Anderson Sweeney Williams Camm Cochran Fry Ohlmann

An Introduction to
Management Science, 15e
Quantitative Approaches to Decision Making

© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Lecture - 2
Chapter 3

Linear Programming:
Sensitivity Analysis and
Interpretation of Solution
2
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Linear Programming: Sensitivity Analysis and
Interpretation of Solution

– Introduction to Sensitivity Analysis

– Graphical Sensitivity Analysis
– Sensitivity Analysis: Computer Solution
– Limitations of Classical Sensitivity Analysis
– The Electronic Communications Problem

3
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Introduction to Sensitivity Analysis

Sensitivity analysis (or post-optimality analysis) is used

to determine how the optimal solution is affected by
changes, within specified ranges, in:
• the objective function coefficients
• the right-hand side (RHS) values
• Sensitivity analysis is important to a manager who
must operate in a dynamic environment with
imprecise estimates of the coefficients.
• Sensitivity analysis allows a manager to ask certain
what-if questions about the problem.

4
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Objective Function Coefficients

• Let us consider how changes in the objective function

coefficients might affect the optimal solution.
• The range of optimality for each coefficient provides
the range of values over which the current solution
will remain optimal.
• Managers should focus on those objective
coefficients that have a narrow range of optimality
and coefficients near the endpoints of the range.

5
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Range of Optimality

Graphically, the limits of a range of optimality are found by

changing the slope of the objective function line within the limits
of the slopes of the binding constraint lines.

6
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Range of Optimality

As long as the slope of

the objective function
line is between the slope
of line A (which coincides
with the cutting and
dyeing constraint line)
and the slope of line B
(which coincides with the
finishing constraint line),
extreme point 3 with
S = 540 and
D = 252 will be optimal.

7
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Range of Optimality

Extreme point 3 will be

the optimal solution as
long as
Slope of line B ≤ slope of
the objective function
line ≤ slope of line A.
In slope-intercept form.
7
Line A: 𝐷=− 𝑆+630
10
3
Line B: 𝐷=− 𝑆+1062
2
8
 Range of Optimality

Let us now consider the general form of the slope of the

objective function line. Let CS denote the profit of a standard
bag, CD denote the profit of a deluxe bag, and P denote the
value of the objective function.
Using this notation, the objective function line can be written as
P  Cs S  C D D

CS P
In slope-intercept form: D   S
CD CD

3 CS 7
Extreme point 3 will be optimal as long as     .
2 CD 10

9
 Range of Optimality

To compute the range of optimality for the standard-bag profit

contribution, we hold the profit contribution for the deluxe bag
fixed at its initial value 𝐶 𝐷=9.
3 CS 7
Therefore    
2 9 10
and 6.3  CS  13.5

The range of optimality for tells Par, Inc.’s management that,

with other coefficients unchanged, the profit contribution for
the standard bag can be anywhere between $6.30 and
$13.50 and the production quantities of 540 standard bags
and 252 deluxe bags will remain optimal.

10
 Range of Optimality

In cases where the

rotation of the
objective function
line about an optimal
extreme point
causes the objective
function line to
become vertical,
there will be either
no upper limit or no
lower limit for the
slope.

11
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Range of Optimality

CS 3
In this case  
CD 2

Following the previous

procedure of holding C
constant at its original
value, CD  9 we have
CS 3
 
9 2

13.5  CS  

12
 Range of Optimality

 Graphically, the limits of a range of optimality are found by

changing the slope of the objective function line within the
limits of the slopes of the binding constraint lines.
 Slope of an objective function line,
Max c1x1 + c2x2, is -c1/c2, and
the slope of a constraint, a1x1 + a2x2 = b, is -a1/a2.

 For our Example slope of objective function will be:

 Max 10S + 9 D 

13
 Simultaneous Changes

If two or more objective function coefficients are changed

simultaneously, further analysis is necessary to determine
whether the optimal solution will change.
When solving two-variable problems graphically, simply
compute the slope of the objective function ( CS CD )
for the new coefficient values. If this ratio is greater than
or equal to the lower limit on the slope of the objective
function and less than or equal to the upper limit, then
the changes made will not cause a change in the optimal
solution.

