Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

Chapter -1

1. Define decision modeling. What are some of the organizations that support the use of the scientific
approach?
Decision Modeling is well known as a scientific approach to the managerial decision making. In
other words, it can be defined as the development of a model (usually mathematical) of a real-world
problem scenario or environment. Resulting model should typically be a decision – making process is not
affected by personal bias, whim, emotions and guesswork. It also helps to provide insights into the
solution of the managerial problem. Decision modeling is also commonly referred as quantitative
analysis, managerial science or operations research.
Organizations such as American Airlines, United Airlines, Google, UPS, FedEx and IBM frequently
use decision modeling to help solve complex problems.

2. What is the difference between deterministic and probabilistic models? Give several examples of
each type of model.
Deterministic model: An assumption made that all the relevant input data and parameters are known
with certainty is known as deterministic model. Best example for this type of model is Dell Corporation,
which makes several types of PC products like desktops and laptops, all of which compete for the same
resources such as labor, hard disks, chips, working capital. Dell knows the specific amounts of each
resource required to make one unit of each type of PC, based on the PC’s design specifications. Further,
based on the expected selling price and cost prices of various resources, Dell knows the expected profit
contribution per unit of each type of PC. In such an environment, if Dell decides on a specific production
plan, it is a simple task to compute the quantity required of each resource to satisfy that production plan.
For example, if Dell plans to ship 50,000 units of a specific laptop model, and each unit includes a pair of
2.0GB SDRAM memory chips, then Dell will need 100,000 units of these memory chips. Likewise, it is
easy to compute the total profit that will be realized by this production plan assuming that Dell can sell all
the laptops it makes.
Probabilistic Model: Unlike deterministic model, this one assumes that some of the input data are not
known with certainty. Best examples that suit this model are Google, Facebook and Twitter etc., where
these firms have yielded great rewards to their investors. A student making their decision in selecting
Majors while entering college also serves as another example for this model. Clearly, there is a great deal
of uncertainty regarding several issues in this decision-making problem: the student’s aptitude for a
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

specific major, his or her actual performance in that major, the employment situation in that major in four
years, etc. Nevertheless, a student must choose a major early in his or her college career.

3. What are the differences between quantitative and qualitative factors that may be present in a
decision model?
Data is the main requirement for any decision modeling process just like a raw material for a factory. The
process of raw data manipulation into meaningful information is considered as the heart of decision
modeling. In dealing with a decision-making problem, managers may have to consider both qualitative
and quantitative factors. For example, suppose we are considering several different investment
alternatives, such as certificates of deposit, the stock market, and real estate. We can use quantitative
factors such as rates of return, financial ratios, and cash flows in our decision model to guide our ultimate
decision. In addition to these factors, we may also wish to consider qualitative factors such as pending
state and federal legislation, new technology breakthroughs, and the outcome of an upcoming election. It
can be difficult to quantify these qualitative factors.
Decision making process may have different roles of quantitative decision modeling due to the presence
of qualitative factors. The result of decision model may automate the decision making process due to lack
of qualitative factors. For example, some companies use quantitative inventory models to determine
automatically when to order additional new materials and how much to order. In most cases, decision
modeling is an aid to the decision-making process. The results of decision modeling should be combined
with other qualitative information while making decisions in practice.

5. What steps are involved in the decision modeling process? Give several examples of this process.
Decision modeling process has three distinct steps involved apart from the size and complexity
of the decision- making problem at hand. They are described as follows
1. Formulation: During this process, each aspect of the problem is translated and expressed in
terms of a mathematical model. This is a completely challenging step due to its results of a poorly
formulated problem will almost surely be incorrect. Main goal of the formulation is to ensure that
mathematical model completely addresses issues relevant to the problem at hand. It is further
classified into 3 sub categories
a. Defining the problem: This helps in developing a clear concise statement of the problem.
b. Developing a Model: Once the selected problem is ready to analyze, though we are not
aware of the issue, these models are being used at your daily routines.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

c. Acquiring Input Data: After developing a model, an input data is required in the model.
Accurate data is required to stop the model resulting in any wrong results caused by improper
data. This situation is called as Garbage in, Garbage out (GIGO).
2. Solution: This is a mathematical expressions resulting from the formulation process are actually
solved to identify the optimal solution. The Solution step is further classified into two categories.
a. Developing a solution: This involves manipulating the model to arrive at the best
solution to the problem. For some issues, we try all possible values for the variables in
the model to get best decision which is known as complete enumeration. Apart from the
approach used, accuracy of the solution depends on the accuracy of input data and
decision model itself.
3. Testing the solution: Prior to analyze and implement a solution, it must be tested completely.
Because solution depends on input data and the model which requires testing. There are multiple
ways to test data, one of them is to get data from different sources and test.
a. Interpretation and Sensitivity Analysis: Assuming that the formulation is correct and
has been successfully implemented and solved, the decision maker’s expertise is called
upon because it is up to him or her to recognize the implications of the results that are
presented. The Solution step is further classified into two categories.
b. Analyzing the Results and Sensitivity Analysis: Analyzing the results starts with
determining the implications of the solution. In most cases, a solution to a problem will
result in some kind of action or change in the way an organization is operating. The
implications of these actions or changes must be determined and analyzed before the
results are implemented. This type of analysis is called sensitivity, or what-if analysis.
Sensitivity analysis is used to determine how much the solution will change if there are
changes in the model or the input data.
c. Implementing the Results: The final part of interpretation is to implement the results.
This can be much more difficult than one might imagine. Even if the optimal solution
will result in millions of dollars in additional profits, if managers resist the new solution,
the model is of no value. Experience has shown that a large number of decision modeling
teams have failed in their efforts because they have failed to implement a good, workable
solution properly.
A few examples of changes that might require an analysis to be modified.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

