Implementation and performance of a matrix organization structure
International Journal of Project Management 21 (2003) 51–62 www.elsevier.com/locate/ijpromanImplementation and performance of a matrix organization structure John A. Kuprenas* Department of Civil Engineering, University of Southern California, Los Angeles, CA 90089-2531, USA Received 23 August 2000; received in revised form 9 January 2001; accepted 14 September 2001Abstract This paper presents a case study in the implementation and use of a matrix organization. This paper describes the matrix orga-nization installed at the City of Los Angeles, Bureau of Engineering. The work explains how the need for a matrix structure wasidentiﬁed, reviews the creation of the matrix, describes the problems associated with the implementation of the new structure, andevaluates the eﬀectiveness of the project management process within the matrix organization. The study ﬁnds that althoughimplementation problems have occurred, the performance of the organization while operating under a matrix structure hasimproved with respect to project delivery. # 2002 Elsevier Science Ltd and IPMA. All rights reserved.Keywords: Managing program; Managing projects; Management structure1. Introduction large-scale consulting, aerospace, and construction industries . Matrix management began in the 1960’s as an organiza- This paper presents a case study in the implementationtional means to meet the needs of the aerospace industries and use of a matrix organization. This paper describes. The government contract selection process required a the matrix organization installed at the City of Losproject-oriented system directly linked to top-manage- Angeles, Bureau of Engineering. The work explains howment. In order to meet this requirement, the aerospace the need for a matrix structure was identiﬁed, reviewsﬁrms established a set of horizontal project groups over the creation of the matrix, describes the problems asso-their traditional vertical functional organizations . In ciated with the implementation of the new structure,creating such an organization, many employees would and evaluates the eﬀectiveness of the project manage-end up working both under a department head and a ment process within the matrix organization. This paperproject manager of an interdisciplinary project team. beneﬁts practicing project managers by presenting aWhen depicted on paper, this crossing of organizational case study in which many matrix organization imple-lines is easily be represented by a grid of a matrix; mentation and operation diﬃculties identiﬁed throughhence, the term Matrix Organization was created . past research were in fact manifested in the organiza- When compared with other organizational forms, a tional shift. Managers can learn how this case studymatrix organization is a mixed form in which traditional organization solved these problems and the eﬀectivenesshierarchy is overlaid by some form of lateral authority, of the solutions, as well as the eﬀectiveness of the entireinﬂuence, or communication. This overlay present in a organizational structure change. Managers and aca-matrix creates two chains of command—one along the demics will also beneﬁt through a discussion of addi-functional lines, the other along project lines . Given tional project management research needs in the areasthe complexity of project management in the actual of matrix structures and organizational performance.business world, the matrix structure is often used wherethe need for strong technical assistance across manyareas is required . It is still particularly popular in the 2. Development of the matrix The case study takes place in the city of Los Angeles. * Tel.: +1-213-740-0603; fax: +1-213-744-1426. With a population of over 3.5 million persons, Los E-mail address: firstname.lastname@example.org Angeles is located along the southern coast of the State0263-7863/02/$22.00 # 2002 Elsevier Science Ltd and IPMA. All rights reserved.PII: S0263-7863(01)00065-5
52 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62of California in the western United States. The City is Consistent to all of these reviews was the speciﬁc468 square miles and has an irregular shape with the recommendation to move from the existing functionalmost expansive areas being in the northern part of the organization toward a project management style ofcity and tapering down to a strip in the southern por- project delivery and organization. The reports assertedtion. Los Angeles experiences a Mediterranean type of that within the current system, little or no ownership ofclimate (dry summers and wet winters) with an average projects existed and no performance measures were intemperature of 18 C. The City of Los Angeles, place to measure project status. As a result, in FebruaryDepartment of Public Works is the City’s third largest 1997, the