14
 Simultaneous Changes

Consider changes in both of the objective function coefficients

for the Par, Inc., problem. Suppose the profit contribution per
standard bag is increased to $13 and the profit contribution per
deluxe bag is simultaneously reduced to $8. Recall that the
ranges of optimality for CS and CD (both computed in a one-at-
a-time manner) are
6.3  CS  13.5
6.67  CD  14.29
For these ranges of optimality, we can conclude that changing
either CS to $13 or CD to $8 (but not both) would not cause a
change in the optimal solution of S = 540 and D = 252.

15
 Simultaneous Changes

However in recomputing the slope of the objective function

with simultaneous changes for both CS and CD, we saw that
the optimal solution did change.
A range of optimality, by itself, can only be used to draw a
conclusion about changes made to one objective function
coefficient at a time.

16
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Simultaneous Changes

 Range of Optimality and 100% Rule

The 100% rule states that simultaneous changes in
objective function coefficients will not change the optimal
solution as long as the sum of the percentages of the
change divided by the corresponding maximum allowable
change in the range of optimality for each coefficient does
not exceed 100%.

17
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Right-Hand Sides

Let us consider how a change in the right-hand side for a

constraint might affect the feasible region and perhaps cause
a change in the optimal solution.
• The change in the value of the optimal solution per unit
increase in the right hand side of the constraint is called
the dual value.
• As the RHS increases, other constraints will become
binding and limit the change in the value of the objective
function.

18
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Right-Hand Sides (Cutting and Dyeing Constraint)

Let us consider what happens if an additional 10 hours

of production time become available in the cutting and
dyeing department of Par, Inc. The right-hand side of the
cutting and dyeing constraint is changed from 630 to
640, and the constraint is rewritten as
7
S  1D  640
10
By obtaining an additional 10 hours of cutting and
dyeing time, we expand the feasible region for the
problem.

19
 Right-Hand Sides (Cutting and Dyeing Constraint)

Using the graphical

solution procedure with
the enlarged feasible
region shows that the
extreme point S = 527.5
and D = 270.75 now
provides the optimal
solution.
The increase in profit is
$7711.75 - $7668.00 =
$43.75.

20
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Right-Hand Sides (Cutting and Dyeing Constraint)

The change in the value of the optimal solution per unit

increase in the right-hand side of the constraint is called the
dual value.
• Here, the dual value for the cutting and dyeing constraint is
$4.375
• If we increase the right-hand side of the cutting and dyeing
constraint by 1 hour, the value of the objective function will
increase by $4.375.
• Conversely, if the right-hand side of the cutting and dyeing
constraint were to decrease by 1 hour, the objective function
would go down by $4.375.

21
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Dual Value – Shadow Price

On Textbook Page 93

22
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Sensitivity Analysis: Computer Solution

Software packages such as LINGO and Microsoft Excel

provide the following LP information:
 Information about the objective function:
• its optimal value
• coefficient ranges (ranges of optimality)
 Information about the decision variables:
• their optimal values
• their reduced costs
 Information about the constraints:
• the amount of slack or surplus
• the dual prices
• right-hand side ranges (ranges of feasibility)

23
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Sensitivity Analysis: Computer Solution

Let’s look at the Par, Inc., problem restated below:

24
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Solution for the Par, Inc., Problem

Optimal Objective Value = 7668.00000

Variable Value Reduced Cost

S 540.00000 0.00000
D 252.00000 0.00000

Constraint Slack/Surplus Dual Value

1 0.00000 4.37500
2 120.00000 0.00000
3 0.00000 6.93750
4 18.00000 0.00000

Variable Objective Coefficient Allowable Increase Allowable Decrease

S 10.00000 3.50000 3.70000
D 9.00000 5.28571 2.33333

Constraint RHS Value Allowable Increase Allowable Decrease

1 630.00000 52.36364 134.40000
2 600.00000 Infinite 120.00000
3 708.00000 192.00000 128.00000
4 135.00000 Infinite 18.00000

25
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Solution for the Par, Inc., Problem
The Dual Value column contains information about the marginal
value of each of the four resources at the optimal solution.
Optimal Objective Value = 7668.00000

