Lesson 3.

1: Alternatives and Models in Decision

Making

Lesson Summary
This lesson covers the alternatives, models, decision theory, multiple criteria,
the decision evaluation display, and decisions under risk and uncertainty.

Learning Outcomes
At the end of the lesson, the students should be able to:
1. Explain this statement that a final and complete alternative seldom appears
in its final state.
2. Differentiate the limiting and strategic factors.
3. Discuss the various meaning of the word model.
4. Describe the difference between mathematical models and traditionally used
in physical science engaged in decision making.
5. Explain the general form of the evaluation function for money flow modeling
and its symbols.
6. Clarify why it is not possible to formulate a model that accurately represents
reality.
7. Apply the multiple-criteria decision in an experience-based situation and the
concepts of the three or five most important criteria.
8. Discuss the level to which the criteria in a multiple-criteria decision situation
are genuinely independent.
9. Evaluate the use of paired comparisons to rank the criteria in order based on
its importance.
10.Determine the best alternative of the following conditions: decision under
assumed certainty, decision making under risk, and decision making under
uncertainty.

Motivation questions
Imagine how the enrollment system of VSU-Isabel works? What do you think the
processes used in the said system?

Discussion
ALTERNATIVE IN DECISION MAKING
 A complete and more comprehensive alternative seldom appears in its final
states. In its final form, an alternative should consist of a thorough description of its
objectives and its requirements in terms of benefits and costs.
The term alternative encompasses different outcomes and methods. These
alternatives are proposed for analysis even though there seems to be little
likelihood that they will prove feasible but not necessarily attainable. The
impression is that it is better to consider many alternatives than to overlook that
might be ideal. Alternatives that are not considered cannot be adopted, no matter
how desirable they may be.

Limiting and Strategic factors

Limiting factors are those factors that stand in the way of achieving
objectives. A vital element of the system engineering process is identifying the
limiting factors restricting the accomplishment of the desired objectives. Those
identified limiting factors are examined to locate strategic factors. Strategic factors
are those factors that can be altered to make progress possible.
Identifying these factors is essential for it allows the decision-maker to
concentrate effort on those areas in which success is attainable. This means that
will achieve the desired objective may consist of the procedure, a technical process,
or mechanical, organizational, or managerial change.
A vital element of defining alternatives is the identification of the limiting
factors restricting the achievement of the desired objective.
Comparing Alternative Equivalently
Alternatives must be translated to a standard measure in order to compare
alternatives equivalently. This conversion to common measures permits comparison
based on equivalence. Models and optimization are necessary for the conversion
process and permit alternatives to be compared equivalently.

MODELS IN DECISION MAKING

Models and the process of the simulation are useful tools in systems analysis.
A model may be used to represent a system to be brought into being or to analyze
a system already in being. This is essential when it is impossible to manipulate
reality because the system is not yet in existence, or when manipulation is
expensive and disruptive as with complex industrial systems.
There is a ultimate difference between models used in science and
engineering. Science is concerned with the natural world, whereas engineering is
concerned primarily with the human-made world. Models are used in Science to
gain an understanding of the way things are in the natural world. Engineering uses
models of the human-made world in an attempt to achieve what ought to be. The
validated model of science is used in engineering to establish bounds for
engineering creations and improve such creations' products.
Classification of Models
 The model implies representation. The model is designed to represent a system
under study, an idealized example of reality, to explain the essential relationship.
They can be classified as:
a) Physical model – looks like what they represent, e.g., globe
b) Analog/Analogue model – behaves like the original, e.g., electric circuit
c) Schematic model – graphically describes a situation and process, e.g.,
organizational chart
d) Mathematical model – symbolically represents the principles of a situation
being studied, e.g., linear programming model, systems dynamic models,
inventory model, etc.
Models and Indirect Experimentation
Models and manipulation provide convenient means of gathering factual
information about the system being designed. The central part of the design
process requires a decision based on a system model rather than decisions derived
from the system itself.
Direct experimentation is an experiment where an object, state, or event and
environment are repeatedly manipulated until the desired results are achieved. For
example, you want to have a new set up of your living room, essentially arranging
the furniture until the desired outcome is observed. This experimentation can be
applied in a manufacturing setting on rearranging the equipment of the factory.
However, this procedure is time-consuming, disruptive, and expensive. With this,
simulation or indirect experimentation is employed.
Indirect experimentation (Simulation) enables the systems analyst to evaluate
the probable outcome of given experimentation without changing the operating
system. This makes it possible to determine how changes in those aspects of the
system under the decision-maker's control affect the modeled system. The primary
use of systems engineering simulation is to explore the effects of alternative system
characteristics on a system performance without actually producing and testing
each candidate system.
Models are useful in design and operations because they take the decision-
maker part way to the decision point. These are for design decisions, and
operational decision-making is an abstraction of the system under-investigated.
Several assumptions involve in models. These are assumptions about the operating
characteristics of components, the behavior of people, and the environment's
nature.
It manipulation can lead to a model modification to reduce the misfit between
the model and the real setting. This validation process has a recurrent pattern
analogous to the scientific method. The concepts consist of three steps:
1. Postulate a model
2. Test the model prediction
3. Modification of model
The validation process should continue until the model is supported reasonably by
evidence from measurement and observation.

