 case. Together, D2 and M3 pro-vide the most robust combination ofinterventions to maximize performancewhile minimizing variance. The drywallsubcontractor should increase its num-ber of workers from 10 to 15 having thesame skill level. Conversely, the mechan-ical subcontractor should increase its

crew skill level to high and increase itsnumber of workers from 10 to 15employees. This would enable the testcase building renovation project toperform robustly across the threedynamic levels of uncertainty. Across thethree noise levels, the project durationfor the third and fourth floors would be179.4 days, 180.1 days, and 181.4 days.

Table 5: Robust design experiment results for each combination of interventions at varying uncontrollable factor levels.

Objective Function

Experiment Resultsyi yi yiNoise Level Noise Level Noise Level

# Interventions 1 2 3 Sm S/N

1 D1M1 1001.02 1001.33 997.62 999988 �59.9999

2 D1M2 1000.47 999.60 998.07 998762 �59.9946

3 D1M3 1000.05 998.49 997.54 997388 �59.9886

4 D1M4 998.39 998.23 998.16 996524 �59.9849

5 D1M5 998.29 996.09 995.21 993073 �59.9698

6 D2M1 999.51 996.12 999.26 996598 �59.9852

7 D2M2 997.44 996.78 999.69 995945 �59.9824

8 D2M3 995.20 995.27 996.90 991597 �59.9634

9 D2M4 995.99 994.82 996.24 991392 �59.9625

10 D2M5 996.72 995.83 997.21 993187 �59.9703

11 D3M1 999.67 999.07 999.18 998615 �59.994

12 D3M2 996.78 997.25 998.49 995018 �59.9783

13 D3M3 997.91 997.62 997.92 995638 �59.981

14 D3M4 997.76 998.71 997.82 996198 �59.9835

15 D3M5 999.43 1001.54 998.26 999484 �59.9978

16 D4M1 998.10 999.94 1000.40 998969 �59.9955

17 D4M2 999.28 1000.56 1000.18 1000011 �60.000

18 D4M3 995.97 999.14 998.35 995643 �59.981

19 D4M4 996.64 1000.90 1001.07 999075 �59.996

20 D4M5 997.92 999.72 999.06 997808 �59.9905

21 D5M1 1006.40 1004.90 1007.33 1012463 �60.0538

22 D5M2 1006.83 1007.26 1006.47 1013753 �60.0593

23 D5M3 1006.24 1006.81 1006.84 1013309 �60.0574

24 D5M4 1006.45 1007.48 1006.91 1013940 �60.0601

25 D5M5 1011.53 1006.53 1005.81 1015985 �60.0689

pmj402_20109.qxd 4/28/09 5:51 PM Page 89



AP

ER

S

The difference between the lowest andhighest project cost across the dynamicuncontrollable factor levels was only1.1%. Finally, the project and functionalrisk index quality indicators were pre-dicted to have FRI values of 49.3%,49.3%, and 49.2% (a range of 0.1%), andPRI values of 34.2%, 35.3%, and 37.3%

(a range of 3.1%) across the dynamicrange of uncertainty levels. These valuesshow performance is consistent acrossuncontrollable factor levels for this com-bination of interventions. The mostrobust project network design for thistest case will result from this combina-tion of controllable factor interventions.

Extending the Capabilities ofProject ManagersThe robust network design approachpresented in this research article findsthat D2 and M3 are the most robustcombination of interventions that PM-A should apply to reliably achieve therenovation project objectives. Followingthe robust project network design effortfor this case, we interviewed 20 experi-enced project managers and askedthem to select the combination of dry-wall and mechanical subcontractorinterventions that would lead to robustperformance in the presence of dynam-ic uncertainties in the project executionenvironment. Nine project managers(45%) correctly selected D2 for thedrywall subcontractor intervention.However, nine other project managers(45%) selected D4, and two projectmanagers (10%) selected D1. Only two(10%) project managers correctlyselected M3 as the most robust mechan-ical crew intervention. Thirteen (65%)project managers selected M4, threeother project managers (15%) selectedM5, and two project managers (10%)selected M2.

The interventions selected most byproject managers for the less task- andcommunication-interdependent dry-wall subcontractor was aligned with therobust project network design approachresults. This provides some validationthat the results of the combinedmethodologies are reasonable. However,achieving the most robust project net-work design requires identifying themost robust combination of factors.Only 5% of the project managers cor-rectly selected both D2 and M3.Because 19 out of 20 project managerswere not able to identify the mostrobust project network design, there isevidence that the approach developedin this article can extend the capabilitiesof project managers to reliably predictproject performance in dynamicallyuncertain project environments. Thefact that the forecasted cost overrunwas U.S. $65,800 for the two floors exam-ined further highlights the importance of

1008

Drywall

1006

Mea

n P

roje

ct P

erfo

rman

ce V

alue

s

1004

1002

1000

998

9961 2 3 4 5D D D D DM M M M M

Mechanical

Figure 3: Factor effect on S/N values of drywall contractor.

Figure 2: Factor effect on mean.

�59.950

�59.970

STN

Val

ues

�59.990

�60.010

�60.030

�60.050

�60.0701 2 3 4 5

D D D D D

pmj402_20109.qxd 4/28/09 5:51 PM Page 90


enabling project managers to designproject networks capable of performingreliably across a range of uncertainties.

ConclusionIn this article, we have introduced anintegrated approach to design robustproject networks capable of performingreliably in dynamically uncertain proj-ect environments. The framework inte-grates validated project organizationdesign simulation research with vali-dated robust design methodologies.This research makes several importantcontributions to knowledge. First, wedemonstrate through the simulationmodeling of two floors of a buildingrenovation project and running arobust design experiment on the simu-lated outcomes that project organiza-tion design simulation tools can becombined with robust design method-ologies. We were able to work with proj-ect managers to derive an objectivefunction for a test case project and a setof interventions to achieve projectobjectives. We ran a robust designexperiment that included two interven-tions (skill level and crew size) in twoorganizations (drywall subcontractorand mechanical subcontractor) in a

project network. Each controllable fac-tor intervention was examined acrossthree possible uncontrollable factor-setting levels, and then a robust designexperiment was conducted to arrive atthe most robust project network design.

Finally, we surveyed 20 projectmanagers and asked them to predictthe most robust combination ofmechanical subcontractor and drywallsubcontractor interventions for the testcase building renovation project thatwould lead to reliable performancegiven dynamic project execution envi-ronment uncertainties. Only 5% of theproject managers were able to identifythe correct combination of factors.Hence, the approach introduced in thisarticle of integrating two validatedmethods can extend the capabilities ofproject managers to design project net-works that will perform reliably inuncertain project environments. Giventhe significant levels of uncertainty thatcan plague construction projects, it iscrucial that new approaches similar tothis one be created. Project managers inall industries suffering from dynamicuncertainties need these tools andmethods to achieve predictable per-formance. Future research should

consider the impact of the rapid global-ization of project networks on thesefindings.

AcknowledgmentsWe would like to thank Gerry Sepe andPat Gorman of ePM for their input andsupport on this project. We are particu-larly indebted to “Project Manager A,”who provided the test case for thisresearch but who remains unnamed toprotect confidentiality. Finally, wewould like to thank the 20 project man-agers who participated as intervieweesduring this project. ■

NotationsThe following symbols are used in thisarticle:yi � objective function value per

combinationn � the number of data pointsSm � sum of the squares of the

meanS/N � signal-to-noise ratioPDi � project duration per combi-

nationPCi � project cost per combinationFRIi � functional risk index per com-

binationPRIi � project risk index per combi-

nation

ReferencesAntony, J., & Antony, F. (2001). Teachingthe Taguchi method to industrial engi-neers. Work Study, 50(4), 141–149.

Baligh, H., Burton, M., & Obel, B. (1996).Organizational consultant: Creating ausable theory for organizational design.Management Science, 42, 1648–1662.

Beary, T., & Abdelhamid, T. (2005,April). Production planning processin residential construction using leanconstruction and Six Sigma principles.Presented at the Construction ResearchCongress, San Diego, CA.

Bendell, A., Disney, J., & Pridmore, W.(1989). Taguchi methods: Applicationsin world industry. Bedford, UK: IFSPublications.

Ben-Haim, Y., & Laufer, A. (1998).Robust reliability of projects withactivity-duration uncertainty. Journal

Figure 4: Factor effect on S/N values of mechanical contractor.

�59.990

�59.995

STN

Val

ues

�60.000

�60.005

�60.0101 2 3 4 5

M M M M M

pmj402_20109.qxd 4/28/09 5:51 PM Page 91



AP

ER

S

of Construction Engineering andManagement, 124(2), 125–132.

Chinowsky, P., & Taylor, J. (2007).Project networks: Learning, leadershipand development. In M. Sexton (Ed.),Revaluing construction (pp. 54–65).Rotterdam, Netherlands: InternationalCouncil for Research and Innovationin Building Construction (CIB).

Galbraith, J. (1973). Designing complexorganizations. Reading, MA: Addison-Wesley.

Galbraith, J. (1977). Organizationdesign. Reading, MA: Addison-Wesley.

Harty, C. (2005). Innovation in con-struction: A sociology of technologyapproach. Building Research andInformation, 33, 512–522.

Hofstede, G. (1991). Cultures andorganizations: Software of the mind,intercultural cooperation and itsimportance for survival. New York:McGraw-Hill.

