Chapter Overview

Chapter 8 - Instructor’s Resource Guide to
Project Management: A Managerial Approach, seventh edition
Chapter Overview
1) Overview – This chapter continues the discussion of project implementation by
covering various scheduling techniques.
2) Background – Per the text, “A schedule is the conversion of a project action plan into
an operating timetable.” Because projects are unique, a schedule is especially
important because there is often no on-going process that simply has to be repeated
on a daily basis. The basic process is to identify all the tasks and the sequential
relationships among them, that is, which tasks must precede or succeed others. There
are a number of benefits to the creation and use of these networks including:
a) It is a consistent framework for planning, scheduling and controlling the project.
b) It can be used to determine a beginning and ending date for every project task.
c) It identifies the activities that if delayed will delay the project.
3) Network Techniques: PERT (ADM) and CPM (PDM) – PERT and CPM are the most
commonly used approaches to project scheduling. Both were introduced in the
1950’s. PERT has been primarily associated with R&D projects, while CPM with
construction projects. Today PERT is not used much since project management
software generates CPM style networks. The primary difference between them is that
PERT uses probabilistic techniques to determine task durations, while CPM relies on
a single duration estimate for each task. Both techniques identify the critical path
(tasks that cannot be delayed without delaying the project) and associated float or
slack in the schedule. In 2005 the Project Management Institute (PMI) deemed it
necessary to change the names of these techniques. According to PMI, PERT is
called ADM/PERT (Arrow Diagram Method) and CPM is PDM/CPM (Precedence
Diagramming Method).
a) Terminology – The following are the key terms associated with the development
and use of networks:
i) Activity – A specific task or set of tasks that have a beginning and end and
consume resources.
ii) Event – The result of completing one or more activities. Events don’t use
resources.
iii) Network – The arrangement of all activities and events in their logical
sequence represented by arcs and nodes.
iv) Path – The series of activities between any two events.
v) Critical – Activities, events or paths which, if delayed, will delay the project.
To construct the network the predecessors and successors of each activity must be
identified. Activities that start the network will have no predecessor. Activities
that end the network have no successor. Regardless of the technique used, it is
Rev: Aug, 2008 Revised: Dwayne Whitten Page 1 of 57

good practice to link the activities with no other predecessor to a START

milestone. Those without any successor should be linked to an END milestone.
PDM/CPM networks identify the activities as nodes in the network: the so-called
Activity on Node (AON) network. The arrows in between the nodes depict the
predecessor/successor relationships among the activities. The ADM/PERT
method, on the other hand, uses Activity on Arrow (AOA) networks. Here the
nodes represent events and the arrows represent the actual activities.
b) Constructing the Network, AON Version – The text illustrates the development of
a simple AON network. All the major project management software packages
will generate this type of network.
c) Constructing the Network, AOA Version – The AOA network has some
development rules that make it somewhat more difficult to construct than the
AON network. The primary rule is that any given activity must have its source in
one and only one node. As a result, some network relationships can only be
depicted with the use of a “dummy” activity. This is an activity that has no
duration and consumes no resources. Its sole purpose is to indicate a precedence
relationship. The text uses various figures to illustrate the use of dummy
activities.
d) Gantt (Bars) Charts and Microsoft® Project (MSP) – The most familiar tool for
depicting project schedules is the Gantt chart, invented by Henry L. Gantt in
1917. The activities are depicted as horizontal bars with their length proportional
to their duration. This method results in an easy to read graphical depiction of the
project schedule. Gantt charts can be difficult to maintain if there are large
changes in the project schedule. Tools like Microsoft® Project (MSP) will
automatically draw the Gantt chart as a by-product of the network entered by the
user. The disadvantage of the Gantt chart is that it typically does not depict the
network relationships.
e) Solving the Network – The text illustrates the development of another AON
network based on the project detailed in Table 8-1.
f) Calculating Activity Times – The sample project in the text has three duration
estimates for each activity: optimistic (a), most likely (m) and pessimistic (b).
Optimistic and pessimistic are defined as the durations that represent 99 percent
certainty. In other words the actual duration of an activity will be less than the
optimistic or greater than the pessimistic only one percent of the time. Then the
expected time (TE) is found with the formula:
TE = (a + 4m + b)/6
where:
a = optimistic time estimate
b = pessimistic time estimate

m = most likely time estimate, the mode

This formula is based on the beta statistical distribution. In spite of a flurry of
discussion in the 1980’s the assumptions used to derive this formula have stood
the test of time. Along with TE, the variance of the durations can be calculated
as:

2
   b a / 6 2
and the standard deviation as:
 2
g) Critical Path and Time – Using the example network, the text describes the
concept of the critical path. For simple projects, the critical path can be found by
determining the longest path through the network.
h) Slack (aka, Float) – In the previous section, the earliest possible dates for each
activity were determined. By starting the analysis at the end of the network and
working through it backwards, the latest possible dates for each activity can be
determined. The difference between the early dates and the late dates is float or
slack. Activities on the critical path have zero float.
i) Precedence Diagramming – The Precedence Diagram Method allows for
additional relationships to be established between activities. They are:
i) Finish to Start – The successor activity cannot begin until the predecessor
finishes. This is the most common relationship depicted in networks.
ii) Start to Start – The successor activity cannot begin until the predecessor
begins.
iii) Finish to Finish – The successor activity cannot finish until the predecessor
activity finishes.
iv) Start to Finish – The successor activity cannot finish until the predecessor
activity starts. This relationship is rarely used.
In addition to these relationships, PDM allows for leads and lags which is the
introduction of a specific time period between the linked activities. For example,
in a Start to Start relationship with five days of lag, the successor activity cannot
begin until five days after the predecessor starts. The critical path and slack
calculation resulting from these relationships can be complicated and counter
intuitive.
j) Once again, Microsoft® Project – The text illustrates the use of MSP for
calculating the most likely project duration using the PERT method.

k) Exhibits Available from Software, a Bit More MSP – The text illustrates the types
of outputs available from MSP.
l) Uncertainty of Project Completion Time – The chance of completing a project
within a given time period can be calculated. The project activities are assumed
to be statistically independent and the variance of a set of activities is equal to the
sum of the variances of the individual activities comprising the set. Then the
chance of meeting a particular project duration can be calculated as:
Z  D   / 
2

where:
D = the desired project completion time
µ = the critical time of the project, the sum of the TEs for activities on the critical
path

