Available online at www.sciencedirect.

com

ScienceDirect
International Journal of Project Management 32 (2014) 1019 – 1041
www.elsevier.com/locate/ijproman

EDM: Earned Duration Management, a new approach to
schedule performance management and measurement
Homayoun Khamooshi ⁎, Hamed Golafshani 1
Department of Decision Sciences, School of Business, The George Washington University, Washington, D.C. 20052, United States

Received 2 February 2013; received in revised form 3 October 2013; accepted 12 November 2013
Available online 19 December 2013

Abstract

The concept of schedule monitoring and control as one of the most important functions of project and program management has not been fully
exploited. One possible explanation could be the dominance of the Earned Value Management System (EVMS, also known as EVM). EVM was
originally developed as a cost management and control tool which was extended to track the schedule as well. EVM and its derivatives (e.g. Earned
Schedule) use cost as a proxy to measure schedule performance to control the duration of the project. While there is a correlation between schedule,
cost, quality, and scope of a project, using cost to control duration has proven to be misleading. In contrast to Earned Value and Earned Schedule,
the authors have developed the Earned Duration Management (EDM) in which they have decoupled schedule and cost performance measures and
developed a number of indices to measure progress and performance of schedule and cost, as well as the efficacy and efficiency of the plan at any
level of the project. These new indices are easy to understand, have wider applications, and can be used by contractors, clients and the scheduling
offices to assess and measure schedule performance. The newly developed duration performance measures are all schedule-based and can be used
for forecasting the finish date of the project.
© 2013 Elsevier Ltd. APM and IPMA. All rights reserved.

Keywords: Earned Value; EVM; Earned Schedule; Earned Duration; Earned Duration Management; EDM; Schedule management; Monitoring and control;
Efficiency; Performance measurement; Forecasting; Estimation index; Performance and progress index; Project management

1. Introduction tion records, and using this information to keep the project on
track. One of the most widely accepted control systems used by
Project management is about being in control, making sure practitioners in project management field is referred to as the
that the risk of failure is minimized by keeping the project in Earned Value Analysis (EVA) or Earned Value Management
alignment with the developed project plan. It is also about System (EVMS) or in short EVM.
making informed decisions for adjustments to the plan when The evolution of EVM, for the most part, has been focused
required. on cost management, control, and financial analysis for decades
Project monitoring and control, on the other hand, is the (Brandon, 1998; Fleming and Koppelman, 2004; Kim et al.,
process of observing the implementation of the project plan, 2003). This analysis mechanism was developed based on a set
collecting data on implementation, conducting analysis by of 35 criteria, called the Cost/Schedule Control Systems Criteria
comparing the planned values against the actual implementa- (C/SCSC). Later the C/SCSC method was modified to accom-
modate ease of use and pragmatism simply because the original
version was too bureaucratic and complicated. Also, the
⁎ Corresponding author at: Department of Decision Sciences, Funger Hall, Suite organizations were reluctant to spend a lot of time and money
415, The George Washington University, Washington, D.C. 20052, United States. on setup costs for its implementation (GAO, Significant Changes
Tel.: +1 202 994 4862; fax: +1 202 994 2736.
E-mail addresses: hkh@gwu.edu (H. Khamooshi), gol@gwmail.gwu.edu
Underway in DOD's Earned Value Management Process, 1997).
(H. Golafshani). The U.S. government and the associated agencies opted for
1
Federal Aviation Administration. making the use of EVM a requirement of government contracting
0263-7863/$36.00 © 2013 Elsevier Ltd. APM and IPMA. All rights reserved.
http://dx.doi.org/10.1016/j.ijproman.2013.11.002

1020 H. Khamooshi, H. Golafshani / International Journal of Project Management 32 (2014) 1019–1041

and bidding for projects (General Service Administration, irrespective of whether the actual duration of the activity
Department of Defense, and National Aeronautics And Space happened before or after its planned completion date. There are
Administration, 2006). There is abundance of literature on EVM other widely used performance measures that calculate project
and its basic principles and definitions (e.g. see Anbari, 2003; efficiency, or forecast a future state of the project, among other
Fleming and Koppelman, 2010). things. These include, Cost Performance Index:
In short, a clear project scope, a well-defined schedule, and
EV
detailed budget lay the foundation for implementing EVM and CPI ¼ ð3Þ
its derivatives for a project. The budget plan and schedule are to AC
be set for the lowest level of Work Breakdown Structure and Schedule Performance Index:
(WBS) elements to be monitored and controlled. Such a plan
will be considered a baseline for the project and the progress EV
SPI ¼ ð4Þ
will be assessed against that baseline (Cioffi, 2006). PV
EVM defines parameters to enable monitoring and controlling
Despite EVM's numerous benefits and its widespread use,
of projects. EVM defines the Planned Value (PV) as the value
there are some concerns with utilizing EVM as the only tool for
that is planned to be spent for executing the work according to the
monitoring and controlling cost and schedule of a project. Paige
original schedule at any point in time. Previously, PV was
in a lucid paper some 50 years ago proposed the separation of
denoted as BCWS or Budgeted Cost of Work Scheduled. Earned
concerns with schedule from concerns with budgetary matters
Value (EV) is the monetary value of the progress made (work
(Paige, 1963). It was not till a few years ago where researchers
completed) at a certain point in time. Previously, EV was denoted
and practitioners brought up the issues with use of EVM for
as BCWP or Budgeted Cost of Work Performed. Actual Cost
schedule management. The cost management component of
(AC) represents the monetary value of what has been expensed to
EVM is considered to be very effective whereas its schedule
achieve the progress made at a certain point in time. Previously,
aspect has been questioned conceptually in the last few years.
AC was denoted as ACWP or Actual Cost of Work Performed.
Fleming and Koppelman (Fleming and Koppelman, 2004)
Prior to the dominance of EVM, PV and AC were mainly the
recommend that Schedule Performance Index should be used
determinant factors in measuring project progress. However,
just as a warning mechanism and not as a real tool to analyze how
these measures were not enough for showing the entire picture
the project is performing with regard to schedule.
and determining whether the project was getting the worth of
Lipke argued (Lipke, 2003) that “from the time of the
what it was expending. EVM provides various performance
development of the EVM indicators, it has been known that the
measures that are designed to assist with monitoring and
schedule indicators are flawed and exhibit strange behavior
controlling the project depending on the type of information
over the final third of the project when performance is poor”.
needed. For instance, Cost Variance (CV) is utilized to represent
He introduced the concept of Earned Schedule, ES, as an
the difference between how much was actually spent and the
extension to EVM. This research shows that SPI is not an
value earned (EV) through the work performed. In that case, CV is
accurate or reliable measure of schedule performance. The reason
defined as:
given is the fact that any finished activity has an SPI equal to one.
CV ¼ EV −AC ð1Þ This perfect score is irrespective of the actual performance of the
completed activity. To improve the performance of SPI, the
Hence, a negative value points out that the project has spent Earned Schedule method converts the Earned Value at a given
more for the executed activities than what it is worth. To the point in time into its equivalent duration (on the PV or planned
contrary, a positive value indicates that the project has spent value graph) required to achieve that Planned Value. Fig. 1
less to gain the value of the executed work. In general, EVM demonstrates this conversion on a conceptual EVM graph.
and its derivatives use positive values to represent favorable Using this approach, the method provides the Earned
signs of progress and negative values as unfavorable signs. The Schedule, ES(t), for the project. Therefore, Earned Schedule
Schedule Variance (SV) per EVM has traditionally been used as could mathematically be defined as adapted from (Jacob, 2006)
an indicator which represents whether the project is on schedule as:
or not:
Find t such that EV ≥ PV t and EV b PV tþ1ðcalendar unitÞ
SV ¼ EV −PV ð2Þ EV −PV t ð5Þ
ES ðt Þ ¼ t þ  1 ðcalendar unit Þ
PV tþ1ðcalendar unitÞ −PV t
Similar to CV, a negative SV means that the project is behind
the planned schedule, whereas a positive SV represents a project where ES(t) is Earned Schedule at the status date, EV is the
that is ahead of schedule. It is to be noted that, this schedule Earned Value at the status date, PVt is the Planned Value at
performance measure is not very reliable when the cost of time instant t, and the calendar unit represents the unit in which
delayed critical activities is a fraction of the cost of other duration t is measured. The corresponding duration from the
activities. Hence, the delay in completion of these critical beginning of the project till status date is generally defined as
activities does not present much of a variance! Additionally, Actual Time (AT) or elapsed time. We use the term Actual
when the activities finish, SV equals to zero as both EV and PV Duration (AD) in place of Actual Time for consistency and
are the same upon completion of the activity. This outcome is accuracy in this paper.