DECISION EVALUATION THEORY

In systems engineering and analysis, decision evaluation is a vital part. It is needed
as a basis for choice among alternatives that arise from design activities and
optimizing systems already in operation.
This section presents two general categories of decision evaluation models: the
money flow modeling and economic optimization modeling.

Evaluation by Money Flow Modeling

Under this approach on evaluating a system, the economic equivalence is
expressed as the present equivalent, annual equivalent, or future equivalent
amount, as well as the internal rate-of-return and the payback period. Each of these
will be discussed in detail in the next lesson.
As a general equivalence function subsuming each equivalence expression may be
stated as
PE , AE ,∨FE=f ( F t , i, n ) −eq .1
Where t=0 , 1 ,2 , … , n∧where
F t= positive r=¿∨negative money flow at the end of the year
i=annual rate of interest
n=number of years
The product life cycle is the underlying money flow generator over the acquisition
and utilization phases, as shown in figure 13.
 Figure 1. Money flow for two modes of acquisition/utilization

Evaluation by Economic Optimization Modeling

Decision evaluation often requires a combination of both money flow modeling and
economic optimization. In determining economic equivalence, it is essential to
optimize those variables such as investment costs, recurring costs, or project life
function decision variables. This optimization is linked to decision evaluation
through one or more money flows. It requires that an evaluation measure be
derived from an economic optimization model.

An economic optimization function is a mathematical model formally linking an

evaluation measure, E, with controllable decision variables, X, and system
parameters, Y, which cannot be directly controlled by the decision-maker. It is
expressed as
E=f ( X , Y )−eq . 2
The optimization function may be extended to both operational and design
decisions involving alternative. The extension involves identifying and isolating
design-decision-dependent system parameters, Yd. from design-or decision-
independent system parameters, Yi. This can be denoted in an equation
E=f ( X , Y d , Y i ) −eq . 3
This equation three can be applied to the decision evaluation function for
procurement and inventory systems to meet the demand for an item available for
one of the various sources.
The design-dependent parameter approach facilitates the life cycle complete
evaluation of alternative system designs. This approach involves the separation of
the system design space. Terms are defined as follows:
1. Design-dependent parameters (Yd.) – are factors with values under the
designer(s) and that are impacted by the specialty disciplines during the
development process.
2. Design-independent parameters (Yi) – these factors beyond the designer(s)
but that impact the effectiveness of all candidate systems or design
alternatives, can significantly alter their goodness or desirability.
3. Design variable (X) – these factors that define the design optimization space.
A summary of the evolution and development of the decision evaluation function
is tabulated in table 1 with citing the references. The succeeding lesson models
are developed under the decision evaluation functions presented in this lesson.
 Table 1. The evolution of forms of the decision evaluation functions

References Functional Form Application

Churchman, E=f ( x i , y i ) Operation
Ackoff and Arnoff
(1957)
E=¿ system effectiveness
x i = variables under direct
control
y i = variables not subject to
direct control
Fabrycky, Ghare, E=f ( X , Y ) ; g ( X ) ¿ B Operations
and Torgersen ¿
(1984)
E=¿ evaluation measure
X = controllable variables
Y = uncontrollable variables
Banks and E=f ( X , Y d , Y i ) ; g ( X , Y d ) ¿ C Procurement
Fabrycky (1987) ¿ operation
E=¿ evaluation measure
X = procurement level and
procurement quantity
Y d = source-dependent
parameters
Y i = source-independent
parameters
 Fabrycky and E=f ( X , Y d , Y i ) ; g ( X , Y d ) ¿ C Design optimization
Blanchard (1991) ¿

E=¿ evaluation measure

X = design variables
Y d = source-dependent
parameters
Y i = source-independent
parameters
The role and where it can be applied to equations 2 and 3 differ in system
analysis and design evaluation. If we consider in terms of the life cycle, then it can
be applied to a significant segment, as shown in Figure 14. Equation 2 can be
applied to the optimization of operation already exist. In contrast, Equation 3 is
useful for choosing among mutually exclusive design alternatives based on design-
dependent parameters.

Figure 2. Economic optimization in design and operations.

In systems engineering, design-dependent parameters are invoked. As you

can see on the fourth entry of Table 1 and Figure 14, particular on the design phase,
the source-dependent parameters are explicitly partitioned. The values of these
parameters are TPMs. They make it possible to distinguish the difference between
mutually exclusive design alternatives.