1. Tax Computation: To help calculate this tax, the Millers would like to set up a spreadsheet-
based decision model. Assume that they have the following information available:
 Their only source of income is from their jobs.
 They would like to put away 5% of their total income in a retirement account, up to a
maximum of $6,000. Any amount they put in that account can be deducted from their total
income for tax purposes.
 They are entitled to a personal exemption of $3,700 each. This means that they can deduct
$7,400 (_ 2 _ $3,700) from their total income for tax purposes.
 The standard deduction for married couples filing taxes jointly this year is $11,600. This
means that $11,600 of their income is free from any taxes and can be deducted from their
total income.
 They do not anticipate having any other deductions from their income for tax purposes.
 The tax brackets for this year are 10% for the first $17,000 of taxable income, 15% between
$17,001 and $69,000 and 25% between $69,001 and $139,350. The Millers don’t believe that
tax brackets beyond $139,350 are relevant for them this year.
2. Excel Notes: In each of our Excel layouts, for clarity, we color code the cells as follows:
 Variable input cells, in which we enter specific values for the variables in the problem, are
shaded yellow.
 Output cells, which show the results of our analysis, are shaded green.
 We have used callouts to annotate the screenshots in this textbook to highlight important
issues in the decision model.
 Wherever necessary, many of these callouts are also included as comments in the Excel files
themselves, making it easier for you to understand the logic behind each model.

6. Why is it important to have an iterative process between the steps of the decision modeling
approach?
Iterative process plays a key role in between the three steps of decision modeling to obtain final solution.
For example, testing the solution might reveal that the model is incomplete or that some of the input data
are being measured incorrectly. This means that the formulation needs to be revised. This, in turn, causes
all the subsequent steps to be changed.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

7. Briefly trace the history of decision modeling. What happened to the development of decision
modeling during World War II?
It is a known fact that Decision modeling was in practice since the beginning of the history, but,
in 1990, Frederick W. Taylor was the one who pioneered the principles of scientific approach to the
management. Several scientific and quantitative techniques were developed during World War II to help
military. These new developments were so successful that after World War II, many companies started
using similar techniques in managerial decision making and planning. Today, many organizations employ
a staff of operations research or management science personnel or consultants to apply the principles of
scientific management to problems and opportunities. The terms management science, operations
research, and quantitative analysis can be used interchangeably, though here we use decision modeling.

10. Define decision variable. Give some examples of variables in a decision model.
A measurable quantity that may vary or which is subject to change is referred as Variable. It can be a
either controllable or uncontrollable. A controllable variable is known as decision variable. Suitable
example for that is to decide number of inventories to be ordered.

11. What is a problem parameter? Give some examples of parameters in a decision model.
A measurable quantity that is inherent in the problem is known as problem parameter such as cost of
placing an order for inventory items. Generally, variables are unknown quantities and parameters which
also referred to input data are known quantities.

12. List some advantages of using spreadsheets for decision modeling.

Due to increased usage of technology in modern times, Computers are playing key role in the decision
modeling process in today’s business environments. Prior to early 90’s, several modeling techniques
needed specialized software packages in order to be solved using a computer. Spreadsheet packages such
as Microsoft Excel have become increasingly capable of setting up and solving most of the decision
modeling techniques commonly used in practical situations. For this reason, the current trend in many
college courses on decision modeling focuses on spreadsheet-based instruction. To continue with this
trend, we discuss the role and usage of spreadsheets during our study of different decision modeling
techniques mentioned here.
In addition to discussing the use of some of Excel’s built-in functions and procedures such as
Goal Seek, Data Table and Chart Wizard etc., we also discuss several add-INS for Excel.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

14. Describe the use of sensitivity analysis, or post optimality analysis, in analyzing the results of
decision models.
Result analysis always starts by determining the implications of the solution. Generally, a solution to a
problem will result in some actions or change in the way an organization is operating. The implications of
these actions or changes must be determined and analyzed before the results are implemented. Because a
model is only an approximation of reality, the sensitivity of the solution to changes in the model and input
data is an important part of analyzing the results. This type of analysis is called sensitivity, post
optimality, or what-if analysis. Sensitivity analysis is used to determine how much the solution will
change if there are changes in the model or the input data. When the optimal solution is very sensitive to
changes in the input data and the model specifications, additional testing must be performed to make sure
the model and input data are accurate and valid.
The importance of sensitivity analysis cannot be overemphasized, because input data may not always be
accurate or model assumptions may not be completely appropriate, sensitivity analysis may become an
important part of decision modeling.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