(2009) also These indices can be used to measure progress and performance conducted a similar research comparing the accuracy and reliability of the project. when ES is used as a measure to predict the remaining indices which are easy to measure. Conceptual EVM graph. we argue that still there are some issues associated with the Index (SPI) and Cost-Schedule Index defined as: use of this method for schedule performance measurement as well. of SPI(t) is the fact that. and the scheduling offices to assess the volume of data used in the forecasting process is large. According to “Schedule is Different” where 2. it measures schedule we have provided a separate table listing all the formulas. We also discuss completion. Review of existing schedule and cost progress and the idea of ES is introduced. They claimed the potential advantages of these measures over Earned Value that SPI has shown to produce reasonably good predictions only and Earned Schedule indices. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1021 Fig. We then introduce a comprehensive set of new Thus. forecast at different stages of the project. The Earned Schedule defines Schedule Performance monetary terms (cost-based) to capture duration-based level of Index denoted by SPI(t) as: progress made in the project. Khamooshi. Additionally. presumably based on past performance. The Earned Schedule performance measure which was the Earned Schedule method uses EV as a proxy to get to introduced by Lipke (W. clients. 1. As an example. we first present a critical evaluation of Earned CSI ¼ SPI  CPI ð7Þ Value and Earned Schedule as a tool for assessing schedule performance. it provides a more reliable with wider applications and implications in managing projects. In this paper. (2009) uses Earned Value which is measured in (AD). whereas efficiency and efficacy of the estimated and planned durations SPI(t) generally performs better. understand and calculate duration or completion of the project. In other words. EV is not necessarily an accurate . ES ðt Þ Vandevoorde and Vanhoucke (2006) and Vandevoorde and SPI ðt Þ ¼ ð6Þ AD Vanhoucke (2007) in their comparative analysis of schedule performance measures came to the conclusion that SPI(t) is a While we agree that ES and consequently SPI(t) is a better better measure of project schedule performance compared to all schedule/duration performance measure of the progress compared the other measures including the classical Schedule Performance to SPI. H. In addition. at different stages of progress toward project compared to the actual duration passed. performance using monetary terms of Earned Value (EV) and Planned Value (PV). The main drawback is given at the end of the paper for ease of reference. conceptual shortcomings and could not be validated as a general The list of acronyms and symbols which are used in this paper method which is applicable to all scenarios. 1 and the overview of Earned Schedule in the previous section. and costs. 2003). the Duration Performance Index of a number of indices which could be used for forecasting the introduced measures the duration/time-based progress of the duration of a project. “the cumulative value of ES(t) performance measures is found by using BCWP to identify in which time increment of BCWS the cost value occurs” (Lipke. The resulting ES(t) is then compared to Actual Duration Lipke et al. it has some Earned Schedule. we develop duration for large projects like those of the United States Department of and cost estimation performance measures which can be used by Defense projects where the duration of the project is long and the contractors. 2003) and later reinforced by corresponding duration. Please refer to Fig. H. Lipke. similar to SPI. Lipke et al. These new measures are then applied to a real life Though SPI(t) is argued to be a preferred option to measure project and the results are compared with Earned Value and duration/time performance compared to SPI.

sections. could be compensated by other elements. longer. therefore. therefore is the critical path for the project. In using such EVM is recognized as the most widely used and accepted models. In spite of this widespread of a project. the project duration performance must EVM's cost-based measures. This activity resembles long lead procurement activities This approach as explained above ignores the dynamics of that happen during the initial stages of a project. H. including activity D which has the highest and duration match. The other path contains the words. and E. 500 ment Baseline. 2009. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 measure of schedule/duration progress made. Schedule of a sample project with high value non-critical activities. or S-curve) is not necessarily enough for assessing Fig. The critical ignoring the build-up of these figures and measures. and reliability of estimates used in planning/scheduling activity with a high PV and short duration could dramatically is frequently within the mix. Only focusing on the duration of the critical path and the associated EV $210. we calculate: scrutinized. It is. suggest that a more rigorous and comprehen- sive assessment of estimation performance is also needed. as failure of projects.1022 H. which use a cost profile to offer a schedule measure will not be It is evident that because a critical activity is not even started accurate. generally considered the best measure of “value earned” AD e 14 monetarily. In such instances. That is. EVM does not take change SPI(t) by overshadowing large delays of inexpensive advantage of the available data to assess the performance and activities on the critical path. . i. Khamooshi. will inherently be deficient in not be at either of these two high performances. the more inaccurate the schedule performance acceptance of EVM. schedule and cost overruns by one element marker on Fig. still a high percentage of projects fail to offered will be. GAO inaccurate results. similar to any method which uses according to the plan. or departments are Using the given information. time and on budget. 2. In this example occurrence of events within the system and has been debated by and for simplicity. This deficiency is illustrated in the accuracy of estimates developed by contractors and their sample project below. SPI(t) and SPI will both show achieve baseline schedule and cost objectives (GAO.e. total duration and cost at each stage. lack of focus on duration planning and shown in the following example: A completed procurement control. “the project status is portrayed as green”. performance measures on EVM is 92% and based on Earned Schedule method is 93%. Consequently. the higher the disparity between time and cost profiles project cost control system today. Nevertheless. In other path activities include A. B. planning offices. Standish Group. Hence. EVM was primarily developed to emphasize In this example. and at times. ES. except activity B that is not started. dealing with duration performance measurement. One path is on the big picture. we have assumed that all activities are on other practitioners and researchers as well (Jacob and Kane. Any section of a project which shows disproportionate 2011). there are two parallel paths. We. Even though we might see correlations between duration and cost profiles throughout the life of a project. SPI(t) could even perform worse Cost Estimating and Assessment Guide. 2 shows the status date. The 2004). these profiles As shown in the example above. This could be fine at EVM output of the Microsoft Project report is provided in the macro level but could create serious issues at the micro level Table 1 below: where the performance of teams. 000 planned cumulative cost curve (known as Performance Measure- EVM → SPI ¼ ¼ ¼ 0:92 PV $227. schedule efficiency based are not generally the same. on the other ES ðt Þ 13 ES → SPI ðt Þ ¼ ¼ 0:93 hand. While there could be many reasons contributing to duration and cost profiles can demonstrate such inaccuracy. as long as the total planned and actual figures of cost remaining activities. than SPI. value.

can be misleading information to project team in assessing the schedule.714 Baseline and where the project stands using EVM parameters (all cost-based. or alternatively. the estimation performance measures defined in Section 3.000 We would like to start by differentiating between two potential groups of duration (schedule) and cost performance measures: 10 162.000 155.000 140.000 169. it could end up showing overall poor in DOD's Earned Value Management Process. as a minimum. we suggest utilizing GAO Cost Estimating and Assessment Guide. The focus on big picture finish the work.000 monitoring processes for benchmarking.857 106.500 186. etc. 2 34. Vanhoucke. Significant Changes Underway score of finishing on time. is valuable and.286 34.429 Few methods have been suggested to measure duration/ schedule performance of a project (Turner.714 87. These measures compare the planned duration or cost to the actual duration or cost spent to the schedule performance of the project.500 138. Also.143 17. The government of the United States of America and many other authorities all around the world have imposed EVM as the We need to make a distinction between these two separate prime tool for managing contracts with regard to meeting performance measures.429 51.571 68. 2010. In most of these measures.857 reliable performance measures. 13 198. schedule and cost dimensions are decoupled and the above ideas are expanded and used to develop some new indices for monitoring schedule and cost separately. We believe that using cost as a proxy for schedule performance could provide the data collected for EVM reporting. These perfor- mance measures are indicators of project being on schedule or on budget.000 150. overemphasis on EVM and based on collected performance data is ignored. These measures consider the accuracy of the original 14 210. Therefore. work package. However.000 estimates and the capability of the project team delivering the work (activity.000 • One set of measures are focused on progress and performance of duration and cost at the project level. H. Measuring Time: Improving Project Performance Using Earned 4 68.000 deliver the project on time and on budget.000 213.000 174.000 184.500 162. SPI (or SPI(t).g.571 68. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1023 Table 1 information about the progress and performance of the project.143 These measures. e. utilized in other ways to provide better and more accurate Therefore. also can offer new beneficial insights about the project which Work EV PV AC were not necessarily available before. 2012). These measures have the potential to be used using S-curve only accounts for some of the monitoring and for benchmarking. reflection on original estimated durations and lessons learned upon completion of activities. New progress and estimation performance measures 3 51. assessing the performance of planning and scheduling 8 138. as it was discussed earlier. duration estimation accuracy.857 106. do not possess the deficiency of EVM and ES mentioned earlier. 1 17. reflecting on benchmarking and documenting lessons learned Also.571 Value Management. estimation process.000 • The second type of measures focus on estimation perfor- mance.000 126.000 baseline against the planned/scheduled progress. Khamooshi. improving controlling aspect of project management. the attention is focused on the Performance Measurement 5 87. etcetera) and CPI. The authors argue that these methods can be improved to provide better and more 6 106.000 198. While a project may have a perfect schedule and cost targets (GAO.143 17. additional measures could be designed to use the available data resulting from existing 7 126. and the ability to 11 174. 9 150. It EVM results. and measuring efficiency and efficacy of estimation methods used.286 3.000 198. to provide more accurate performance measures there .714 87. 1997. documenting lessons learned. H.000 departments.429 51. ahead or behind schedule or budget.500 210.4 in mostly all cases. GAO.000 227. These measures usually compare the actual progress of the project on the performance measurement 12 186. day In the next section.286 34. or larger elements) within the allocated time or cost. the use of this control system as a tool for which provide such insight and fill the current void. monetary measures). example project. if not all. 2009).000 126.