DECISIONS INVOLVING MULTIPLE CRITERIA

This section presents some formal approached for dealing with multiple
criteria that incorporate both qualitative knowledge and judgment.
Multiple Criteria Considerations
 This method ensures recognizing and considering the number of useful
concepts before developing and presenting some formal and structured methods for
choosing among multiple alternatives.
Methods and approaches to follow for the decision-maker when multiple criteria
are present.
 Selecting the criteria to be considered.
It is essential that the identified, selected, and characterized alternatives
were already being compared on importance criteria. Moreover, the process
of identification and selection is always a good step in meeting toward a
choice of the best alternative. Note that the consideration in the selection of
criteria is that they are independent of each other.
 Differences are the basis for decisions.
Decisions between alternatives should be made only based on their actual
difference across all criteria. Only actual differences are essential in decision
making. All identical attributes and of the same sign may be canceled out.
However, make sure that the canceled attributes are genuinely identical and
significance.
Direct Ranking Methods
This approach stated that the decision-maker must present all chosen
alternatives for ranking. Ranking methods can be applied to criteria /alternatives.
This can be applied one at a time since it is difficult to apply both simultaneously.
A process that may make the task of ranking easier and more effective is
referred to as the method of paired comparison. This method suggests the
submission of criteria and alternatives to the decision-maker two at a time. It also
provides a check on the internal consistency of value judgments provided by the
decision-maker.
For illustration, consider five essential criteria that pertain to the design of
office equipment. The criteria, designated 1-5, are to be rank-ordered: (1) better, (2)
cheaper, (3) faster, (4) repairable, and (5) disposable. These criteria are to be used
in a succeeding analysis to determine the best mutually exclusive design for
equipment.
When there are N criteria/alternatives to be ranked, N(N-1)/2 pairs must be
compared. For the five equipment design criteria, assume that the results of the
paired comparisons are as follows (where > means “preferred to” and = means
“equally preferred to”):
1<2 2>3 3>4 4>5
1>3 2>4 3=5
1>4 2>5
1=5
Table 2. The exhibit of Preference Comparison

Criteria 1 2 3 4 5 Times
Preferred
1 - P P = 2+
 2 P - P P P 4
3 - P = 1+
4 - P 1
5 = - ++
Note: the “+” sign denoted on the criteria having “=” sign.
A display of the pairwise comparison is given in Table 2. A P is entered for
each pair where the row criterion is preferred to the column criterion. There should
not be any diagonal entries in that the given criterion cannot be compared with
itself. To verify that all pairs of criteria have been considered, check to see if there
is an entry for each pair with two entries being made for equally ranked pairs.
The number of times a given criterion in each row is preferred in the right-
hand column of Table 2. A simple rule of ranking is to count the number of times
one criterion is preferred over others. For these examples, the rank order of
preference is 2>1>3>4=5. It means that Cheaper is most important, with Better-
ranked second. Faster is ranked ahead to Repairable and Disposal. These criteria
are ranked in the last place and considered approximately equally preferred.
This approach presented assume the transitivity of preferences. That is, if
1>2 and 2>3, then 1 must be >3. Suppose two or more criteria are preferred an
equal number of times (except for ties). In that case, there is evidence of a lack of
consistency. However, there is no evidence of a lack of consistency in this
illustration. If there were, the remedy would be to question and reconsider the
preference choices pertaining to the affected criteria.
Systematic Elimination Methods
These are among the most straightforward approaches available for choosing
from among alternatives in the face of multiple criteria. These methods are useful
when values and outcomes can be specified for all criteria and all alternatives. The
values should be measurable or at least ordinal.
There are two limitations of elimination methods: (1) they do not consider
weights that might be applicable to the criteria or attributes, and (2) they are
noncompensatory in that they do not consider possible trade-offs among the criteria
across alternatives.
An example will be used to illustrate three systematic elimination approaches
(comparing alternatives against each other, comparing alternatives against
standards, and comparing criteria across alternatives). Consider alternative A-D in
Table 3, with each having an estimated scalar or ordinal outcome. The right-hand
columns specify the ideal and minimum standard value for each criterion.
Table 3. Estimated criterion values for alternatives