Bureau of Engineering executive staﬀ and theDepartment with over 5500 employees and is respon- Board of Public Works mandated a dramatic shift tosible for construction, renovation, and the operation of improve project delivery. Despite well documented dif-City facilities and infrastructure. The Bureau of Engi- ﬁculties in the use of a matrix structure [9–11] all of theneering, with over 900 employees (with over 550 engi- Bureau divisions were to be organized into a matrixneers), is a part of the City of Los Angeles’ Department organization with a project manager being the focus forof Public Works and is responsible for the planning, project delivery with project conception-to-conclusiondesign, and construction management of capital responsibility. Several organizational options were con-improvement projects for the city including municipal sidered before the ﬁnal matrix structure was selected, butfacilities, stormwater, sewer, street and other infra- in the ﬁnal analysis, a matrix organization was selected.structures projects. A matrix was selected because past research has Although the quality of the Bureau of Engineering’s shown that of the existing conditions and requirementsdesigns has always been excellent, some past programs within the Bureau are the classic elements that wouldwithin the Bureau experienced signiﬁcant delays and make the structure a strong choice. The Bureau condi-cost overruns in the design and construction of projects. tions ofIn 1993 a new Mayor was elected for Los Angeles, and . outside pressure for dual focus (between projectthese past overruns were no longer tolerated. The delivery and state of the art design);Mayor felt that many groups within the City’s Munici- . pressures for high information processing (multiplepal Government had become entrenched in bureau- diverse projects and reporting requirements); andcracy, and in order to save the city money, some . pressures for shared resources (for all non Civilorganizations (including the Bureau of Engineering) design disciplines);could be privatized or reorganized to be run more eﬃ-ciently at a lower cost. Hence, in the time period from were all identiﬁed by Tatum as key basic conditions for1994 to 1996 several groups were commissioned by the selection of a matrix . Other reasons also identiﬁedMayor to analyze the Department of Public Works and to favor the use of the matrix were the necessary inher-the Bureau of Engineering. The groups studied the ent control of client group projects/programs (theBureau’s past performance with respect to capital pro- Bureau’s principle function) in the structure whileject design and compared the costs to similar engineer- maintaining functional authority levels (needed for aing organizations within the United States. The groups public organization such as the Bureau). Other optionsalso studied the Bureau’s organizational structure and considered but not used are shown in Fig. 1.interviewed dozens of Bureau staﬀ to better understand Even after selecting a matrix structure, the Bureauthe reporting and communication relationships still needed to determine which type of matrix to use All study ﬁndings were similar. The ﬁrst study identi- since the matrix structure can operate in may diﬀerentﬁed the need for a comprehensive project control sys- ways. Larson and Gobeli  deﬁned three-matrixtem, a coordinated plan of Bureau programs, and a new types—functional, balanced, and project. The amountmanagement structure to create this plan . The sec- of authority of the functional manager diﬀerentiatesond report identiﬁed the need to organize the Bureau between the three types. A summary of the types ofaround its key programs, to create a project manage- matrix is shown below:ment organization to improve accountability for each Functional Matrix: in a functional matrix, staﬀproject, and to correct or improve the senior engineers’ involved in the delivery process remain under controlhigh degree of autonomy that had made it diﬃcult to of the functional manager, while project managerseﬀectively manage projects and programs that cross are formally designated to oversee the project acrossdivisional boundaries . The third report identiﬁed the diﬀerent functional areas. As a result, project man-need to establish a group of project managers (separate agers have limited authority over functional staﬀ and,from project engineers) to manage the design and con- therefore, primarily plan and coordinate the project.struction phases of capital projects and the need for a Under this form of matrix, functional managersmulti-year capital improvement program (CIP) with retain primary responsibility for their speciﬁc seg-priorities to be used to identify type and amount of ments of the project .required resources .
J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 53 Fig. 1. Organizational options considered.