Horii, T., Jin, Y., & Levitt, R. (2005).Modeling and analyzing cultural influ-ences on project team performance.Computational and MathematicalOrganization Theory, 10, 305–321.

Ibrahim, R., & Nissen, M. (2007).Discontinuity in organizations:Developing a knowledge-basedorganizational performance modelfor discontinuous membership.International Journal of KnowledgeManagement, 3(1), 18–36.

Jin, Y., & Levitt, R. (1996). The virtualdesign team: A computational modelof project organizations. Computationaland Mathematical OrganizationTheory, 2, 171–196.

Kanter, R. (1991). The future ofbureaucracy and hierarchy in organi-zational theory: A report from the field.In P. Bordieu & J. Coleman (Eds.),Social theory for a changing society(pp. 63–90). Boulder, CO: Westview.

Khosraviani, B., Levitt, R., & Koza, J.(2004). Organization design optimiza-tion using genetic programming. InM. Keijzer (Ed.), Late breaking papers at

the 2004 Genetic and EvolutionaryComputation Conference, distributed onCD-ROM at GECCO-2004, Seattle, WA.

Kolltveit, B., Karlsen, J., & Gronhaug,K. (2004). Exploiting opportunities inuncertainty during the early projectphase. Journal of Management inEngineering, 20(4), 134–140.

Kunz, J., Christensen, T., Cohen, G.,Jin, Y., & Levitt, R. (1998). The virtualdesign team. Communications of theACM, 41(11), 84–91.

Kwon, K., & Lin, R. (2005). Robustdamage location in structures usingTaguchi method. Journal of StructuralEngineering, 131, 629–642.

Lawrence, P., & Lorsch, J. (1967).Differentiation and integration incomplex organizations. AdministrativeScience Quarterly, 12(1), 1–30.

Levitt, R., Thomsen, J., Christiansen,T., Kunz, J., Jin, Y., & Nass, C. (1999).Simulating project work processesand organizations: Toward a micro-contingency theory of organizationdesign. Management Science, 45,1479–1495.

Love, P., Mandal, P., & Li, H. (1999).Determining the casual structure ofrework influences in construction.Construction Management andEconomics, 17, 505–517.

Peace, G. (1993). Taguchi methods:Hands-on approach. New York:Addison-Wesley.

Pekericli, M., Akinci, B., & Karaesmen,I. (2003, November). Modeling informa-tion dependencies in construction proj-ect network organizations. Presented atthe 4th Joint Symposium on IT in CivilEngineering, Nashville, TN.

Powell, W. (1990). Neither market norhierarchy: Network forms of organiza-tion. Research in OrganizationalBehavior, 12, 295–336.

Roy, R. (1990). A primer on the Taguchimethod. Dearborn, MI: Society ofManufacturing Engineers.

Salem, O., Solomon, J., Genaidy, A., &Minkarah, I. (2006). Lean construction:

From theory to implementation.Journal of Management in Engineering,22(4), 168–175.

SimVision User Guide. (2005). Austin,TX: ePM, LLC.

Taguchi, G., & Clausing, D. (1990).Robust quality. Harvard BusinessReview, 68(1), 65–75.

Taylor, J. (2007). The antecedents ofsuccessful 3D CAD implementationin design and construction networks.Journal of Construction Engineeringand Management, 133, 993–1002.

Taylor, J., & Levitt, R. (2004).Understanding and managing systemicinnovation in project-based industries.In D. Slevin, D. Cleland, & J. Pinto(Eds.), Innovations: Project manage-ment research 2004 (pp. 83–99).Newtown Square, PA: ProjectManagement Institute.

Taylor, J., & Levitt, R. (2007).Innovation alignment and projectnetwork dynamics: An integrativemodel for change. Project ManagementJournal, 38(3), 22–35.

Taylor, J., Levitt, R., & Villarroel, A.(2006). Simulating learning in inter-organizational networks: The insidiousrole of task interdependence and rela-tional instability on system-level learn-ing. Presented at the North AmericanAssociation for Computational Socialand Organizational Sciences AnnualConference, Notre Dame, Indiana.

Thompson, J. (1967). Organizations inaction: Social science bases of adminis-trative theory. New York: McGraw-Hill.

Thomsen, J., Levitt, R., & Nass, C. (2004).The Virtual Team Alliance (VTA):Extending Galbraith’s information-processing model to account for goalincongruency. Computational andMathematical Organization Theory, 10,349–372.

Vaziri, K., Carr, P., & Nozick, L. (2007).Project planning for construction underuncertainty with limited resources.Journal of Construction Engineering andManagement, 133, 268–276.

pmj402_20109.qxd 4/28/09 5:51 PM Page 92


David A. Ortiz de Orue is a recent MS degreegraduate from the University of Texas at Austin.His master’s degree research focused on devel-opment of a plan for implementing buildinginformation modeling (BIM) in the ArchitecturalEngineering Curriculum. Currently, he is a proj-ect controls engineer at The Shaw Group Inc.,where he is actively involved in the E-P-C sched-uling for large energy and chemical projects.

John E. Taylor, PhD, is an assistant professorin Columbia University’s School of Engineeringand Applied Science. He received his PhD fromStanford University. His research explores howfirms in project networks adapt to disruptivechanges such as new technology implementa-tion, the international outsourcing of complexengineering services, and the adoption of ener-gy efficient practices. As a Fulbright scholar,

he developed his early interest in project networkdynamics while working with the United Nationsto understand and minimize fraudulent freighttraffic across national borders. He developscomputational simulation models in his researchto explore concepts derived from his in-depthfield and laboratory research. By grounding hissimulation work in field and experimental data,he is able to test, extend, and develop new orga-nizational theory for project networks. He is thedirector of the Project Network Dynamics Labat Columbia University. He is the founder of twotechnology start-ups serving project-basedindustries and has 5 years of experience work-ing as a project manager.

Arpamart Chanmeka is a PhD candidate at theUniversity of Texas at Austin and is workingfor the Construction Industry Institute’s (CII)Benchmarking and Metrics program as a

research assistant. At CII, she analyzes CIIannual safety data and is the account managerfor the Canadian Oil Sands Benchmarking. Herresearch focuses on Alberta major project per-formance and productivity benchmarking. Sheholds an ME in structural engineering from theAsian Institute of Technology in Thailand. Prior toher doctoral studies, she held several engineer-ing positions and managed construction projectsat the Department of Drainage and Sewerage forthe Bangkok Metropolitan Administration.

Runi Weerasooriya is a consultant at SpireConsulting Group, LLC, in Austin, Texas, whereshe provides forensic construction services andlitigation support to a wide range of clients. Sheis a recent MS graduate from The University ofTexas at Austin, where she researched lessonslearned programs in conjunction with theConstruction Industry Institute.

pmj402_20109.qxd 4/28/09 5:51 PM Page 93

PA

PE

RS


INTRODUCTION ■

Researchers and practitioners have been adapting different academ-ic disciplines to contribute to the body of knowledge in projectmanagement. From the management science field, network sched-uling techniques, program evaluation and review technique (PERT)

applications (Miller, 1962), critical path method (CPM) computations(Crowston & Thompson, 1967), decision-making tools (Kunreuther, 1969),resource allocation problems (Schussel & Price, 1970; Zangwill, 1966), opti-mization approaches (Charnes & Cooper, 1957), and computer applications(Hoffman, 1982; Maxfield, 1981; Vazsonyi, 1982; Wasil & Assad, 1988) aresome of the examples of theory adopted in project management. Morerecently, risk and uncertainty management (Dillon, Paté-Cornell, &Guikema, 2003; Gustafsson & Salo, 2005; Gutierrez & Paul, 2000) were high-lighted. These topics and the application of decision sciences and risk man-agement gained strong interest in the project management community.

Organizational dynamics theories provided insights by discussing the project manager (Gaddis, 1959), research and development (R&D) proj-ect applications (Roman, 1964), project organization (Middleton, 1967),project control (Howell, 1968), and project leadership (Hodgetts, 1968). Morerecently, subjects on the benefits of project-driven organizations have beenrecognized by presenting papers on virtual teams (Cascio, 2000; Karl, 1999;Malhotra, Majchrzak, & Benson, 2007; Townsend, DeMarie, & Hendrickson,1998), and team-based organizational performance (Cianni & Wnuck, 1997;Forrester & Drexler, 1999; Scott & Einstein, 2001).

Strategic alignment with project objectives (Bunch, 2003; Gottschalg &Zollo, 2007; Miles, Snow, Mathews, Miles, & Coleman, 1997), project manage-ment implementation to new product development projects (Northcraft &Wolf, 1984), and several strategic and practical applications approaches were developed to adopt allied disciplines to project management (Gerwin,2004; Gerwin & Ferris, 2004; Jassawalla & Sashittal, 2002; Kessler & Bierly,2001; Leifer, O’Connor, & Rice, 2001; Li, Bingham, & Umphress, 2007). Casestudies such as the Sydney Olympics (Pitsis, Clegg, Marosszeky, & Rura-Polley, 2003), research-oriented organizations (O’Connor, Rice, Peters, &Veryzer, 2003), and systems development teams (Kirsch, 1996) aimed to bet-ter understand the significance of organizational alignment with projectobjectives.