Alternative

A B C D Idea Minimum Standard

Criterion
l
1. Better 40 35 50 30 50 30
2. 90 80 75 60 100 70
Cheaper
3. Faster 6 5 8 6 10 7
 4. Gone G P VG E E F
In the table above, criteria 1-3 are estimated in scalar terms with higher
numbers being better. Criterion 4 is rank-ordered with E being excellent, VG being
excellent, G being good, F being fair, and P being poor.
Three systematic elimination methods are:
Comparing alternatives against each other: the most apparent
elimination method is to check for dominance by making mutual comparisons
across alternatives. Dominated alternatives can be safely eliminated from further
consideration. If an alternative is better than or equal to other alternatives across all
criteria, then the alternative is dominated.
To illustrate, refer to Table 3, as you can see the Alternative A dominates
Alternative B. This is due to 40>30 for Better, 90>80 for Cheaper, 6>5 for Faster,
and G>P for Gone. No other pair can be found in which dominance occurs.
Dominance is the only elimination technique. It will not be necessary to lead
to the selection of a preferred alternative unless other criteria are applied.
Comparing alternatives against a standard: comparisons against the
standard (also called satisficing) are commonplace and easy to explain. These
comparison methods may be applied at one of two extremes, or at any point in
between. Extremes provide the following rules:
Rule 1: An alternative may be retained (not eliminated) only if it meets the
standard for at least one criterion.
Rule 2: An alternative may be retained (not eliminated) only if it meets the
standard for all criteria.
Its application of these rules and variants may be accomplished by setting up
a table to show the derived results, as shown in Table 4. An X is used to indicate the
minimum criterion values that are violated by each alternative. Just refer to Table 3
for the data given.
Table 4. Alternatives not meeting criteria standards

Criterion

1 2 3 4
Alternativ
e
A X
B X X
C
D X X
A thorough inspection of the result in the tabulated data leads to the following
conclusions about the alternatives under consideration:
1. All alternatives meet the standards for at least one criterion, and under Rule
1, all may be retained for further evaluation.
2. Alternative C meets the standard for all criteria, and under Rule 2, it is the
only one that may be retained.
 Comparing criteria across alternatives: comparisons across alternatives can
be made in two ways after relevant criteria are ranked based on its importance:
1. For the most important criterion, choose the alternative (if any) that is
best. A tie between two or more alternatives is broken by using the
second important criterion as a basis for your choice. Continue until an
available alternative survives or until all criteria have been considered.
2. Examine one criterion at a time, comparisons among the alternatives.
Eliminate alternatives that do not meet the minimum standard value.
Continue until all alternatives except one have been eliminated, or until all
criteria have been examined.
For example, assume that the importance ranking for criteria is 2>1>3>4, as
determined by the direct ranking method. Using the minimum standards in Table 3.
The comparison rules give the following findings:
1. Criterion2 (Cheaper) is most essential, and Alternative B is best when
evaluated on this criterion alone. However, this alternative does not meet the
minimum standard for one of the other criteria (Faster).
2. Examination of one criterion at a time and eliminating those that do not meet
minimum standard results in the following:
Criteri Elimina Remaini
on te ng
1 None A,B,C,D
2 D A,B,C
3 A, B, D C
4 None C
For the above-mentioned method, under rule 1, Alternative A could be
chosen of violation of the minimum for Criterion 3 (faster) could be overlooked.
Otherwise, Alternative C could be chosen as determined by the application of the
second rule.
Weighting methods of evaluation
This refers to consider direct and subjective techniques for assigning weights
or importance ratings to criteria. There two approaches in weighting methods
evaluation: the tabular additive and graphical additive. These techniques are useful
aids to the selection of the preferred alternative that is based on a combination of
criterion weights and outcome ratings of the alternatives.
Weighting of criteria or attributes. Make sure to apply the resulting
weights for criteria for all alternatives on whatever conditions. Then, it is essential
that the sum of the resulting weights to 1.00 or 100% for each criterion.
For the illustration, assume that Better is weighted 35%, Cheaper is weighted
50%, and faster is weighted 15%, with the total weights of 100%. If the weight of
one criterion is changed, the others must be changed too. Therefore, directly
assigning weights may be thought of as the same as allocating 100 points to the
criteria under consideration. These weights criteria will apply to all design
alternatives.
 The tabular additive method
 To further illustrate this approach, refer to Table 5 from the prior section. It
identifies criteria, weights (W), the performance rating (R) of each alternative
over each criterion, the product of weights and rating (W x R), and the
aggregate totals of W X R. these aggregate totals may be used in comparing
alternatives.
Table 5. Tabular additive method computation

Alternative A Alternative B
Criterion Weights Rating (R) WxR Rating (R) WxR
(W)
Better 0.35 6 2.10 7 2.45
Cheaper 0.45 10 4.50 6 2.70
Faster 0.20 5 1.00 3 0.60
1.00 7.60 5.75

 The graphical additive method

A graphical version of the information given in Table 5 is shown in the figure
below. This is a stacked bar chart, with the total height of each bar
corresponding to the total score for each alternative.
 Figure 3. Bar chart for weighted evaluation