in time. 2005b). i¼1 Considering that APIi measures the schedule progress of an activity. physical progress could be used as a medium (e. is the authorized budget assigned to the scheduled work to Baseline Planned Duration of scheduled activity i: BPDi. Planned Value of scheduled activity i: PVi.e. particular point in time. is the value of work performed expressed as proportion for the following two categories: of the approved duration assigned to that work for activity i (e. EVi. For EDM purposes. days). at any point in It is to be reiterated that PDi. Some may refer to it as the baseline duration for time. it will always have a value of less than or equal to where n is the number of in-progress and completed activities one. We understand that there could be challenges in measuring At the macro level we define and use the following terms: progress. EVM descriptions are adapted activity i. 3. one can use progress of an activity Baseline Planned Duration: BPD. and equivalent to EV of an activity in EVM. measures along with their analogues EVM (cost-based) counter. is the sum of PDi for all the planned then: activities at that time according to baseline plan. at any point in At micro. With that in completion. 2005a. calculated as: Each of these two types of performance measures can be EDi ¼ BPDi  API i ð9Þ defined at various levels of project within telescopic Work Breakdown Structure. Some may from A Guide to the Project Management Body of Knowledge refer to it the baseline cost for activity i.) (Project Management Institute. progress. approved budget assigned to that work for a schedule activity as Planned Duration of scheduled activity i: PDi. we developed the following duration-based performance completion of the activity. only the ADi duration of the activity to the left of the vertical status date line API i ¼ ð8Þ ADi þ EDT C i counts toward TPD. measures the progress of activity. BPDi is independent of the Earned Value of scheduled activity i: EVi. for EDM method is the duration counterpart or equivalent to AC of an activity in EVM. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 is a need to decouple schedule and cost dimensions. Notations used at the micro level Baseline Planned Value of scheduled activity i: BPVi. etc. an alternative method of calculating assigned to the scheduled work to be accomplished for the schedule progress is through estimating the remaining duration entire project irrespective of the status date. is the time in calendar units between the actual start of the The following definitions and notations are used to formulate activity and either that point in time if the activity is in progress the equations for the proposed approach which we call Earned or the actual finish date if the activity is complete. at any for scheduled activity i — EDTCi) by subject matter experts. at any point defined and used in EVM. be accomplished for the activity up to that time as defined in is the authorized duration assigned to the scheduled work to EVM.g. 3.1. at any point in status date. at any point in micro or activity level and macro or project level. This variable for EDM is the duration counterpart or a) Progress and performance measures.1024 H. The schedule progress counterparts or equivalent twins of PVi. 2013). is the authorized duration assigned to the scheduled Actual Cost of scheduled activity i: ACi. Khamooshi.g. EDi. level we define and use: time. Earned Duration of scheduled activity i: EDi. BPVi is independent of the status date. H. This measure Duration Management (EDM). at any point in time. time. . for the project. Calculating APIi using this equation is completely duration X n TPD ¼ PDi ð10Þ based. This variable for EDM is the total costs actually incurred and recorded for accomplishing method is the duration counterpart or equivalent to PV of an work performed during the ADi for the activity as defined by activity in EVM. to complete activity i (i.EDi can normally be b) Estimation performance measures. and ACi for EVM of an activity could be defined in two ways. indepedant of the status date and are the authorized duration one can assume percentage complete to be an accurate measure of and cost of activity i per baseline schedule. However. BPDi and BPVi are progress made on the activity has a linear relationship with time. is the authorized duration similar to EVM. at any point in part measures to more accurately present schedule and cost status time. EVM. Notations used at the macro level lines of code. is the value of work performed expressed in terms of the activity i. activity. we focus on two levels only. It needs to be said that for activities crossing over the status date. However. work to be accomplished for activity i. Estimated Duration To Complete Total Planned Duration: TPD. (Project Management Institute. material quantities. Actual Duration of scheduled activity i: ADi. for activity i: APIi. Activity Progress Index.2. is the authorized budget assigned to the work to be accomplished for In defining these notations. its value approaches one. Value of APIi starts from zero and as the activity nears up to that time. and ADi for EDM are time. In cases where the and are all dependent on the status date. labor hours. and becomes one upon mind. be accomplished for activity i. In this paper. when this is not the case.

Actual Cost for the TAD ¼ ADi ð13Þ project can be defined and calculated as: i¼1 X n where n is the number of in-progress and completed activities AC ¼ AC i ð16Þ up to that time. is the sum of EDi for all the in-progress and EDM results for the example project. Earned Value for the project levels. at any particular Table 2 point in time. Khamooshi. at any point in time. is the period of time from the start of the project till that X n time. i¼1 Planned Value: PV. Table 2 depicts ED(t) calculation for the example project. Or time. Or Day Total Total Total earned planned actual X n duration duration duration TED ¼ EDi ð11Þ i¼1 1 2 2 2 where n is the number of in-progress and completed activities up 2 4 4 4 to that time. H. for the project at any particular point as: in time. 4 presents a conceptual graph which depicts 14 21 28 21 calculation of ED(t). where n is the number of planned activities up to that time. (an EVM terminology) at any point in activities at that time. the activity. Micro level progress and performance measures for i¼1 duration and cost Duration Performance Index. for the project is the total costs actually incurred and X n recorded in accomplishing work performed. Fig. it shows how well the activity is . is the value of work performed expressed in terms of the approved budget assigned to all the in-progress and The measures defined below. for activity i: DPIi. Earned Value: EV. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1025 Total Earned Duration: TED. which mathematically could be 5 10 10 10 expressed as: 6 12 12 12 Find t such that TED ≥ TPDt and TED b TPDtþ1ðcalendar unitÞ 7 14 14 14 TED−TPDt EDðt Þ ¼ t þ  1 ðcalendar unitÞ 8 15 16 15 TPDtþ1ðcalendar unitÞ −TPDt ð12Þ 9 16 18 16 10 17 20 17 where ED(t) is Earned Duration at status date or Actual Duration (AD). 3 6 6 6 Earned Duration: ED(t). and can be defined and calculated as: control of duration/schedule and cost progress and performance and the ability to deliver the project on time and on budget. PV ¼ PV i ð14Þ It is to be noted that TPD. for the project (as used in EVM terminology) at any point in time. 4 8 8 8 is the duration corresponding to Total Earned Duration (TED) on Total Planned Duration S-curve. for the project. evaluation. Progress and Performance Measures for Duration and Cost any point in time. accomplished up to that point. for the project is the sum of ADi for all the in-progress and completed Actual Cost: AC. completed activities at that time. classified into micro and macro completed activities at that time. for the project (an EVM terminology) at 3. Total Actual Duration: TAD. In other words. EV and AC for EVM. It can be defined and calculated Actual Duration: AD. and the calendar unit 12 19 24 19 represents the unit in which time instant t is measured. X n EV ¼ EV i ð15Þ 3.1. At the end of the project ED(t) is the same as BPD (Baseline Planned 13 20 26 20 Duration). TED and TAD for EDM are i¼1 counterparts or equivalent twins of PV.3. H.3. is a measure of schedule performance in delivering that time. for the project. TED is the Total Earned Duration at AD and TPDt is the 11 18 22 18 Total Planned Duration at time instant t. will assist project managers in monitoring. is the authorized budget where n is the number of in-progress and completed activities at assigned to all the activities or scheduled work to be that time. at any where n is the number of in-progress and completed activities at point in time. at any point in time.