Allied disciplines have influenced project management, and in return itinfluenced them. Therefore, innovative theories, trends, and challengeslearned from investigating allied disciplines of project management couldhave important implications for the future of project management. To clarify

Availability-Impact Analysis of ProjectManagement Trends: PerspectivesFrom Allied DisciplinesYoung Hoon Kwak, The George Washington University, Washington, DC, USAFrank T. Anbari, The George Washington University, Washington, DC, USA

ABSTRACT ■

The goal of this research was to analyze currentand future trends of the allied management dis-ciplines that influence project management. Weasked academicians and practitioners aboutthe availability of knowledge and the potentialimpact of allied disciplines on project manage-ment. We were able to assess where the allieddisciplines currently stand in terms of availabil-ity and impact as well as make predictionsabout the future. The thoughts and visions col-lected from this research provide valuableinsights for identifying trends in the allied disci-plines and their impact on the future of projectmanagement, as well as opportunities, chal-lenges, and obstacles.

KEYWORDS: project management; allieddisciplines; trends analysis; survey analysis





DOI: 10.1002/pmj.20111

pmj402_20111.qxd 4/28/09 4:25 PM Page 94


the importance of project managementas a theoretical and practical manage-ment discipline, we addressed the fol-lowing questions:• What future trends in the allied disci-

plines might significantly impactproject management?

• How would the allied disciplines’trends change project management?

• How would project managers have tochange their mind-set because of theallied disciplines’ trends impact?

• How do we behave proactively tomeet challenges of trends in the allieddisciplines?

Research Objectives andMethodologyPrevious research analyzed trends ofvarious disciplines including operationsresearch and management science(Corbett & Van Wassenhove, 1993;Ormerod & Kiossis, 1997; Reisman &Kirschnick, 1994), production and oper-ations management (Barman, Hanna, &LaForge, 2001; Barman, Tersine, &Buckley, 1991; Goh, Holsapple, Johnson,& Tanner, 1997; Olson, 2005; Prasad &Babbar, 2000; Soteriou, Hadjinicola, &Patsia, 1999; Vokura, 1996), decisionanalysis (Keefer, Kirkwood, & Corner,2004), technology management (Liao,2005; Linton & Thongpapanl, 2004;Merino, Pereira do Carmo, & Alvarez,2006; Reisman, 1994), and project man-agement (Anbari, Bredillet, & Turner,2008; Artto, Martinsuo, Gemunden, &Murtoaro, 2009; Artto & Wikström, 2005;Betts & Lansley, 1995; Bredillet, 2002,2006; Crawford, Pollack, & England, 2006;Kloppenborg & Opfer, 2002; Kwak &Anbari, 2008a, 2008b; Söderlund, 2004;Themistocleous & Wearne, 2000). It isimportant to note that research publica-tions on project management in topacademic management journals as wellas in project management–specific jour-nals have been increasing substantiallysince the 1950s, and it is expected thatthis strong upward trend will continue inthe future (Kwak & Anbari, in press).

We designed and administered a sur-vey to the project management community

to collect data about the availability ofknowledge and information regardingthe allied disciplines and their impact onproject management in the present andthe future. We also collected data onthoughts and opinions on the impact,trends, and future of project manage-ment. The goal of this research was toanalyze current and future trends of theallied disciplines that influence projectmanagement. To do so, the followingpoints were addressed and examined.• Find and plot the current and future

availability-impact relationships ofallied disciplines related to projectmanagement. For each allied disci-pline, we developed a 2 � 2 matrix: Onthe horizontal (X) axis, we showed theavailability of knowledge and infor-mation, and on the vertical (Y) axis,we showed the impact of each allieddiscipline on project management.The results were plotted for eachallied discipline.

• Identify how the allied disciplines’trends change project management.Project management practitionersand researchers were surveyed onproject management trends.

• Conduct a structured survey of how theproject management community needsto look forward at the allied disciplines’trends, and how that will improve theunderstanding of what kind of mind-set project managers should have ordevelop. Based on extensive analysis ofthe survey results of the project man-agement community and evaluationof experts’ subjective judgment, weidentified how the project manage-ment community as a whole couldrespond and meet the opportunitiesand challenges that lie ahead in thefuture. The research team incorporat-ed all the reviews and recommenda-tions and compiled proactive responsestrategies toward these macro trendsthat are happening now and areexpected to happen in the future.

Research ApproachWe conducted a survey to collect acade-micians’ and practitioners’ perceptions

of the trends, impact, and challengesrelevant to project management. Weasked the respondents to reflect on theirbest understanding of the availabilityand impact of research in the allied dis-ciplines on project management appli-cations at present and in the future(Kwak & Anbari, 2008b).

We reviewed major themes andprior works on project managementresearch and identified eight broadallied disciplines to organize researchpublications related to the projectmanagement field and to gain betterunderstanding of trends in projectmanagement–related research in thesedisciplines (Kwak & Anbari, 2008a, inpress). We believe that the followingeight categories represent the allieddisciplines where one can find projectmanagement research:1. Operations Research/Decision Sciences/

Operation Management/ Supply-ChainManagement (OR/DS/OM/SCM) refersto the discipline associated with quan-titative decision analysis and manage-ment principles, including variousoptimization tools and techniques,network analysis, resource leveling,simulation, and so on.

2. Organizational Behavior/Human Re-source Management (OB/HR) refers tothe discipline associated with organi-zational structure, organizationaldynamics, motivation, leadership,conflict management, and so on.

3. Information Technology/InformationSystems (IT/IS) refers to the disciplineassociated with the use of computersand computer systems to process,transmit, store, and retrieve informa-tion for better management decisions.

4.Technology Applications/Innovation/New Product Development/Researchand Development (TECH/INNOV/NPD/R&D) refers to the disciplineassociated with the concepts of mak-ing innovative and technologicalimprovements and the research anddevelopment of entirely new prod-ucts, services, and processes.

5. Engineering and Construction/Contracts/Legal Aspects/Expert Witness

pmj402_20111.qxd 4/28/09 4:25 PM Page 95


Perspectives from Allied DisciplinesP

AP

ER

S

(EC/CONTRACT/LEGAL) refers to thediscipline associated with the useand application of a broad range ofprofessional expertise to resolveissues related to engineering andconstruction, contracts, expert wit-ness, and their legal implications.

6. St ra t e g y / In t e g ra t i o n / Po r t f o l i o Management/Value of Project Manage-ment/Marketing (STRATEGY/PPM)refers to the concepts of organizingand managing resources to maximizeprofit, minimize cost, and support theoverall strategy of the organization.

7. Performance Management/EarnedValue Management/Project Financeand Accounting (PERFORM/EVM)refers to the concepts and techniquesthat measure project progress objec-tively by combining measurementsof technical performance, scheduleperformance, and cost performance.

8. Quality Management/Six Sigma/Process Improvement (QM/6SIGMA/PI) refers to the concepts of improv-ing processes, minimizing defects,and reducing cost by implementingcontinual improvement principlesand specific measures and metrics.

In our survey questionnaire, wedescribed each of the eight allied disci-plines just identified and asked respon-dents to rank the impact on projectmanagement and the availability ofinformation/knowledge/research foreach discipline on a scale from 1 (low)to 7 (high) in the past, present, andfuture. We asked respondents open-ended questions dealing with trends ofthe allied disciplines and their potentialimpact on project management, as wellas how the project management com-munity should change its mind-set inresponse to trends of the allied disci-plines in the future. These importantquestions were intended to collect andanalyze respondents’ views of trends inthe allied disciplines and their impact onproject management in the past, as wellas their predictions for the future of proj-ect management. We also asked ques-tions about respondents’ demographics

concerning their education, work expe-rience, and occupation (Kwak & Anbari,2008b).

We narrowed the survey participantsto focused groups who are part of theproject management community eitheras practitioners, consultants, scholars,or other professionals working in theproject management environment. Wewere able to collect usable responses tothe structured survey questions as wellas rich, qualitative information andopinions from participants by askingopen-ended questions regarding thetrends and future of project manage-ment and its allied disciplines.

We also conducted face-to-face andtelephone in-depth interviews to col-lect thoughts from project manage-ment scholars and practitioners con-cerning trends in the allied disciplinesand their potential impact on the futureof project management. The thoughtscollected in the survey of the projectmanagement community and the inter-views provided valuable insights foridentifying trends in the allied disci-plines and their impact on the future ofproject management, as well as theopportunities, challenges, and obsta-cles that are likely to persist or material-ize in the future.

Survey Respondents’DemographicsOverall, we received 82 responses over aperiod of 5 months. Approximately 98% ofthe respondents had earned a bachelor’s

degree or higher. Bachelor’s degree hold-ers were 32%, master’s degree holderswere 50%, and doctoral degree holderswere 16% of the respondents. Table 1shows the distribution in detail. We askedrespondents about the years of theirwork experience. Almost 90% worked formore than 5 years in the project manage-ment field, and among them about 46%worked for more than 20 years. Table 2summarizes respondents’ years of experi-ence. The respondents’ occupations werediverse yet focused on the project man-agement field. Twelve percent were aca-demic scholars, 11% were project man-agement graduate students, 44% wereproject management practitioners, and16% were project management consult-ants/trainers. Others included federalgovernment capital planning specialists,engineers, program analysts, and proj-ect directors. Table 3 summarizes thedistribution of respondents’ occupa-tions. To summarize, project manage-ment practitioners with a master’sdegree who have 20 years or more ofexperience are the ones that best repre-sent the respondents.