THE DECISION EVALUATION DISPLAY

This is based on the premise that differences between alternatives and the
degree of compliance with multiple criteria are appropriately displayed. It is also
recommended as a means for simultaneously exhibiting the difference that multiple
alternatives create in the face of multiple criteria.
The general form and basic structure of the decision evaluation displays are
shown in Figure 16. It is one way of simultaneously displaying the elements
pertaining to one or more mutually exclusive alternatives in the face of economic
and non-economic criteria. These elements are as follows:
 Figure 4. General decision evaluation display

1. Alternatives ( A, B, C) – two or more alternatives appear as vertical lines

in the field of the decision evaluation display.
2. Equivalent cost or profit – the horizontal axis represents present equivalent,
annual equivalent, or future equivalent cost or profit.
3. Other criteria (X, Y, Z) – vertical axes on the left represents one or more
criteria, usually of a noneconomic mature.
4. Other criteria thresholds – horizontal lines emanating from the vertical axes
represent threshold or limiting values for noneconomic criteria.
5. Predicted and estimated values – anticipated outcomes for each alternative
are entered in circles placed above, on, or below the thresholds.

DECISION UNDER RISK AND UNCERTAINTY

In this part, a decision evaluation matrix is introduced to accommodate
decisions under assumed certainty, risk, and uncertainty. Then each of these
decision situations is treated separately, with emphasis on risk and uncertainty.
The Decision evaluation matrix
A decision evaluation matrix, as presented in Figure 17, is a formal way of
displaying the interaction of a finite set of alternatives and a finite set of possible
futures (state of nature), which means the courses of action from among which a
decision-maker expects to choose. The states of nature are not pertaining to natural
events such as rains or snow, but are a wide variety of future outcomes over which
the decision-maker has no direct control.
The general evaluation matrix is a model depicting the positive and negative
results that may occur for each alternative under each possible future. Its symbols
are defined as follows:
Ai = an alternative available for selection by the decision-maker, where
i =1, 2,.,n
F j= a future not under the control of the decision-maker, where j =1,2,..n
 P j= the probability that the jth future will occur, where j =1,2,..n
Eij = evaluation measure (+/-) associated with the ith alternative and the jth
future determined from the equation1 or 3.

Figure 5. The decision evaluation matrix

Various assumptions underlying the application of this decision evaluation

matrix model to decision making under assumed certainty, risk, and uncertainty.
Primarily is that all alternatives have been considered, and all possible futures have
been identified.
Evaluation measures in the matrix model are associated with the results that
may be objective or subjective. For example, the payoffs may be profit expressed in
dollars, yield expressed in pounds, cost expressed in dollars, or other undesirable
measures. Subjective outcomes, conversely, are those that are valued on an ordinal
or ranking scale. Examples are expressions of preference, such as an excellent
corporate image being preferred to a low image, higher quality outputs being
preferred to those of lower quality, and so forth.
Other assumptions of importance in the evaluation matrix representation of
decisions:
1. The occurrence of one future precludes the occurrence of any other future
(futures are mutually exclusive).
2. The occurrence of a specific future is not influenced by the alternative
selected.
3. The occurrence of a specific future is not known with certainty, even though
certainty is often assumed for analysis purposes.
The decision under assumed certainty
Boyle’s law, Ohm’s law, and other physical laws can be applied with a higher
degree of certainty to specific instances. They are supplemented by many models
for physical phenomena that enable conclusions to be reached about the physical
environment that matches the facts with narrow limits.
Nonetheless, the primary aim of operations research and management
science is to bring the scientific approach to bear to a maximum feasible extent.
This is done with the help of conceptual simplifications and the model of reality, the
most common being the assumption of a single known future.
 The evaluation matrix for decision making under assumed certainty is a
vector with as many evaluations as there are alternatives, with the outcomes
constituting a single column. The prior figure is a particular case of a matrix of
decision vector. This appears as in succeeding figure with the payoffs represented
by Ei, where i=1, 2,.,m. The single future is assumed to occur with certainty actually
carries a probability of unity in the matrix. All other futures are suppressed by
carrying probabilities of zero (P=0.0).