Chapter – 2

1. It is important to understand the assumptions underlying the use of any quantitative analysis model.
What are the assumptions and requirements for an LP model to be formulated and used?
Technically, there are four additional requirements of an LP problem of which you should be aware:

1. We assume that conditions of certainty exist. That is, numbers used in the objective function and
constraints are known with certainty and do not change during the period being studied.
2. We also assume that proportionality exists in the objective function and constraints. This means
that if production of 1 unit of a product uses 3 hours of a particular resource, then making 10
units of that product uses 30 hours of the resource.
3. The third assumption deals with additivity, meaning that the total of all activities equals the sum
of the individual activities. For example, if an objective is to maximize profit = $8 per unit of the
first product made plus $3 per unit of the second product made, and if 1 unit of each product is
actually produced, the profit contributions of $8 and $3 must add up to produce a sum of $11.
4. We make the divisibility assumption that solutions need not necessarily be in whole numbers.
That is, they may take any fractional value. If a fraction of a product cannot be produced, an
integer programming problem exists.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

4. Under what condition is it possible for an LP problem to have an unbounded solution?

During LP Problem solving process, four special situations might be encountered. They are
(1) Redundant constraints,
(2) Infeasibility,
(3) Alternate optimal solutions and
(4) Unbounded solutions.

A fine solution is not possible when a problem has an unbounded feasible region. When the
solution is unbounded in a maximization problem, the objective function value can be made infinitely
large without violating any constraints. When a LP model has a bounded feasible region, as in the Flair
Furniture example, it has an identifiable optimal corner point solution.
As per Holiday Meal Turkey Ranch example, if the feasible region is unbounded in one or more
directions, then the model may or may not have a fine solution. In the Holiday Meal problem, for
example, we were able to identify a finite solution because the optimal corner point existed on the
bounded side. Now suppose that instead of minimizing cost, the owner of the ranch wants to use a
different objective function. Specifically, based on his experience with the feeds and their fattening
impact on his turkeys, assume that the owner estimates that brand A feed yields a “fattening value” of 8
per pound, while brand B feed yields a fattening value of 12 per pound. The owner wants to find the diet
that maximizes the total fattening value. The objective function now changes from “Minimize $0.10A +
$0.15B” to “Maximize 8A + 12B.” Because this is now a maximization problem and the feasible region is
unbounded in the direction in which profit increases, the solution itself is unbounded. That is, the profit
can be made infinitely large without violating any constraints. In real-world situations, the occurrence of
an unbounded solution usually means the problem has been formulated improperly. That is, either one or
more constraints have the wrong sign or values, or some constraints have been overlooked. In Holiday
Meal’s case, it would indeed be wonderful to achieve an infinite fattening value, but that would have
serious adverse implications for the amount of feed that the turkeys must eat each month.

6. The production manager of a large Cincinnati manufacturing firm once made the statement, “I
would like to use LP, but it’s a technique that operates under conditions of certainty. My plant doesn’t
have that certainty; it’s a world of uncertainty. So LP can’t be used here.” Do you think this statement
has any merit? Explain why the manager may have said it.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

LP has been extensively applicable in all the areas such as transportation, operations, financial,
marketing, accounting, and human resources, medical and agricultural problems. Regardless of the size
and complexity of the decision-making problem at hand in these diverse applications, the development of
all LP models can be viewed in terms of the three distinct steps:
 Formulation
 Solution
 Interpretation
The manager’s decision for not including the LP programming was because it does not have
enough merit as in any company or a firm there is a minimum or moderate amount of uncertainty. In
order to make calculated risks, or make qualitative decisions we would need linear programming, this has
been the approach of many manufacturing companies and have been successfully implemented in
aeronautical industry, manufacturing industry and so on.

7. The mathematical relationships that follow were formulated by an operations research analyst at the
Smith-Lawton Chemical Company. Which ones are invalid for use in an LP problem? Why? Maximize
profit = 4X1 + 3X1X2 + 8X2 + 5X3 subject to the constraints
2X1 + X2 + 2X3 … 50
X1 - 4X2 Ú 6
1.5X21
+ 6X2 + 3X3 Ú 21
19X2 - 0.33X3 = 17
5X1 + 4X2 + 32X3 … 80

The equations 3 and 5 are opposing the Fourth LP property: Mathematical relationships are linear, which
explains that equations should be linear,

The objective and constraints in LP problems must be expressed in terms of linear equations or
inequalities. In linear mathematical relationships, all terms used in the objective function and constraints
are of the first degree. Hence, the equation 2A + 5B = 10 is a valid linear function, whereas the equation
2A^2 + 5B^3 + AB = 10 is not linear because the variable A is squared, the variable B is cubed, and the
two variables appear as a product in the third term.
Assignment -1 IT – 608 Managerial Decision Modeling Durga Lakshman Bhaganagarapu

321996

As per the above explanation contradicts to the linear equations concept these 2 equations which have this
functionality are termed as invalid.