date in consideration of the critical path. is a duration-based measure of overall work performed in the activity is in progress and cost-wise has achieved more than terms of Earned Duration. PDi Duration Performance Index. this end value of DPIi for activity against the baseline planned duration as: i represents the accuracy of the original duration estimate. At the same time. Similar to SPI(t). could be formulated as: AD The value of DPI will be less than one. at any point in time. or one which suggests perfor. similar to APIi. In other words. PPI value will be one EDI i ¼ ð18Þ upon completion of the project. as: ADi DPIi can have values of greater than one presenting a better EV i performance than planned. Duration Performance Adapting directly from EVM. And DPI starts and reaches maximum value of one when the work is will be greater than one when the project schedule is overall complete. Adapting directly from EVM. the project will be seen as on schedule. PPI's value starts from zero and as the project nears time. we argue that SPIi is not a direct TED EDI ¼ ð23Þ schedule progress or performance measure. measures the monetary value of the work performance on the critical path and toward completion of done compared to the plan up to that time. due the end date of the project. represents the overall schedule progress performance less. SPIi is actually more similar to Earned Duration Index. at macro level. at any point in time. at any point At any point in time. the value of ED(t) approaches to BPD which results in the value of PPI approaching one. Hence. It can be calculated necessarily provide the insight that either APIi. for activity i. Thus. or lower than one when the activity is up to that point in time. if project is overall performing at the same level as The value of SPIi. for activity i: SPIi. ED(t) is the same as BPD. At the end of the EDi project. At the macro. respectively. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 doing in achieving the target completion date. at any point in It is to be noted that because PPI represents the schedule time.3. rather. Thus. or equal to one. Earned Duration Index. SPIi does being ahead of the schedule. an activity might have achieved more. is a measure performance in delivering the activity compared to the plan. if the project EV i SPI i ¼ ð19Þ schedule is not performing at the same level as planned. we next define: EDi Cost Performance Index. in time. Another difference project. This characteristic might suggest SPIi is dependencies and is irrespective of their interactions to achieve similar to Duration Performance Index for activity i. Khamooshi. or the same amount of work in comparison with the plan toward the completion of the project. Earned Duration Index. However. However. The higher or lower performance will result in the activity being seen as ahead or behind schedule 3. We define: completion. level we define: Upon completion of the activity: DPIi represents the overall Project Progress Index: PPI. DPIi. H. in comparison with the work planned what has been planned. for project: DPI. at any point EDðt Þ DPI ¼ ð22Þ in time.g. for project: EDI at any point in Unlike APIi.1026 H. for activity i: EDIi. SPIi can take on values of greater than one when time. its value is always less than or equal to compared to what was planned to be done up to that point in one. DPIi is an indicator of the activity being on. this measure can have values of how well the project is doing in achieving the target completion higher than one. number of days) progress.2. or behind schedule. DPIi. hence PV i it will be perceived as being behind schedule. Duration not measure the overall duration focused progression of the Performance Index for project provides the measure of progress activity. It. Macro level progress and performance measures for respectively. EDIi. is a measure of duration earned (e. between these two indices is in their range of possible values. at the activity level. It measures ED(t) From a different perspective. lower than one. we Index is defined as: define: Schedule Performance Index. DPI will be equal to one. of overall duration progress of the project. duration and cost ahead. Hence. lower than one indicating worse CPI i ¼ ð20Þ performance compared to plan. or EDIi as: offer. Thus. We define: to the resemblance in the values possible. This measure does not consider activity underperforming. it shows up to that point in time. despite the similarity in starting and performing better than planned which will result in project finishing values of these two performance indices. it may not TPD . for scheduled activity i: CPIi (an DPI i ¼ ð17Þ EVM terminology). AC i mance is according to plan. project. We call the value of DPIi at completion of activity i the Duration EDðt Þ PPI ¼ ð21Þ Estimation Index for that activity (DEIi) and define that index BPD in the next section and explain its applications. is zero before the activity planned.

Construction (EPC) type contracts. H. From Table 2 one can see that TED at the status date or day The progress and performance measures defined above. or the same amount of work in comparison EVM index defined as: with the work planned to be achieved by that time. of day 11. is the classical achieved more. Forecasting and other uses DPI will be calculated as: Similar to EVM. As an example. may not necessarily be an accurate prediction.92 for this entire contract will be behind schedule because the engineering example project. or EV CPI ¼ ð24Þ equal to one respectively. 4 is its EDM counterpart showing TPD. the project was planned to earn this duration by the end accurate information on the status of a project and its activities. Hence.e. if the stages Fig. Khamooshi. H. 3 is the familiar EVM or ES budget. and EDI is EDI. shows that DPI is a measure similar to SPI(t). If the index goes beyond these limits. and TAD for the example project above in to maximize and take advantage of an opportunity or to minimize Table 2 below: the risks. in those cases. Considering that the actual duration passed from the start of the project is 14 days. That is. time and cost savings for the overall contract. 3.93. in Engineering. if on budget. one other potential usage of the progress performance measures is forecasting the future state of the EDðt Þ 11 project. Thus. Comparing The authors' experience shows that spending a little bit more time these values with DPI = 0. suggested by ES or EVM is at 92 or 93 percent level. 14 (i. TPD. ED(t) is roughly 11 days. this measure can have values of greater than one. and TED in an EDM graph. Such schedule compared to SPI(t) and SPI for a project where a critical minimal delay.3. assuming the Earlier. We believe that the performance of the project at each DPI ¼ ∼ ¼ 0:79 AD 14 stage is a function of the nature of the project and the context and content of that stage (characteristics of the S curve) of project as For this sample project we can calculate the Earned Duration supported by Bower (2007).3. AC To demonstrate the difference between DPI and SPI(t) on one hand. 3. either present TED. we have prepared two The value of CPI will be less than one. and LCL). could reduce the cost and duration of installation and commis- DPI and EDI present much more valid information about the sioning activities (part of the construction stage) drastically.79 and EDI = 0. AD = 14) is equal to 21 days. could actually help achieve huge activity is severely behind schedule. lower than one. if the project is over similar conceptual graphs. TPD 28 Procurement. and greater than one when graph whereas Fig. Comparing these two graphs Project managers can use control charts to monitor DPI. then generally To demonstrate the use of these indices we calculate and there is a need for investigation and possibly intervention. CPI. and CPI over time with upper and lower control limits (UCL similar to SPI. According to the TPD separate duration and cost performance measures to deliver more column. we found that SPI(t) = 0.75 respectively and and resources on the engineering stage when deemed necessary knowing the status of the project. Fig. equal to one. . Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1027 At any point in time duration-wise a project might have Cost Performance Index: for project. Thus. and SPI = 0. Conceptual EVM graph. under budget. We also believe that using any Index as: performance measure developed at one stage of a project as a TED 21 proxy to forecast performance of the project many periods away EDI ¼ ¼ ¼ 0:75 in the future is questionable. schedule or duration performance stage is delayed. and EDI and SPI on the other. TAD. less. it can be seen that in both cases.