2

= the variance of the critical path, the sum of the variances of activities on the
critical path
Z = the number of standard deviation of a normal distribution (the standard
normal deviate)
The weakest element of this technique is that it is difficult to account for the
possibility that other paths through the network may become critical due to
variation in their duration. Simulation techniques using tools like Crystal Ball®
or Risk +® can be used to better model this situation.
m) Toward Realistic Time Estimates – The traditional PERT method uses optimistic
and pessimistic duration estimates at the 99% confidence level. The calculations
can be adjusted for lower confidence levels that estimators may be more
comfortable in predicting such as 90 or 95%.
4) Risk Analysis Using Simulation With Crystal Ball® – Tools like Crystal Ball® can
be used to model the project and determine the likelihood of completion within a
certain time. Since Crystal Ball® works with Microsoft® Excel, the project network
must be modeled in the spreadsheet. This involves creating cells that calculate the
early and late dates for each activity. Then a distribution (commonly triangular) with
the appropriate parameters can be assigned to each duration. Finally, Crystal Ball®
runs its simulation and the results are displayed.
a) Traditional Statistics or Simulation? – With the advent of inexpensive and easy to
use tools, simulation is the recommended way to model uncertainty in project
durations. Both methods require the development of three durations for each
activity. The simulation method, however, does a much better job of handling the
possibility that the critical path will shift due to variation in durations of the

activities. This issue can be analyzed with traditional statistics, but it takes
considerable manual effort on the part of the analyst.
5) Using These Tools – The text gives an example of the use of the tools on a specific
project.
Teaching Tips
The authors correctly applaud the advent of the many user-friendly and powerful project
management tools available for PC’s. They have performed a great service by integrating
the use of tools like Microsoft® Project into the subject matter. My experience in
teaching these tools, however, reminds me of the story about IBM’s new programming
language. The story goes that IBM, in their marketing campaign for their new language,
said that the language was so easy to use that it would virtually eliminate the position of
programmer. Anybody would be able to use this language to easily create computer
programs. The punch line is that the new language was FORTRAN, a 1960’s era product
long ago supplanted by other “user friendly” products. Even today, with all the
marvelous tools we have available, the position of programmer has not disappeared. The
point of all this is that in the classroom students will have a wide range of skills and
abilities that may or may not be applicable to Microsoft® Project and similar tools. My
experience is that it is unreasonable to expect students to pick up MSP and use it
successfully based solely on the printed examples in the text. Unfortunately, even
students who claim to be experienced in the use of the tool often know how to draw a
Gantt chart and little else. This is due to three reasons:
 MSP on the surface may look like a fancy spreadsheet, but “under the hood”
it’s is a very complex tool. Many of its processes are dependent on complex
algorithms, controlled by a seemingly endless series of settings with
mysterious titles.
 The training provided by most organizations ranges between non-existent
and abominable. If there is training, it’s usually administered by someone
who has never managed a project and doesn’t understand much more than
the students he or she is teaching. One of my very computer literate
colleagues will probably become homicidal if he is forced to go to yet
another training session that concentrates on issues vital to the PM like
changing the color of the fonts. In my twenty years of both managing
projects and teaching project management, I have encountered one person
who is both an accomplished user of MSP on actual projects and a capable
instructor.
 MSP training, if it exists, is usually done out of context. It is taught as a
standalone computer tool, without any of the concepts of project
management to put its use in the proper perspective.
The opportunity then, for any project management instructor, is to provide both the
concept and meaningful tools training in one package. Ideally, lectures on concept

should be alternated with lab sessions using the tool. If this is not possible then, at a
minimum, the instructor must set aside class time to demonstrate the key elements of
MSP needed to complete the homework problems. Then, a week later, be prepared for
the questions and frustrations that will erupt from the class.
A good reference case for this chapter follows:
Ivey cases:
9A97D001 Note on Introduction to Project Management. An introduction to projects
with a simple AON problem.
9A97D002 Gadget Toy Company. A simple AON problem that also introduces
Microsoft Project (MSP).
9A95D015 H.M.S. Pinafore. A moderately longer AOA problem.
9A98D020 Procter & Gamble Canada: Dayquil Sampling Operations. A realistic
problem based on an actual summer intern’s experience involving a quick decision on a
new product line that requires a number of tasks to be executed.
Univ. of Virginia Darden case:

UVA-OM-0803 Tastee Snax Cookie Company. A straightforward but more involved
network problem.
Material Review Questions

Question 1:
1) Activity: Activities have a start and finish and consume resources.
2) Event: Events do not consume resources. Typically, events designate the start or the
completion of an activity or of a path.
3) Path: A path traces the predecessor-successor relationships that exist among a set of
interconnected activities and events. The boundaries of a path use a starting event
and a completion event to designate the path’s endpoints.
4) Dummy Activity: Dummy activities are tasks of zero duration that have no resources
assigned to them. They are used to maintain predecessor/successor relationships in
AOA networks.
Question 2:
Activities along the critical path cannot be delayed without delaying the completion of
the project.

Question 3:
1) Gantt Chart: The Gantt chart compares planned and actual progress for the
detailed tasks in a project.
2) Master Schedule: The Gantt chart format (bars to represent progress over time)
may be used to display data regarding the master project schedule, but the master
schedule is oriented towards overall management of the project and will only
focus on the major project tasks. For example, the Gantt view in MS-Project can
be filtered to only show summary tasks at a particular level of the WBS hierarchy.
Question 4:
Total slack vs. free slack:
Total slack is the difference between the calculated earliest finish time of the very last
activity and the project’s required completion time.
Free slack is the time an activity can be delayed without affecting the start time of any
successor activity.
So, the total slack deals with the relationship between the current activity and the total
project completion time, while free slack relates to the next activity.
Question 5:
The authors of the text have suggested that PERT/CPM are very similar. Therefore, the
terms PERT and CPM have been used interchangeably throughout most of the textbook
when explaining basic educational concepts about project schedules. The following
guidelines are suggested regarding when to use each type of scheduling technique
discussed in this chapter.
1) PDM/CPM should be used when the control of costs associated with expediting
work is an important concern. PDM networks should be used when the project
requires the use of leads and lags between activities. PDM is easier to draw than
ADM, is used in most project software applications, and tends to be preferred
when CPM is used.
2) ADM/PERT should be used when the activity times are estimated using
probability distributions in order to evaluate the range of uncertainty around the
expected project duration. ADM networks should be used when it is desirable to
show completion events as part of the scheduling network, though nothing
prevents the use of Start and Finish events in a PDM/CPM network.
3) The Gantt chart is a useful tool for displaying the schedule regardless of what
method is used to derive it. The Gantt chart can be used directly to develop small
project schedules.
4) A less-known approach: GERT should be used when the project plan is complex
enough to require loop backs and/or the use of multiple probability distributions
associated with branching options in the relationships between activities.

