BAF 204 Decision Making Techniques

About this Decision Making Techniques Module
The Decision Making Techniques Module has been produced by National Institute of Public
Administration (NIPA). All modules produced by the Institute are structured in the same way,
as outlined below.
How this Decision Making Techniques Module is structured
The Module overview

The module overview gives you a general introduction to the module. Information contained
in the module overview will help you determine:
 What you expect from the course.
 How much time you will need to invest to complete the course.
The overview also provides guidance on:

 Study skills.
 Where to get help.
Assignments and assessments
Activity icons
 Units.
We strongly recommend that you read the overview carefully before starting your
study.
The Module content

The Module is broken down into ten (5) units. Each unit comprises of:
 An introduction to the unit content.
 Unit outcomes.
 New terminology.
 Core content of the unit with a variety of learning activities.
 A unit summary.
 Assignments and/or assessments, as applicable.
For those interested in learning more on this subject, we provide you with a list of additional
resources at the end of this module; these may be books, articles or web sites.
National Institute of Public Administration- Outreach Programmes Division

Your comments
After completing this Decision Making Techniques Module, we would appreciate it if you
would take a few moments to give us your feedback on any aspect of this course. Your
feedback might include comments on:
 Content and structure.
 Reading materials and resources.
 Assignments and Assessments.
 Duration.
 Support (assigned tutors, technical help, etc.)
Your constructive feedback will help us to improve and enhance this course.

Welcome to Decision Making Techniques Module
This Decision Making Techniques Module equip students with requisite knowledge and skills of decision
making techniques.
Module learning outcomes

Upon completion of this Module, you will be able to:
 Explain business decisions and the decision-making process
 Apply business modelling techniques including regression analysis,
Linear programming and time series
 Apply business planning techniques including Material Requirements
Planning (MRP) systems,
Inventory management and aggregate planning
 Demonstrate the use of probability in decision making
 Apply estimation techniques and hypothesis tests to aid the decision-making
process
Expected duration of this Module is 6 months
Formal study time required is 4 weeks before the beginning of the semester
Self-study time recommended is 4 hours per week
Time Frame

Study skills As an adult learner your approach to learning will be different to that from your
school days: you will choose what you want to study, you will have professional
and/or personal motivation for doing so and you will most likely be fitting your study
activities around other professional or domestic responsibilities.
Essentially you will be taking control of your learning environment. As a
consequence, you will need to consider performance issues related to time
management, goal setting, stress management, etc. Perhaps you will also need to
reacquaint yourself in areas such as essay planning, coping with exams and using the
web as a learning resource.
Your most significant considerations will be time and space i.e. the time you dedicate
to your learning and the environment in which you engage in that learning.
We recommend that you take time now—before starting your self-study—to
familiarize yourself with these issues. There are a number of excellent resources on
the web. A few suggested links are:
 http://www.how-to-study.com/
The “How to study” web site is dedicated to study skills resources. You will find
links to study preparation (a list of nine essentials for a good study place), taking
notes, strategies for reading text books, using reference sources, test anxiety.
 http://www.ucc.vt.edu/stdysk/stdyhlp.html
This is the web site of the Virginia Tech, Division of Student Affairs. You will find
links to time scheduling (including a “where does time go?” link), a study skill
checklist, basic concentration techniques, control of the study environment, note
taking, how to read essays for analysis, memory skills (“remembering”).
 http://www.howtostudy.org/resources.php
Another “How to study” web site with useful links to time management, efficient
reading, questioning/listening/observing skills, getting the most out of doing (“hands-
on” learning), memory building, tips for staying motivated, developing a learning
plan.
The above links are our suggestions to start you on your way. At the time of writing
these web links were active. If you want to look for more go to www.google.com and
type “self-study basics”, “self-study tips”, “self-study skills” or similar.