Survey Results and AnalysisWe asked about the availability of knowl-edge (articles, literature, and experts) andthe potential impact of allied disciplinesrelated to project management. We wereable to assess where the allied disciplinescurrently stand in terms of availabilityand impact as well as make meaningfulpredictions about the future. We used a 1

Education Response % Response Count

High school degree 1.2 1

Two-year associate’s degree 1.2 1

Four-year bachelor’s degree 31.7 26

Master’s degree 50 41

Doctoral degree 15.9 13

Total 100 82

Table 1: Respondents’ education.

pmj402_20111.qxd 4/28/09 4:25 PM Page 96

to 7 Likert scale to capture the respon-dents’ assessment. Then we used a 2 � 2matrix to understand where each allieddiscipline currently stands relative toother disciplines. We summarized theresults so that each quadrant has eitherlow or high availability and impact. Table4 shows the four quadrants we used toportray the results. Based on the respons-es, we plotted the eight disciplines rela-tive to each other using this availability-impact scheme. Table 5 shows the currenttrends of allied disciplines as they relateto project management.

OB/HR, TECH/INNOV/NPD/R&D,STRATEGY/PPM, and OR/DS/OM/SCMwere the four disciplines that respondents

perceived as low on availability of relatedknowledge and low on impact related toproject management. This could be inter-preted as these four disciplines still havepotential for more research (availability)and could have potential for greater influ-ence (impact) in the future. There was nodiscipline in the low-availability, high-impact quadrant. PERFORM/EVM andEC/CONTRACT/LEGAL were the twoareas that respondents perceived as highin availability of related knowledge butlow on impact related to project manage-ment. This could be interpreted as there is

plenty of related information availablecurrently, but the impact on project man-agement is low. QM/6SIGMA/PI andIT/IS were the two areas that were in thehigh-availability, high-impact quadrant.This could be interpreted as these twoareas are the hot fields in project manage-ment currently.

We then plotted the eight allied dis-ciplines relative to each other to visual-ize the future trends of allied disciplinesin project management (Table 6). OB/HR, TECH/INNOV/NPD/R&D, and EC/CONTRACT/LEGAL were the three dis-ciplines that respondents predict to below on availability of related knowledgeand low on impact related to projectmanagement in the future. There wasno discipline that was in the low-avail-ability, high-impact quadrant.OR/DS/OM/SCM was the area thatrespondents predicted to be high inavailability of related knowledge butlow on impact related to project man-agement in the future. PERFORM/EVM,QM/6SIGMA/PI, STRATEGY/PPM, andIT/IS were the four areas in the high-availability, high-impact quadrant.

By carefully analyzing Table 6, weobserve that the project managementcommunity expects that in the future(1) IT/IS and OB/HRM will have sub-stantially more impact than availabilityof knowledge, (2) TECH/INNOV/NPD/R&D, STRATEGY/PPM, PERFORM/EVM, QM/6SIGMA/PI, and OR/DS/OM/SCM will have somewhat moreimpact than availability of knowledge,and (3) EC/CONTRACT/LEGAL willhave substantially more availability ofknowledge than impact on project

Table 2: Respondents’ years of work experiences.

Years of Work Experience Response % Response Count

Less than 2 years 4.9 4

2 to 5 years 7.3 6

6 to 10 years 11.0 9

11 to 15 years 15.9 13

16 to 20 years 14.6 12

More than 20 years 46.3 38

Total 100 82

Table 3: Respondents’ occupations.

Occupation Response % Response Count

Academicians (professor or researcher) 12.2 10

Student 11.0 9

PM practitioner 43.9 36

Consultant/Trainer 15.9 13

Other (please specify) 17.1 14

Total 100 82

Table 4: Availability-impact matrix.

Low Availability High Availability

High Impact High Impact

Low Availability High Availability

Low Impact Low Impact

Table 5: Current trends of availability-impact matrix.

Low Availability, High Impact High Availability, High ImpactNone QM/6SIGMA/PI

IT/IS

Low Availability, Low Impact High Availability, Low ImpactOB/HR PERFORM/EVM

TECH/INNOV/NPD/R&D EC/CONTRACT/LEGALSTRATEGY/PPMOR/DS/OM/SCM


pmj402_20111.qxd 4/28/09 4:25 PM Page 97



AP

ER

S

management. These observations indi-cate that in general, there will be moreimpact than availability of knowledge,calling for further research and dissemi-nation of knowledge in most of the allieddisciplines. The observations also high-light areas that are particularly expectedto have more impact than availability ofknowledge from the perspective of theproject management community.

Analysis of the data clearly showsthat all eight disciplines have an upwardtrend in terms of the availability andimpact starting in the 1960s and into thefuture. However, there are inherent lim-itations on analyzing and forecasting

the future of the eight disciplines relat-ed to project management. Realizingthat the values of current and future ofthe eight disciplines related to projectmanagement are simply relative to eachother, we normalized the two sets ofdata (current and future) and over-lapped the plots of current and future ofthe eight disciplines to visualize thetrends and opportunities of the eightdisciplines in the future, as shown inFigure 1. This analysis provides us withclear and meaningful information tohighlight research opportunities andimpact of the allied disciplines asshown in Table 7.

The analysis indicates that STRATE-GY/PPM, OB/HR, IT/IS, PERFORM/EVM, and TECH/INNOV/NPD/R&Dwould have more availability in knowl-edge and information and have greaterimpact in the future. EC/CONTRACT/LEGAL and QM/6SIGMA/PI wouldhave less availability in knowledge andinformation and have less impact in thefuture. OR/DS/OM/SCM would havemore availability in knowledge andinformation, but the impact will remainthe same in the future. It is interestingto note that except for OR/DS/OM/SCM, current research availability asso-ciates strongly with the impact of allieddisciplines on project management inthe future. This shows that researchbeing published in various outlets maybe setting the trends of impact on proj-ect management and its allied disci-plines in the future.

Thoughts From ProjectManagement Scholars andPractitionersWe conducted in-depth interviews tocollect thoughts from project manage-ment scholars and practitioners con-cerning trends in the allied disciplinesand their potential impact on the futureof project management. The thoughtscollected from interviews provide valu-able insights in identifying trends in theallied disciplines and their impact onthe future of project management, aswell as the opportunities, challenges,and obstacles that are likely to persistor materialize in the future. Some ofthese thoughts are presented in the sec-tions that follow.

Thoughts From Scholar AThe impact of decision sciences onproject management was high until the1960s. Project management had itsgenesis in operations research in the1960s. However, since the 1970s the rel-ative importance, knowledge, andresearch in this area have been decreas-ing. Research in HR on project manage-ment started in the 1970s and has beengrowing since then. The EVM impact onproject management was low in the

Table 6: Future trends of availability-impact matrix.

Low Availability, High Impact High Availability, High ImpactNone PERFORM/EVM

QM/6SIGMA/PISTRATEGY/PPM

IT/IS

Low Availability, Low Impact High Availability, Low ImpactOB/HR OR/DS/OM/SCM

TECH/INNOV/NPD/R&DEC/CONTRACT/LEGAL

QM/6Sigma/PI

IT/IS

EC/Contract/Legal

Perform/EVM

OR/DS/OM/SCM

Strategy/PPM

Tech/Innov/NPD/R&D

OB/HR

IT/IS

Strategy/PPM

QM/6Sigma/PIPerform/EVM

OR/DS/OM/SCM

OB/HR

Tech/Innov/NPD/R&D

EC/Contract/Legal

CurrentFuture

Availability

Impact

Figure 1: Future opportunities of allied disciplines in project management.

pmj402_20111.qxd 4/28/09 4:25 PM Page 98


1970s and 1980s but has been growingsince then. Innovation within projectmanagement was not strong in the1970s, because project managementwas basically used as a tool for themeasurement of innovation. Innovationand R&D are not hugely impacted byproject management. The impact priorto the 1980s was low because projectmanagement was associated with con-struction. IT and innovation/R&Dimpact on project management startedin the early 1960s when IBM used proj-ect management in the development ofits products. However, an appreciablegrowth in the impact of innovation/R&Don project management was witnessedfrom the 1980s. In the late 1980s, theimpact of organizational strategy onproject management was low but hasshown an appreciable increase in thelate 1990s and now. Quality was lowuntil the early 1980s and has beengrowing since then. In the future, proj-ect management will be more sophisticated in that it will becomeless task-focused and more people-focused. Project management needs tobe recognized in the management fac-ulty and accorded an equal status withother traditional management sciencedisciplines. To do so, project manage-ment professionals and academiccommunities need to sell themselvesbetter.

Thoughts From Scholar BIf the word gets out about research inproject management, then it would not

be long before project management isestablished as an area in managementscience, such as happened in stochasticmodeling, which emerged as a quiterelevant set of topics and areas in thelast 10 to 15 years. If researchersbecome aware of the type of problemsthat arise in project management, thenthere will probably be more interest inthe academic community. The questionis whether researchers in operationsmanagement have recognized prob-lems in project management. There isas much potential for developing newtheories in project management, as hasbeen seen in other focus areas.

Thoughts From Scholar CThere are very interesting developmentsin project management by comparisonto traditional management. Based on anindustrial economy, we traditionallythink in terms of managing a big, costlyfactory with yearly model changes, ormanaging a hotel, or managing a facility,and so on. However, products arebecoming obsolete faster. So, we findthat we need to construct a new factory,a new building, satellites, network tow-ers, an enterprise resource planning sys-tem, and other change projects. Writingan article is a smaller project. For com-panies, it is a larger model of manage-ment replacing traditional manage-ment. Software can be a great aid toproject management to help with coor-dination. It is nice to have technologyavailable. Projects are becoming anelectronic era phenomenon.