Question 6:
AON (activity on node) places the activities or tasks on a rectangle (node), whereas the
AOA places the activities on arrows connecting nodes. Typically the AON provides more
information per activity in the diagram itself because more information can be placed on
the node itself (start time, finish time, etc)
Question 7:
Simulation requires the project schedule be modeled mathematically, which happens to
be a by-product of any of the network scheduling techniques. Once the model is
established, simulation involves inputting appropriately distributed random numbers into
the independent variables and analyzing the resulting distribution of the dependent
variables (those calculated by the model). To make the result meaningful, hundreds if not
thousands of trials are run, to build a statistically significant output distribution. Once the
output distribution is established, probabilities of various outcomes can be calculated.
Question 8:
Networks are drawn from left to right. Arrowheads indicate the direction of flow in the
network. The flow designates the precedence relationships between activities in the
network
Question 9:
The early start time and the early finish time are computed by performing the “forward
pass” calculation in a network. The late start time is computed by performing the
“backward pass” calculation in a network.
1) Late start time: Given the precedence relationships in a network, this is the latest
time that an activity can begin without extending the time required to complete
the entire project.
2) Early start time: Given the precedence relationships in a network, this is the
earliest time that an activity can begin. In order to begin, all predecessor
constraints must have been satisfied.
3) Early finish time: Given the precedence relationships in a network and the
activities duration, this is the earliest time that an activity can be completed if all
predecessor constraints are satisfied.
Question 10:
The critical path is determined by performing the “forward pass” and backward pass
calculations. Float is calculated by subtracting the early dates from the late dates,
specifically the early start from the late start, or the early finish from the late finish. If an
activity has zero float, then it is on the critical path because any delay would extend the
project’s completion date.

Question 11:
The reason slack is important is for two reasons: 1) Slack tells us that we can be a bit
more forgiving about delays on paths with slack whereas our primary attention should be
directed to the critical path. 2) If we need additional resources for some reason (such as a
delay on the critical path), the first place to look is at the resources on paths with slack in
case they might be available for use.
Class Discussion Questions

Question 12:
The network diagram could serve as a rough process flowchart showing the steps in a
manufacturing cycle. The direct and indirect costs for each step could be identified and
scheduled for each iteration of the total manufacturing cycle. However, the basic
networks would not be sophisticated enough to capture costs associated with variables
such as production yields, rework loop backs, and branching logic commonly associated
with control points to assure quality.
Question 13:
1) Benefits:
a) Illustrates task interdependencies
b) Establishes the sequence of activities (precedence)
c) Highlights critical and near-critical paths and their tasks
d) Highlights activities that contain float
2) Disadvantages:
a) Emphasizes time at the expense of other dimensions of project success
b) Large networks are difficult to print in a convenient format and they may
require significant wall space to view the entire network
c) As the network technique becomes more complex, its effectiveness as a
control tool is reduced
Question 14:
This is a great question to kick off a lively class discussion. There are no black and white
answers to this question but here are a couple thoughts:
 It’s easy to become obsessed by the use of even more sophisticated tools and lose
sight of the big picture. Project management tools are only useful if they help
projects achieve their cost, schedule and scope goals. Just because a tool is more
sophisticated doesn’t mean that it will yield a better result for the business.
 Organizations have to clearly articulate the goals of a project, put together some
kind of a plan, and then meticulously monitor its execution. Many organizations

gloss over the monitoring part because they believe it smacks of micro-
management. In spite of what Dilbert thinks, managers must have a mechanism
for knowing where their project is every day. This allows corrective action to be
taken before the problem grows beyond recovery. This attention to detail is
boring and repetitive, but it’s far more fundamental to the success of the project
than the sophistication of the simulation tools used to model the plan.
Question 15:
Both methods are of significant value because they force the PM to consider the
relationships among the project activities. Then using these relationships, both methods
produce a schedule for those activities. In addition, both methods can be used for
analysis of variances and problems when the schedule is executed.
Question 16:
There are many ways to deal with uncertainty. The most common in the scheduling
process are:
 Adding buffer or padding to the duration of each activity
 Adding buffer to the overall project schedule
 Developing schedules based on a range of activity durations
 Calculating probabilities of completion using statistical or simulation techniques
 Taking specific actions to reduce the uncertainty in duration for some or all the
activities
Question 17:
The “free slack” as it is called, is the slack along a path in the project and is the minimum
of all the slacks on that path. Thus, if the path of interest is A-B-C and the slacks on A
and B are each 3 while the slack on C is 2, the free slack on the path is 2.
Question 18:
Activity times are generally estimated in a manner similar to budgets. For example, they
can be individually estimated by the participants, or calculated based on rate data from an
earlier project.
Question 19:
Yes and no. Critical path activities deserve closer scrutiny because if they run late the
project is sure to be late. In a situation where scant resources have to be allocated to help
late activities recover, for example, the critical path activities would get the resources
before the non-critical path activities.
This, however, does not absolve the PM from monitoring the non-critical path items.
Items off the critical path may feed the critical path, so if they are late they could delay
the project indirectly. Also if non-critical path activities get late enough then the critical
path may shift to them, again delaying the entire project.

Question 20:
I’m not aware of any network relationship that can’t be built through some combination
of the PDM relationships with leads and lags. As the text points out though, the
relationships can become quite complicated, leading to anomalies in the critical path.
Questions for Project Management in Practice
Replacing the Atigun Section of the TransAlaska Pipeline
Question 21:
The pipeline does have sections that go below ground, allowing animals to end up on top.
The bear in this picture hasn’t found a way to get on top yet.
Question 22:
Petroleum engineers built a bypass system that allows diverting the oil flow temporarily
for repairs without interrupting it.
Question 23:
The environment for this project was very hostile. In addition to limited sunlight (3 hours
per day), temperatures were as low as -60 degrees during winter. Unless robots could be
used, shifts were likely to be limited to the time one could withstand the temperatures and
still avoid frostbite.
Hosting the Annual Project Management Institute Symposium
Question 24:
One unique aspect of this project is its length. The Gantt chart shows that planning for
the symposium began more than four years before the event and continued for a year
after. This means that several symposia are in the continuously in the planning process
throughout the United States.
Question 25:
The symposium took place in September 1992 and the supporting project completed in
April of 1993.
Question 26:
The activities after completion of the symposium are tasks associated with project
closeout work. This would include tasks related to contract closeout and administrative
closeout, to include creation of a project archive and a summary report of lessons learned.
For this project, closeout lasted approximately 6 months. This question points out the
importance of establishing a common understanding of when any project is actually done.
The answer may not be obvious, and it can come back to bite the PM when he or she least
expects it.

Problems
NOTE: Many of the AON graphics in this solutions set depict the start day of the
successor activity to be the same day as the completion of the predecessor. This is
consistent with the presentation in the text. It is not consistent with the result that would
be obtained using Microsoft® Project, where the start day of the successor is always the
next working day after the completion of the predecessor.
Problem 1:
Problem 2:
Problem 3:
1) The arrows cannot form a loop such as the one shown between nodes 2, 3, and 5.
2) The dummy arrow between nodes 6 and 7 is not required because 6 precedes 5
and 5 precedes 7.

3) Nodes 8 and 9 do not have successors, so it appears that this network has two
final termination nodes. That is not a conventional diagramming technique. An
arrow from 8 should point to 9.
Problem 4:
a) The critical path is B-E-G.
b) 23 work periods.
Problem 5:
Initial PDM Diagram
Adjusted PDM Diagram
Problem 6:
PDM Diagram 6a

PDM Diagram 6b
Problem 7:
Figure 7a is ADM format.
Figure 7b is PDM format.