Need Help? In case you need help, you can contact NIPA at the following website, phone number
or you can email.
www.nipa.ac.zm
NIPA-Main Campus – Outreach Programmes Division
Phone Numbers:+260-211-222480
Fax:
e-mail address:opd@nipa.ac.zm
The teaching assistant for routine enquiries can be located from the Outreach Division
from 08:00 to 17:00 or can be contacted on the numbers and email address indicated
above.
Library
There is a library located at the main campus along Dunshabe Road. The library
opens Monday to Friday from 08:00 to 17:00.
Assignments There shall be two assignments given for this module.

The assignments should be sent by post or emailed to the provided email addressed
to the Outreach Programmes Division – Nigeria Hall.
Assignments should be submitted to Outreach Programmes Division Registry.
Assessments There shall be a minimum of two (02) assessments given to the students undertaking
this subject
These assessments shall be teacher marked assessments.
The assessments shall be determined and given by the course tutors after you have
covered a number of topics
The teacher/tutor shall ensure that the assessments are marked and dispatched to the
student.

Getting around the Decision Making Techniques Module
Margin icons
While working through this Decision Making Techniques module, you will notice the
frequent use of margin icons. These icons serve to “signpost” a particular piece of text, a new
task or change in activity; they have been included to help you to find your way around this
Decision Making techniques module.
A complete icon set is shown below. We suggest that you familiarize yourself with the icons
and their meaning before starting your study.
Activity Assessment Assignment Case study
Discussion Group Help Note it!

activity
Outcomes Reading Reflection Study skills
Summary Terminology Time frame Tip

Learning tips
You may not have studied by distance education before. Here are some guidelines to help you.
How long will it take?

It will probably take you a minimum of 90 hours to work through this study guide. The time
should be spent on studying the module and the readings, doing the activities and self-help
questions and completing the assessment tasks.
Note that units are not all the same length, so make sure you plan and pace your work to give
yourself time to complete all of them.
About the study guide

This study guide gives you a unit-by-unit guide to the module you are studying. Each unit
includes information, activities, self-help questions and readings for you to complete. These
are all designed to help you achieve the learning outcomes that are stated at the beginning of
the module.
Activities, self-help questions and assessments

The activities, self-help questions and assessments are part of a planned distance education
programme.
They will help you make your learning more active and effective, as you process and apply
what you read. They will help you to engage with ideas and check your own understanding. It
is vital that you take the time to complete them in the order that they occur in the study guide.
Make sure you write full answers to the activities, or take notes of any discussion.
We recommend you write your answers in your learning journal and keep it with your study
materials as a record of your work. You can refer to it whenever you need to remind yourself
of what you have done.
Unit summary
At the end of each unit there is a list of the main points. Use it to help you review your
learning. Go back if you think you have not covered something properly.

Studying at a distance
There are many advantages to studying by distance education – a full set of learning materials
as provided, and you study close to home in your own community. You can also plan some of
your study time to fit in with other commitments like work or family.
However, there are also challenges. Learning at a distance from your learning institution
requires discipline and motivation. Here are some tips for studying at a distance.
1. Plan – Give priority to study sessions with your tutor and make sure you allow enough
travel time to your meeting place. Make a study schedule and try to stick to it. Set specific
days and times each week for study and keep them free of other activities. Make a note of the
dates that your assessment pieces are due and plan for extra study time around those dates.
2. Manage your time – Set aside a reasonable amount of time each week for your study
programme – but don’t be too ambitious or you won’t be able to keep up the pace. Work in
productive blocks of time and include regular rests.
3. Be organized – Have your study materials organized in one place and keep your notes
clearly labeled and sorted. Work through the topics in your study guide systematically and
seek help for difficulties straight away. Never leave this until later.
4. Find a good place to study – Most people need order and quiet to study effectively, so try
to find a suitable place to do your work – preferably somewhere where you can leave your
study materials ready until next time.
5. Ask for help if you need it – This is the most vital part of studying at a distance. No
matter what the difficulty is, seek help from your tutor or fellow students straight away.
6. Don’t give up – If you miss deadlines for assessments, speak to your tutor – together you
can work out what to do. Talking to other students can also make a difference to your study
progress. Seeking help when you need it is a key way of making sure you complete your
studies – so don’t give up.