54 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 Balanced Matrix: in a balanced matrix, the functional 3. Implementation challenges manager and the project manager share responsibility for the project resources. Under this form of matrix, Research has shown that successful implementation project managers are assigned to oversee the project of a matrix structure in any form of organization can be and interact on an equal basis with functional man- expected to be diﬃcult. The matrix implementation agers. Functional managers and project managers within the Bureau of Engineering did present several jointly direct project work and approve technical and implementation diﬃculties speciﬁcally predicted in a operational decisions . review of the previous research literature. Table 1 shows Project Matrix: in a project matrix, the functional a summary of implementation diﬃculties and how the managers authority is the smallest, with functional Bureau implementation overcame the speciﬁc challenges managers only assign resources for the project and pro- predicted by past research on matrix organization vide technical consultation on an as-needed basis. Pro- implementation [4,5,13–17]. The ﬁrst column in the ject managers are assigned to oversee the project and are table shows the implementation challenge identiﬁed responsible for the completion of the project . through past research studies. The second column shows how these theoretical diﬃculties were in fact In the analysis of the matrix organizational options, manifested in the implementation of the matrix organi-the Bureau opted for a hybrid functional matrix where zation at the Bureau of Engineering. Additional detailsthe project and functional manager authorities are on each of the challenges and solutions follows.deﬁned as for a classic functional matrix, but the matrixstructures (between functional groups and project man- 3.1. Roles and responsibilitiesagers) are constructed around projects within programswith the program manager having complete authority As expected based upon past research, confusion overover all projects in his or her program. Four main pro- roles and responsibilities is common within a matrixgrams exist in the new matrix structure—Wastewater, organization [13–15]. In turn, the Bureau experienced aStreet, Stormwater, and Municipal Facilities. The new signiﬁcant amount of confusion and conﬂict over rolesorganizational structure of the Bureau is shown in Fig. 2 and responsibilities between functional managers lead-and was designated by the Bureau to be called a pro- ing design teams and project managers overseeing pro-gram based matrix. ject performance. This was manifested through angryTable 1Implementation challenges and case study solutions necessityImplementation challenge identiﬁed through past research Bureau of Engineering case study implementation solution(1) (2)Confusion and conﬂict over roles and responsibilities between Creation of summary lists of roles and responsibilities for functional andfunctional managers leading design teams and project project managersmanagers overseeing project performance [13–15]Need for a reporting system to monitor functional Creation of project design cost templatesfor all types of projects donemanager commitments  within the Bureau and implementation of new project reporting and control systemFunctional manager politicization of assignment of scarce Development and use of project prioritization protocol (policy approvedresources between projects leading to project delays/changes by the City Engineer which became a standard Bureau practice)and to changes in project prioritization [13,16]The dual authority of a matrix requires people who are All Bureau staﬀ were trained in human relations training speciﬁcallyadaptive and comfortable with ambiguity to prevent pertaining to change, communication, and working in teamsnegative inﬂuences to motivation and job satisfaction[4,16] numerous interfaces inherent in a matrix structurerequire strong communication skills and an ability towork in teams [4,17]A development program speciﬁc to project managers is Weekly mentoring sessions were held for all project delivery team membersneeded to establish a common language and understanding (functional or project). Monthly project manager roundtables were heldof management processes  with the project managers to share problems and solutionsFunctional side of the organization becomes more powerful Formalization of an annual project planning process that evaluates functionalthat the project side; functional managers do not gain a group performance based upon project based goals (number of projectsproject focus  completed and labor hours required to complete the projects)
J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 55 Fig. 2. Program based matrix structure.