Figure 6. Decision evaluation vector

When the outcomes, Ei, are stated in monetary terms, the decision rule or
principle of choice is simple. If the alternatives are equal in all other aspects, one
will choose the alternative that minimizes cost profit. In this case, it can be
expressed as
min { E i } for i=1 , 2 ,… . , m
i
For-profit, one would choose
max { E i } for i=1 , 2 , … ., m
i
It is often possible to accept the premise that only the cost or the profit
differences are essential, with intangibles and complex having little or no effect.
Unquantifiable nonmonetary factors may be significant enough to outweigh
calculated costs or profits differences among alternatives. In other terms, the
outcome is not easily expressed in monetary terms, or even in quantitative teems
of some other evaluation measure, such as time, percentage of the market, and
others. Valid quantitative comparisons may be made when the quantitative
outcomes cannot stand alone, and when the outcomes are nonquantitative.
Decision making under risk
Under this decision condition, it specifies that there is a little probability of
anticipating futures that will coincide with actual futures. The physical and
economic elements on which a course of action depends may vary from their
estimated values because of chance causes. The lack of uncertainty about the
future makes decision making one of the most challenging tasks faced by individual,
industry, and government.
Decision making under risk occurs when the decision-maker does not
suppress acknowledged ignorance about the future but make it evident through
assigning of probabilities. These probabilities may be based on experimental
evidence, expert opinion, subjective judgment, or the combination of these. This
approach is also using the concept of dominance.
Suppose a computer system firm has the opportunity to bid on two related
contracts being advertised by a municipality. The first pertains to the selection and
installation of hardware for a central computing facility together with the required
software. The second involves the development of a distributed computing network
involving the selection and installation of hardware and software. The firm may be
awarded either Contract 1 (C1), or contract 2(C2), or both contract (C1 + C2). Thus,
there are three possible futures for this problem, given that the payoff values and
probabilities are shown in Table 6. Assume that the profit level is set to be at least
$400,000.00, and the loss level is set to be no more than $100,000.00.

Table 6. Decision evaluation matrix (profit; in $ 000s)

Probabilit (0.3) (0.2) (0.5)

y:
Future: C1 C2 C1 + C2
A1 100 100 400
A2 -200 150 600
Alternati
ve A3 0 200 500
A4 100 300 200
A5 -400 100 200
Compute the best alternative using the following approaches:
a. Aspiration level criterion
b. Most probable future criterion, and
c. Expected value criterion

Solution:
a. Aspiration level criterion
In this method, some form of aspiration level exists in most personal
and professional decision making. An aspiration level is some desired level of
achieving such as profit, or some undesirable result level to be avoided, such
as loss. This involves selecting some level of achievement that is to be met,
followed by the selection of the alternative, which maximizes the probability
of achieving the stated aspiration level.
For the problem above, the decision-maker must set a minimum
aspiration level for-profit and possibly a maximum aspiration level for loss.
Given the aspiration level of the problem of which profit level is at least
$400,000.00 and the loss level is no more than $100,000.00.
 Therefore, under these aspiration level choices, Alternative A 1 – A3
qualify as to profit potential, but alternative A 2 fails the loss and must be
eliminated. The choice could be now between A 1 and A3 by some other
criterion, even though both satisfy the aspiration level criterion.
b. Most probable future criterion
This approach to decision making suggests that all except the most
probable future be disregarded. Although somewhat equivalent to decision
making under certainty, this criterion works well when the most probable
future has a significantly high probability so as to dominate partially.
Under the most probable future criterion, the computer system firms
would focus its selection process from among four alternatives on the profit
associated with the future designated by C 1 + C2. This is because the
probability of this future occurring is 0.5, the most probable possibility.
Alternative A2 is preferred by this approach. The most probable future
criterion could be applied to select between A 1 and A3, as identified under the
aspiration level criterion. If this is done, the firm will choose alternative A 3.
c. Expected value criterion
The expected outcome criterion is viewed with caution only when the
payoff consequences of possible outcomes are disproportionately large,
making an outcome that deviates from the expected outcome a distinct
possibility.
For the calculation of the expected values requires weighing all payoffs by
their probabilities of occurrence. For the problem above, it is calculated as
follows:
A1=$ 100 ( 0.03 )+ $ 100 ( 0.2 ) + $ 400 ( 0.5 )=$ 250
A2=−$ 200 ( 0.03 )+ $ 150 ( 0.2 )+ $ 600 ( 0.5 )=$ 270
A3 =$ 0 ( 0.03 ) +$ 200 ( 0.2 )+ $ 500 ( 0.5 )=$ 290
A 4=$ 100 ( 0.03 ) + $ 300 ( 0.2 ) +$ 200 ( 0.5 )=$ 190
For analysis, it is clear that alternative A 3 would be selected since it
yields the highest expected value.
For the final analysis, you may compare all the results among the three
approaches.
a. Aspiration level criterion = A1 or A3
b. Most probable future criterion = A2
c. Expected value criterion = A3
Thus, by carefully looking at the results, the best alternative is A3 arising from
the use of the three criteria.
Decision making under uncertainty
Under this situation, decisions are made in a more abstract environment. Several
decision criteria will be applied to the example, as tabulated in Table 6. The formal
approaches that are available are:
a. Laplace criterion
 Referred to the computer system problem, suppose that the said company is
unwilling to assess the futures in terms of probabilities for Contract 1 (C 1),
contract 2(C2), and both contract (C1 + C2). In the absence of these
probabilities, the reason might be that each possible state of nature is as
likely to occur as any other. The rationale of this assumption is that there is
no stated basis for one state of nature to be more likely than any other,
which is referred to as the Laplace principle or the principle of insufficient
reason based on the philosophy that nature is assumed to be different.
Under the Laplace principle, the probability of occurrence of each future state
of nature is assumed to be 1/n, where n is the number of possible future
states. To illustrate this approach, try to look at the Table7.
Table 7. Computation of Average Profit ($ 00s)