Unit 1– Explaining the business and decision making process
.
1.1 Unit learning outcomes
After studying this unit, the learner will be able to:
 Describe types of business decisions
 Describe the decision making process
 Apply simple decision making techniques
1.1 Types of business decision

The nature of business decisions.
A business decision is simply a decision in business. Business decisions are taken by the
owners of a business organization, or by managers within the organization.
As a result of a business decision, the organization may commit resources to a business
activity. It is therefore highly desirable that the decisions that are taken are good decisions. If
they are bad, they could weaken its competitive position.
Decisions are made when there is more than one option that can be taken. Decisions are
usually taken under conditions of uncertainty. Decisions might be improved, even for skilled
and experienced managers, if suitable decision-making techniques are used to help with the
decision.

Structured and unstructured business decisions.
A structured decision is a decision that is repetitive and routine in nature.
An unstructured decision is a decision where the decision-maker uses his insight, experience
and intuition to identify and define the nature of the problem and to consider alternative
solutions. The problems are non-routine.
Strategic, operating and control decisions
Strategic decisions are decisions concerned with deciding the objectives, resources and
policies of the organization. They are often unstructured non-routine decisions. They are made
by top management.
Operational decisions are about planning production or the provision of services to meet
demand and delivery requirements of customers.
Control decisions are decisions about what to do when something goes wrong, or how to
prevent something from going wrong.
RN Anthony identified three levels of decision within an organization.
1. Strategic decisions: These are decisions about the objectives of the organization and
plans for achieving them, mostly they are over the long term
2. Tactical decisions: These are decisions concerned with shorter-term plans for
achieving medium-term objectives, and control decisions concerned with achieving
these shorter term plans.
3. Operational decisions: These are decisions concerned with detailed planning of day-to-
day operations and control activities, often at a supervisor level or junior management.
1.2 The decision making process
Decisions making in business can be explained as a series of stages.
Stage 1: Identify the problem or challenge. The first step is to recognize that a problem or a
challenge exists, and identify what the problem is.
Stage 2: Gather and evaluate information. Information should be gathered. Information is
sometimes needed in order to become aware of the problem.
Stage 3: Consider alternative solutions and their possible implications. Having established the
nature of the problem and the decision that is required, the next step is to consider a range of
different possible ways of dealing with the matter and the possible outcomes from each should
be assessed.
Stage 4: Choose the best alternative. The next step is to select the preferred option from all
the alternatives that have been considered.
Stage 5: Implement and monitor what happens. When a decision has been taken, someone
should make sure that it is put into effect. This is usually done by management and they
should also monitor the events after the decision has been implemented, to check whether the
decision has had its intended effect.

1.3 Decision making techniques.
Pareto analysis: This is a simple technique that helps the decision-maker to focus on the most
important problems. It is based on the so-called 80:20 rules. This states that in any situation,
20% of an item represents 80% of its importance or value and the remaining 80% represents
only 20% of value or importance. Therefore, managers should focus on the 20% that is most
important and spend less times and effort on the 80% that is not so important. The proportions
80% and 20% are rough estimates only.
Plus/minus/interesting (PMI) analysis: PMI is a method that can be used for creative
thinking, because the decision-maker uses to write down ideas before reaching a decision. The
technique is simple, it involves three columns as shown below:
Plus Minus Interesting
Ideas in favour of a decision ideas against a decision ideas that are not obviously
either . plus, or minus but
are interesting
Cost-benefit analysis: This a relatively simple technique where the value of the benefits of a
course of action and the costs associated with it are estimated. Costs are either one-off or
maybe ongoing. Benefits are most often received over a period of time. The cost and benefits
are compared and a decision is made on the basis of the expected net benefit.
Grid analysis: decision matrices analysis
This is a technique that helps a decision maker to choose between alternative options. It relies
on judgment and opinion. Although some mathematics are involved, it is not really a
mathematical model for decision-making.
The technique involves identifying the factors that matter to the decision-maker when
choosing between different alternatives, and giving each factor a weighting. Each alternative
is then given a score for each factor and this is multiplied by the weighting for the factor. The
weighted scores for all the factors are added and the alternative with the highest weighted
score is selected.