Thomas Friedman, in The World IsFlat (2005), speaks of a platform anddiscusses globalization of countries(1492–1800), companies (1800–2000),and individuals (2000 and beyond).This involves the Internet, outsourcing,and collaboration beyond the bound-aries of the organization. Academicglobalization is allowing scholars tocollaborate on a variety of projects. Theacademic world is undergoing interest-ing changes. IT/IS is a preamble, andwe are now starting to utilize it. Oncesoftware is developed, people aroundthe world can start using it. More andmore people get involved and they cre-ate a new world. We are just beginningto use this platform and have notthought through various implications.We need to keep these huge changes inmind. Projects are a new direction fromthe manufacturing perspective of man-agement. Increasing use of Englishflattens the world and enhances collab-oration among multinational teams inprojects. There is an ideological shiftdissolving political, government, union,and other differences for greater eco-nomic prosperity.

Thoughts From Practitioner AProject management has always been aproduct of operations research, andoperations research has waned inrecent times. Performance measure-ment and Six Sigma are important toolsin project management. Six Sigma isimportant; however, it is all about get-ting the right measurement of theprocess. It has gone somewhat too far,such that government and the privatesector have gone measurement-crazy.Portfolio/program management willbecome more important in the future.The next generation of project man-agers will have more profound projectmanagement experience and probablyportfolio management experience.Integrative skill is a very good skill thatneeds to be developed by all projectand portfolio managers. The ability tointegrate all the facets of a project orgroup of projects to ascertain their

Table 7: Future trends of allied disciplines in project management.

Availability ImpactAllied Disciplines Current Future Current Future

OR/DS/OM/SCM Low High Low LowOB/HR Low High Low HighIT/IS Low High Low HighTECH/INNOV/NPD/R&D Low High Low HighEC/CONTRACT/LEGAL High Low High LowSTRATEGY/PPM Low High Low HighPERFORM/EVM Low High Low HighQM/6SIGMA/PI High Low High Low

pmj402_20111.qxd 4/28/09 4:25 PM Page 99



AP

ER

S

cross-impact is a much-needed skillamong project/portfolio managers.Project management is highly strategicbecause it has always been a means toan end and not an end itself, so the out-put of a project is the starting point forachieving the outcome of a businessand the outcome is basically what isstrategic. Once a corporate strategy hasbeen determined, project/programmanagement is the best tool to helporganizations gain some kind of disci-pline in the achievement of its strategicobjectives.

For any organization to survive,there must be a value that is represent-ed at all levels of the organization. Inthe future, there will be more organi-zations striving to be more ethical.There will be more corporate socialresponsibility, all of which will have animpact on the nature of projects. Asproject management becomes more ofthe transformational way of movingforward, ethics and corporate socialresponsibility will be integrated in all facets of project managementpractice.

Thoughts From Practitioner BIt is going to be a project-based worldin the future to respond to increasingcomplexity, requirements for accounta-bility, globalization, and the need fortalent and competence. The projectmanagement discipline is needed todeal with challenges in a more com-plex, uncertain world. Project manage-ment is the next stage of managementand leadership to meet the demandsfor efficiency and successful outcomes.It is difficult to think of aspects of man-agement that do not need projectmanagement. The world is very differ-ent, and projects are very different.Knowledge and skills need to be adapt-ed to grow the project managementfield effectively. Theory and practiceneed to adapt to provide the next gen-eration of leadership and management.There are complex, multidirectionalforces that shape government, industry,and universities. Project management

needs to keep evolving and adapting—not deterred and not locked in.

Future Trends and Implicationsof Project ManagementTrends of Allied Disciplines andImpact on the Future of ProjectManagementMore and more organizations areembracing project management prac-tices and incorporating the allied disci-plines into these practices to resolvecomplex management issues. Allied dis-ciplines are becoming more sophisticat-ed in their theories, tools, and applica-tions and will continue to add value toproject management concepts, theories,and applications. More collaborationamong allied disciplines and projectmanagement will happen due to theproven efficiency and effectiveness ofvarious disciplines. Project managementand allied disciplines will continue tomake recognizable contributions to themanagement community. Eventually,project management will be the lan-guage of good management practice asaccounting is the language of business.

The availability of research in proj-ect management and its allied disci-plines has exploded over the last 20years. OB/HR is a critical discipline inthe project management world. Somuch of what organizations do is man-aging internal and external businessand human resources and relation-ships. Fundamentally, project manage-ment today is more about peoplemanagement than about task orienta-tion. However, this trend of shiftingfrom OR/DS/OM/SCM to OB/HR haspossibly reached its peak. IT/IS willcontinue to provide enhanced tools forproject management and will have ahigh impact in the future. As the IT/ISdiscipline matures, project manage-ment tools and techniques will contin-ue to become available to general users,which makes it easier to implementproject management principles.

PERFORM/EVM and related meth-ods for measuring project perform-ance will grow rapidly, particularly due

to governmental regulations requiringsystematic cost/schedule evaluation inmanaging government projects. Relevantmetrics in all of the allied disciplines and measurement activities will becritical to the performance of manage-ment, which will lead to enhanced prob-ability of project success in the globaleconomy.

STRATEGY/PPM and QM/6SIGMA/PI should have a growing impact onproject management, as businessstrategies are developed and qualitiesare measured and analyzed to plan andimplement effective project manage-ment. For many organizations, thebiggest gains they see come from adapt-ing and implementing STRATEGY/PPMand QM/6SIGMA/PI.

TECH/INNOV/NPD/R&D and PERFORM/EVM are poised to makemajor breakthroughs, given the recentorganizational interest and institutionaldetermination on achieving project suc-cess. OR/DS/OM/SCM, PERFORM/EVM,IT/IS, and TECH/INNOV/NPD/R&Dwill work together to deliver tools andtechniques to allow the “science” ofplanning, scheduling, and cost controlto function in a real project-deliveryenvironment.

In the future, project managementas a discipline will be getting closer tothe general management and organiza-tion theory fields because project man-agement needs the theoretical advancesthat exist within these fields, andbecause management and organization-al theory need project management, asprojects are becoming an integral partof modern management in a number ofindustries and sectors. The trends of theallied disciplines and their potentialimpact on project management willdepend on the successful integration oforganizational strategies. In otherwords, if organizations embrace thealignment of allied disciplines and proj-ect management, then the impact onproject management will be fully recog-nized, and project management shouldbecome an integral component ofmainstream management thought and

pmj402_20111.qxd 4/28/09 4:25 PM Page 100


practice. Obviously, the allied disci-plines will bring great value to projectmanagement as the interdisciplinaryconcepts and practices become morewidely used and understood by practic-ing project managers and by opera-tional, general, and executive manage-ment.

Project Management Community’sMind-Set and the Trends of AlliedDisciplinesProject management should strengthenits theoretical and philosophical foun-dations to include ideas in the sociolog-ical, cultural, political, and ethicalrealms. The project management com-munity should be engaged in adaptingnew ideas and changes and look at howsocial networks and collaborative think-ing work in project management. Theproject management community mustunderstand that the allied disciplinesare needed and must embrace andapply them as part of project manage-ment. To do so, the project managementcommunity should be engaged inacquiring knowledge and participatingin major activities of the allied disci-plines to better understand the emerg-ing and promising practices for futureproject endeavors. Project managementshould never forget the human aspectssuch as trust, teamwork, and pride.

Trends and strategies will continueto evolve as project managementbecomes the norm. It will be an ever-changing field where highly skilled,multifaceted project managers will bein high demand. To ensure that projectmanagement flourishes, there is a needfor conducting research, asking ques-tions, and developing intellectualresources. Project management needsto revamp itself. The mind-set shouldmove toward a collaborative attituderather than being separate from all oth-ers. Each of the allied disciplines shouldbe respected for its knowledge, expert-ise, and wisdom. The role of projectmanagement should be viewed as thefocal point of information gatheringand dissemination to other areas of

related disciplines. The emerging gener-ation of project management researchershas a very open mind-set to the integra-tion of knowledge from allied fields. Theproject management communityshould focus on enhancing its capabili-ties in both the hard and soft disci-plines. It should embrace all researchthat improves the interaction of peopleto people, people to technology, andtechnology to technology.

Project management practitionersand researchers need to increase theirleadership and presence in industrythrough continued education andinformation sharing. They should leadthe way in driving the implementationof business strategies with knowledgeof STRATEGY/PPM and QM/6Sigma/PIas well as better understanding ofbehavioral issues and complexity of theproject world. An open mind-set is par-ticularly needed. The project manage-ment profession is continuously evolv-ing, so the project management com-munity should be receptive to newideas, and also be sensitive to theyearning of the public and professionalcommunity so as to model projectmanagement practices to meet theirexpectations. The best suggestion wecan make is to aggressively educatemore business professionals on projectmanagement as a discipline.

AcknowledgmentsThis research was partially funded by aresearch grant from the ProjectManagement Institute, whose supportis greatly appreciated. Any opinions,findings, conclusions, or recommenda-tions expressed in this paper are thoseof the authors and do not necessarilyreflect the views of the ProjectManagement Institute. The authorswould like to thank all participants inthe survey and the face-to-face inter-views, as well as all others who support-ed this research project. ■

ReferencesAnbari, F. T., Bredillet, C. N., & Turner,J. R. (2008). Perspectives on research in

project management. In Best PapersProceedings, Academy of Management2008 Meeting [CD]. Anaheim, CA:Academy of Management.