Variable Value Reduced Cost

S 540.00000 0.00000
D 252.00000 0.00000

Constraint Slack/Surplus Dual Value

1 0.00000 4.37500
2 120.00000 0.00000
3 0.00000 6.93750
4 18.00000 0.00000

The dual values of 4.375 for constraint 1 (cutting and dyeing)

and 6.9375 for constraint 3 (finishing) tell us that an additional
hour of cutting and dyeing time increases the value of the
optimal solution by $4.37, and an additional hour of finishing
time increases the value of the optimal solution by $6.94.
26
 The Solution for the Par, Inc., Problem

The reduced cost is equal to the dual value for the

nonnegativity constraint associated with the variable.
Optimal Objective Value = 7668.00000

Variable Value Reduced Cost

S 540.00000 0.00000
D 252.00000 0.00000

The reduced cost on variable S and D is zero. The

nonnegativity constraint is S ≥ 0. The current value of S is 540,
so changing the nonnegativity constraint to S ≥ 1 has no effect
on the optimal solution value. Because increasing the right-hand
side by one unit has no effect on the optimal objective function
value, the dual value (i.e., reduced cost) of this nonnegativity
constraint is zero.

27
 The Solution for the Par, Inc., Problem

Variable S, which has a current profit coefficient of 10, has an

allowable increase of 3.5 and an allowable decrease of 3.7.
Variable Objective Coefficient Allowable Increase Allowable Decrease
S 10.00000 3.50000 3.70000
D 9.00000 5.28571 2.33333

Therefore, as long as the profit contribution associated with the

standard bag is between $10 - $3.7 = $6.30 and $10 - $3.5 =
$13.50, the production of S = 540 standard bags and D = 252
deluxe bags will remain the optimal solution.

28
 The Solution for the Par, Inc., Problem

The final section of the computer output provides the allowable

increase and allowable decrease in the right-hand sides of the
constraints relative to the dual values holding.
Constraint RHS Value Allowable Increase Allowable Decrease
1 630.00000 52.36364 134.40000
2 600.00000 Infinite 120.00000
3 708.00000 192.00000 128.00000
4 135.00000 Infinite 18.00000

As long as the constraint right-hand side is not increased

(decreased) by more than the allowable increase (decrease),
the associated dual value gives the exact change in the value of
the optimal solution per unit increase in the right-hand side.

29
The Solution for the Par, Inc., Problem (Excel Output)

30
 Relevant Cost and Sunk Cost

• A resource cost is a relevant cost if the amount paid for it

is dependent upon the amount of the resource used by
the decision variables.
• Relevant costs are reflected in the objective function
coefficients.
• A resource cost is a sunk cost if it must be paid
regardless of the amount of the resource actually used by
the decision variables.
• Sunk resource costs are not reflected in the objective
function coefficients.

31
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Cautionary Note on the Interpretation of Dual Values

When the cost of a resource is sunk, the dual value can be

interpreted as the maximum amount the company should
be willing to pay for one additional unit of the resource.
When the cost of a resource used is relevant, the dual
value can be interpreted as the amount by which the value
of the resource exceeds its cost.

32
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Modified Par, Inc., Problem

The graphical solution procedure is useful only for linear

programs involving two decision variables. In practice, the
problems solved using linear programming usually involve a
large number of variables and constraints.
For instance, the Management Science in Action, Determining
Optimal Production Quantities at GE Plastics, describes how a
linear programming model with 3100 variables and 1100
constraints was solved in less than 10 seconds to determine
the optimal production quantities at GE Plastics.

33
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Modified Par, Inc., Problem

In this section we discuss the formulation and computer

solution for two linear programs with three decision variables.
In doing so, we will show how to interpret the reduced-cost
portion of the computer output.

34
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Modified Par, Inc., Problem

Suppose that management is also considering producing a

lightweight model designed specifically for golfers who prefer
to carry their bags.
• The design department estimates that each new lightweight
model will require 0.8 hours for cutting and dyeing, 1 hour
for sewing, 1 hour for finishing, and 0.25 hours for
inspection and packaging.
• Because of the unique capabilities designed into the new
model, Par, Inc.’s management feels they will realize a profit
contribution of $12.85 for each lightweight model produced
during the current production period.