UNIT 2: Applying business modelling techniques including regression
analysis linear programming and time series
Learning Outcomes
After studying the unit, you will be able to:
 Apply regression analysis

 Apply correlation
 Apply time series analysis
 Apply linear programming (LP)
 Apply simulation
 Apply replacement analysis
 Use decision trees
 Apply queuing theory
 Use assignments
1.1 Correlation and regression analysis

Correlation
Two variables are said to be correlated if a change in the value of one variable is accompanied
by a change in the value of another variable. This is what is meant by correlation.
Examples of variables which might be correlated are:
 A person’s height and weight.
 The distance of a journey and the time it takes to make the journey.
Scatter diagrams

One way of showing the correlation between two related variables is on a scatter diagram.
This simply takes historical measurements for the two variables and plots them on a chart or
graph. For example, a scatter diagram showing monthly selling costs against the volume of
sales for a 12-month period might be as follows:
Selling
costs,
Sales
Volume
Independent and dependent variables
When we draw a scatter diagram, we often assume that the value of one of the variables
depends on the value of the other. For example, in the diagram above, we assume that the
amount of selling costs depends on the volume of sales. Selling costs are the dependent
variable in this example, and sales volume is the independent variable.
The independent variable (the cause) is plotted on the horizontal (x) axis
The dependent variable (the effect) is plotted on the vertical (y) axis.
The diagram suggests there is some correlation between selling costs and sales volume.
Degrees of correlation
Two variable might be perfectly correlated, partly correlated or uncorrelated. Correlation can
be positive or negative.
These differing degrees of correlation can be illustrated by scatter diagrams.
Perfect correlation
(a) Y (b) Y
X X

All the pairs of values lie on a straight line. An exact linear relationship exists between the
two variables,
Partial correlation
(c) Y (d) Y
X X
In (c) although there is no exact relationship, low values of X tend to be associated with low
values of Y, and high values of X with high values of Y.
In (d) again, there is no exact relationship, but low values of X tend to be associated with high
values of Y and vice versa
No correlation
The values of these two variables are not correlated with each other.
Positive and negative correlation
Correlation whether perfect or partial, can be positive or negative.
Positive correlation means that low values of one variable are associated with low values of
the other, and high values of one variable are associated with high values of the other.
Negative correlation means that low values of one variable are associated with high values of
the other, and high values of one variable with low values of the other.
The correlation coefficient and the coefficient of determination

The degree of correlation between two variables is measured by Pearson’s correlation
coefficient, r. This has a value in the range -1 to +1. The nearer r is to +1 or -1, the stronger
the relationship or correlation.
Pearson’s correlation coefficient, r (also known as the product moment correlation coefficient)
is used to measure how strong the connection is between two variables, known as the degree
of correlation.
It is calculated using the formula
Where x and y represent pairs of data for the two variables x and y, n = the number of pairs of
data used in the analysis.
r must always fall between -1 and +1.
 r = +1 (means variables are perfectly positively correlated)
 r = -1 (means variables are perfectly negatively correlated)
 r = o (means variables are uncorrelated)
CLASS EXERCISE: The cost of output at a factory is thought to depend on the number of
units produced. Data have been collected for the number of units produced each month in the
last six months, and the associated costs.
Month Output(‘000s of units) Cost (K million)
X Y
1 2 9
2 3 11
3 1 7
4 4 13
5 3 11
6 5 15
Required: Assess whether there is any correlation between output and cost.
The coefficient of determination r2

The coefficient of determination r2 measures the proportion of the total variation in the value
of one variable that can be explained by the variations in the value of the other variable.
Unless the correlation coefficient r is exactly (or very nearly) +1, -1 or 0, its meaning is a little
unclear. For example, if the correlation coefficient for two variables is +0.8, this would tell us
that the variables are positively correlated, but the correlation is not perfect. It would not
really tell us much else. A more meaningful analysis is available from the square of the
correlation coefficient, r, which is called the coefficient of determination, r 2.
Interpreting r2: Suppose that we measure the correlation coefficient between the volume of
output produced in a month and the total costs for the month, and find that the correlation
coefficient is +0.985. If r = +0.985, r2 = 0.9702. This would mean that 97.02% of variations in
costs can be explained by variations in the volume of output, leaving 0.028 (less than 3%) of
variations to be explained by other factors.
Note, however, that r2 = 0.9702, we would say that 97.02% of the variations in y can be
explained by variations in x. We do not necessarily conclude that 81% of variations in y are
caused by the variations in x. We must beware of reading too much significance into our
statistical analysis.
Correlation and causation.
If two variables are well correlated, either positively or negatively, this may be due to pure
chance or there may be a reason for it. The larger the number of pairs of data collected, the
less likely it is that the correlation is due to chance, though that possibility should never be
ignored entirely. For example, sales of ice cream and of sunglasses are well correlated, not
because of a direct causal link but because the weather influences both.
Spearman’s rank correlation coefficient
Spearman’s rank correlation coefficient is used to measure the correlation between two
variables when data is given in terms of order or rank, rather than actual values.
Coefficient of rank correlation, r =
Where n = number of pairs of data

D = the difference between the rankings in each set of data
The coefficient of rank correlation can be interpreted in exactly the same way as the ordinary
correlation coefficient. Its values can range from -1 to +1.

Student Statistics placing Economics placing
A 2 1
B 1 3
C 4 7
D 6 5
E 5 6
F 3 2
G 7 4
Assess whether the position of the students in statistics correlates with their position in
economics
Linear regression analysis.
Correlation enables us to determine the strength of any relationship between two variables but
it does not offer us any method of forecasting values for one variable, y given values of
another variable, x.
Linear regression analysis (the least squares method) is one techniques for estimating a line of
best fit, once an equation for a line of best fit has been determined, forecasts can be made.
The least squares method is a very common method of estimating a line of best fit because it
does not rely on judgement to make the estimate.
The least squares method of linear regression analysis involves using the following formulae
to find a and b in Y= a + bX
Where n = the number of pairs of data

X bar = mean value of all the pairs of data.
Y bar = the mean value of all the pairs of data.
The value of b must be calculated first as it is needed to calculate a

CLASS EXERCISE: Suppose we have the following pairs of data about output and costs
Month Output(‘000s of units) Cost (K million)
X Y
1 2 9
2 3 11
3 1 7
4 4 13
5 3 11
6 5 15
(a) Use linear regression analysis to estimate the relationship between output volume and
costs
(b) Find the cost of producing 22,000 units of output
(c) Calculate the degree of correlation between output and costs by calculating the
correlation coefficient.
USING EXCEL FOR LINEAR REGRESSION ANALYSIS (PRACTICAL IN

COMPUTER LAB)

2.2 TIME SERIES ANALYSIS
A time series is a series of figures or values recorded over time. Any pattern found in the data
is then assumed to continue into the future and an extrapolative forecast is produced.
Examples of time series include:
 Output at a factory each day for the last month.
 Monthly sales over the last two years.
There are four components of a time series: trend, seasonal variations, cyclical variations and
random variations.
The trend
The trend is the underlying long-term movement over time in the values of the data recorded.
Preparing time series graphs and identifying trends
Year Output per labor hour Cost per unit (K) Number of employees
(units)
2001 30 1.00 100
2002 24 1.08 103
2003 26 1.20 96
2004 22 1.15 102
2005 21 1.18 103
2006 17 1.25 98
(A) (B) (C)
(a) In time series (A) there is a downward trend in the output per labour hour. Output per
labour hour did not fall every year, because it went up between 2002 and 2003, but the
long term movement is clearly a downward one.

(b) In time series (B) there is an upward trend in the cost per unit. Although unit costs
went down in 2004 from a higher level in 2003, the basis movement overtime is one of
rising costs.
(C) In time series (C) there is no clear movement up or down, and the number of
employees remained fairly constant around 100. The trend line is therefore a static, or
level one.
Seasonal variations
Seasonal variations are short-term fluctuations in recorded values, due to different
circumstances which affect results at different times of the year, on different days of the week
or at different times of day etc.
Examples of seasonal variations

(a) Sales of ice cream will be higher in summer than in winter
(b) Shops might expect higher sales shortly before Christmas
(c) Sales might be higher on Friday and Saturday than on Monday.
(d) The telephone network may be heavily used at certain times of the day (such as
midmorning and mid-afternoon) and much less used at other times (such as in the
middle of the night)
Cyclical variations
Cyclical variations are medium-term changes in results caused by circumstances which repeat
in cycles.
In business, cyclical variations are commonly associated with economic cycles, successive
booms and slumps in the economy. Economic cycles may last a few years. Cyclical variations
are longer term than seasonal variations.
Though you should be aware of the cyclical component, you will not be expected to carry
out any calculation connected with isolating it in this course. The mathematical models
which we will use, therefore exclude any reference to cyclical variations.
Summarizing the components

The components of a time series combine to produce a variable in one of two different ways.
Additive model: Series = Trend + Seasonal + Random
Y=T+S+R
Multiplicative model: Series = Trend X Seasonal X Random
Y=TXSXR
Methods of finding the trend

Main methods of finding a trend
(a) A line of best fit (the trend line) can be drawn by eye of the graph
(b) Linear regression analysis can be used
(c) A technique known as moving averages can be used.

Methods of finding the trend
A moving average is an average of the results of a fixed number of periods.
The moving averages method attempts to remove seasonal variations from actual data by a
process of averaging.
EXAMPLE
Moving averages of an odd number of results
Year Sales Units
2000 390
2001 380
2002 460
2003 450
2004 470
2005 440
2006 500
Required: Take a moving average of the annual sales over a period of three years.
Solution:
Year Sales Units Moving total of 3 Moving average of
years sales 3 years sales
2000 390
2001 380 1230 410
2002 460 1290 430
2003 450 1380 460
2004 470 1360 453
2005 440 1410 470
2006 500
Over what time period should a moving average be taken?

The above example averaged over a three-year period. Over what period should a moving
average be taken? The answer to this question is that the moving average which is most
appropriate will depend on the circumstances and the nature of the time series. Note the
following points
(a) A moving average which takes an average of the results in many time periods will
represent results over a longer term than a moving average of two or three periods
(b) On the other hand, with a moving average of results in many time periods, the last
figure in the series will be out of date by several periods. In our example, the most
recent average related to 2005, With a moving average of five years results, the final
figure in the series would relate to 2004

(c) When there is a known cycle over which seasonal variations occur, such as all the days
in the week or all the seasons in the year, the most suitable moving average would be
one which covers on full cycle
Moving averages of an even number of results

In the previous example, moving averages were taken of the results in an odd number of
time periods, and the average then related to the mid-pont of the overall period. If a
moving average were taken of results in an even number of time periods, the basic
technique would be the same, but the mid-point of the overall period would not relate to a
single period. For example, suppose an average were taken of the following four results
Spring 120
Summer 90
Average 115
Autumn 180
Winter 70
The average would relate to the mid-point of the period, between summer and autumn. The
trend line average figures need to relate to a particular time period: otherwise, seasonal
variations cannot be calculated. To overcome this difficulty, we take a moving average of the
moving average. An example will illustrate this technique.
EXAMPLE
Moving averages of an even number of results
Year Quarter Volume of sales( ‘000 units)
2005 1 600
2 840
3 420
4 720
2006 1 640
2 860
3 420
4 740
2007 1 670
2 900
3 430
4 760
Required :calculate a moving average of four periods

Solution
Year Quarter Volume of Moving total Moving average Mid-point of 2
sales of 4 quarters of 4 quarters Moving averages
sales sales Trend line
2005 1 600
2 840
2580 654
3 420 650
2620 655
4 720 657.50
2640 660
2006 1 640 660
2640 660
2 860 662.50
2660 665
3 420 668.75
2690 672.5
4 740 677.50
2730 682.5
2007 1 670 683.75
2740 685
2 900 687.50
2760 690
3 430
4 760
Finding the seasonal variations

Finding the seasonal component using the additive model
Once a trend has been established, by whatever method, we can find the seasonal variations
Step 1: The additive model for time series analysis is Y = T + S + R
Step 2: If we deduct the trend from the additive model, we get Y – T = S + R
Step 3: If we assume that R, the random component of the time series is relatively small and
therefore negligible, then S = Y – T.
Therefore, the seasonal component, S = Y – T (the detrended series).

Exercise
Output at a factory appears to vary with the day of the week. Output over the last three weeks
has been as follows.
Week 1 (‘000 units) Week 2 (‘000 units) Week 3 (‘000 units)
Mon 80 82 84
Tue 104 110 116
Wed 94 97 100
Thur 120 125 130
Fri 62 64 66
Find the seasonal variation for each of the 15 days, and the average seasonal variation or each
day of the week using the moving averages method (take a five-day period)
Finding the seasonal component using the multiplicative model
The method of estimating the seasonal variations in the additive model is to use the
differences between the trend and actual data. The additive model assumes that the
components of the series are independent of each other, an increasing trend not affecting the
seasonal variations for example. The alternative is to use the multiplicative model whereby
each actual figure is expressed as a proportion of the trend. Sometimes this method is called
the proportional model.
When to use the multiplicative model
The multiplicative model is better than the additive model for forecasting when the trend is
increasing or decreasing over time. In such circumstances, seasonal variations are likely to be
increasing or decreasing too. The additive model simply adds absolute and unchanging
seasonal variations to the trend figures whereas the multiplicative model, by multiplying
increasing or decreasing trend values by constant seasonal variation factor, takes account of
changing seasonal variations.
Summary
Step 1: calculate the moving total for an appropriate period.
Step 2: calculate the moving average (trend) for the period. (calculate the mid-point of two
moving averages if there are an even number of periods
Step 3: calculate the seasonal variation. For an additive model, this is Y – T. For a
multiplicative model, this is Y/T
Step 4: calculate an average of the seasonal variations
Step 5: Adjust the average seasonal variations so that they add up to zero for an additive
model. When using the multiplicative model, the average seasonal variations should add up to
an average of 1.

Seasonally-adjusted data
Seasonally adjusted data (deseasonalised) are data which have had any seasonal variations
taken out, so leaving a figure which might indicate the trend. Seasonally-adjusted data should
indicate whether the overall trend is rising, falling or stationary.
Exercise
Actual sales figures for four quarters, together with appropriate seasonal adjustment factors
derived from previous data, are as follows
Seasonal adjustments
Quarter Actual sales (K’000) Additive model Multiplicative model
(K’000)
1 150 +3 1.02
2 160 +4 1.05
3 164 -2 0.98
4 170 -5 0.95
Required Deseasonalise these data

Residuals
A residual is the difference between the results which would have been predicted (for a past
period for which we already have data) by the trend line adjusted for the average seasonal
variation and the actual results. The residual is therefore the difference which is not explained
by the trend line and the average seasonal variation. The residual gives some indication of
how much actual resukts were affected by other factors. Large residuals suggest that any
forecast is likely unreliable.
Random variation.
This is the variability of a process caused by irregular and erratic fluctuations that cannot be
anticipated or detected (and therefore cannot be eliminated). For example, why coffee shop
sales vary from one day to another may be down to a number of predictable factors, such as
the days of the week and the weather – but will also be subject to random variations.