A C F
G
Start D H End
B E

Problem 8:
Please see note about network depiction preceding Problem 1
a) The critical path activities are A, C, E, and G.
b) The project’s duration is 22 days.
c) Yes, activity B can be delayed one day without delaying the completion of
the project.
Problem 9:
Task a m b Expected Variance Std Dev.
A 6.5 7.5 14.5 8.5 1.78 1.33
B 8.5 10.5 12.5 10.5 0.44 0.67
C 2.5 3.5 4.5 3.5 0.11 0.33
D 6.5 7.5 8.5 7.5 0.11 0.33
E 5.5 5.5 9.5 6.2 0.44 0.67
F 5.5 7.5 9.5 7.5 0.44 0.67
G 4.5 6.5 8.5 6.5 0.44 0.67
H 2.5 3.5 3.5 3.3 0.03 0.17

Sum of
Expected Variances
Desired Project Critical
Duration Duration Path Z Probability
21 24.7 2.78 -2.2 a) 1.4%
22 24.7 2.78 -1.6 b) 5.5%
25 24.7 2.78 0.2 c) 57.9%

Problem 10:
a) The critical path is AC CB BE EF.

b) The only event with slack is “D” at 3 days.
c) If “D” were the final event in the network, then the critical path would be AC CB
BD.
d) The following spreadsheet excerpt illustrates the calculation of the probability of
completion in 14 days:
AB 3 6 9 6.0 1.00 1.00
AC 1 4 7 4.0 1.00 1.00
CB 0 3 6 3.0 1.00 1.00
CD 3 3 3 3.0 0.00 0.00
CE 2 2 8 3.0 1.00 1.00
BD 0 0 6 1.0 1.00 1.00
BE 2 5 8 5.0 1.00 1.00
DF 4 4 10 5.0 1.00 1.00
DE 1 1 1 1.0 0.00 0.00
EF 1 4 7 4.0 1.00 1.00

Sum of
Expected Variances
14 16.0 4.00 -1 15.9%
e) If CD slips to six days the critical path is unchanged but slack on D is reduced. If
CD slips to seven days then there are two critical paths: AC CB BE EF and AC
CD DF. If CD slips to eight days then the critical path shifts to AC CD DF and
the project duration extends to 17 days.

Problem 11:
10 28 36
B G 8 I
17 28 41
1
3 10
9
4 2
0 7 18 37 43
A 3 C 2 F 11 H 6 J
0 7 18 37 43
1
4 5
6 3
3
3 13
D 8 E
3 13
Figure 8.11 shows duration on the arrow in matching the “(i,j)” notation used to
define the problem’s source data.
b) The critical path is A, D, C, E, F, G, H, J.

c) The completion time is 43 days.

Problem 12:
Figure 8.12a shows the PDM network for the data from Table A of Problem
8-12 assuming that the data were applied as shown in Figure 8.12b.
1) The critical path is 2,3,4,5,7,8,9.
2) The slack for activity 1 is 11.7 days. The slack for activity 6 is 4 days.
2) The following table shows the calculation of the expected completion time:
Activity a m b Expected
1 8 10 13 10.2
2 5 6 8 6.2
3 13 15 21 15.7
4 10 12 14 12.0
5 11 20 30 20.2
6 4 5 8 5.3
7 2 3 4 3.0
8 4 6 10 6.3
9 2 3 4 3.0
Expected
Project
Duration
66.4

Problem 13:
Figure 8.13 shows the network for problem 13.
1) The critical path is A, B, E, I, L, M, N, P.
2) The completion time is 75 months.
Problem 14:
Figure 8.14a shows the original network diagram for problem 14.

1) The critical path activities are A, D, G, and J. Activities B and E should be closely
monitored as a near critical path.
Figure 8.14b shows the impact of the project’s performance to date.

2) The project will be completed in 12.5 days instead of the 13 days originally
expected. The near critical path (B, E, G, J) is now critical. Activities A, D, F, and H are
now near critical activities.
Problem 15:
Figure 8.15a shows the network.

Using critical path analysis results in the following table:
Activity a m b Expected Variance Std Dev.

1,2 5 11 11 10.0 1.00 1.00
1,3 10 10 10 10.0 0.00 0.00
1,4 2 5 8 5.0 1.00 1.00
2,6 1 7 13 7.0 4.00 2.00
3,6 4 4 10 5.0 1.00 1.00
3,7 4 7 10 7.0 1.00 1.00
3,5 2 2 2 2.0 0.00 0.00
4,5 0 6 6 5.0 1.00 1.00
5,7 2 8 14 8.0 4.00 2.00
6,7 1 4 7 4.0 1.00 1.00
Sum of
Expected Variances
17 21.0 6.00 -1.63299 5.1%
18 21.0 6.00 -1.22474 11.0%
22 21.0 6.00 0.408248 65.8%
24 21.0 6.00 1.224745 89.0%
24.13 21.0 6.00 1.277817 89.9%
The probability of completion at 17 weeks is quite poor, only 5.1%.

Management can be sure that the project will complete 90% of the time in slightly over
24 weeks.
This analysis suggests that 11% of the time the project will achieve the bonus and 34.2%
(100 – 65.8) of the time the project will incur the penalty. Since the project is expected to
generate no profit or loss independent of the bonuses and penalties the expected value
calculation yields:
11% X $10,000 + 34.2% X $(-5000) = $(-604)

Strictly based on the expected return management should “no bid” the project. Because
the loss is so small, however, the firm may wish to apply other criteria to decide whether
to pursue the project.
Problem 16:
Using critical path analysis with the data provided gives the following table:
Activity Expected Std Dev. Variance

a 2.0 2.00 4.00
b 3.0 1.00 1.00
c 4.0 0.00 0.00
d 2.0 3.00 9.00
e 1.0 1.00 1.00
f 6.0 2.00 4.00
g 4.0 2.00 4.00
h 2.0 0.00 0.00
Expected Sum of
Desired Project Variances
Duration Duration Critical Path Z Probability
12 13.0 9.00 -0.33 36.9%
13 13.0 9.00 0.00 50.0%
16 13.0 9.00 1.00 84.1%
17.3 13.0 9.00 1.43 92.5%
For this problem the variance has to be calculated from the standard deviation, and the
durations provided are assumed to be the expected durations. As can be seen there is
about an 84% chance of completing the project within the drop dead time. If a little more
than a week is added to the duration, the chance of completing the project on time rises to
92.5%.

Problem 17:
Figure 8.17 shows the network for problem 17.

a) There are two critical paths: C, G and C, F, I.
b) The earliest complete date is 16 days.
c) E = 6d slack, F = 0d slack, and H = 1d slack.

Problem 18:
C, F, H is the critical path as indicated in the network. The slack times are as indicated for
all activities.