Duration Variance for the project to bring it back on track. Forecasting words. To measure the deviation from the plan on the critical path. (EDAC). Considering that we used cost parameters similar to EVM. We define: CV i ¼ EV i −AC i ð31Þ Estimated Duration at Completion. Moslemi Naeni et al. Therefore. And at the project level Total Duration Variance for Project. and AD assessing the performance of planning and scheduling depart- EDAC ¼ ð26Þ PPI ments. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 Fig. Conceptual EDM graph. Thus. (TDV). the performance of the project in the future is not necessarily constant and the same as the past but DV i ¼ EDi −PDi ð28Þ most probably is flexible and dynamic and responsive to past performance (Cooper. it is a measure of total deviation of all the activities project completion cost using CPI which uses EV has shown to be planned up to any point in time. one TDV ¼ TED−TPD ð29Þ could reasonably use any of the many simple forecasting models (e. a relatively good choice (Vandevoorde and Vanhoucke. we can define a duration plan. DPI PPI However. and or more TDV shows the sum total of all the deviations from the plan advance and complex approaches (fuzzy time series etcetera) to which is a measure of the accuracy of the planning. similar to EVM. 2006). the case. we we define Duration Variance for a project. 4. there is no rationale for assuming past performance is a for project. DV. as: suggest utilizing DPI which is based on EDM data and does not have the computational difficulties and potential errors of cost/ DV ¼ EDðt Þ−AD ð30Þ value-based models (EVM. exponential smoothing). lessons learned. (DVi). as: And at the project level: EDAC ¼ BPD ð25Þ CV ¼ EV −AC ð32Þ DPI 3. and consequently resulting in a activity i. which normally is Or we can calculate the Estimated Duration To Complete. Despite the above. Khamooshi. 2006. Hence. one can formulate a forecasting equation there will be no changes to the cost variance performance for EDM similar to the classical forecasting approaches for measures.4. Cioffi.. Duration and cost estimation performance measures Alternatively. one aspect of control is to use the collected information for benchmarking. is defined as: In spite of the difficulties and issues mentioned earlier. EDTC as: good predictor of the future.1028 H. of the project are different and heterogeneous. ES) as discussed earlier. On the other hand. H. In other predict the value of the index for the next few periods.g. we use the data collected as part of monitoring . to forecast the duration of the project. 2011). At the activity level. as a minimum the project manager normally tries to influence variances for EDM. can be calculated as: change in performance. moving average. In all these forecasts we assume the current performance BPD AD  ð1−PPI Þ EDTC ¼ −AD ¼ ð27Þ continues and remains the same for the remainder of the project. we know that if the performance is not according to the Additionally. one can use the following formula to estimate the duration at completion: As discussed earlier. at the activity level: EVM (considering a constant performance for the remainder of the project). 1994.

For these indices. One can use other methods such In addition to providing more accurate information about the as weighted average (based on duration. is the cost counterpart Duration Estimation Index of scheduled activity i: DEIi. could provide very such as engineering. Possible values of defined estimation measures Measures The value of DEIi. there weighted average version of Duration Estimation Index for will be no additional costs for data collection to produce EDM project could be formulated as: reports. Macro Level Duration and Cost Estimation Performance 3. as a effort. The project management teams for generating EVM reports. DEI ¼ i¼1 ð35Þ Using a project management software with built-in func- m tionality to input and process performance management data. 3. measure for the activities within similar functions or departments i. separating homogenous activities. all the aforementioned measures can be developed and reported We use the average function at the project level. ðwi  DEI i Þ and macro levels. etcetera. these indices are usually less than one projects we expect higher reliability of duration and cost due to notional rules such as Parkinson's Law which states that estimates and less risk of delays or over-runs.g. Engineering.4. One can use DEIi of all the completed activities (m) at any stage of the control charts to monitor DEI and C EI over time with upper and project. another benefit .4. At the completion of the project. Khamooshi. H. department. performance variation is expected (Khamooshi and Cioffi. CEI could be higher or lower Now that estimation performance measures at the micro than one. 2013). development. level we define: Cost Estimation Index for project: CEI. the previously executed projects. etc. or part of project. cost. DEI. This averaging produces better performance section.1.3. we define: the element types. total number of activities in the Project. with some by filtering the relevant activities or work packages for each reservations. As this required data is currently available to Index at higher levels depending on the activity types. DEIi compares the Xm CEI i duration originally planned (authorized) for the activity versus CEI ¼ i¼1 ð37Þ what was the actual duration upon completion of the activity: m BPDi where m is the number of activities which are finished. DEIi and CEIi should be used to report duration and cost Knowing that the planned durations are not always one estimation accuracy or performance for activities only after hundred percent reliable. cost. In which case the following formula can be used: measure to gain a better understanding of the accuracy of the Xm original estimated cost of the activity: ðy  CEI i Þ i¼1 i CEI ¼ ð38Þ m BPV i CEI i ¼ ð34Þ where yi is the designated weight for the activity i and m is the AC i number of activities which are finished. This is in contrast “work expands so as to fill the time available for its completion”. For deterministic Based on experience. one can use weighted averages to get a more ADi accurate Cost Estimation Index at higher levels depending on Similarly. We define: development of plan.e. a value of more than one is a sign of level have been established. DEI ¼ i¼1 ð36Þ m 3. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1029 and controlling processes of project management to develop most appropriate) to get a more accurate Duration Estimation these measures. This index can be valuable information for assessing the schedule and estimation used for benchmarking or any kind of comparative analysis with performance of any particular section or department. as duration status of the project and its activities. we discuss macro or project level better performance compared to what has been agreed on during estimation performance measures. To the contrary. This process could provide us with the opportunity for taking Xm advantage of any opportunity for improving the estimation DEI i process and minimizing the risks associated with estimates. and Develop- These indices could occasionally be more than or equal to one as ment projects) in which the estimates are less reliable and well. as the average indicative of worse than expected performance. breakdown or based on duration. CEIi.2. original estimated duration of the activity. activity. for DEI and CEI for different types of projects. design. Xm Once again. m is equal to M or lower control limits (UCL and LCL) for evaluation of estimation. H. This filtering operation. Micro level duration and cost estimation performance where wi is the designated weight for the activity i and m is the measures number of activities which are finished. these measures can be defined at both micro. to probabilistic projects (Research. as a of DEI and measures average cost estimation performance for the measure to gain a better understanding of the accuracy of the project based on all completed activities (m). where m is the number of activities which are finished. At the micro. effort. Cost Estimation Index of scheduled activity i: CEIi. etc. we expect some variation in the values they are completed. values less than one are Duration Estimation Index for project: DEI. size. DEI i ¼ ð33Þ Similar to DEI. e.4.

A in Appendix 1 presents the same data in a tabular tion) progress of the project compared with the baseline planned format). 5. boat handling and maintenance facility. Table 3.000 $600. and improve the estimation for future projects. We have proportionally According to the TPD column. We as: show this information at the end of week 37 as well. garage boat management facility to that of a modernly equipped TED at the status date of week 18 is equal to 180 days. $1.000 $1.1030 H. ED(t) is roughly 16 weeks.000 $0 EV PV AC Fig. Thus. the project was planned to earn reduced the costs to maintain the confidentiality of project data this duration during week 16. TPD. document lessons learned by comparing estimates week 11. and EDM data when 18 start of the project is 18 weeks. the project was planned to earn this value roughly during project.600. we can produce the EDM graph as shown in Fig. This The same data can be shown in a tabular format presenting project upgrades the capacity and capability of an old storage TED.000. According to the PV performance measures provide us better tools to control the column. These end of week 18 is equal to $467. ES. DPI ¼ ¼ ¼ 0:89 The calculations below are fairly straightforward and are AD 18 TED 180 accommodated by using a project management software package EDI ¼ ¼ ¼ 0:91 TPD 198 (Microsoft Project in this case). Considering that the actual number of weeks passed from the We will first demonstrate EVM. to a real life United States Coast Guard (USCG) project. 890 The status of the project according to EVM at the end of week Using Table 1.789. Khamooshi.000 $200. 5 for ES. EDM empowers the project manager to have a better handle on project duration control and higher chance of ES ðt Þ 11 ES → SPI ðt Þ ¼ ¼ 0:61 delivering a project successfully. 6. Later. This figure depicts a noticeably different profile for EDM In this section we apply the suggested performance measures compared with the one presented in Fig. EVM graph at the end of week 18. 7.000 $1. we EDðt Þ 16 will show the performance measures at project completion. Considering against actuals. Fig. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 of EDM is its estimation performance measures. 789 PPI ¼ ¼ ¼ 0:35 EVM → SPI ¼ ¼ ¼ 0:68 BPD 46 PV $685.400.000 $400. 5 shows EVM graph at the end of week 18 for this project The Project Progress Index measures the time-based (dura- (Table 1.A in Appendix 2).000 $1. EDM applied to the United States Coast Guard Using USCG project data available for generating EVM (USCG) project graphs. H. while keeping the baseline schedule data intact.A in Appendix 3 presents the Project EVM report makes it easier to calculate ES(t). DPI and EDI will be calculated out of 46 weeks of the planned project duration has passed. duration and is calculated as follows: EVM calculates SPI as: EDðt Þ 16 EV $467. AD e 18 4. ES(t) is approximately 11 weeks. and TAD (Table 2.A containing the data extract of the Microsoft 37 is presented in Fig.000 $800.000 $1.200. SPI(t) will be calculated as follows: Consequently. . EV at the same data in a tabular format. Thus.800. that AD is 18 weeks.