Artto, K. A., Martinsuo, M., Gemunden,H. G., & Murtoaro, J. (2009).Foundations of program management:A bibliometric view. InternationalJournal of Project Management, 27,1–18.

Artto, K. A., & Wikström, K. (2005).What is project business? InternationalJournal of Project Management, 23,343–353.

Barman, S., Hanna, M. D., & LaForge,R. W. (2001). Perceived relevance andquality of POM journals: A decadelater. Journal of OperationsManagement, 19, 367–385.

Barman, S., Tersine, R. J., & Buckley,M. R. (1991). An empirical assessmentof the perceived relevance and qualityof POM-related journals by academi-cians. Journal of OperationsManagement, 10, 194–212.

Betts, M., & Lansley, P. (1995).International Journal of ProjectManagement: A review of the first tenyears. International Journal of ProjectManagement, 12(4), 207–217.

Bredillet, C. N. (2002). Mapping thedynamics of project management field:Project management in action. InProceedings of PMI ResearchConference (pp. 157–169). NewtownSquare, PA: Project ManagementInstitute.

Bredillet, C. N. (2006). The future ofproject management: Mapping thedynamics of project management fieldin action. In D. I. Cleland & R. Gareis(Eds.), Global project managementhandbook (2nd ed., pp. 3-1–3-24) NewYork: McGraw-Hill.

Bunch, J. F. S. (2003). Connecting thedots: Aligning projects with objectivesin unpredictable times. Academy ofManagement Executive, 17(4), 157–159.

Cascio, W. F. (2000). Managing a virtualworkplace. Academy of ManagementExecutive, 14(3), 81–90.

pmj402_20111.qxd 4/28/09 4:25 PM Page 101



AP

ER

S

Charnes, A., & Cooper, W. W. (1957).Management models and industrialapplications of linear programming.Management Science, 4(1), 38–91.

Cianni, M., & Wnuck, D. (1997).Individual growth and team enhance-ment: Moving toward a new model ofcareer development. Academy ofManagement Executive, 11(1), 105–115.

Corbett, C. J., & Van Wassenhove, L. N.(1993). The natural drift: What hap-pened to operations research?Operations Research, 41, 625–640.

Crawford, L., Pollack, J., & England, D.(2006). Uncovering the trends in proj-ect management: Journal emphasisover the last 10 years. InternationalJournal of Project Management, 24,175–184.

Crowston, W., & Thompson, G. L.(1967). Decision CPM: A method forsimultaneous planning, scheduling,and control of projects. OperationsResearch, 15, 407–426.

Dillon, R. D., Paté-Cornell, E. M., &Guikema, S. D. (2003). Programmaticrisk analysis for critical engineeringsystems under tight resource con-straints. Operations Research, 51,354–370.

Forrester, J., & Drexler, A. B. (1999). Amodel for team-based organizationalperformance. Academy of ManagementExecutive, 13(3), 36–49.

Friedman, T. L. (2005). The world is flat:Brief history of the twenty-first century.New York: Farrar, Straus, and Giroux.

Gaddis, P. O. (1959). The project man-ager. Harvard Business Review, 37(3),89–97.

Gerwin, D. (2004). Coordinating newproduct development in strategicalliances. Academy of ManagementReview, 29, 241–257.

Gerwin, D., & Ferris, J. S. (2004).Organizing new product developmentprojects in strategic alliances.Organization Science, 15(1), 22–37.

Goh, C. H., Holsapple, C. W., Johnson,L. E., & Tanner, J. R. (1997). Evaluatingand classifying POM journals. Journal

of Operations Management, 15(2),123–138.

Gottschalg, O., & Zollo, M. (2007).Interest alignment and competitiveadvantage. Academy of ManagementReview, 32, 418–437.

Gustafsson, J., & Salo, A. (2005).Contingent portfolio programming forthe management of risky projects.Operations Research, 53, 946–995.

Gutierrez, G., & Paul, A. (2000).Analysis of the effects of uncertainty,risk-pooling, and subcontractingmechanisms on project performance.Operations Research, 48, 927–938.

Hodgetts, R. M. (1968). Leadershiptechniques in project organization.Academy of Management Journal, 11,211–219.

Hoffman, G. H. (1982). Project cost con-trol: Dynamic risk analysis of randomlyordered sequential decisions underuncertainty. Interfaces, 12(3), 45–51.

Howell, R. A. (1968). Multiproject con-trol. Harvard Business Review, 46(2),63–70.

Jassawalla, A. R., & Sashittal, H. C.(2002). Cultures that support produc-tion-innovation processes. Academy ofManagement Executive, 16(3), 42–54.

Karl, K. A. (1999). Mastering virtualteams. Academy of ManagementExecutive, 13(3), 118–119.

Keefer, D. L., Kirkwood, C. W., &Corner, J. L. (2004). Perspective ondecision analysis applications, 1990-2001. Decision Analysis, 1, 4–22.

Kessler, E. H., & Bierly III, P. E. (2001).Vasa syndrome: Insights from a 17th-century new-product disaster.Academy of Management Executive,15(3), 80–91.

Kirsch, L. J. (1996). The management ofcomplex tasks in organizations:Controlling the systems developmentprocess. Organization Science, 7(1), 1–21.

Kloppenborg, T., & Opfer, W. (2002).The current state of project manage-ment research: Trends, interpretation,and predictions. Project ManagementJournal, 33(2), 5–18.

Kunreuther, H. (1969). Extension ofBowman’s theory on managerialdecision-making. ManagementScience, 15(8), B-415–B-439.

Kwak, Y. H., & Anbari, F. T. (2008a).Project management research trendsof allied disciplines. Presented at the2008 PMI Research Conference,Warsaw, Poland.

Kwak, Y. H., & Anbari, F. T. (2008b).Impact on project management ofallied disciplines: Trends and future ofproject management practices andresearch. Newtown Square, PA: ProjectManagement Institute.

Kwak, Y. H., & Anbari, F. T. (in press).Analyzing project managementresearch: Perspectives from top man-agement journals. InternationalJournal of Project Management.

Leifer, R., O’Connor, G. C., & Rice, M.(2001). Implementing radical innova-tion in mature firms. Academy ofManagement Executive, 15(3), 102–113.

Li, H., Bingham, J. B., & Umphress,E. E. (2007). Fairness from the top:Perceived procedural justice and col-laborative problem solving in newproduct development. OrganizationScience, 18, 200–216.

Liao, S. (2005). Technology manage-ment methodologies and applications—A literature review from 1995 to 2003.Technovation, 25, 381–292.

Linton, J. D., & Thongpapanl, N.(2004). Perspective: Ranking the tech-nology innovation management jour-nals. Journal of Product InnovationManagement, 21, 123–139.

Malhotra, A., Majchrzak, A., & Benson,R. (2007). Leading virtual teams.Academy of Management Perspectives,21(1), 60–70.

Maxfield, M. W. (1981). Sequencingand scheduling in real time—Quickly.Interfaces, 11(3), 40–43.

Merino, T. G., Pereira do Carmo, L., &Alvarez, V. S. (2006). 25 years ofTechnovation: Characteristics and evo-lution of the journal. Technovation, 26,1303–1316.

pmj402_20111.qxd 4/28/09 4:25 PM Page 102


Middleton, C. J. (1967). How to set up aproject organization. Harvard BusinessReview, 45(2), 73–82.

Miles, R. E., Snow, C. S., Mathews, J. A.,Miles, G., & Coleman, H. J., Jr. (1997).Organizing in the knowledge age:Anticipating the cellular form.Academy of Management Executive,11(4), 7–20.

Miller, R. W. (1962). How to plan andcontrol with PERT. Harvard BusinessReview, 40(2), 93–104.

Northcraft, G. B., & Wolf, G. (1984).Dollars, sense, and sunk costs: A lifecycle model of resource allocationdecisions. Academy of ManagementReview, 9, 225–234.

O’Connor, G. C., Rice, M. P., Peters, L., &Veryzer, R. W. (2003). Managing inter-disciplinary, longitudinal researchteams: Extending grounded theory-building methodologies. OrganizationScience, 14, 353–373.

Olson, J. E. (2005). Top 25 businessschool professors rate journals in oper-ations management and related fields.Interfaces, 35, 323–338.

Ormerod, R., & Kiossis, I. (1997).OR/MS publications: Extension of the analysis of US flagship journals to theUnited Kingdom. Operations Research,45, 178–187.

Pitsis, T., Clegg, S. R., Marosszeky, M., &Rura-Polley, T. (2003). Constructingthe Olympic dream: A future perfectstrategy of project management.Organization Science, 14, 574–590.

Prasad, S., & Babbar, S. (2000).International operations managementresearch. Journal of OperationsManagement, 18, 209–247.

Reisman, A. (1994). Technology man-agement: A brief review of the last 40years and some thoughts on its future.IEEE Transactions on EngineeringManagement, 41, 342–346.

Reisman, A., & Kirschnick, F. (1994).The devolution of OR/MS: Implicationsfrom a statistical content analysis ofpapers in flagship journals. OperationsResearch, 42, 577–588.