Problem 19:
Figure 8.19 shows the network diagram for problem 19.
Legend for Solution
3 0 10 10 5 14 Early
Slack
Early
Start Finish
C H Activity Critical Path

0 0 3 Late Late
Duration
3 7 10 15 4 19 Start Finish
A
0 3 3
5 3 7 10 0 16
D E 16 0 19
Start 19
8 2 10 10 6 16 I End
19
0 6 5
16 3 19
B 5 2 9
6 5 11 F 5 2 13
7 4 11 G
11 8 19

Problem 20:

a) The critical path is A, D, E, G, I, J.
b) The slack on process confirmation (F) is 20 days.
c) The slack on test pension plan (C) is 61 days.
d) The slack on verify debt restriction compliance (H) is 20 days.

Problem 21:
2 0 5 5 7 9
0 0 2 D H
A 2 3 5 5 0 13 12 4 16
0 2 2 I
2 6 7
5 8 13
E
0 9 4 8 5 13 13 1 15
16
Start B J End
16
9 4 13 3 4 9 14 2 16
F
13 0 16
7 6 13
0 4 3 L
C 3 6 7 13 3 16
4 3 7 G
Legend for Solution
9 4 13 Early Early
Slack
Start Finish
Activity Critical Path

Late Late
Duration
Start Finish

a) The ES and LS for each activity are shown in Figure 8.21.
b) The website’s early completion time is 16 weeks.
c) The slack on each activity is shown in Figure 8.21
d) The critical activities are A, D, I, L plus J.
e) The critical path is A, D, I, L.

Problem 22:

b) The critical path is B, F, H.
c) Week 9.
d) If activity E requires one extra week, the time will be absorbed in free float and
will not affect any other activity. If activity E requires two extra weeks, then a second
critical path will be created for activities B, E, G. If activity E requires three weeks,
negative float will be created and the project cannot complete in nine weeks. The new
completion time will rise to 10 weeks.

Problem 23:
Figure 8.23 shows the network diagram solution for problem 23.
The critical path is A, B, F, H, I as indicated above.
Problem 24:
Figure 8.24a shows the network diagram for problem 24a.

b) The critical path is B, E, G, H.
Figure 8.24c shows the network diagram solution to problem 24c.

d) Given a float value of 6 weeks, activity F seems to be the best candidate to supply
resources needed to crash the project. Since the float is almost 50% of the activity’s
duration, using its resources to work other activities is unlikely to convert activity F into a
near-critical activity. Since activity D is both critical and concurrent to activity F, the
resources should be transferred there.

Problem 25:
Following the discussion in the text, the calculations of TE and  need to be modified.
For a confidence level of 95%, the calculation for variance would be:

2
 b a  / 3.29
' '
2
Using these values, the following table can be calculated:
95% Confidence Level

AB 3 6 9 6.0 3.33 1.82
AC 1 4 7 4.0 3.33 1.82
CB 0 3 6 3.0 3.33 1.82
CD 3 3 3 3.0 0.00 0.00
CE 2 2 8 3.0 3.33 1.82
BD 0 0 6 1.0 3.33 1.82
BE 2 5 8 5.0 3.33 1.82
DF 4 4 10 5.0 3.33 1.82
DE 1 1 1 1.0 0.00 0.00
EF 1 4 7 4.0 3.33 1.82
Sum of
Expected Variances
14 16.0 13.30 -0.54833 29.2%
Note that the expected durations for the critical path remain unchanged, however, the
variances are quite a bit larger indicating that there is a larger chance that any given
duration will fall outside of the optimistic/pessimistic limit. Similarly for the 90%
confidence level, the following equations would be used:

2
 b a  / 2.56
' '
2

and the following table can be calculated:
90% Confidence Level

AB 3 6 9 6.0 5.49 2.34
AC 1 4 7 4.0 5.49 2.34
CB 0 3 6 3.0 5.49 2.34
CD 3 3 3 3.0 0.00 0.00
CE 2 2 8 3.0 5.49 2.34
BD 0 0 6 1.0 5.49 2.34
BE 2 5 8 5.0 5.49 2.34
DF 4 4 10 5.0 5.49 2.34
DE 1 1 1 1.0 0.00 0.00
EF 1 4 7 4.0 5.49 2.34
Sum of
Expected Variances
14 16.0 21.97 -0.42667 33.5%

Problem 26:
Figure 8.26a shows the network, critical path and slack times.
The critical path in the above diagram is 1-4, 4-5, 5-6, 6-7, 7-10, 10-11 with a duration of
59 days.
Tabulating the calculations for expected durations and probability looks like this:

1-2 6 8 10 8 0.44 0.67
1-3 5 6 7 6 0.11 0.33
1-4 6 6 6 6 0.00 0.00
2-6 0 0 0 0 0.00 0.00
2-7 10 11 12 11 0.11 0.33
3-6 12 14 16 14 0.44 0.67
4-5 5 8 11 8 1.00 1.00
4-9 7 9 11 9 0.44 0.67
5-6 8 10 12 10 0.44 0.67
5-9 0 0 0 0 0.00 0.00
6-7 14 15 16 15 0.11 0.33
6-8 10 12 14 12 0.44 0.67
7-10 9 12 15 12 1.00 1.00


8-10 0 4 14 5 5.44 2.33
9-11 5 5 5 5 0.00 0.00
10-11 7 8 9 8 0.11 0.33
Desired Expected Project Sum of Variances

Duration Duration Critical Path Z Probability
61.69 59.0 2.67 1.65 95.0%
The next longest path is 1-3, 3-6, 6-7, 7-10, 10-11 at 55 days. It will only be a concern if
under some circumstances; its duration exceeds the actual critical path of 59 days. Using
the same technique for calculating the probability of exceeding a particular duration gives
the following table for this path:
Probability for path 1-3-6-7-10-11
Expected
Desired Path Sum of Path
Duration Duration Variances Z Probability
59 55.0 1.78 3.00 99.9%
Clearly the chance of exceeding 59 days is quite small. The same technique can be
applied to the next longest path 1-4, 4-5, 5-6, 6-8, 8-10, 10-11 which while relatively
short has high variance:
Probability for path 1-4-5-6-8-10-11
Expected
Desired Path Sum of Path
Duration Duration Variances Z Probability
59 49.0 7.44 3.67 100.0%
Again it is clear that it is unlikely that this path will cause problems with the overall
project duration.

Problem 27:
Figure 8.27a shows the PDM network for problem 27.

The critical path in the above figure is 1, 4, 6, 8, 12, 15 with a duration of 52 days.
b) See Figure 8.27a for the critical path and its early completion time.
c) See the following table for the probabilities:

1 2 4 6 4 0.44 0.67
2 5 5 5 5 0.00 0.00
3 3 5 7 5 0.44 0.67
4 7 10 13 10 1.00 1.00
5 11 12 13 12 0.11 0.33
6 5 6 7 6 0.11 0.33
7 9 10 11 10 0.11 0.33
8 5 7 9 7 0.44 0.67
9 7 9 11 9 0.44 0.67
10 3 3 3 3 0.00 0.00


11 15 17 19 17 0.44 0.67
12 6 8 10 8 0.44 0.67
13 7 8 15 9 1.78 1.33
14 12 14 16 14 0.44 0.67
15 16 17 18 17 0.11 0.33
Desired Expected Sum of Variances

Duration Project Duration Critical Path Z Probability
53.35 52.0 2.56 0.84 80.1%
51.6 52.0 2.56 -0.25 40.1%
d) Since the play is supporting an Independence Day event, a one day delay would
be severe.

