Quantitative Techniques for Decision Making

Introduction:
The quantitative approach is to make an optimal decision by using mathematical and statistical models in a situation when the probability of all
outcomes is uncertain.

Quantitative approach to decision making produces the best results when the problem is clearly defined, several alternatives exist, and decision
outcomes are easily measurable. However, in the case that many external factors are outside of the decision maker’s control and their probability
is unknown, quantitative methods can be unreliable.

Rationale in using quantitative techniques

Quantitative techniques, especially those that rely on statistical software, have the advantage of suggesting the best solution to a problem without
even identifying all possible alternatives. This feature is quite useful in problems where the number of possible alternatives is very large although
only a few are worth considering for selection. Once the problem and conditions are defined, the decision making process becomes quick.

Quantitative Decision-Making Tools

1) Decision Tree- used as a tool for selecting an optimal decision among alternatives. A decision tree usually starts with recording the
problem graphically as a starting node or root of diagram and the alternative solutions as its branches. In the case that a solution requires
further decisions to be made, more branches are added at a lower level. The probability, cost and value of each scenario are recorded.

2) Probability Analysis- probability is simply how likely something is to happen. Whenever we are unsure about the outcome of an event,
we can talk about the probabilities of certain outcomes—how likely they are. The analysis of events governed by probability is called
statistics. The best example for understanding probability is flipping a coin. There are two possible outcomes: heads or tails.

Notes:
a) The probability of an event can only be between 0 and 1 and can also be written as a percentage.
b) Two events are mutually exclusive if they cannot occur simultaneously (e.g., heads and tails cannot both occur on a single toss of
coin)
c) The joint probability for two events is the probability that both will occur.
d) The conditional probability of two events is the probability that one will occur given that the other has already occurred.
e) Two events are independent if the occurrence of one has no effect on the probability of the other.
 3) Payoff (decision) tables- A payoff table represents the outcome (payoff) of specific decisions when certain states of nature (events not
within the control of the decision maker) occur. The payoff table is a helpful tool for identifying the best solution given several decision
alternatives and future conditions that involve risk.

Illustration: Mr. B cooks and sells cookies. Each box of cookies sells for P50 during regular hours,
from 10 am to 8 pm. If every box is sold by 8 pm, Mr. B calls it a day. However, all unsold boxes by 8 pm
are sold at half the regular price up to 9 pm. The variable cost per box is P30. The CM per box is shown
below:

From 10 am to 8 pm-- Selling price P50

Variable cost 30
CM per box 20

After 8 pm: Selling price P25

Variable cost 30
Loss per box 5
Past experience has shown that the daily sales demand (up to 8 pm) and their probabilities are as follows:
Sales per day Probability
500 boxes 0.2
600 boxes 0.7
700 boxes 0.1
If the sales demand will be the same as in the past, a payoff table showing the contribution margin (conditional
value) for the possible sales quantities under each production level strategy, as well as the expected value of
such contribution margin can be constructed as follows:

Expected
CONTRIBUTION MARGIN FOR POSSIBLE SALES QUANTITIES: Value of CM
Prepare 500 boxes: 500 boxes (500 x P20) P 10,000
600 boxes (500 x P20) 10,000
700 boxes (500 x P20) 10,000 P10,000 (1)

Prepare 600 boxes: 500 boxes (500 x P20) - (100 x P5) P 9,500
600 boxes (600 x P20) 12,000
700 boxes (600 x P20) 12,000 P11,500 (2)

Prepare 700 boxes: 500 boxes (500 x P20) - (200 x P5) P 9,000
600 boxes (600 x P20) - (100 x P5) 11,500
700 boxes (700 x P20) 14,000 P11,250 (3)

(1) (P10,000 x 0.20) + (P10,000 x 0.70) + (P10,000 x 0.10) = P10,000

(2) (P9,500 x 0.20) + (P12,000 x 0.70 ) + (P12,000 x 0.10)= P11,500
(3) (P9,000 x 0.20) + (P11,500 x 0.70) + (P14,000 x 0.10) = P11,250
Based on the above payoff table, the best course of action is to prepare 600 boxes of cookies per day
because it has the highest expected value of contribution margin of P11,500.
4) Expected value of perfect information

Perfect information is the knowledge that a future state of nature (event) will occur with certainty. In this case, it is assumed that the
probability distribution is an accurate representation of the relative frequency of future demand and that the decision maker knows
exactly when each possible event will occur.

Expected value of perfect information is the difference between the expected value without perfect information and the result if the
best action is taken given perfect information.

Using the previous example, assume that Mr.B knew the daily sales demand for cookies with certainty, he would prepare exactly the
number of boxes demanded. The expected value of perfect information is computed as follows:

Daily sales CM per box Total CM Probability Expected

Demand Value

500 boxes P20 P10,0000.20 P2,000

600 boxes 20 12,000 0.70 8,400
700 boxes 20 14,000 0.10 1,400
Expected CM given perfect information 11,800
Less expected value of CM using the best course of
action without perfect information (from
Previous illustration) 11,500
Expected value of perfect information 300
=====

5) Gantt Chart (or Bar Chart)- It is one of the oldest and the simplest methods in planning and programming construction projects. Its
beginning dates back to 1915 when it was invented by Henry Gantt. Commonly used in project management, it is one of the most
popular and useful ways of showing activities (tasks) displayed against time. On the left of the chart is a list of the activities and along the
top is a suitable time scale. Each activity is represented by a bar; the position and length of the bar reflects the start date, duration and end
date of the activity.

Advantages of a Gantt Chart:

a) It is simple to construct and use, requiring no special tools or mathematics. It can be used on all types of projects.
b) It is a very useful control tool. As the project progresses, actual completion time can be compared with the plan.
c) It can be used to monitor the activities in a project. It shows which activity should be in progress as of a certain date and how close it
is to completion time.

Disadvantages of Gantt Chart:

A gantt chart does not show the interrelationships among the activities in a project. Only simple relationships can be shown in the chart.
Illustration:
6) Program Evaluation and Review Technique (PERT)
It is a networking technique used for planning and controlling the activities in a project. It provides management pertinent information
about a project, such as:
a) Expected completion time of a project
b) When each activity in a project is scheduled to start and finish
c) Which part of the project must be finished on time so avoid making the whole project late
d) How resources may be shifted from one part to another part of the project without affecting the overall completion time of the
project
e) The progress of each part of a project as of a certain date.

PERT DIAGRAM- an arrow diagram on a network showing the relationships or interdependence of the various activities of a project.
A PERT diagram incorporates probabilistic time estimates and identifies crirical path.

Types of Activities shown in a Pert Diagram:

1) Series- an activity cannot be performed unless its predecessor activity is finished.
2) Parallel- activities that can be performed simultaneously.

Formula for Calculating PERT

PERT is determined using three points: Optimistic (O), Most Likely (M), and Pessimistic (P). PERT combines probability
theory and statistics to derive a formula for the average activity from the three point estimates. PERT estimate formula is:
(O + 4M + P)/6

Optimistic time- the least amount of time it can take to complete a task.
Pessimistic time- the maximum amount of time it should take to complete a task
Most likely time- assuming there are no problems, the best or most reasonable estimate of how long it should take to complete
a task.

Example of a PERT diagram:

7) Linear Programming- a quantitative technique used to find the optimal solution to short term resource allocation problems, such as:
a) Maximization of the revenue, contribution margin , or profit functions; and
b) Minimization of cost function, subject to constraints( e.g. limited resources, production capacity levels, etc).

Other business applications of linear programming:

1) Determination of the best product mix
2) Determination of the optimum materials mix
3) Assignment of jobs to manufacturing equipment
4) Workforce scheduling
5) Determination of transportation routes

Steps in formulating a linear program:

1) Identify the decision variables. Decision variables are the unknowns used to construct the objective and constraint
functions. The values of such variables are the outputs from the linear programming process.

2) Express the objective and constraint functions in terms of the decision variables identified in Step 1.

Methods for solving linear programming problems

1) Graphical method- limited to problems with only 2 variables.

Step 1: Formulate the LP (Linear programming) problem.

Step 2: Construct a graph and plot the constraint lines.
Step 3: Determine the valid side of each constraint line.
Step 4: Identify the feasible solution region.
Step 5: Plot the objective function on the graph.
Step 6: Find the optimum point.

2) Simplex method is an approach to solving linear programming models by hand using slack variables, tableaus, and
pivot variables as a means to finding the optimal solution of an optimization problem. Simplex tableau is used to perform
row operations on the linear programming model as well as for checking optimality.
 8) Learning Curve (or Experience Curve)- a mathematical expression of the phenomenon that incremental unit costs to produce (or
incremental unit time used to produce) decrease as managers and labor gain experience from practice.

The learning curve model shows a constant percentage reduction (usually from 20% to 40%) in the average direct labor input time
required per unit as the cumulative output doubles.

For example,
- if the cumulative average time were 1,000 hours at the production of the first unit, 700 hours at the production of the
second, 490 hours at the fourth, 343 hours at the eighth and so on, this would be a 70 percent learning curve.

9) Sensitivity Analysis- the study of how the outcome of a decision process changes as one or more of the assumptions change.

Illustrative example:
With sales of 100 units, the company’s projected profit is P500.
Sales P5,000
Less variable costs 3,000
Contribution margin 2,000
Less fixed costs 1,500
Profit 500
—-------

How much would profit be if:

1) sales in units increases by 20%
2)variable costs per unit decreases by 5%
3)fixed costs decreases by P500

Present data Units increase by VC decreases by Fixed costs

20% 5% decreases by P500
Sales (100 units) 5,000 6,000 5,000 5,000
Less Variable costs 3,000 3,600 2,850 3,000
Contribution 2,000 2,400 2,150 2,000
margin
Less fixed costs 1,500 1,500 1,500 1,000
Profit 500 900 650 1,000

In the above sensitivity analysis, different scenarios are presented, showing the expected profit as the profit factors are changed.