Alternati Average Payoff

ve
A1 ($100+$100+$400)÷ 3=$200
A2 (-$200+$150+$600)÷ 3=$183
A3 ($0+$200+$500)÷ 3=$233
A4 ($100+$300+$200)÷ 3=$200
To select the best alternative, choose an alternative that yields the maximum
profit based on the computed arithmetic average. In the above-mentioned
problem, A3 with $ 233,000 would be selected.

b. Maximin and Maximax criteria

Another approach can be applied to dealing with decisions under uncertainty.
The first is the maximin rule, based on an overly pessimistic view of the
outcome of nature. The use of this rule will be justified if it is judged that
nature does its worst. The second one is the maximax rule. This is based on a
too optimistic view of the future. The use of this rule is justified if it is judged
that nature will do its best.
Because of the pessimism embraces the maximin rule, its application will
lead to the alternative that assures the best of the worst outcomes. If Eij is
used to represent the payoff for the ith alternative and the jth state of nature,
the required computation is
max {min Eij }
ij
Table 8. Profit by the Maximin Rule ($ 000s)

Alternati Minimum Eij

ve ($)
A1 100
A2 -200
A3 0
A4 100
The best alternative under this condition is A1 or A4. This assures that the firm
of a profit of at least $100,000 regardless of the future.
 The optimism of the maximax rule is in sharp contrast to the pessimism of
the maximin rule. Its application will choose the alternative that assures the
best of the best possible outcomes. As before, if Eij represents the payoff for
the ith alternative and the jth state of nature, the required computation is
max {max Eij }
ij
Table 9. Profit by the Maximax Rule ($ 000s)

Alternati Maximum Eij

ve ($)
A1 400
A2 600
A3 500
A4 300
With this condition, based on this rule, we will choose alternative A2, having a
profit of $600,000 if the future is benign.
As a decision-maker who chooses the maximin rule only the worst possible
occurrence of alternative and selects the alternative that promises the best
of the worst possible outcomes, as shown in Table 9 where A1 was chosen,
the firm would be assured of a profit of at least $100,000, but could not
receive profit any more significant than $400,000. Alternatively, is A 4 would
be chosen, the firm could not receive a profit not greater than $300,000.
However, if we choose the A 2 having the highest profit offer, the firm faces
the possibility of a loss of $200,000 while seeking a profit of $600,000.
c. Hurwicz criterion
The third approach to decision making under uncertainty involves an index of
relative optimism and pessimism. It is called the Hurwicz rule.
An integration between optimism and pessimism is embraced in the said
approach
max {α ¿ ¿
ij
As an example of the Hurwicz rule, consider the payoff matrix of Table 6 with
α =0.2 . The required computations are shown in Table 10, and Alternative A1 would
be chosen as the firm having the highest computed results.
Table 10. Profit by the Hurwicz rule with α = 0.2 ($ 000s)

Alternati α ¿ ($)
ve
A1 0.2($400)+0.8($100)=$16
0
A2 0.2($600)+0.8(-$200)=-
$40
A3 0.2($500)+0.8(0)=$100
A4 0.2($300)+0.8($100)=$14
0
 To choose the best alternative under decision making under uncertainty, just
do it, what we did in comparing the findings of the different approaches.
To compare, the following results are shown:
Laplace criterion: A3
Maximin criterion: A1 or A4
Maximax criterion: A2
Hurwicz criterion: A1

Examination of the selections recommended by the five decision rule

indicates that each has its own merit. Several factors may influence a decision
maker’s choice of a rule in a given decision situation. The decision maker’s attitude
toward the uncertainty and his personal utility function are essential influences;
thus, the choice of a particular decision rule for a given situation must be based on
subjective judgment.

Learning Task
1. Make a blog/video describing your understanding of how to do the following
methods or alternatives in decision making being applied in a specific
business firm. Make sure to articulate clearly in a simple manner, and you
may cite an example/s for better understanding. Be honest about doing this
activity. You output will be graded according to the following criteria below.