Problem 28:
Figure 28a shows the PDM network diagram for problem 28.
The duration in the figure above is 1, 3, 6 with a duration of 41 days.
The following table tabulates the variances and probability for this project:

1 6 10 14 10 1.78 1.33
2 0 1 2 1 0.11 0.33
3 16 20 30 21 5.44 2.33
4 3 5 7 5 0.44 0.67
5 2 3 4 3 0.11 0.33
6 7 10 13 10 1.00 1.00
7 1 2 3 2 0.11 0.33
8 0 2 4 2 0.44 0.67
9 2 2 2 2 0.00 0.00
10 2 3 4 3 0.11 0.33
11 0 1 2 1 0.11 0.33
12 1 2 3 2 0.11 0.33
Sum of
Expected Variances
44 41.0 8.22 1.05 85.2%


Problem 29:
To simulate the network in Crystal Ball®, a spreadsheet must be prepared that calculates
the duration of each possible path through the network:
Activities
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
4.0 5.0 5.0 10.0 12.0 6.0 10.0 7.0 9.0 3.0 17.0 8.0 10.0 14.0 17.0
Paths Project
1-5-10-13-15 1-4-9-13-15 1-4-6-8-12-15 2-6-8-12-15 2-3-8-12-15 2-3-7-11 2-3-7-14 Completion
46.0 50.0 52.0 29.0 44.0 37.0 34.0 52.0
The path durations are calculated by summing the expected duration of each participating
activity. The Project Completion is calculated as the maximum duration of any of the
possible paths.
Once the spreadsheet is setup, Crystal Ball® can be used for the simulation. Each of the
Expected durations is set to a triangular distribution with minimum and maximum equal
to the optimistic and pessimistic durations respectively using the Define Assumption
function. The setup for activity 1 looks like this:

Then a forecast cell is setup for the Project Completion to view how it changes due to the
simulation. The resulting histogram looks like this:
The statistics that correspond to this graph look like this:
By adjusting the sliders on the histogram, the 80% and 40% confidence levels can be
checked. They look like this:

These graphs show that a 54-day duration can be achieved with an 80% confidence level
and a 52-day duration with a 40% confidence level. These are slightly worse than the
results calculated with the variance method. The reason is that the simulation properly
takes into account the possibility that the critical path shifts for some possible
combinations of activity durations.

Problem 30:
The setup for problem 30 is similar to that for problem 29. First the spreadsheet in Excel
is prepared with the calculations for the paths:
Activities
1 2 3 4 5 6 7 8 9 10 11 12
10 1 22 5 3 10 2 2 2 3 1 2
Paths Project
1-2-4-5 1-2-4-7-8-11-12 1-2-4-7-9-10-12 1-3-9-10-12 1-3-6 Completion
19 23 25 39 42 42
Then, similar to problem 29, triangle distributions are established to calculate the
durations for all activities except 9 (no variation in the estimate).
The resulting forecast for the duration of the project and corresponding statistics are:

Note that the probability of completing the project in 44 days has dropped to about 70%.
Problem 31:
Dec '05 Jan '06 Feb '06 Mar '06 Apr '06
ID Task Name Duration Start Finish 4 11 18 25 1 8 15 22 29 5 12 19 26 5 12 19 26 2 9
1 Start 0 days Sun 12/11/05 Sun 12/11/05 12/11
2 a 2 wks Mon 12/12/05 Fri 12/23/05
3 b 2 wks Mon 12/12/05 Fri 12/23/05
4 c 4 wks Mon 12/26/05 Fri 1/20/06
5 d 3 wks Mon 1/23/06 Fri 2/10/06
6 e 1 wk Mon 1/23/06 Fri 1/27/06
7 f 2 wks Mon 2/13/06 Fri 2/24/06
8 g 3 wks Mon 2/27/06 Fri 3/17/06
9 h 1 wk Mon 3/20/06 Fri 3/24/06
10 End 0 days Fri 3/24/06 Fri 3/24/06 3/24
The figure shows the default Gantt chart view of the problem, with a project start day of
Sunday December 11, 2005. Note that MSP moves the beginning of the first task to the
first workday of Monday the 12th. This display shows the default calendar of 5 day 40 hr.
weeks with no holidays. A “Start” and “End” milestone have been inserted to insure that
all activities have at least one predecessor and successor.

The default “Tracking Gantt” view can be used to display the critical path:
Dec '05 Jan '06 Feb '06 Mar '06 Apr '06
ID Task Name Duration Start 4 11 18 25 1 8 15 22 29 5 12 19 26 5 12 19 26 2 9
1 Start 0 days Sun 12/11/05 12/11
2 a 2 wks Mon 12/12/05 0%
3 b 2 wks Mon 12/12/05 0%
4 c 4 wks Mon 12/26/05 0%
5 d 3 wks Mon 1/23/06 0%
6 e 1 wk Mon 1/23/06 0%
7 f 2 wks Mon 2/13/06 0%
8 g 3 wks Mon 2/27/06 0%
9 h 1 wk Mon 3/20/06 0%
10 End 0 days Fri 3/24/06 3/24
The project duration (in workdays) can be displayed in the “Project” >> “Project
Information” >> “Project Statistics” window, which looks like this:
Problem 32:
The “Pert Entry Form” in Microsoft® Project is used to enter the three durations. After
they are in the “Calculate Pert” button is clicked to populate the Duration field with the
expected durations. Note that MSP uses the non-standard terminology “Expected” in lieu
of “Most Likely.”

Using the calculated durations, the Gantt chart looks like this:
Dec '05 Jan '06 Feb '06
ID Task Name Duration Start Finish Predecessors 4 11 18 25 1 8 15 22 29 5 12 19 26
1 Start 0 days Sun 12/11/05 Sun 12/11/05 12/11
2 a 7.5 days Mon 12/12/05 Wed 12/21/05 1
3 b 8 days Mon 12/12/05 Wed 12/21/05 1
4 c 6 days Wed 12/21/05 Thu 12/29/05 2
5 d 14.5 days Thu 12/22/05 Wed 1/11/06 2,3
6 e 7 days Wed 1/11/06 Fri 1/20/06 4,5
7 f 11.5 days Fri 1/20/06 Mon 2/6/06 3,6
8 g 8 days Tue 2/7/06 Thu 2/16/06 7
9 End 0 days Thu 2/16/06 Thu 2/16/06 8 2/16
The figure shows the default Gantt chart view of the problem, with a project start day of
Sunday December 11, 2005. Note that MSP moves the beginning of the first task to the
first workday of Monday the 12th. This display shows the default calendar of 5 day 40 hr.
weeks with no holidays. A “Start” and “End” milestone have been inserted to insure that
all activities have at least one predecessor and successor.
The tracking Gantt view can be used to display the critical path:
Dec '05 Jan '06 Feb '06
ID Task Name Duration 4 11 18 25 1 8 15 22 29 5 12 19 26
1 Start 0 days 12/11
2 a 7.5 days 0%
3 b 8 days 0%
4 c 6 days 0%
5 d 14.5 days 0%
6 e 7 days 0%
7 f 11.5 days 0%
8 g 8 days 0%
9 End 0 days 2/16