EVM graph at the end of week 37. EVM calculates SPI as: with the one presented in Fig. H. and TAD Table 4. 280 TED at the status date of week 37 is equal to 282 days. 8.800. the project was planned to earn Using Table 3. Thus. TPD. 6. The same data is shown in a tabular format presenting TED. This TED 282 EDI ¼ ¼ ¼ 0:69 figure depicts a noticeably different profile for EDM compared TPD 408 $1.000 $600. 915 Appendix 4.915.600.000 $200. DPI and EDI will be calculated column. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1031 600 500 400 300 200 100 0 TED TPD TAD Fig. EDM graph at the end of week 18. the total duration (y axis is in days) whereas the project duration/progress is in weeks. Project EVM report makes it easier to calculate ES(t). EV at the Considering that the actual number of weeks passed from the end of week 37 is equal to $870.000 $1.000 $1.200.000 $400. Thus. 7. According to the PV start of the project is 37 weeks.A.000 $1. EVM : SPI ¼ ¼ ¼ 0:63 PV $1. SPI(t) will be calculated as follows: EDðt Þ 30 ES ðt Þ 29 DPI ¼ ¼ ¼ 0:81 ES : SPI ðt Þ ¼ ¼ 0:78 AD 37 AD e 37 EDM graph at the end of week 37 is shown in Fig. ED(t) is roughly 30 weeks. the project was planned to earn this value roughly during as: week 29. 390.000.400. Khamooshi. EV $870. 7 for ES.000 $0 EV PV AC Fig. ES(t) is approximately 29 weeks. Considering that AD is 37 weeks. According to the TPD column.000 $800.A containing the data extract of the Microsoft this duration during week 30.000 $1. H. .

Khamooshi. H. As explained in the previous sections. DEI EDðt Þ BPD 46 is given by: PPI ¼ ¼ ¼ ¼1 BPD BPD 46 X1 DEI i DEI ¼ m ¼ 0:82 Each of these measures depicts a different aspect of the m project and its activities. for project as: USCG project shows that EDM provides a more realistic schedule assessment. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 500 450 400 350 300 250 TED 200 TPD TAD 150 100 50 0 Week 3 Week 1 Week 5 Week 7 Week 9 Week 11 Week 13 Week 15 Week 17 Week 19 Week 21 Week 23 Week 25 Week 27 Week 29 Week 31 Week 33 Week 35 Week 37 Week 39 Week 41 Week 43 Week 45 Fig. EDI at the end of the project will results for micro (activity) level measures are presented in always be equal to one. the performance data for the future projects.A. 8. 57 as opposed to the scheduled week 46. Appendix 5. could be used in place of DPI as a good schedule measure. . EDM graph at the end of week 37. and combined together. estimation methods and improving the accuracy of planning similar to the results at the end of week 18. we expect the Project Progress Index at the same. this is mainly due to cost and schedule profiles getting closer to each This means that the estimating team had an overall accuracy other at this stage of the project. and the calculated DEIi and CEIi we can calculate Duration Estimation Index for the USCG project. end of the project to be 100% or equal to one as follows: Using this data. In case of USCG project SPI(t) is a much EDðt Þ BPD 46 better measure than SPI but as it was discussed above. Had These findings could be used for review and assessment of the cost and schedule profiles not been close at this stage. the total duration (y axis is in days) whereas the project duration/progress is in weeks. of course. expected from a comprehensive monitoring and control system. Using the USCG project data upon completion. the calculated As mentioned earlier. the duration performance and estimation indices will be the And. The Project Progress Index is calculated as follows: And CEI is calculated as follows: EDðt Þ 30 X1 PPI ¼ ¼ ¼ 0:65 CEI i BPD 46 CEI ¼ m ¼ 0:89 This time around.1032 H. As explained earlier EDI ¼ ¼ ¼1 TPD 484:5 and can be seen in Table 5.A. we calculate The result from both the earlier example project and the the Duration Performance Index. the activity progress index for each activity is TED 484:5 at 100% showing the activity is completed. hence. SPI(t) is much closer to the performance m suggested by DPI. We can calculate this index as: Table 5. provide a This means that the overall duration estimation accuracy was comprehensive understanding of the status of the project as about 82%. or that SPI(t) project. upon completion of the activities. the DPI ¼ ¼ ¼ ¼ 0:8 AD AD 57 duration-based index is a more rational alternative and natural choice when dealing with duration or time. suggested by SPI(t) would have been different and inaccurate Considering that the USCG project was completed in week compared to the value suggested by DPI. And is not a sign of SPI(t) of about 89% in estimating the cost for the activities of this being a good schedule performance measure. As expected.

The PV Planned Value newly introduced Project Progress Index (PPI). Schedule Performance Micro/macro (6) Index CSI = SPI × CPI Cost-Schedule Index Micro/macro (7) API i ¼ ADi þEDT ADi Ci Activity Progress Index. LCL Upper and Lower Control Limits ACWP Actual Cost of Work Performed. Total Actual Duration graph is a very PD Planned Duration powerful tool focused on duration (called EDM graph) which PPI Project Progress Index could play the same role as EVM graph does for cost. TED Total Earned Duration TPD Total Planned Duration. The new estimation indices can be used for EV Earned Value benchmarking. and time DV Duration Variance similar to EVM. Construction schedule related performance of a project provides a number ES Earned Schedule (method) of advantages. Conclusion and recommendations AD Actual Duration APIi Activity Progress Index (for activity i) There are hundreds of projects failing to be delivered as BCWP Budgeted Cost of Work Performed scheduled each year. Schedule Variance Micro/macro (2) CPI ¼ EVAC Earned Value. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1033 5. and re-evaluation of EVA Earned Value Analysis the models and processes used for each part. department EVM Earned Value Management (method) or the project as a whole in the future. Table of formulas. The Total Planned Duration. Cost Performance Index Micro/macro (3) SPI ¼ EV PV Earned Value. we CSI Cost-Schedule Index provided a new approach and recommendations to decouple CV Cost Variance the schedule and cost dimensions in EVM by adding Earned DEI Duration Estimation Index Duration while maintaining the unique interaction of the three DPI Duration Performance Index major project management elements of scope. Formula Description Measure Level Equation number CV = EV − AC Earned Value. Schedule Performance Index control and management of project schedules. Realizing that there is CEI Cost Estimation Index normally some correlation between time and cost. Procurement. section. and using cost as a proxy for measuring BPD Baseline Planned Duration duration-related performance of the project in EVM have been C/SCSC Cost/Schedule Control Systems Criteria identified in this paper as an issue. Over-emphasis on cost. Focusing on duration from estimation to delivery EDM Earned Duration Management (method) not only helps to manage the schedule more effectively. cost. EVMS Earned Value Management System Total Earned Duration. A new model to emphasize the duration/schedule EDAC Estimated Duration at Completion aspect of the project called Earned Duration Management or EDM EDI Earned Duration Index is introduced. Khamooshi. for activity i Micro (8) EDi = BPDi × APIi Earned Duration of scheduled activity i Micro (9) n Total Planned Duration Macro (10) TPD ¼ ∑ PDi i¼1 (continued on next page) . and Earned Duration Index (EDI) SPI Schedule Performance Index are better duration-based performance measures to be used in SPI(t) Earned Schedule. it could EDTC Estimated Duration To Complete (for project) also lead to cost saving as well. List of acronyms and symbols USCG United States Coast Guard AC Actual Cost UCL. lack of focus on BCWS Budgeted Cost of Work Scheduled accuracy of estimation. These measures TAD Total Actual Duration overcome the deficiencies of cost-based schedule performance TDV Total Duration Variance indices. H. documenting lessons learned. H. to improve CPI Cost Performance Index duration-focused performance measurement accuracy. Schedule Performance Index Micro/macro (4) Find t such that EV ≥PV t and EV b PV tþ1ðcalendar unitÞ Earned Schedule Micro/macro (5) EV −PV t ES ðtÞ ¼ t þ PV tþ1ðcalendar unitÞ −PV t ñ1 ðcalendar unit Þ SPI ðt Þ ¼ ESADðtÞ Earned Schedule. Cost Variance Micro/macro (1) SV = EV − PV Earned Value. Duration SV Schedule Variance Performance Index (DPI). WBS Work Breakdown Structure. Using duration to assess EPC Engineering.