56 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62memos to supervisors, non-attendance at project meet- with respect to equal opportunity outreach to minorityings by functional managers, slowed delivery of design and woman owned businesses. When completed, thesestatus reports to project managers, and lack of comple- various Handshake Agreements become the commit-tion of certain administrative tasks with subsequent ﬁn- ments of the functional managers. Handshake Agree-ger-pointing. ments for all Bureau project design processes on all The solution to the problem was creation of written projects were ﬁnalized by 1 June 1998. Most agreementsroles and responsibilities for both project and functional were simple memorandum documents copied to allmanagers. A ﬁrst attempt to create a comprehensive list members of the project team and the project teamof all tasks for both types of managers resulted in per- supervisors.haps more confusion than assistance, as hundreds of Although simple in form, the Handshake Agreementtasks were identiﬁed and literally thousands of more commitments from the functional managers are basedspeciﬁc duties could certainly have been identiﬁed for upon templates of historical functional team processeither position. A simpler and much more eﬀective performance. Unreasonable expectations cannot besolution to the confusion and conﬂict was the compila- imposed upon the functional managers by the projecttion of a list of ten fundamental tasks for each position. managers since all handshake agreement baseline valuesThese lists are shown in Fig. 3. All parties understood are negotiated from 21 pre-established project designthat these lists were by no means all-inclusive; rather templates which give speciﬁc dollar value and timethey provided a foundation for each party’s responsi- duration values for every Bureau design sub-elementbilities. After the publication of these simple lists and 1 based upon size and type of project. These templatesmonth of learning, the manifestations of confusion eliminate the temptation for Project managers to ‘‘lowover roles and responsibilities reduced to practically ball’’ the design team to agree to an unrealistic perfor-zero with the managers having taught themselves mance standard, and prevent the temptation of thewho was to perform what speciﬁc tasks, often times functional managers to over-inﬂate their estimates inwith much compromise depending upon individual order to appear productive. The Handshake Agree-workloads. ments can also be used to help eliminate bottlenecks created by poor resource leveling. Future agreements3.2. Reporting system will include a master project schedule so that the func- tional managers know at what time each project is A second implementation diﬃculty identiﬁed thor- committed to be delivered and what functional team(s)ough past research literature and manifested within this will conduct the work. Hence, the functional managerscase study was the need for a reporting system to will be less likely to over-commit his/her staﬀ.monitor functional manager commitments . Projectmanager monitoring and control of functional team 3.3. Politicization of projects and resourcesdesign progress was practically non-existent since nowritten estimates of design task cost and schedule were The review of past research on matrix structuresavailable. Hence, the project managers had no control indicated that functional manager politicization ofother than the functional manager’s assurance that assignment of scarce resources between projects caneverything was ﬁne. Reporting project performance is lead to project delays and changes in project prioritiza-now accomplished by a new Project Management Con- tion [13,16]. This was true within this study. In the past,trol System (PMCS) created to monitor and control all even before the establishment of the matrix, manyprojects within the Bureau. The PMCS tracks progress Bureau employees would make arbitrary changes toof all functional manager design teams. This tracking project prioritization based upon from which client theyuses traditional measures of earned value and is repor- last received a phone call or to increase their stature inted at the project manager, and functional manager, and the eyes of city politicians. Upon the shift to a matrixprogram manager levels. structure, the Bureau created, published and began use Most important to this tool and to its success within of a formal project prioritization process under the sig-the matrix organization is that the project manager nature of the City Engineer. The process is shown inensures from the outset of any project that expectations, Fig. 4. The process assigns each project within theroles and responsibilities are established through for- Bureau a speciﬁc rank. Without Program Managermalized speciﬁc project agreements (called Handshake approval (in essence re-prioritization), no work is to beAgreements) in which all functional managers commit done on a lower rank project until the higher rank pro-to project scope, budgets, and schedules for the various ject is complete. Project templates Handshake Agree-components of project delivery. Functional team duties ments are still used to establish functional teamvary based on the team, and they range from detailed performance measures, but politicization of the teamcivil engineering design to compiling environmental eﬀort by the functional manager is eliminated by theimpact documentation to review of bidder compliance new prioritization protocol.
J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 57 Fig. 3. Ten fundamental tasks list.
58 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 Fig. 4. Project prioritization process.