35
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 The Modified Par, Inc., Problem

After adding L to the objective function and to each of the four

constraints, we obtain the following linear program for the
modified problem:

36
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Solution To The Modified Par, Inc., Problem
Optimal Objective Value = 8299.80000

Variable Value Reduced Cost

S 280.00000 0.00000
D 0.00000 –1.15000
L 428.00000 0.00000

Constraint Slack/Surplus Dual Value

1 91.60000 0.00000
2 32.00000 0.00000
3 0.00000 8.10000
4 0.00000 19.00000

Variable Objective Coefficient Allowable Increase Allowable Decrease

S 10.00000 2.07000 4.86000
D 9.00000 1.15000 Infinite
L 12.85000 12.15000 0.94091

Constraint RHS Value Allowable Increase Allowable Decrease

1 630.00000 Infinite 91.60000
2 600.00000 Infinite 32.00000
3 708.00000 144.63158 168.00000
4 135.00000 9.60000 64.20000

37
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Solution To The Modified Par, Inc., Problem

The optimal solution calls for the production of 280 standard

bags, 0 deluxe bags, and 428 of the new lightweight bags;
the value of the optimal solution is $8299.80.
Optimal Objective Value = 8299.80000

Variable Value Reduced Cost

S 280.00000 0.00000
D 0.00000 –1.15000
L 428.00000 0.00000

38
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Solution To The Modified Par, Inc., Problem

Excel Output

39
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Solution To The Modified Par, Inc., Problem

Let us now look at the Reduced Cost column. The reduced

costs are the dual values of the corresponding nonnegativity
constraints.
Optimal Objective Value = 8299.80000

Variable Value Reduced Cost

S 280.00000 0.00000
D 0.00000 –1.15000
L 428.00000 0.00000

The reduced costs for S and L are zero because these decision
variables already have positive values in the optimal solution.
However, the reduced cost for decision variable D is –1.15. The
interpretation of this number is that if the production of deluxe
bags is increased from the current level of 0 to 1, then the
optimal objective function value will decrease by 1.15.
40
 Solution To The Modified Par, Inc., Problem

The dual values for constraints 3 and 4 are 8.1 and 19,
respectively, indicating that these two constraints are binding in
the optimal solution.
Constraint Slack/Surplus Dual Value
1 91.60000 0.00000

2 32.00000 0.00000
3 0.00000 8.10000
4 0.00000 19.00000

Each additional hour in the finishing department would

increase the value of the optimal solution by $8.10, and each
additional hour in the inspection and packaging department
would increase the value of the optimal solution by $19.00.

41
 Solution To The Modified Par, Inc., Problem

Because of a slack of 91.6 hours in the cutting and dyeing

department and 32 hours in the sewing department,
management might want to consider the possibility of utilizing
these unused labor-hours in the finishing or inspection and
packaging departments.
Constraint Slack/Surplus Dual Value
1 91.60000 0.00000

2 32.00000 0.00000
3 0.00000 8.10000
4 0.00000 19.00000

For example, some of the employees in the cutting and dyeing

department could be used to perform certain operations in
either the finishing department or the inspection and packaging
department.

42
 Limitations of Classical Sensitivity Analysis

Classical sensitivity analysis obtained from computer output

can provide useful information on the sensitivity of the solution
to changes in the model input data.
Classical sensitivity analysis provided by most computer
packages does have its limitations. Three such limitations are:
1. Simultaneous changes in input data
2. Changes in constraint coefficients
3. Nonintuitive dual values

43
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Limitations of Classical Sensitivity Analysis

Limitation: Simultaneous changes in input data

• The sensitivity analysis information in computer
output is based on the assumption that only one
coefficient changes; it is assumed that all other
coefficients will remain as stated in the original
problem. Thus, the range analysis for the objective
function coefficients and the constraint righthand
sides is only applicable for changes in a single
coefficient.

44
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Limitations of Classical Sensitivity Analysis

Limitation: Changes in constraint coefficients

• Classical sensitivity analysis provides no
information about changes resulting from a
change in the coefficient of a variable in a
constraint. Instead, we must simply change the
coefficient and rerun the model.

45
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Limitations of Classical Sensitivity Analysis

Limitation: Nonintuitive Dual Values

• Constraints with variables naturally on both the
left-hand and right-hand sides often lead to dual
values that have a nonintuitve explanation.

46
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 End of Presentation: Chapter 3

47
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.