for project (an Progress and performance Macro (24) EVM terminology) measures EDAC ¼ DPI BPD p Estimated Duration at Completion Progress and performance Macro (25) measures EDAC ¼ PPI AD Estimated Duration at Completion Progress and performance Macro (26) measures EDTC ¼ DPI BPD p −AD ¼ ADñðPPI 1−PPI Þ Estimated Duration To Complete. for project Progress and performance Macro (23) measures CPI ¼ EV AC Cost Performance Index. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 (continued) Formula Description Measure Level Equation number n Total Earned Duration Macro (11) TED ¼ ∑ EDi i¼1 Find t such that TED≥TPDt and TED b TPDtþ1ðcalendar unitÞ Earned Duration Macro (12) EDðtÞ ¼ t þ TPDtþ1TED−TPD t ðcalendar unitÞ −TPDt  1 ðcalendar unitÞ n Total Actual Duration Macro (13) TAD ¼ ∑ ADi i¼1 n Planned Value (an EVM terminology) Macro (14) PV ¼ ∑ PV i i¼1 n Earned Value (an EVM terminology) Macro (15) EV ¼ ∑ EV i i¼1 n Actual Cost (an EVM terminology) Macro (16) AC ¼ ∑ AC i i¼1 DPI i ¼ EDi ADi Duration Performance Index. for activity i Progress and performance Micro (18) measures SPI i ¼ EV i PV i Schedule Performance Index. for activity Progress and performance Micro (19) i Cost (an EVM terminology) measures CPI i ¼ EV i AC i Cost Performance Index. for activity i Progress and performance Micro (20) Cost (an EVM terminology) measures ðt Þ PPI ¼ ED BPD Project Progress Index Progress and performance Macro (21) measures ðt Þ DPI ¼ ED AD Duration Performance Index. for project Progress and performance Macro (22) measures EDI ¼ TED TPD Earned Duration Index. H. Khamooshi. for Progress and performance Macro (27) project measures DVi = EDi − PDi Duration Variance for activity i Progress and performance Micro (28) measures TDV = TED − TPD Total Duration Variance for Project Progress and performance Macro (29) measures DV = ED(t) − AD Duration Variance for Project Progress and performance Macro (30) measures CVi = EVi − PVi Cost Variance for activity i Progress and performance Micro (31) measures CV = EV − PV Cost Variance for project Progress and performance Macro (32) measures DEI i ¼ BPD ADi i Duration Estimation Index of activity i Duration and cost Micro (33) estimation performance measures CEI i ¼ BPV AC i i Cost Estimation Index of activity i Duration and cost Micro (34) estimation performance measures ∑m i¼1 DEI i DEI ¼ m Duration Estimation Index for project Duration and cost Macro (35) estimation performance measures ∑m i¼1 ðwi ñDEI i Þ DEI ¼ m Duration Estimation Index for project Duration and cost Macro (36) (weighted) estimation performance measures ∑m i¼1 CEI i CEI ¼ m Cost Estimation Index for project Duration and cost Macro (37) estimation performance measures ∑m i¼1 ðyi ñCEI i Þ CEI ¼ m Cost Estimation Index for project Duration and cost Macro (38) (weighted) estimation performance measures . for activity i Progress and performance Micro (17) measures EDI i ¼ EDi PDi Earned Duration Index.1034 H.

Table 1. .A EVM results. at the end of week 18. Khamooshi. H. USCG project. H. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1035 Appendix 1 Table 1.A below presents EVM/ES results for the USCG project at the end of week 18.

Khamooshi. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 Appendix 2 Table 2. .A below presents EDM results for the USCG project at the end of week 18. Table 2. at the end of week 18.A EDM results. H. USCG project.1036 H.

at the end of week 37. Table 3.A below presents EVM/ES results for the USCG project at the end of week 37. USCG project. . H. Khamooshi. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1037 Appendix 3 Table 3.A EVM results. H.

Golafshani / International Journal of Project Management 32 (2014) 1019–1041 Appendix 4 Table 4. Khamooshi. .1038 H. H. Table 4.A below presents EDM results for the USCG project at the end of week 37. at the end of week 37. USCG project.A EDM results.

00 $29.870.000.480.00 100% 100% 100% 92% 79 0.00 100% 100% 100% 77% 39 3 days 3 days $13.00 100% 94% 94% 100% 24 3 days 4 days $5639.850.440.00 100% 100% 100% 85% 67 3 days 3 days $12.000.000.400.00 100% 67% 67% 82% 47 3 days 4 days $16.000.00 $2.00 $7.A below presents activity actual duration and cost upon completion of the USCG project along with the performance measures for each activity.00 $3.262.00 $55.00 100% 100% 100% 100% 69 1 day 1.000.0 100% 75% 75% 76% (continued on next page) .000.5 days $368.00 100% 80% 80% 100% 23 15 days 16 days $70.824.20 100% 75% 75% 100% 25 10 days 13 days $8672.00 $7.00 100% 67% 67% 100% 22 12 days 15 days $35.00 100% 100% 100% 100% 63 2 days 2 days $16.00 100% 100% 100% 91% 9 4 days 6 days $43.00 $10.000.00 $30.000.00 100% 75% 75% 86% 33 2 days 4 days $1800.00 100% 43% 43% 45% 11 5 days 8 days $30. Table 5.00 $5.112.500.00 $20.00 $14.000.00 100% 100% 100% 85% 30 5 days 9 days $30.928.00 100% 50% 50% 80% 7 3 days 3 days $11.000.000.00 100% 91% 91% 86% 43 7 days 9 days $14.20 100% 86% 86% 100% 18 2 days 3 days $69.00 $2.00 100% 100% 100% 94% 80 20 days 25 days $50.00 $368.391.000.980.639.00 $4.000.00 $6.00 100% 100% 100% 91% 50 4 days 4 days $6128.716.00 100% 100% 100% 94% 75 3 days 3 days $32.00 $6.00 $9.00 100% 63% 63% 100% 13 3 days 4 days $4839.000.000.000.954.000.300.00 100% 83% 83% 92% 83 4 days 5 days $16.00 100% 78% 78% 77% 46 2 days 3 days $9784. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1039 Appendix 5 Table 5.00 100% 67% 67% 63% 10 3 days 7 days $4548.00 $40.000.432.5 days $512.00 100% 100% 100% 93% 62 2 days 2 days $4992.200.00 100% 67% 67% 100% 87 3 days 4 days $3186. H.888.196.000.920.00 $2. Khamooshi.00 100% 63% 63% 100% 37 3 days 5 days $5832.00 $1.000.00 100% 100% 100% 83% 65 1 day 1 day $1888.00 $23.00 $7.00 $18.20 $20.000.304.00 $70.672.00 $1.081.000.600.610.00 $7.A Activity actual duration and cost along with performance measures for the USCG project.904.00 100% 80% 80% 82% 84 20 days 30 days $26.00 100% 75% 75% 69% 14 3 days 5 days $29.00 100% 100% 100% 80% 74 2 days 2 days $13.000.00 100% 100% 100% 87% 41 2 days 2 days $1728.00 $8.856.500.00 100% 100% 100% 86% 71 3 days 3 days $17.000.00 100% 60% 60% 93% 38 4 days 4 days $18.20 $7.00 100% 50% 50% 68% 20 1 day 1.20 $6.112.5 days $1872.00 100% 40% 40% 62% 56 2 days 2 days $3776.00 $12.00 $16.339.000.00 100% 100% 100% 100% 54 2 days 5 days $4312.00 $2.00 $5.00 100% 100% 100% 100% 66 2 days 2 days $2136.00 $36.610.600.00 $90.000.796.500.00 $20.00 $1.00 $19.00 100% 67% 67% 86% 53 3 days 3 days $12.20 $5. H.00 $3.00 100% 60% 60% 100% 15 3 days 3.00 $26.00 100% 100% 100% 77% 35 5 days 8 days $23.00 100% 75% 75% 106% 49 2 days 2 days $6832.00 100% 67% 67% 51% 70 2 days 2 days $2736.00 $25.00 100% 100% 100% 99% 73 2 days 2 days $4016.000.00 $10.00 100% 80% 80% 92% 81 5 days 6 days $5502.00 $13.800.00 $35.610.00 100% 100% 100% 102% 52 4 days 6 days $7724.000.824.00 100% 77% 77% 100% 29 4 days 4 days $8544.000.000.00 100% 100% 100% 54% 57 5 days 6 days $6860.00 $6.00 $35.560.00 100% 63% 63% 84% 61 1 day 1 day $3248.00 100% 83% 83% 98% 58 5 days 8 days $12.00 $70.5 days $6391. ID Baseline Planned Duration (BPDi) Actual Duration (ADi) Baseline Cost Actual Cost APIi DPIi DEIi CEIi 5 5 days 7 days $39.00 $7.00 $24.546.5 days 0.20 $15.00 $15.200.00 100% 100% 100% 86% 42 10 days 11 days $21.000.00 100% 67% 67% 77% 19 1 day 2 days $60.00 100% 50% 50% 97% 34 2 days 2 days $1936.00 $12.196.776.500.000.00 100% 71% 71% 98% 6 2 days 4 days $15.00 $13. upon completion of the project.00 $90.980.400.000.536.000.00 100% 56% 56% 85% 31 6 days 8 days $6504.600.

000.00 $6..000.000. GAO cost estimating and assessment guide (Washington).500.00 100% 86% 86% 87% 111 5 days 5 days $25.274.00 100% 100% 100% 102% 122 9 days 10 days $16. Lipke.000.00 100% 40% 40% 91% 141 2 days 3 days $3504. Proj. Manag.00 $3. Res.000.40 $4. but It's not RMIT University. A model for projects? The Measurable News.00 100% 100% 100% 100% 90 8 days 10 days $8224.00 100% 80% 80% 98% 109 15 days 15 days $81. The $2000 hour: how managers influence project estimates: fusion of probabilistic forecasts and deterministic scheduling.00 $10. 1–5. William G..00 100% 100% 100% 82% 110 12 days 14 days $52. Newtown Square.985.400.00 100% 100% 100% 93% 144 12 days 14 days $51. Federal Acquisition Regula- Bower.00 $5.00 $19.00 100% 75% 75% 96% 117 2 days 3 days $3240.000..00 100% 80% 80% 97% 92 10 days 12 days $6068.020. New directions in project performance and progress tion. pp.00 $16. U. General Services Administration. J. Significant changes underway in DOD's earned value management process. Cioffi. Property and Project Management. 488–497. Coll.00 100% 80% 80% 98% 99 2 days 2 days $3040. 5 (March).00 100% 67% 67% 54% 116 3 days 4 days $2894.00 $17. Proj..000. Washington. Denis. Fleming. Manag.488. GAO. Daniel M.40 $56. D. J.00 100% 100% 100% 61% 149 10 days 11 days $49. 2003..264. 2009. Completing projects according to plans: an earned value value metrics…revisited.5 days $11.000. Earned value project management method and General Service Administration.00 100% 91% 91% 90% References GAO.800.00 100% 83% 83% 91% 139 4 days 5 days $14. Quentin W.00 100% 80% 80% 96% 140 2 days 5 days $6384. J.00 100% 75% 75% 100% 127 10 days 12 days $13.000. Fleming.00 100% 100% 100% 98% 125 20 days 20 days $17.020. 2010.60 $20.968. Constr.00 100% 100% 100% 100% 107 4 days 5 days $19. School of Construction..000.00 100% 100% 100% 99% 112 5 days 6 days $24.432.000.00 100% 92% 92% 80% 97 4 days 5 days $20. Manage.00 $24. 1997.00 $50. Duffey.60 $18. 290–295. EunHong.000.40 $3.000. Forecasting schedule completion using earned Cioffi. PA. Soc.00 100% 75% 75% 94% 136 2 days 3 days $1641.696. third edition.000.00 $26. Proj.. effective implementation of Earned Value Management. Implementing earned value easily and Measurable News.450.00 $15.00 100% 67% 67% 93% 119 4 days 5 days $6822. 139 (5).00 100% 80% 80% 97% 120 2 days 2 days $1024. M.S. Michael R.000. J. H.600.521.200.. Koppelman.60 $1.00 100% 67% 67% 86% 137 5 days 6 days $8184.440. 12–23.00 $22.80 $21. Anbari.204. D.00 $1. Khamooshi.600. J. Schedule is different.00 100% 100% 100% 93% 126 3 days 4 days $2700.800. necessarily the premier indicator for assessing schedule performance.00 $7.564. management.000.60 $3. Perform. J. evaluation.00 $7. Inst. 2003. 2003. .000.604.000.000. 2007. and National extensions. The Brandon Jr. Manag. Jacob.00 $3.900.F. 1994. Koppelman.000.000.00 $2. Department of Defense. performance through the rework cycle. Walt.00 100% 83% 83% 100% 115 1 day 1.515.00 $8. 2006. 2013.700.000. pp. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 Table 5.00 $92.. Aeronautics and Space Administration. GAO.500.000. Manage. 1998.00 $13.20 $12.S.00 $9. 11–24. K.00 100% 100% 100% 98% 132 2 days 4 days $3238.040.60 $8. 2004. Proj. D.000.00 100% 75% 75% 99% 105 4 days 5 days $17. Eng.601.00 100% 90% 90% 97% 123 2 days 2 days $3424.156.C.A (continued) ID Baseline Planned Duration (BPDi) Actual Duration (ADi) Baseline Cost Actual Cost APIi DPIi DEIi CEIi 88 5 days 5 days $10.00 100% 86% 86% 93% 146 1 day 2 days $3976.G. Uncertainty in task duration and cost Cooper. 10–17.652. Int. H.00 100% 100% 100% 94% 135 3 days 4 days $5628.00 100% 50% 50% 66% 147 1 day 2 days $2429.800.. Joel M.00 $34.488. Douglas.00 $3. 2004. 2006. Project Management Institute.1040 H.00 100% 80% 80% 100% 106 14 days 14 days $92.00 100% 67% 67% 70% 143 10 days 10 days $46.00 100% 80% 80% 96% 101 15 days 16 days $33.520.00 $5. 2006. Jacob.00 $6. 80–85.000.00 $55.00 100% 94% 94% 98% 102 2 days 3 days $4560.00 $25. Proj. Frank T. 25.00 100% 50% 50% 81% 133 3 days 3 days $23. Manage.00 $6. effectively. The Measurable News. If EVM is so good… why isn't it used on all Kim.000.00 100% 83% 83% 100% 130 5 days 5 days $16. (57).000.274.80 $60. Is “Earned Schedule” an unreliable indicator? No. Khamooshi. Oper.416.00 $100.00 100% 100% 100% 100% 100 4 days 5 days $11. improvement index.00 $55.00 100% 83% 83% 98% 96 12 days 13 days $44. Manag.440.000.000..00 100% 67% 67% 91% 104 3 days 4 days $8485. Earned value project 375–382. Kane.. Wells Jr..00 $16.00 $15.00 100% 100% 100% 99% 131 5 days 5 days $16.. D.500.00 100% 50% 50% 81% 148 1 day 1 day $9212.

Manag. Practice standard for Earned Value Vanhoucke. W. 2011.. Vanhoucke. CHAOS manifesto 2011. Proj. . 87–95.W. 2013. H. Turner. Anbari. J. 764–772. Golafshani / International Journal of Project Management 32 (2014) 1019–1041 1041 Lipke. 2005a. Standish Group. Int. 1361–1374. 41 (6). Res. Shadrokh.. Evolution of project management as evidenced by papers Moslemi Naeni. Frank. Amir. Salehipour. Mario. 1–6. Harv. 289–302. Mario. 24 (4). Proj. Stephan.. Project Management Institute. Proj. Soc. Manag. Int. Ofer. 29 (6). 2006. A simulation and evaluation of earned Management. 2011. J. Project Management Institute. second edition. A guide to the project management body of project outcome: the application of statistical methods to earned value knowledge (PMBOK® guide). Oper. Prediction of Project Management Institute. Henderson. J. J. Practice standard for Earned Value Vandevoorde. How PERT-cost helps the general manager. Leila. J. J. Vanhoucke.R. Zwikael. management and earned schedule performance indexes. 27 (4).. project duration forecasting methods. Int. approach for the earned value management. A comparison of different Bus. Project Management Institute. Shahram. M. H. 2010. Hilliard W. S. Manag. Int. 2009. Kym. 2005b. Manag. Paige. A fuzzy published in the International Journal of Project Management. 1963. Rev. Measuring time: improving project performance Management. 400–407. fifth edition. 2007. Proj. value metrics to forecast project duration. 58 (10). using earned value management.. Khamooshi. Vandevoorde. 2012.