J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 59 With the use of this new prioritization protocol, new as a result of the organizational shift. Hence, as part ofprojects can still be added to a program. When a new or the implementation, all Bureau staﬀ were immediately‘‘rush’’ project now enters a Bureau program, the trained in human relations training speciﬁcally pertain-importance of the project is assessed by the program ing to coping with change, communication, and work-manager using the ﬂowchart to determine whether any ing in teams. This training focused on the personal andongoing design work should be stopped. If the ongoing practical staﬀ needs identiﬁed in the research literaturework is stopped or resources reallocated, then clear in order for the organization to be successful in the newdocumentation exists for the switch in the functional matrix structure. A summary of the human relationsmanager resource’s eﬀort and the change is at the dis- training program is shown in Fig. 5.cretion of senior executive level staﬀ rather than the Training on adapting to change and helping othersfunctional manager. adapt to change explored how disorientation, a normal reaction to change, can aﬀect individuals and teams.3.4. Need skills training The success of the Bureau during the restructuring depended on how people react and adapt. These mod- The numerous interfaces inherent in a matrix struc- ules focused on the crucial role leaders and managersture require strong communication skills and an ability play in eﬀectively exploring change, introducing change,to work in teams, while the dual authority of a matrix and helping others overcome resistance typically asso-requires people who are adaptive and comfortable with ciated with change. Staﬀ learned skills for conductingambiguity in order to prevent negative inﬂuences to eﬀective change discussions that minimized the poten-motivation and job satisfaction [4,16,17]. The Bureau tially negative eﬀects of change on morale, processes,obviously did not wish to experience decreased morale and productivity. Fig. 5. Summary if human relation training.
60 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 The communication training modules were designed works with other Bureau staﬀ on day-to-day routineto help participants understand the impact of eﬀective problems. Both the mentoring and roundtable were wellinteraction skills during communication. Bureau staﬀ received and are planned to be continued. Informationlearned to recognize and overcome communication presented at these sessions is also incorporated into abarriers and interact eﬀectively with Bureau co-workers, new, and continually expanding, project manager pro-managers, and customers. The working in teams’ mod- cedures guideline.ules taught Bureau staﬀ to recognize the personal,interpersonal, and organizational advantages of team- 3.6. Lack of project level focuswork and cooperation—all skills identiﬁed to be criticalto success within a matrix. Participants learned to iden- Another potential diﬃculty of matrix implementationtify and understand the stages of team development and is that the functional side of the organization becomesthe six factors that make teams eﬀective. This training more powerful that the project side. Hence, functionalmodule also oﬀered tips to make Bureau teams more managers do not gain a project focus . Given that thesuccessful, identiﬁed key team member skills, and Bureau was implementing a hybrid function matrix, thisintroduced an action plan for participants. predicted diﬃculty was manifested. When the matrix structure was ﬁrst established, many functional man-3.5. Continued project manager development agers took the view that as long as they still had super- visory control of their staﬀ, no real change had taken Research literature indicates that for a matrix struc- place. They felt that the organization remained a tradi-ture to thrive, a development program speciﬁc to project tional functional organization. This problem wasmanagers is needed to establish a common language and recognized and was corrected by formalization of anunderstanding of management processes . In the annual project planning process.case study of this implementation, this certainly was Functional manager recognition of the importance oftrue. One urgent problem in the implementation was project delivery is now maintained through this newuncertainty (beyond the list of roles and responsibilities) process—a yearly work program plan called the Workover the position of project manager since the new Program Resource Report (WPRR). This plan includesposition had no history and little recognition. The solu- all projects that each functional team within a particulartion used was the Bureau was as prescribed by the program are expected to complete. Based upon the pre-research literature—provide mentoring. Mentoring sup- deﬁned project design templates that provide a laborplied a forum for the new project mangers to receive hour estimate for each project, functional team staﬀdirection and encouragement in performing their new allocations for each year are made and checked versustasks as well as educate key staﬀ within the Bureau at to project design performance. In the years since estab-the duties of the new position. Weekly mentoring ses- lishment of the matrix, by evaluating functional groupsions were initiated for the new project mangers and performance based upon project based goals (number ofselected functional managers. The focus of these ses- projects completed and labor hours required to com-sions was on Department and City processes and how plete the projects), the Bureau has seen a dramaticthe project manager should participate in these pro- increase in WPRR performance (as reported later).cesses. Topics discussed in these training sessions haveincluded: 4. Performance under the matrix 1. Federal funding projects. 2. Environmental documents and requirements. Performance improvements experienced by the 3. The bid process—City requirements. Bureau as a result of functioning under the matrix 4. The award process—City requirements. structure are diﬃcult to assess since performance can be 5. Supplier/Designer Handshake Agreements. measured at several levels. Studies and reports of a pri- 6. Funding of projects through the City Adminis- vate sector engineering ﬁrm reorganized into a matrix trative Oﬃcer (CAO). structures found increased communication and ﬂex- 7. Resolution Authority process (funding of staﬀ ibility while maintaining organizational accountability positions). , more eﬃciency in multiple project deign work, as 8. Role of the City’s Contract administration well as entrepreneurial stability for the ﬁrm , and inspector. reduced unbillable time and improved marketing . In addition this mentoring, when the matrix was ﬁrst As a public sector organization, performance measurescreated, project managers held monthly roundtables. are hard to deﬁne. Nonetheless, the Bureau has devisedThese roundtables were hosted by the Bureau programs some measures and project delivery has improved. Toand were meant to facilitate information exchange on date, the Bureau has measured performance across twothe role of the project manager and how the manager levels:
J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 61Table 2Performance assessment under the matrix structureLevel of performance measurement Performance assessment under the matrix structure(1) (2)Project Level For a t-test (two-sample assuming unequal variances with hypothesized mean diﬀerence=0 and alpha=0.05) the null hypothesis was proven true; therefore, the Bureau found no statistical diﬀerence in design costs between project completed in a matrix structure or under the old functional structure (sample size 243 projects)Program Level 40% Reduction in time from opening of work order to award of construction contract in the Street Program 50% Improvement in annual capital program delivery in the Street Program (delivered 39 out of 41 projects with a budget of $60 million) 50% Improvement in annual capital program delivery in the Stormwater Program (delivered 25 out of 25 projects with a budget of $6 million, in addition to 10 urgent necessity/emergency projects with a budget of $1 million) Creation and use of a standardized Memorandum of Understanding (MOU) to establish performance requirements of project (i.e. manner and method of payment, scope, budget, schedule, funding etc.) within the Municipal Facilities Program 1. Project level performance—how has the individual 5. Conclusions and future research project performance in terms of meeting design cost objectives changed since the implementation Past research on matrix organizational structure of the matrix? implementation has identiﬁed several implementation 2. Program level performance—how has the program diﬃculties. The implementation of the matrix structure performance in terms of meeting annual goals/ within the case study organization of this work has commitments changed since implementation of the shown the past research to be accurate, with the docu- matrix? mented research diﬃculties in fact being manifested in the case study. To address the diﬃculties experienced in Table 2 shows the results of the implementation the case study organization, the organization leadersacross these two levels. The table shows that while the developed polices implemented tools, and conductedcost of completing any particular project has not statis- training within the organization. This study has foundtically changed (based upon two-sample t-test with that despite these implementation problems, the perfor-alpha=0.05) with the new matrix structure, the pro- mance of the organization while operating under agram commitments made during the annual funding/ matrix structure has improved.budget cycle are now being satisﬁed much more The beneﬁt of this work to managers is the doc-completely. umentation of this implementation. Speciﬁc beneﬁts include information relating to: . Additional performance measures beyond these two measures are continually developed and tested . how the need for a matrix structure was identiﬁed; by the Bureau. Tools currently under development . steps in the creation of the matrix and the organi- include: zational options also considered; . Client (i.e. City Department) review of Bureau . process problems associated with the implementa- performance through written surveys or ques- tion; tionnaires. . tangible, tested solutions to process problems . Benchmarking of Bureau performance (project associated with the implementation; and and program) against similar municipal agencies. . evaluation tools to measure the eﬀectiveness of the . Statistical process control charting of Bureau project management process within the matrix functional unit design costs. organization. Additional tools will inevitably be developed as the Future research should continue to document thisuse of the matrix structure continues within the Bureau case study for years to come. Research should monitor
62 J.A. Kuprenas / International Journal of Project Management 21 (2003) 51–62 . whether if any additional implantation diﬃculties  Norby, L. Bond acceleration program benchmark report. Report arise; by the Oﬃce of the City Controller, City of Los Angeles, January 1995. . how the new tools and process described in this  Barrington-Wellesley, Group, Inc. Benchmarking study of the work evolve and perform with time; and bureau of engineering capital improvement program. Report to . whether the performance enhancements docu- the City Administrative Oﬃce of the City of Los Angeles, Feb- mented in this work continue to be achieved by the ruary 1996. organization.  Capital Partnerships, Inc. Alternative organization and staﬃng of DWP’s engineering and other CIP functions. Report to the Los Angeles City Council Ad Hoc Committee to Restructure the Additional research should attempt to identify whe- Public Works Department, September 1996.ther any of the processes used within this case study  McCollum JK, Sherman JD. The Matrix structure: bane or ben-implementation were used, or could be used, in other eﬁt to high tech organizations. Project Management Journalimplementations (matrix or otherwise), and how the 1993;24(2):23–6.eﬀectiveness of the tools varied from implementation to  McCollum JK, Sherman JD. The eﬀects of matrix size and num- ber of project assignments on performance. IEEE Transactionsimplementation. Additional research should also be on Engineering Management 1991;38(1):75–8.conducted with respect to matrix performance. This  Laslo Z, Goldberg AI. Matrix atructures and performance: thework has focused on implementation. Enhanced per- search for optimal adjustment to organizational objectives. IEEEformance as a result of the implementation appears to Transactions on Engineering Management 2001;48(2):144–56.be taking place, but additional performance measures  Tatum CB. New matrix organization for construction manager. Issues in Engineering Journal of Professional Activities 1981;(new/established and quantitative/qualitative) need to 107(4):255–67.be monitored and reported through research publica-  Pitagorsky G. The project manager/functional manager partner-tions for many years to come. ship. Project Management Journal 1998;29(4):7–16.  Kuehn RR, Khandekar RP, Scott CR. The eﬀects of marginality ad reward on matrix conﬂict. Project Management Journal 1996; 27(3):17–26.  Johns TG. On creating organizational support for the project management method. International Journal of Project Manage- ment 1999;17(1):47–53.References  Babcock DL. Managing engineering and technology. Englewood Cliﬀs, New Jersey: Prentice Hall, 1991 (ISBN 0-13-552233-1).  Larson EW, Gobeli DH. Matrix management: contradictions  Turner SG, Utley DR, Westbrook JD. Project managers and and insights. California Management Review 1987;29(4):126–38. functional managers: case study of job satisfaction in a matrix  Knight, K, editors. Matrix management. Gower Press, Teakﬁeld organization. Project Management Journal 1998;29(3):11–19. Limited, Aldershot, UK, 1977.  Bates GD, Roenker G, Junker DJ. A new matrix organization for  Gobeli DH, Larson EW. Matrix management: more than and a design ﬁrm. Issues in Engineering Journal of Professional fad. Engineering Management International 1986;4:71–6. Activities 1981;107(4):237–46.  El-Najdawi MK, Liberatore MJ. Matrix management eﬀective-  Birrell GS. Multiproject multidesigner work by matrix format. ness: an update for research and engineering prganizations. Pro- Journal of Professional Issues in Engineering 1984;110(1):37–53. ject Management Journal 1997;28(1):25–31.  Cardinal RJ. Matrix management may meet your needs for the  Lawler EE. Rethinking organization size. IEEE Engineering 80s. Issues in Engineering Journal of Professional Activities 1980; Management Review 1999;27(2):3–10. 106(1):27–39.