Roman, D. D. (1964). Project manage-ment recognizes R&D performance.Academy of Management Journal, 7(1),7–20.

Schussel, G., & Price, S. (1970). A casehistory in optimum inventory schedul-ing. Operations Research, 18(1), 1–23.

Scott, S. G., & Einstein, W. O. (2001).Strategic performance appraisal inteam-based organizations. Academy ofManagement Executive, 15(2), 107–116.

Söderlund, J. (2004). On the broaden-ing scope of the research on projects: Areview and a model for analysis.International Journal of ProjectManagement, 22, 655–667.

Soteriou, A. C., Hadjinicola, G. C., &Patsia, K. (1999). Assessing productionand operations management relatedjournals: The European perspective.Journal of Operations Management, 17,225–238.

Themistocleous, G., & Wearne, S. H.(2000). Project management topic cov-erage in journals. International Journalof Project Management, 31(2), 7–11.

Townsend, A. M., DeMarie, S. M., &Hendrickson, A. R. (1998). Virtualteams: Technology and the workplaceof the future. Academy of ManagementExecutive, 13(3), 17–29.

Vazsonyi, A. (1982). Computer-supported Gedanken experiments.Interfaces, 12(4), 34–41.

Vokura, R. J. (1996). The relativeimportance of journals used in opera-tions management research: A citationanalysis. Journal of OperationsManagement, 14, 345–355.

Wasil, E. A., & Assad, A. A. (1988).Project management on the PC:Software, applications, and trends.Interfaces, 18(2), 75–84.

Zangwill, W. I. (1966). A deterministicmultiproduct, multifacility productionand inventory model. OperationsResearch, 14, 486–507.

Young Hoon Kwak, PhD, is a project manage-ment faculty member at The George WashingtonUniversity’s School of Business. His researchinterests are in the area of strategic issues inproject management, risk management, technol-ogy management, and engineering-construction-infrastructure project management. For moreinformation, please visit his website athttp://home.gwu.edu/�kwak.

Frank T. Anbari (PhD in project managementand quality enhancement, MBA, MS in engineer-ing, PMP, PE, and Six Sigma Black Belt) is a facul-ty member and past director of the ProjectManagement Program at The George WashingtonUniversity. He held leadership positions inindustry and taught at Drexel University, PennState University, and the University of Texas atDallas. For more information, please visithttp://home.gwu.edu/�anbarif.

pmj402_20111.qxd 4/28/09 4:25 PM Page 103

Cover to CoverKenneth H. Rose, PMP, Book Review Editor


One consistent critical success factor

for every endeavor is the ability to

adapt to an ever-changing market.

The fourth edition of the Project

Management Institute’s (PMI’s) PMBOK ®

Guide has achieved this objective.

In the introduction, a quick summary of

the relationships among projects, programs,

and portfolios, along with that of a project

management office, is given, thereby setting

the stage for the other associated PMI® stan-

dards. Another addition is the broadened

project life cycle, which allows for a more

adaptive application of the project life cycle

from that of a linear phased approach to one

that is able to accommodate overlapping

and more iterative and incremental life

cycles. This adaptation is further reinforced

by the clarification that the Process Groups

are not project phases, and that the Process Groups are to be

considered a part of each life-cycle phase.

The fourth edition has several notable changes and clarifica-

tions. Some processes have been deleted, and new processes

have been added (e.g., Identify Stakeholders). For the project

management processes (versus the product-oriented processes),

several points are stressed, including that these processes,

although they are used across regions, industries, and other

project differentiators, are not to be applied as a “one-size-fits-

all” rigid approach. Similarly, although the processes are pre-

sented in a sequential fashion, their application is not to be

defined as strictly linear, and they are not to be considered as

separate entities. Although this seems obvious, the fourth edition

refocuses the view of the PM standard back to its roots as a

framework rather than a rigid protocol. Another subtle, but effec-

tive change is that of using a verb-noun format, thereby accentu-

ating the end deliverable for each process.

The fourth edition has added Process Group data flow dia-

grams that are exceptionally helpful, focusing the reader’s

attention on the inputs/activity/outputs in a clear and concise

manner. This approach dovetails well with the intent of letting

the project team determine what artifacts are necessary for

their project, and to what level of detail.

The appendices really add value on sever-

al fronts. Appendix A clearly outlines the major

changes for each of the Knowledge Areas, as

well as the overall PM process. A third versus

fourth edition comparison chart for each

Knowledge Area has been included that

enables the reader to clearly see the exact

changes. Also, there is a short clarification sec-

tion regarding the differences and relationship

between the project charter and the project

scope statement. The project charter is one of

the inputs into the development of the project

scope statement, and the project scope state-

ment focuses on the tactical approach for the

project. For those readers who prefer a short,

but focused outline of the new standard,

Appendix F provides an excellent summary of

the standard, identifying and providing a con-

cise one- or two-line definition or description

of the activities within each Knowledge Area. Similarly, for those

interested in the evolution of the standard, Appendix B provides a

short overview of the progression of the standard, noting the major

changes for each of the four editions.

Appendix D is very short but opens the door for the concept

of developing industry-specific extensions to the standard. This

extension concept has been developed for the construction

industry, as well as for government work. The refocusing of

the PMBOK ® Guide back to its framework legacy, along with the

accommodations for different life cycles, will also encourage

the development of other industry extensions.

Appendix G is a wonderful addition to the standard. It focuses

on interpersonal skills needed in the project management arena.

Traditionally, most of the emphasis on project management has

been more on the technical side (e.g. work breakdown structures,

schedules, budgets, etc.), and while those items are critical to plan-

ning and executing a project, project managers achieve results

through others. Interpersonal and other “soft” skills are equally crit-

ical and arguably the most important for the success of our projects.

Although this new fourth edition is a good resource for those

studying for Project Management Professional (PMP®)certification,

a solid review is strongly recommended for even the most seasoned

PM professional, if even just to review the changes from previous

editions and to serve as the foundation for the other PMI standards.

Reviewed by Greg Indelicato, PMP, Director, Delivery Management, ATX Group,Inc., Irving, TX, USA.

A Guide to the Project Management Body of Knowledge (PMBOK® Guide), Fourth Editionby Project Management Institute

Project Management Institute, 2008,ISBN: 9781933890517, paperback, 485pp., $49.50 Member, $65.95 Nonmember.

Project Management Journal, Vol. 40, No. 2, 104© 2009 by the Project Management InstitutePublished online in Wiley InterScience (www.interscience.wiley.com)DOI: 10.1002/pmj.20125

PMJ402_20125.qxd 4/28/09 4:57 PM Page 104


Co

ver

to C

ove

r

S ince many organizations have emb-

raced project management and are

operating in a management-by-proj-

ects context, it is only natural that

program management has emerged as a dis-

cipline of its own. Through program manage-

ment and effective grouping of projects into

programs to obtain greater benefits than

when projects are managed in a stand-alone

fashion, organizations are able to increase

effectiveness and efficiency, not to mention

customer satisfaction. Dr. Paul Sanghera has

written an outstanding reference guide to

program management complementing the

Project Management Institute’s (PMI’s) The

Practice Standard for Program Management

(2004). Following the issuance of the stan-

dard, PMI launched a Program Management

Professional (PgMP®) credential in 2007.

Sanghera’s book provides many insights into

the concepts discussed in the standard and is especially useful

with the examples that are provided.

Each chapter contains learning objectives, an introduction, a

fundamental triplet (the three most important points covered in

the chapter), a summary, key terms, and discussion exercises.

Relevant notes with examples, such as the meaning of an order-of-

magnitude estimate in Chapter 3 and program milestones in

Chapter 5, are inserted as appropriate. He also includes “alerts,”

such as one in Chapter 6 on controlling cost in which he states that

the traditional approach is to focus on cost overruns, but underruns

also need consideration to be able to return funds to the organiza-

tion if possible. “Tips” are interspersed in many chapters, such as

one in Chapter 8 on developing the program team that advises read-

ers to ensure there is a balance between the needs of the program

and those of an individual team member’s career path. “Cautions”

are described as well. For example, in Chapter 3 Sanghera warns

that projects may exist before the program is established but also

may be authorized throughout all program phases except closing.

This approach makes the book an excellent complement to a short

course or university course on program management.

Sanghera begins with an overview of project management in

Chapter 1 to set the stage for the rest of the book, which focuses

on program management. Chapter 2 presents an introduction

describing differences between project and program management,

the three program management themes

(stakeholders, benefits, and governance), the

program life cycle, and the program Process

Groups. As it may be somewhat difficult to

grasp the relationship between the phases in

the life cycle and the Process Groups, Sanghera

explicitly explains this concept. He also adds to

the standard in several places in the book, and

in this chapter in particular, through a discus-

sion of applicable program management met-

rics. Chapter 2 also introduces a fictional case

study, which is used in the remaining chapters

to explain the key concepts covered.

Chapters 3–11 are organized by the nine

Knowledge Areas in the third edition of PMI’s

A Guide to the Project Management Body of

Knowledge (PMBOK® Guide) but with a pro-

gram perspective. Each chapter contains an

initial discussion of the “big picture” followed

by inputs, tools and techniques, and outputs

of the various processes in these Knowledge Areas, with examples

showing the importance to program management. Insightful

descriptions of various methods for program and project selec-

tion are included, as well as an in-depth treatment of the key

earned-value formulas and why they are useful in program per-

formance measurement. His approach, with Knowledge Areas

and tools and techniques, is consistent with that now in the sec-

ond edition of PMI’s Standard for Program Management (2008).