 Criteria for grading on learning task

Criteria 8-10 6-7 (Good) 4-5 (Fair) 1-3 (Needs Ratin
(Excellent ) Improvemen g
t)
Delivery Holds attention Consistent Displays Holds no eye X2
of entire use of direct minimal eye contact with
audience with eye contact with the audience,
the use of contact with the as an entire
direct eye the audience, report is
contact, seldom audience, while read from
looking at notes but reading notes
• Speaks with still returns mostly from • Speaks in
fluctuation in to notes the notes low volume
volume and • Speaks • Speaks in and/
inflection to with uneven or
maintain satisfactory volume monotonous
audience variation of with little or tone,
interest volume and no inflection which causes
and emphasize inflection the audience
key points to
disengage
Content/ Demonstrates Is at ease Is Does not have X4
Organization full with uncomfortab a grasp of
knowledge by expected le with information
answering answers to information and cannot
all class all questions, and is able answer
questions without to questions
with elaboration answer only about
explanations • Has rudimentary subject
 and somewhat questions • Does not
elaboration clear • Attempts clearly define
• Provides clear purpose and to define the subject and
purpose and subject; purpose purpose;
subject; some and subject; provides weak
pertinent examples, provides or no
examples, facts, and weak support of
facts and statistics examples, subject; gives
statistics; that support facts, and/ insufficient
supports the or statistics, support for
conclusions/ide subject; which do not ideas
as includes adequately or conclusions
with evidence some data support the
or evidence subject;
that includes very
supports thin
conclusions data or
evidence
Timeliness of The project Highly Successful Requirements X2
the turned in before successful submitted on are not met,
presentation the deadline +2 submitted on due date but the project
per week due date and not on time submitted late
time. 2 weeks or
more after the
due date
Enthusiasm/ Demonstrates Shows some Shows little Shows no X2
Audience strong enthusiastic or mixed interest in the
Awareness enthusiasm feelings feelings topic
about topic about the about the presented
during the topic topic • Fails to
entire • Raises being increase
presentation audience presented audience
• Significantly understandin • Raises understanding
increases g and audience of
audience awareness of understandin knowledge of
understanding most points g and topic
and knowledge knowledge
of the topic; of some
convinces an points
audience to
recognize the
validity and
importance of
the subject
Comments: Score:

Total 100.0
Possible 0
Highest
Points

Assessment
Answer the following questions based on your understanding of the lesson.
1. Explain this statement “a complete and all-inclusive alternative rarely
emerges in its final state.”
2. Differentiate the limiting and strategic factors.
3. Discuss the various meaning of the word model.
4. How do mathematical models directed to decision situations differ from those
traditionally used in physical science?
5. Write the general form of the evaluation function for money flow modeling
and define its symbols.
6. Why is it not possible to formulate a model that accurately represents reality?
7. Identify a multiple-criteria decision situation with which you have experience.
Select three or five most important criteria.
8. Discuss the degree to which you think the criteria you selected in question 7
are genuinely independent. Weight each criterion, check for consistency, and
then normalize the weight so that the total sums to 1.00
9. Use the method of paired comparisons to rank the criteria in order of
decreasing importance.
10.The values for three alternatives considered against four criteria are given
(with higher values being better). What can you conclude using the following
systematic elimination methods?
a. Comparing the alternative against each other (dominance)
b. Comparing the alternatives against standard
c. Comparing criteria across alternatives (criteria ranked 2>3>1>4).
Criteri Alternative Ideal Minimu
on A B C m
Standa
rd
1. 6 5 8 10 7
2. 90 90 75 100 70
3. 40 35 50 50 30
4. G P VG E F
11. A specialty software development firm is planning to offer one or four new
software products and wishes to maximize profit, minimize risk, and increase
market share. A weight of 65% is assigned to annual profit potential, 20% to
probability risk, and 15 % to market share. Use the tabular additive method
for this situation and identify the product that would be best for the firm to
introduce.
New Product Profit Profit Risk Market
Potential ($) ($) Share
SW I 100K 40K High
SW II 140K 35K Medium
SW III 150K 50K Low
SW IV 130K 45K Medium
12.The cost of developing an internal training program for office automation is
unknown but described by the following probability distribution:
Cost ($) Probability of Occurrence
80,000 0.20
95,000 0.30
105,000 0.25
115,000 0.20
130,000 0.05
 What is the expected cost of the course? What is the most probable cost?
What is the maximum cost that will occur with a 95% assurance?
13.Daily positive and negative payoffs are given for five alternatives and five
futures in the following matrix. Which alternative should be chosen to
maximize the probability of receiving a payoff at least 9? What choice would
be made by using the most probable future criterion?
(0.15) (0.20) (0.30) (0.20) (0.15)
F1 F2 F3 F4 F5
A1 12 8 -4 0 9
A2 10 0 5 10 16
A3 6 5 10 15 -4
A4 4 14 20 6 12
A5 -8 22 12 5 9
14.Refer to the given of problem 13, assumed that there are no available
probabilities of receiving payoffs, which alternative would be chosen under
Laplace principle, the maximin rule, the maximax rule, and the Hurwicz rule
with the alpha of 0.75?

Instructions on how to submit student output

Refer to the course policies and content plan attached at the back of this module.