The network diagram can be displayed directly from MSP using the “Network Diagram”
view. A portion of it with the default format settings looks like this:
a
Start
Start: 12/12/05 ID: 2
Milestone Date: Sun 12/11/05 Finish: 12/21/05 Dur: 7.5 days
ID: 1 Res:
b
Start: 12/12/05 ID: 3
Finish: 12/21/05 Dur: 8 days
Res:
The slack values are automatically calculated by MSP. They can be revealed in a number
of different views:
Nov '05 Dec '05 Jan '06 Feb '06 Mar '06 Apr '06
ID Task Name Start Finish Late Start Late Finish Free Slack Total Slack 27 4 11 18 25 1 8 15 22 29 5 12 19 26 5 12 19 26 2 9 16 23
1 Start Sun 12/11/05 Sun 12/11/05 Mon 12/12/05 Mon 12/12/05 0 days 0 days
2 a Mon 12/12/05 Wed 12/21/05 Mon 12/12/05 Wed 12/21/05 0 days 0.5 days
3 b Mon 12/12/05 Wed 12/21/05 Mon 12/12/05 Wed 12/21/05 0 days 0 days
4 c Wed 12/21/05 Thu 12/29/05 Tue 1/3/06 Wed 1/11/06 9 days 9 days 9 days
5 d Thu 12/22/05 Wed 1/11/06 Thu 12/22/05 Wed 1/11/06 0 days 0 days
6 e Wed 1/11/06 Fri 1/20/06 Wed 1/11/06 Fri 1/20/06 0 days 0 days
7 f Fri 1/20/06 Mon 2/6/06 Fri 1/20/06 Mon 2/6/06 0 days 0 days
8 g Tue 2/7/06 Thu 2/16/06 Tue 2/7/06 Thu 2/16/06 0 days 0 days
9 End Thu 2/16/06 Thu 2/16/06 Thu 2/16/06 Thu 2/16/06 0 days 0 days
This view shows the View “Detail Gantt” combined with the “Schedule” Table. Note that
the Gantt chart also displays the slack as a green line.

Incidents for Discussion

Yankee Chair Company
This incident is a good opportunity to engage the students in a discussion of what’s
important in the planning and scheduling process. The spirit of the case is to not do
dumb things like apply scheduling techniques that aren’t understood by the users.
Gort’s approach won’t work, because he doesn’t understand the scheduling process that
he intends to apply. The sophistication of the technique is not the issue. If he doesn’t use
PERT he can still come up with a useable schedule with any one of several techniques.
The issue is that as the PM he needs to understand why the schedule comes out the way it
does. He needs to be able to explain the schedule to others. He needs to be able to
understand the implications of uncertain data, and he needs to be able to understand the
impact of a reschedule. If he is using a scheduling technique that he doesn’t understand,
then he won’t be able to perform any of these important management tasks.
Cincinnati Software
This incident, like the previous one, is an excellent opportunity to engage the students in
a discussion of what’s really important in the project management process. I have seen
discussions similar to this incident take place many times in business situations. The
flavor of the discussion is, “what’s the latest technology and how much can we apply to
this project?” The real question should be, “ which technology is appropriate for the goal
we are trying to achieve?” In business we fall prey to the fallacy that the sophistication
of the technology will make up for our lack of planning and discipline. As the text has
discussed the keys to a good planning and scheduling process are:
 Solid understanding of senior management’s and the customer’s goals in
pursuing the project
 A complete listing of all the activities and products the project must accomplish
 A schedule that provides a start and complete date for each identified task
 A budget for each task
 A clear understanding of the outcome of each task
 A system for monitoring the progress of each task
 A process for identifying and managing change to the project scope
Note that the schedule is only one element, and as the text has explained, there are
several ways to get there. I will take a project that has every one of these in place, but
creates the schedule on a white board over a project that uses other methods, but has no
clear idea of what the customer’s expectations are.

That being said the options look like this:

 I would reject the Gantt chart as being too simplistic for the large plan and
schedule that will have to be maintained for a project as complicated as an ERP
system.
 I would work with the PERT and CPM camps to reach a compromise and use
the best of both worlds. That would be a CPM based network using PERT for
estimating durations.
CASE: The Sharon Construction Corporation

Question 1:
The first element necessary to analyze this situation is a project schedule for the base
plan. The Gantt chart from Microsoft® Project for the plan without any alterations looks
like this:
Figure 1: Base Schedule

The following assumptions were used in preparing this schedule:
1) The start date was chosen as February 15, 1999 so that the project penalty date would
be Tuesday February 15, 2000.
2) A five day, forty-hour work calendar was used with no holidays.
3) The durations were designated in weeks, and predecessors established per the
Appendix.

As can be seen, the project would end on 1/14/00 well before the penalty would kick in.
This date will be used as the baseline for all subsequent analysis.
The next step is to understand the impact of a strike on the schedule. For the first
scenario eight weeks of non-working time was added to the calendar between 11/29/99
and 1/21/00. The new schedule looks like this:
Figure 2: 8-Week Strike

Note that the project completion is delayed until 3/10/00. This causes the following
additional costs to be incurred:
Overhead costs: 8 weeks at $500/week = $4000
Penalty costs: 4 weeks at $15,000/week = $60,000

Next the 12-week strike will be analyzed:
Figure 3: 12 Week Strike

Note that now the completion date is delayed until 4/7/00. This causes the following
additional costs to be incurred:
Overhead costs: 12 weeks at $500/week = $6000
Penalty costs: 8 weeks at $15,000/week = $120,000
Next, the expected value of each possible outcome is calculated. To determine expected
value, the payoff is multiplied times the chance of occurrence. The following chart was
prepared following the format presented in Spreadsheet Modeling and Decision Analysis
by Cliff T. Ragsdale:

Figure 4: Project Expected Value Without Mitigations
The numbers below the horizontal line and to the left represent the cost or payout of that
individual scenario. Item 1 is the overall payout for the project if no penalties are
incurred. Similarly Item 2 is the cost of an 8-week strike as previously discussed. Item 3
shows the total of all the options on that branch. In this case it is the payout for the
project, less the penalty costs for an 8-week strike. Note that if there is a strike, no
additional costs are incurred for weather problems in December since the strike prevents
work during that month. On the branch with no strike, an additional heating cost of
$2000 would be incurred for an unusually cold December to mitigate the temperature
impact on pouring the seats. The expected value is calculated by working from right to
left starting with the total payout for the branch and is displayed just to the left and below
the node point. For example Item 5 is calculated as the sum of:
0.33 X 236,000 + 0.67 X 236,000 = 236,000.