Fundamentals of Effective Program Management is an excel-

lent reference to assist individuals in preparing for the PgMP®

examination. I only wish it had been available when I took and

passed the examination portion of the PgMP® in 2007. However,

this book is based on the first edition of the PMI standard, and,

unfortunately, much of its content will be out of date when the

exam changes on August 31, 2009, to reflect the second edition of

the standard. At that time, this book will lose its stature as a ref-

erence for those studying for the PgMP® exam; but, it will con-

tinue to be an important guide for practicing program managers,

or students in program management, as they further their skills

and competencies in this emerging and important field.

ReferencesProject Management Institute. (2004). The practice standardfor program management. Newtown Square, PA: Author.

Project Management Institute. (2008). The standard for pro-gram management (2nd ed.). Newtown Square, PA: Author.

Reviewed by Dr. Ginger Levin, PMP, PgMP, Project ManagementConsultant and Educator, Lighthouse Point, FL, USA.

Fundamentals of Effective Program Management:A Process Approach Based on the GlobalStandardby Dr. Paul Sanghera, PMP

J. Ross Publishing, 2008, ISBN:9781932159691, hardcover, 376 pp.,$59.95 list price.


PMJ402_20127.qxd 4/28/09 5:24 PM Page 105

Co

ver

to C

ove

r


J.Rodney Turner is one of the premier

thought leaders in project manage-

ment today. Back in the days when

many people were complaining

about the absence of academic interest in project

management, Turner and a handful of other aca-

demics (including, among others, Peter Morris

and Jeff Pinto) were plugging away, doing first-

rate research on projects, project management,

project organization, and the determinants

of project success and failure. I suspect that the

dearth of scholarly work was rooted in the dearth

of project management programs in universities.

However, in the past 10 years, there has been a

dramatic growth of degree programs in project

management, and with this growth there has

been a corresponding increase in research activi-

ty. Even with the big increase in players, Turner

continues to stand out from the crowd.

When I first got hold of The Handbook of Project-Based

Management, I was puzzled about why it was called a hand-

book. When first leafing through its pages, I felt it was organ-

ized and presented as a standard textbook. Once I got into

the book, however, I saw that it was indeed a handbook in the

sense that you do not begin reading the work on page 1 and

continue through to the end on page 410. It is best read by

thinking about a topic you need information on then going to

that topic directly. Or if you are not looking for elucidation of a

specific topic, it is fun to read by jumping around different sec-

tions. It is also a handbook in the sense that it is chock full of

exhibits, templates, and assessment tools that project teams

can productively employ on their projects. Chapter 18, titled

“Governance of the Project-Based Organization,” contains use-

ful checklists of success elements on projects. They can be

employed to assess the extent to which a given project is pro-

grammed for success . . . or failure. Appendices A and B contain

a raft of exhibits illustrating the use of project management

tools for a hypothetical project.

The book is divided into four parts: Part 1: Managing the

Context; Part 2: Managing Performance; Part 3: Managing the

Process; and Part 4: Governance of Project-Based

Management. Together, these four parts cover most

of what you need to know about the theory and

practice of project management.

I was pleased to see that the third edition is a

bona fide new edition and not simply a cosmetic

rehash of previous editions. It is up to date. Most of

the citations to other works are no more than a few

years old. Turner has worked hard to keep the

handbook current. In the preface, Turner also

points out that the third edition is substantively dif-

ferent from previous editions in that he has added

a brand new section titled “Governance of Project-

Based Management.” In this section he takes a con-

temporary look at project governance, including an

examination of the principal–agent relationship

governing project manager/project sponsor rela-

tionships, bounded rationality, and development

of organizational capability.

My only disappointment with the handbook is that it offers

almost no coverage of the agile and iterative project manage-

ment techniques employed so heavily in software development

projects these days. My disappointment no doubt reflects my

bias as someone whose roots are in software development.

These techniques (including Scrum, Application Prototyping,

Time-Boxed Scheduling, and RUP) offer a dramatic shift in per-

spective from the traditional project planning techniques that

emphasize getting all your ducks in a row before beginning proj-

ect work. With them, detailed pre-project planning is replaced

with a plan-as-you-go outlook, which contradicts the project

management wisdom that has prevailed since the Manhattan

Project. Their successful use raises interesting questions: Are the

agile and iterative approaches a “disruptive technology” that will

change the way we carry out projects in the future (including

construction and engineering projects)? Or are they flukes that

only apply to a limited range of IT projects? I would enjoy read-

ing Turner’s opinion—as an accomplished thought leader—on

the place of the agile and iterative perspectives.

Overall, the book is a worthwhile addition to the project

manager’s reference library.

Reviewed by J. Davidson Frame, PhD, PMP, Academic Dean, University of

Management and Technology, Arlington, VA, USA.

The Handbook of Project-Based Management,Third Editionby J. Rodney Turner


McGraw-Hill Professional,2009, ISBN: 9780071549745,hardcover, 452 pp., $85.45Member, $89.95 Nonmember.

PMJ402_20129.qxd 4/28/09 4:53 PM Page 106


Co

ver

to C

ove

r

P rojects represent a major way of intro-

ducing change into organizations

and have thus become a critical fac-

tor in the growth of modern society.

However, despite the existence of project man-

agement methodologies, bodies of knowledge,

maturity models, and lists of typical success

factors, many projects continue to flounder.

Journalists prey on such failures, governments

are exasperated by the publicity, and many

organizations attempt to hide their existence.

One of the key reasons for many of the failures

is the lack of clearly defined targets, objectives,

and measures. Even where these can be articu-

lated, it is still necessary to determine their

achievement. A focus on obtaining visibility

and transparency during project development

can therefore become crucial to the successful

completion of projects. Project Reviews, Assurance and Governance

by Graham Oakes attempts to step into the breach by focusing

attention on the role of project reviews in delivering organiza-

tional objectives. Indeed, if we don’t know how a project is per-

forming, we will be unable to guide it toward success. Reviews

and assurance provide a prism through which we can assess the

performance of projects, the risks they carry, and the potential

impact on the organization, enabling us to adjust the portfolio in

response.

Reviews come in all shapes and sizes. The main purpose of

reviews is to ground people in reality by providing clear and

independently validated information. Many groups rely on the

outputs of reviews. Project managers need insights into perform-

ance, sponsors require a broader view to manage commitments,

senior executives vie for information to prioritize and manage

their portfolio commitments, and stakeholders seek assurances

regarding the outputs they depend upon. Moreover, reviews can

either emphasize the assurance aspect, where the likelihood of

success is assessed, or the audit perspective, where compliance

in accordance with relevant policies and standards is confirmed.

It therefore becomes imperative to understand the purpose of a

project review before determining the timing (event-based or

periodic), independence (independent assurance, peer review,

or self-assessment), and formality (informal checkpoint or for-

mal gateway reviews and audits) required.

The book is divided into four parts. The

first part introduces project reviews and their

role in assurance and governance. One of the

examples provided is of the Office of Govern-

ment Commerce in the United Kingdom,

which mandated the use of the Gateway

Review Process in major public-sector proj-

ects. The six types of gateways encompass all

activities from strategic assessment and busi-

ness justification to readiness for service and

postdelivery benefits evaluation. Key findings

are conveyed to the senior responsible owner,

and the project is assigned a green, amber, or

red status. It is estimated that gateway reviews

now contribute around $1.5 billion of cost

savings per annum in the United Kingdom.

Oakes maintains that there are three

key challenges to establishing reviews and

dedicates the remainder of the book to addressing these in turn.

Given the limited time and resources in most organizations,

Part II offers a clear model for running reviews focusing energy on

defining the objectives and gathering evidence. Part III responds to

the difficulty of getting people to act on the findings by looking at

how reviews inform organizational decision making. Aligning the

outputs of reviews to the audience’s interest and scope of control

maximizes the likelihood that findings will be relevant and subse-

quently acted upon. Part IV is concerned with the final hurdle of

persuading the organization to invest systematically in reviews

and embed them into governance and assurance structures.

Oakes makes a strong case for reflective organizations that can

derive long-term benefits from the insights afforded through

reviews. The core of the book provides practical models, directions,

and structures for review processes and governance. A key strength

of the book is the 11 examples of practices in different industries.

The writing style is somewhat informal, yet passionate, and

the book would have benefited from additional diagrams.

Nonetheless, Oakes delivers tools for keeping in touch with reali-

ty rather than obsessively managing plans. The book is really

about creating a culture that will benefit from introspective

insights and continuous organizational learning. Visibility and

transparency are needed to make projects responsive to new real-

ities. As we strive to improve our project delivery track record, we

need to learn to make sense of the past and adjust the present.

Reviewed by Darren Dalcher, PMP, Chair Professor of Software ProjectManagement at Middlesex University and Director of the NationalCentre for Project Management, London, United Kingdom.

Project Reviews, Assurance and Governanceby Graham Oakes

Gower, 2008, ISBN: 9780566088070,hardcover, 288 pp., $114.95 list price.


PMJ402_20126.qxd 4/28/09 5:57 PM Page 107

Documents

19 Virtual Team Concepts in Projects: A Case Study - pmi.org · Workgroups Through Social Networks ... William Leban,PMP, DeVry University ... 19 Virtual Team Concepts in Projects: