evs 002

8/6/2019 evs 002

http://slidepdf.com/reader/full/evs-002 1/13

Jon of Contct Mngement / Smme 2007 21

The Contton of

Ened Ve Mngement to Poject Sccess onContcted EffotsPjct gs ffct ts tcqs t t t cgs f t pfss.o sc t t ffct s gt (evm). Ts tc

psts qttt sttstcs ppc wt ppt f xpc pctts.

BY roBErt A. MArSHALL

AbaEarned value management (EVM) is believed to bean eective project management methodology onexternal projects under contract. A number o qualitative case studies crossing multiple projectmanagement processes have validated its positivecontribution. Tis research departs rom, as well ascomplements, existing case analyses by taking aquantitative statistical approach. Utilizing Pearson’s

product moment correlation, bivariate linearregression analysis, and canonical discriminant

analysis, this research (1) investigates the directrelationship o the principles o EVM to projectsuccess on contracted eorts; (2) investigates themoderating eect contract type may have on therelationship; and (3) investigates between-groupdierences (xed-price versus cost-plus contractarrangements) with respect to the contribution o EVM mechanics to project procurement manage-ment items. Quantitative techniques are used toanalyze data rom a cross sectional survey o 145experienced earned value practitioners. Te ndingssuggest that stronger implementations o the

principles o EVM result in greater levels o projectsuccess on contracted eorts, with contract typehaving a moderating aect. Te results urther

suggest that EVM mechanics positively contributeto project contract development and administrationand are not dependent on contract type.

inunProject management is challenging. Whether inconstruction, inormation technology, or sowaredevelopment, odds are high that project success will be

jeopardized. Consider that cost overruns o 25–33 percent are not unusual in the construction industry

(Telen Reid & Priest, 2004). Te same industry alsohas seen substantial growth in projects ending in eitherdispute or litigation (Levin, 1998). In the inormationtechnology industry, “at-risk” projects have beencommonplace in the public sector (GAO, 2006).Moreover, 18 percent o projects in the sowareindustry are prematurely canceled (Standish Group,2004). Further still, 53 percent o soware projects will

Abu h Auh

roBErt A. MArSHALL, MBA, MS.PM, PMP, is chai o

the Association o the Adancement o Cost Engineeing(AACE) eaned alue management special inteest

goup, and a pogam manage with the Fedeal Acqui-

sition Seice o the Geneal Seice Administation.

Send comments about this aticle to [email protected].

8/6/2019 evs 002


22 Smme 2007 / Jon of Contct Mngement

dvpmn eVM M

Wk Beakdwn Suue (gaphal)

The ConTribuTion oF earned value manaGemenT

FiGurE 1.

Pj

WP2 WP3 WP4 WP5

1.1 1.2 2.1 2.2 3.1 3.2 4.1 4.2 5.1 5.2

WP1

Wk Beakdwn Suue (abula)

Wk Pakag Numb BcWs/PV BcWs/PV BcWs/PV

1 100 80 90

2 200 175 185

3 300 150 275

4 100 TBD TBD

5 50 TBD TBD

Peiod Total 600 405 550

Poject Total 750 TBD TBD

Pemane Mes

M dfnn Vau h P

BCWS=Pv=Planned value 600

BCWP=Ev=Eaned value 405

ACWP=AC=Actual Cost 550

BCWP - ACWP=Cv=Cost vaiance -145

BCWP (Ev) - BCWS (Pv)=Sv=Schedule vaiance -195

BCWP (Ev) / ACWP (AC)=CPI=Cost Peomance Index .74

BCWP (Ev) / BCWP (Pv)=SPI=Scheduled Peomance Index .67

BAC= åBCWS (Pv) 750

EAC=BAC/CPI 1013vAC=BAC - EAC -263

TCPI=BAC - BCWP (Ev) / BAC 1.72

750

600

550

405 V a l u e

tm

S-cuves

Cuent Peiod(Though Wok Package #3)

8/6/2019 evs 002




exceed their cost, schedule, or scope constraints (Ibid).Tere is no shortage o data about under-perorming projects, and even more requent stories in the workplace about troubled and—worse still—ailed projects. Tereore, project managers need eectivetools and techniques to meet the challenges o their

proession. One such method believed to be eective isearned value management (EVM).

EVM is a comprehensive methodology used tomanage projectized eorts. Te methodology addressesmany project management areas, including projectorganization, planning, scheduling and budgeting,accounting, analysis, reporting, and change control(Fleming and Koppelman, 1996). EVM also incorpo-

rates specic mechanics to include the use o the workbreakdown structure (WBS), perormance curves(S-curves), as well as a dened set o perormance metrics.EVM is among the rst project management methodolo-gies to be codied by national standards setting organizations (ANSI/EIA, 1998; ASI, 2003).

A distinguishing characteristic o EVM is itsunique metrics. EVM integrates a project’s scope,schedule, and cost into a unied set o prescribedmetrics or the purpose o monitoring and orecasting

project perormance. Te building blocks o all EVMmetrics are the ollowing three elements:

Budgeted cost of work performed1.(BCWP or EV 1)=Earned Value

Actual cost of work performed2.(ACWP or AC)=Actual Cost

Budgeted cost of work scheduled3.(BCWS or PV)=Planned Value

FiGurE 1 (on page 22) depicts a simplied project indemonstrating the sequence o EVM implementation.Te methodology begins with planning the entire project scope in the orm o a WBS. Each o the work packages contained in the WBS is cost-estimated,scheduled, and rolled up to the next higher-level WBSelement. An S-curve, bounded by zero and totalcumulative cost as end-points, is created next using BCWS (PV) data to orm a visual baseline. Values orBCWP (EV) and ACWP (AC) are calculated as work progresses, orming their own unique S-curves tocompare against the baseline. Higher-level EVMmetrics are derived rom the three primary values, andare used to monitor and control the project, as well as

to provide perormance inormation to stakeholders.Putting EVM aside and relying on the traditional

accounting viewpoint o this sample project, onemight erroneously conclude that the project is

perorming well. Given that 550 have been expendedagainst the total project budget o 750, it appears the

project is 73 percent complete. Or more optimistically,given that 550 have been expended against a budget todate o 600, the project appears to be perorming below cost.

Closer examination using EVM metrics, however,reveals the opposite. Not only is the project behind inscheduled scope completion [schedule variance(SV)=-195], the project is also over-running costs [cost

variance (CV)=-145]. Only 405 worth o scope havebeen completed despite that 550 have been spent.Moreover, the estimate at completion (EAC) or this project at the current rate o perormance is 1,013—acost overrun o 263. o end on time and on schedule,the project would have to perorm at a rate o 1.72times that originally planned [to complete perormanceindex (CPI)=1.72].

FiGurE 1 oers a simple yet illustrative example o EVM’s unique and eective integration o scope,schedule, and cost data into a unied set o metricsbased on a thoroughly planned project WBS.

lau rvwA review o literature uncovered three categories o knowledge regarding the contribution o earned valuemanagement to project success. Ample literatureoers rational support or EVM’s positive contribu-tion (Presutti, 1993; Fleming and Koppelman, 1996,2002; Christenson, 1998; Abba, 2001; Antvik et al,

2001; Anbari, 2003). Works o this type suggest thebenet o EVM across major project management processes including planning, executing, monitoring,and controlling (PMBOK, 2004). While logical andmeaningul, these works are sources o propositionalknowledge and, in and o themselves, limited to

providing theoretical belies and assertions. Tey donot provide evidence, per se, o EVM’s contribution to project success.

Another source o knowledge is rom the experien-tial accounts o project practitioners (Mukho andLisanti, 1982; Adamczyk, 1989; Chen, 1991; Yu,1996; Antvik, 2001; Kauman, Keating, andConsidine, 2002). Tese works are important in that,unlike theory, they oer empirical evidence based on

8/6/2019 evs 002



rah Famwk

Pj

Pumn

Managmn

personal experience and observations. Tey generallyillustrate EVM’s contribution in one or more projectmanagement processes. o the extent the author’s

projects were successul, a contribution to projectsuccess can be concluded rom these experientialaccounts.

A third source o knowledge about EVM’scontribution to project success blends both theoryand experience (Marrella, 1973; Kim, 2000; Vargas,2003). Tese works oer the highest level o precision in both their methodology and ndingsand can be said to oer scientic knowledge. Worksin this category are research-oriented, and to daterely on qualitative methods to make relevant points.All emphasize EVM’s contribution to project control(Marrella, 1973; Kim, 2000; Vargas, 2003). Addi-

tionally, two o the three have emphasized EVM’scontribution to project planning along with projectcontrol, yet diered on the relative strength o each(Marrella, 1973; Vargas, 2003). In Marrella’s work,the author ound that “…C/SCSC (EVM’s predeces-sor) have aected a signicant improvement in

planning and a positive but less signicant improve-ment in control.” In Vargas’s, the author concludedEVM to be eective in the control o projects, yetdependent on a project’s rst having establishedeective plans. What we know collectively romthese research works is (1) EVM contributes to

project planning; (2) EVM contributes to projectcontrol, but perhaps more or less so than planning;and (3) since EVM contributes to these project

management processes, and these processes arebelieved to positively infuence project perormance,it can be inerred that EVM contributes to projectsuccess.

Among all o the available sources o knowledge,only certain works specically address contract type

with respect to EVM. Teoretical as well as research works support EVM’s contribution to projects undercost-plus contract arrangements (Fleming andKoppelman, 1996; Marrella, 1973; Kim, 2000).Similarly, propositional, empirical, as well as research works, exist in support o EVM’s contribution to projects under xed-price arrangements (Fleming and Koppelman, 1996; Yu, 1996; Antvik, 2001;Kauman, Keating, and Considine, 2002; Vargas,2003). Qualitative justication is available in

support o the belie that EVM contributes to project success generally, and under alternativecontract types, in particular. Conspicuously missing rom the literature is a quantitative approach. Teresearch at hand lls the gap.

Famwk an HyphTe theoretical ramework or this research is thebelie that EVM directly contributes to projectsuccess; and contract type may moderate EVM’soverall contribution as well as the contribution o EVM’s mechanics to contract ormation andadministration items. FiGurE 2 depicts the theoreticalramework or this research.

FiGurE 2.


Pj su

eVM Mhan

Metics

S-Cue

WBS

Pnp eVM

cna typ

Tpe 1: Fixed Pice

Tpe 2: Cost Plus

Contact

Fomation

Contact

Administation

H1H 2

H 3

H 2

H 3

H 4 & H 5

H 4 H 5

H 3

H 4

H5

+

+

+

+

+

Diect elationship Mode ating elationship+

Positie elationship

8/6/2019 evs 002


8/6/2019 evs 002



TablE 1.

S-curves with respect to a single2. contract administration item.

Monitoring work (s-curve)•

Perormance metrics—(earned values,3. schedule variance, cost variance, schedule perormanceindex, cost perormance index, estimate atcompletion, to complete perormance index(CPI)—with respect to multiple contract

administration items.Controlling schedule (metric1),•

Controlling scope (metric2),•

Controlling cost (metric3),•

Evaluating and processing change orders•

(metric4),Evaluating and processing payment•

requests (metric5),Evaluation and analyzing delays (metric6),•

Evaluating and analyzing claims (metric7),•

Acceptance o completed work (metric8),•

Contract close-out (release o claims)•

(metric9), andPost-project audits (metric10).•

Vaable #3: Pje SuessMeeng shedule and budge gals—1. how

well the project satised the resource con-straints o schedule and budget.

Benef usme—2. how well the projectsatised the unctional requirements andtechnical specications o the organizationthat benet rom the project’s realization.

cmmeal suess—3. how well the projectsatises the desired business results (includesreturn on investment or perorming organiza-tion; or enhancing the prots o the beneting organization).

Pepang he uue—4. the contribution the project made to the perorming organization indeveloping new business opportunities;technologies or increasing core competencies.

rah MhgyTe research methodology used is purposivesampling using a structured survey. In that theresearch is to test actual EVM usage with respect to

project success, respondents must have had directexperience using EVM on one or more projects. Fourspecic groups were targeted because o their unique

ocus on EVM and high likelihood o having experienced practitioners as members: Te ProjectManagement Institute’s College o PerormanceManagement, Deense Acquisition University’searned value management community, the Associa-tion o Project Management’s earned value specialinterest group, and the Association or the Advance-ment o Cost Engineering’s earned value specialinterest group.

Each group was solicited on two separate occasionsand directed to an online survey. Te decision to usethe Web or conducting the survey was made because o the ease o administration, its ability to reach a global

population, relatively low expense, and ast delivery o results (Sekaran, 2003).

inu rpn n samp Ppuan

Fequeny Peen Vald

Peen

cumulave

Peen

valid Hea Constuction 3 2.1 2.1 2.1

Light Constuction (residential and Commecial) 2 1.4 1.4 3.4

Goenment/Public Woks 19 13.1 13.1 16.6

Milita/Deense 67 46.2 46.2 62.8

Inomation Technolog 16 11.0 11.0 73.8

Poessional Seices (Banking, Consulting,Management., etc)

10 6.9 6.9 80.7

Aeospace 15 10.3 10.3 91.0

Sotwae 7 4.8 4.8 95.9

Othe 6 4.1 4.1 100.0

tal 154 100 100


8/6/2019 evs 002



Survey questions were created by presenting armative declarations and asking respondents toselect an answer that best refects their level o agreement with the statement as it applies to theirspecic project. For example, with regard to principlesof EVM , the rst item became, “All work scope was planned or the project to completion.” With regardto project procurement management items, the rstitem became “EVM’s work breakdown structurecontributed to contract scope development.” Withregard to project success, the rst item became, “Te

project satised schedule and budget constraints.”Each o these declarations was ollowed by anidentical seven-point Likert scale anchored on the ar

le end with “strongly disagree,” and on the ar right with “strongly agree.” A neutral point was included inthe center o the scale. Te seven possible selections

were designed to correspond to an interval levelmeasurement scale (one to seven) required orstatistical analysis o continuous data.

saa Anay thnquTe statistical techniques used to test the hypothesesare (1) Pearson’s product moment correlation analysis,(2) bivariate linear regression analyses, and (3) canoni-cal discriminant analysis (CDA). Te techniques o correlation-regression are used to measure the relation-ship between dependent and independent variablesrequired to test H1 and H2. In order to perorm thebivariate correlation-regression, two single continuous

variables are required. Single continuous variables arecreated by totaling all responses o a single respondentto all questions comprising the variable and dividing bythe number o questions. Te resulting composites arearithmetic mean scores or each variable. Composites

not only acilitate the statistical technique, they havethe added value o oering higher measurementreliability (Nunnally, 1967; Biemer et al, 1991) thansingle elements or a given construct. o investigatedierences between groups required or testing H3, H4,and H5, the technique o CDA using a backwardstepwise approach based on multiple discriminating

variables (independent variables) and a dichotomouscriterion variable (xed-price and cost-plus contracttype) is used. Te result is a discriminant model thatmaximizes dierences between groups and quantiesthe relative importance o each discriminant variable

with respect to the criterion variable.

ruDuring the period o July 6, 2005, through March 10,2006, a total o 256 respondents logged on to thesurvey Web site. O the total log-ons, 148 completedthe entire survey yielding an online completion rate o 57.8 percent. Surveys that were not complete were

discarded. O the 148 responses, three demonstratedextreme outlier characteristics (standardized residual value > 3 σ) when pre-testing H1 with linear regres-sion analysis. Tese three responses were removed inorder to ocus on the inerring characteristics o thebasic regression model (Barnett and Lewis, 1994). Teremaining 145 responses were analyzed using SPSS

predictive analytics soware.2 Te majority o respondents worked in the military/

deense industry (46.2 percent). Te government/public works industry was the second largest group represented(13.1 percent) and inormation technology was third (11 percent). Nine industries were represented in all, asshown in TablE 1 on page 26. Tirty percent o respon-dents indicated their project was perormed under


Bvaa can (lna ru)

All esponses (n=145)

celans

evmmp suessmp

e v m o m p

Pesons Coelation 1 .474*

Sig. (2-tailed) .000

N 145 145

s u e

s s o m p Pesons Coelation .474* 1

Sig. (2-tailed) .000

N 145 145

*Coelation is signifcant at the 0.01 leel (two-tailed)

cefensªUnsandadzed cef-

ens

Sandadzed

cefens

M o d e l

B Std. Eo Beta T Sig

1 (Constant) 2.945 .375 7.849 .000

emcomp .458 .071 .474 6.443 .000

ªDependent aiable: sucesscomp

TablE 2.

8/6/2019 evs 002



sPss Bivaiate linea regein output

(Fx-P cna)(contact tpe=1.00)(n=43)

cefensª,b

Unsandadzed

cefens

Sandadzed

cefens

M o d e l

B Std. Eo Beta t Sig

1 (Constant) 2.183 .714 3.057 .004

emcomp .626 .135 .585 4.622 .000

ªDependent aiable: sucesscompbSelecting onl cases o which contactpe=2.00

TablE 3.

Bvaa lna rgn oupu

(c-Pu cna)(contact tpe=2.00)(n=102)

cefensª,b

Unsandadzed

cefens

Sandadzed

cefens

M o d e l

B Std. Eo Beta t Sig

1 (Constant) 3.385 .434 7.791 .000

emcomp .365 .082 .405 4.432 .000

ªDependent aiable: sucesscompbSelecting onl cases o which contactpe=2.00

TablE 4.

X–Y sap Pnp eVM an Pj su cmp daAll esponses (n=145)

FiGurE 3.


P r o j e t S u e s s c o m p o s i t e

Pnples EVM cmpse

8/6/2019 evs 002



xed-price contract arrangements (n=43). Seventy percent indicated using cost-plus contracting (n=102).Fieen years o project management experience was mostrequently reported (15.2 percent). Sixty-ve percent o the respondents had 15 or more years o experience. Terange o experience was rom one year to 40 years. Temajority o respondents reported having two to threecontractors working on their project (19.3 and 14.5

percent, respectively). Sixty-our percent had betweenone and ve contractors working on their project. Anexpected project length o ve years was most oenreported (19.6 percent). Tree years and 10 years reportedequal requency (12.8 percent). Te majority o respon-dents reported an expected project length o between

our months and ve years (69.6 percent). Te range o expected project lengths was our months to 20 years.

resuls eed supp H1. Linear correlation-regression showed a moderately strong relationshipbetween the independent variable principles of EVM and dependent variable project success (r =.474; b=.474,t =6.443, n =145). Te results were signicant setting

p at p < .05 (a=.000). Te independent variableexplains 22.47 percent o the variation in thedependent variable (R²=.2247). As predicted,

principles of EVM are signicant positive predictors o project success, as shown in TablE 2 and FiGurE 3.A scatter-plot o the response data is shown in FiGurE 3 on page 28. Te X-Y plot refects a positive orientation,as well as the strength o the relationship between

principles of EVM and project success.

resuls dd n e supp H2. Contrary to the prediction, a dierence exists between xed-price andcost-plus contracted projects with respect to the

contribution o principles of EVM to project success,measured separately. Te results avor xed-pricecontracted projects. Linear correlation-regressionshowed a moderately strong positive relationshipbetween the independent variable principles of EVM and the dependent variable project success or thecost-plus group (r =.405; b=.405, t =4.432, n=102);and a moderately strong and relatively higher positiverelationship or the xed-price group (r =.585; b=.585,t =4.622, n=43). Te independent variable explains16.40 percent o the variation in the dependent variable in cost-plus contracted projects and 34.22 percent in xed-price contracted projects. Te resultsare statistically signicant setting p at p < .05 ( a=.000),separately, as shown in TablES 3 and 4 on page 28.

spw canna dmnan

Funn (Wk Bakwn suu

Contract Formation im)

Egenvalues

Funns Egenvalue %

Vaane

cumulave

%

cannal

celan

1 .068a 100.0 100.0 .253

ªFist 1 canonical disciminant unctions wee used in the

analsis

Wlks’ Lambda

tes

Funns

Wlks’

Lambda

ch-

squae

d Sg.

1 .936 9.389 2 .009

Sandadzed cannal Dsmnan

Funn cefens

Funn

1

wbs4 -.710

wbs5 1.022

Suue Max

Funn

1

wbs5 -.710

wbs4 -.323

wbs7a .151

wbs3a .136

wbs6a .127

wbs2a .067

wbs1a .041

Pooled within-goups coelations between disciminating

aiables and standadized canonical disciminant unctions.vaiables odeed b absolute size o coelation withinunction.aThis aiable not used in analsis.

TablE 5.


8/6/2019 evs 002



resuls eed supp H3, wh exepns. As predicted, there are no signicant dierences betweenxed-price and cost-plus contracted projects on thecontributions o EVM’s WBS to contract formation items, with the exception o wbs4 (schedule planning)and wbs5 (payment planning) (Wilks’ λ=.936,χ²=9.389, d2, p=.009)(eigenvalue=.068; canonicalcorrelation=.253). Five items were removed leaving asingle canonical root with two discriminant coe-cients (wbs4 and wbs5). No signicant dierencesbetween xed-price and cost-plus contracts wereobserved with respect to the contribution o EVM’sWBS to contract formation items or wbs1, 2, 3, 6, and7. Te analysis results are shown in TablE 5 on page 29.

resuls eed supp H4. As predicted, there isno signicant dierence between xed-price andcost-plus contracted projects on the contribution o

EVM’s S-curve to contract administration items (singleitem included in H5 analysis).

resuls eed supp H5, wh exepns.

Te results show no signicant dierence betweenxed-price and cost-plus contracted projects on thecontributions o EVM’s performance metrics to contract

administration (Wilks’ λ=.943, χ²=8.434, d1, p=.004)(eigenvalue=.061; canonical correlation=.240), with theexception o metric5 (evaluating and processing

payment requests). A total o 10 items were removedduring the analysis leaving a single canonical root withone discriminant coecient (metric5). No signicantdierences between xed-price and cost-plus contracts

were observed with respect to the contribution o EVM’s metrics to contract administration items ors-curve, metric1, 2, 3, 4, 6, 7, 8, 9, and 10. Te results are

shown in TablE 6.

dunTe purpose o this research was three-old: (1) toinvestigate the direct relationship o the principles o earned value management (EVM) to project successon contracted eorts; (2) to investigate any moderat-ing eect contract type may have on the relationship;and (3) to investigate dierences between groupsusing xed-price versus cost-plus contract types withrespect to the contribution o specic EVM mechan-ics to project procurement management items.

Consistent with the results o previous qualitativeresearch eorts (Marrella, 1973; Kim, 2000; Vargas,

sPss summay spw canna

dmnan Funn (s-cuv/eVM

M Contract Administration)

Egenvalues

Funns Egenvalue %

Vaane

cumulave

%

cannal

celan

1 .068a 100.0 100.0 .240

ªFist 1 canonical disciminant unctions wee used in the

analsis

Wlks’ Lambda

tes Funns

Wlks’Lambda

ch-squae

d Sg.

1 .943 8.434 1 ,004

Sandadzed cannal Dsmnan

Funn cefens

Funn

1

metic5 1.00

wbs5 1.022

Suue Max

Funn

1

metic5 -.710

metic7a -.323

metic4a .151

metic8a .136

metic2a .127

metic9a .067

metic10a .041

metic6a

scuea

metic1a

metic3a

Pooled within-goups coelations between disciminatingaiables and standadized canonical disciminant unctions.vaiables odeed b absolute size o coelation withinunction.aThis aiable not used in analsis.

TablE 6.


8/6/2019 evs 002



2003), this quantitative research also suggests thatEVM is an eective project management methodol-ogy. EVM is a signicant positive predictor o projectsuccess on contracted eorts. EVM also has shownitsel to be an even greater positive predictor o projectsuccess when using xed-price versus cost-pluscontracts. Tis nding o the relative strength o EVM avoring xed-price contracts is unique and previously unasserted. While the results herein aresuggestive and do not establish causation, it can bereasoned that to the extent that EVM methodology iseective in driving success on the higher risk projectsnormally associated with cost-plus contracted eorts,EVM would logically drive success on the less risky

eorts associated with xed-price eorts. With regard to EVM’s WBS, the results show thatis positively contributes to the ormation o projectcontracts. Te relative contribution o the WBS to

projects under xed-price versus cost-plus arrange-ments was not signicantly dierent with respect toscope development, risk assessment, and the develop-ment o should-cost estimates. Moreover, no signi-cant dierence was observed or EVM’s WBS contribution to evaluating bids, negotiating withbidders, and the awarding o a air contract. However,an important dierence between xed-price andcost-plus contracted projects was observed withrespect to the WBS’s contribution to schedule

planning in cost-plus contracts and payment planning in xed-price contracts. One possible explanation orthe dierences is risk management. In contractualarrangements whereby all costs incurred are paid, acareully planned schedule provides a measure o eciency and eectiveness during project peror-mance. Following the logic “time is money,” the

relatively larger contribution o the WBS to schedule planning in cost-plus contracts directly serves tomanage costs by planning perormance. In contrac-tual arrangements whereby prices are xed, peror-mance is more at risk. Te relatively larger contribu-tion observed or the WBS contribution to payment

planning in xed-price contracted projects serves toboth incentivize perormance and guarantee workaccomplishment. Te relative strength o EVM’sWBS contribution to schedule planning and payment planning refects the inherent risks associated withcost-plus and xed-price contracts, respectively.

Te EVM S-curve is an important contributor tothe administration o project contracts, but notsignicantly dierent between xed-price and

cost-plus contracted projects. EVM metrics also areimportant contributors to the administration o

project contracts. Te relative contribution o EVMmetrics to projects under xed-price versus cost-pluscontracts was not signicantly dierent with respectto controlling schedule, scope, or costs; evaluating and

processing change orders; analyzing delays and claims;the acceptance o completed work; contract close-out;or post-project audits.

However, an important dierence was observed with respect to the contribution EVM metrics hadon evaluating and processing payment requests inxed-price contracts. One possible explanation or therelatively larger benet in xed-price contracts is

related to the payment structure, which was previ-ously discussed. Careully evaluating and processing payment requests is critical to upholding the integrityo the payment structure designed to incentivize

perormance, guarantee work accomplishment, andmitigation o perormance risk associated withxed-price contracted projects.

Tese results have important implications or project practitioners, corporate and governmental policy makers, as well as uture researchers. EVMshould be considered or all projects—not only or its

positive contribution to project procurementmanagement, but or its contribution to projectsuccess as well, regardless o contract type. Contracttype should not be the sole determining actor in thedecision whether or not to use EVM. Te use o EVMmechanics should be used in all projects. Payment

planning should be emphasized in xed-pricecontracts using EVM in order to mitigate peror-mance risk. Schedule planning should be emphasizedin cost-plus contracts using EVM in order to mitigate

nancial risk.

lman rahA primary limitation o the research is in the use o statistical analysis itsel. While statistical analysistechniques may serve to either explain or predict therelationship among one or more variables, an exactcausal connection can never be absolutely deter-mined. Statistical analysis provides only explanatoryor predictive utility. Moreover, the unexplained

portion o the relationships between variables isinherently outside the scope o this research.

Witness the removal o three outliers—the act thatthree data points did not adhere to the model


8/6/2019 evs 002



suggests that the model itsel can be improved byurther investigating unexplored dimensions o either the principles of EVM , project success, or both.Lastly, a larger and broader sample population isalways preerable. Future studies should attempt toachieve greater levels o project diversity within aneven larger sample population.

dn Fuu rahTe ndings o this quantitative study point out several possible directions or uture research on EVM. As alogical extension o the heretoore unasserted belie that the principles o EVM contribute relatively more

to the success o xed-price contracted projects, anin-depth research case study is warranted in order touncover the signicant characteristics o a xed-pricecontracted project environment. A long duration studyon the use o EVM in a xed-price environment wouldserve to discover new variables and new theories.Another area o uture research only touched on in thisstudy is with respect to extreme cases. Like the threeoutliers that did not t the model, research aimed atdescribing extreme cases o both unsuccessul as wellas successul EVM implementations would make acontribution to EVM knowledge by way o compari-son. Lastly, and more broad in scope, study should bedirected at replacing the oen-used contract-orientedramework surrounding EVM in avor o a morerelevant way o thinking about the methodology, givenEVM’s benecial contribution to projects regardless o contract type. Perhaps a more meaningul way torame EVM would be in the terms and dimensionsrepresentative o project risk management. JCM

WorKS citED

Abba, Wane, 2001, “How Eaned value Got toPime Time: A Shot Look Back and GlanceAhead”, The Measurable News, Sping.

Adamczk, W. F., 1989, “Ev- Not Onl o LagePojects”, AACE Transactions, J.2.1–J.2.5.

Ameican National Standads Institute/Electonic Industies Alliance, 1998, Earned

Value Management Systems: EIA-748-A-1998, Washington D.C.: ANSI/EIA.

Anbai, Fank T., 2003, “Eaned value PojectManagement Method and Extensions”, Project

Management Journal, 34(4), 12–23.

Antik, S., 1998, “Eaned value Management: A 200yea Pespectie”, Long Beach 29th Annual PojectManagement Institute Semina & Smposium.

Antik, S. et al., 2001, “Eaned value Managementin 10 Basic Steps,” The Measurable News, Fall.

Antik, S., 2001, “Wh was Eaned value Managementso Impotant to the Swedish Goenment in theGipen Poject?” The Measurable News, Sping.

Banett, v. and T. Lewis, 1994, Outliers in

Statistical Data, 3d edition, Chicheste,U.K.: John Wile and Sons, 17.

Bieme, Paul P. et al, 1991, Measurement Errors in

Survey , New yok: John Wile and Sons, 143.

Chen, Mak T., 1991, “Appling Eaned valuePocedue to Engineeing Management,”

AACE Transactions, 0.4.1.

Chistenson, Daid S., 1998, “The Costs andBenefts o the Eaned value ManagementPocess,” Acquisition Review Quarterly , Fall.

Ciof, Dennis F., 2002, Managing Project Integration,vienna, vA: Management Concepts, 47.

Fleming, Quentin W. and Koppelman, Joel M., 1996,Earned Value Project Management, NewtownSquae, PA: Poject Management Institute.

Fleming, Quentin W. and Koppelman, Joel M.,2002, “Eaned value Management: Mitigating

risks Associated with Constuction Pojects,”Project Manager , Mach-Apil, 90–95.

Jugde, Kam and Mulle, ral, 2005, “A retospectieLook at Ou Eoling Undestanding o PojectSuccess,” Project Management Journal, 36(4), 19.

Kauman, Paul, Chales Keating and CaolConsidine, 2002, “Using Eaned valueMethods to Substantiate Change-o-ScopeClaims,” Engineering Management, 14(1), 13.

Kim, E., 2000, “A Stud on the EectieImplementation o Eaned valueManagement Methodolog,” Ph.D. diss.

Geoge Washington Uniesit.

Kline, Paul, 2000, A Psychometrics Primer ,London, U.K.: Fee Association Books, 37.

Lein, Paul, 1998, Construction Contract

Claim, Changes and Dispute Resolution,2nd edition, reston, vA: ASCE Pess.

Maella, L. S., 1973, “The Eect o Cost/Schedule Contol Sstems Citeia (C/SCSC)on Contacto Planning and Contol,” Ph.D.diss. Geoge Washington Uniesit.

Mukho, S. and Lisanti, J., 1982, “Application o

Eaned value o Small Poject Contol: Panaceao Bane,” AACE Transactions, A.1.1.–A.1.5.

Nunnall, Jim C., 1967, Psychometric

Theory , New yok: McGaw-Hill, 57.


8/6/2019 evs 002


evs 002

Documents