Similarly Item 6 is calculated as:
0.7 X 236,000 + 0.3 X 174,000 = 217,400
Finally the expected value of the project with no mitigating activities against strikes and
weather is $258,370. This is in fact the expected value of the fifth proposal, doing
nothing.
The next step is to evaluate the expected value of the other four proposals. Proposal one
improves the project completion time to 12/3/99 with no strike, 1/28/00 if there is an 8-
week strike and to 2/25/99 for a 12 week strike. This reduces the additional costs to

$1000 for the 8-week (two weeks of overhead and zero weeks of penalty) and $33,000
for the 12-week (six weeks of overhead and two weeks of penalty). Note that once again
a strike and cold weather problems in December are mutually exclusive. Also, because of
the earlier completion with no strike, no concrete pouring activity will take place in
December and in fact the overhead costs are reduced because the project completes six
weeks before the baseline date. This allows a cost savings for the no strike branches of 6
X $500 or $3000
The updated expected value chart looks like this:
Figure 5: Project Expected Value with Expedited Gallery Pour

Note that the total maximum value of the project has been reduced to $280,000 because
proposal one requires an additional $20,000 expense for all possible scenarios. The
expected value of the project, however, rises to $276,200 with this proposal.
For proposal two, examination of Figures 2 and 3 shows that filling the field is not on the
critical path. Hence this amount of money would be wasted. Specifically, it lowers the
expected value of proposal one by $10,000 to $266,200.
Similarly proposal three also reduces the duration of a task off the critical path. As in
proposal two, this expense simply lowers the expected value of proposal one by $9,000 to
$267,200.
The fourth proposal is a little more challenging to evaluate. If there were an 8 or 12-
week strike the same two activities would benefit from expediting; Pour Seats and Paint
Seats. For the 8-week strike a smaller amount of expediting will be used to reduce the

extra cost incurred. Specifically, Pour Seats could be reduced to 2 weeks from 4 weeks
for a cost of $6000 and Paint Seats could be reduced to 1 week from 3 weeks for a cost of
$6000. The Gantt chart for the 8-week strike looks like this:
Figure 6: 8-Week Strike with Proposal Four Mitigations

For the 12-week strike, the maximum possible expediting would be used, specifically
Pour Seats would be reduced to 1.3 weeks for a cost of $8000 (2.66X$3000) and Paint
Seats would be reduced to 1 week for a cost of $6000:
Figure 7: 12-Week Strike with Proposal 4 Mitigations

The delay for the 8-week strike plus the expediting costs would cause costs of:

4 wks X $500/week (overhead) + 0 wks penalty + $12,000 expediting costs = $14,000

For the 12-week strike the delay plus expediting would be:
7.3 wks X $500/week (overhead) + 3.3 wks X $15,000/week (penalty) + $14,000 =
$67,666
If there is no strike, but a cold winter, than four weeks of heating costs will be incurred as
in the base scenario. Placing these values in the appropriate place on the expected value
chart gives the following:
Figure 8: Project Expected Value for Proposal 4

This analysis shows that proposal four has the best expected value and should be
selected.
Question 2:
Other solution techniques drawn from decision analysis such as the minimax decision
rule can be used. To employ this technique a table will be built of the possible payouts
for proposals 1, 4 and 5 (the others can be safely eliminated as no better than one of these
three):

Using minimax the best (biggest) worst case is in proposal 1($247,000), so it would be
selected.
Another technique is the minimax regrets. Using this rule we compare each proposal to
the best one in that column (state of nature) to determine how much would be lost if that
proposal were selected instead of the best one. For example, in the “8-Week Strike”
column the best outcome is $286,000 so the regret for choosing proposal 1 is $279,000-
$286,000 = $(-7,000). In other words it we choose proposal one and an 8-week strike
occurred we would regret our choice because we would have been $7000 better if we had
chosen proposal 4. The complete regrets table looks like this:
Using the minimax regrets criteria we compare the maximum regrets of each proposal
and choose the smallest one. In this case, proposal 4 has the smallest maximum regret at
$(-14,666) so it would be chosen.
Question 3:
There are a number of non-financial factors that could be considered in making the
decision including:
 The impact on long term labor relations of “working around” a strike.
 The political and marketing implications of not completing the job on time.
 The organization’s ability to quickly change and then execute an updated project
plan in response to events.
 Sharon Construction’s track record in predicting the outcome of labor problems.
 The company’s overall ability to execute any project, let alone one with problems
Question 4:
As the president I would execute proposal 4, based on its highest expected value.

Chapter Overview

Uploaded by

Copyright:

Available Formats

Chapter Overview

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter Overview

Uploaded by

Copyright:

Available Formats

Chapter 8 - Instructor’s Resource Guide to

Project Management: A Managerial Approach, seventh edition

Rev: Aug, 2008 Revised: Dwayne Whitten Page 1 of 57

good practice to link the activities with no other predecessor to a START

Rev: Aug, 2008 Revised: Dwayne Whitten Page 2 of 57

m = most likely time estimate, the mode

and the standard deviation as:

Rev: Aug, 2008 Revised: Dwayne Whitten Page 3 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 4 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 5 of 57

Univ. of Virginia Darden case:

Material Review Questions

Rev: Aug, 2008 Revised: Dwayne Whitten Page 6 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 7 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 8 of 57

Class Discussion Questions

Rev: Aug, 2008 Revised: Dwayne Whitten Page 9 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 10 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 11 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 12 of 57

Adjusted PDM Diagram

Rev: Aug, 2008 Revised: Dwayne Whitten Page 13 of 57

Figure 7b is PDM format.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 14 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 15 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 16 of 57

a) The critical path is AC CB BE EF.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 17 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 18 of 57

b) The critical path is A, D, C, E, F, G, H, J.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 19 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 20 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 21 of 57

Figure 8.14b shows the impact of the project’s performance to date.

Figure 8.15a shows the network.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 22 of 57

Using critical path analysis results in the following table:

Activity a m b Expected Variance Std Dev.

The probability of completion at 17 weeks is quite poor, only 5.1%.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 23 of 57

Activity Expected Std Dev. Variance

Rev: Aug, 2008 Revised: Dwayne Whitten Page 24 of 57

Figure 8.17 shows the network for problem 17.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 25 of 57

Rev: Aug, 2008 Revised: Dwayne Whitten Page 26 of 57

C H Activity Critical Path

Please see note about network depiction preceding Problem 1

Rev: Aug, 2008 Revised: Dwayne Whitten Page 27 of 57

Figure 8.20 shows the network diagram for problem 20.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 28 of 57

Activity Critical Path

Figure 8.21 shows the network diagram for problem 21.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 29 of 57

Figure 8.22 shows the network diagram for problem 22.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 30 of 57

Figure 8.24a shows the network diagram for problem 24a.

Rev: Aug, 2008 Revised: Dwayne Whitten Page 31 of 57

b) The critical path is B, E, G, H.

Figure 8.24c shows the network diagram solution to problem 24c.