2.3 LINEAR PROGRAMMING
Many applications in mathematics involve systems of inequalities/equations. In this unit, we

shall apply the systems of linear inequalities/equations to solve some real life problems of the
type as given below:
A furniture dealer deals in only two items–tables and chairs. He has K 50,000 to invest and
has storage space of at most 60 pieces. A table costs K2500 and a chair K500. He estimates
that from the sale of one table, he can make a profit of K250 and that from the sale of one
chair a profit of K75. He wants to know how many tables and chairs he should buy from the
available money so as to maximise his total profit, assuming that he can sell all the items
which he buys.
Such type of problems which seek to maximise (or, minimise) profit (or, cost) form a general
class of problems called optimisation problems. Thus, an optimization problem may involve
finding maximum profit, minimum cost, or minimum use of resources etc.
A special but a very important class of optimisation problems is linear programming

problem. The above stated optimisation problem is an example of linear programming
problem. Linear programming problems are of much interest because of their wide
applicability in industry, commerce, management science etc. In this chapter, we shall study
some linear programming problems and their solutions by graphical method only, though
there are many other methods also to solve such problems.
Linear Programming Problem and its Mathematical Formulation

We begin our discussion with the above example of furniture dealer which will further lead to
a mathematical formulation of the problem in two variables. In this example, we observe
(i) The dealer can invest his money in buying tables or chairs or combination thereof. Further
he would earn different profits by following different investment strategies.

(ii) There are certain overriding conditions or constraints., his investment is limited to a
maximum of K50,000 and so is his storage space which is for a maximum of 60 pieces.
The dealer can invest his money in different ways and he would earn different profits by
following different investment strategies. Now the problem is: How should he invest his
money in order to get maximum profit? To answer this question, let us try to formulate the
problem mathematically.
Mathematical formulation of the problem

Let x be the number of tables and y be the number of chairs that the dealer buys. Obviously, x
and y must be non-negative, i.e.,
x≥0 …(1)
(Non-negative constraints)
y≥0 …(2)
The dealer is constrained by the maximum amount he can invest (Here it is K50,000) and by
the maximum number of items he can store (Here it is 60).
Stated mathematically,
2500x + 500y ≤ 50000 (investment constraint)
or 5x + y ≤ 100 …(3)
and x + y ≤ 60 (storage constraint) …(4)
The dealer wants to invest in such a way so as to maximise his profit, say, Z which stated as a
function of x and y is given by:
Z = 250x + 75y (called objective function) …(5)
Mathematically, the given problems now reduces to:

Maximise Z = 250x + 75y
subject to the constraints:
5x + y ≤ 100
x + y ≤ 60
x ≥ 0, y ≥ 0

So, we have to maximise the linear function Z subject to certain conditions determined by a
set of linear inequalities with variables as non-negative. There are also some other problems
where we have to minimise a linear function subject to certain conditions determined by a set
of linear inequalities with variables as non-negative. Such problems are called Linear
Programming Problems.
Thus, a Linear Programming Problem is one that is concerned with finding the optimal value
(maximum or minimum value) of a linear function (called objective function) of several
variables (say x and y), subject to the conditions that the variables are non-negative and
satisfy a set of linear inequalities (called linear constraints).
The term linear implies that all the mathematical relations used in the problem are linear
relations while the term programming refers to the method of determining a particular
programme or plan of action.
Before we proceed further, we now formally define some terms (which have been used above)
which we shall be using in the linear programming problems:
Objective function Linear function Z = ax + by, where a, b are constants, which has to be
maximised or minimized is called a linear objective function.
In the above example, Z = 250x + 75y is a linear objective function. Variables x and y are
called decision variables.
Constraints The linear inequalities or equations or restrictions on the variables of a

linear programming problem are called constraints. The conditions x ≥ 0, y ≥ 0 are called non
negative restrictions. In the above example, the set of inequalities (1) to (4) are constraints.
Optimisation problem A problem which seeks to maximise or minimise a linear function
(say of two variables x and y) subject to certain constraints as determined by a set of linear
inequalities is called an optimisation problem. Linear programming problems are special
type of optimisation problems. The above problem of investing a given sum by the dealer in
purchasing chairs and tables is an example of an optimization problem as well as of a linear
programming problem.

We will now discuss how to find solutions to a linear programming problem. In this course,
we will be concerned only with the graphical method.
Graphical method of solving linear programming problems

Let us refer to the problem of investment in tables and chairs discussed. We will now solve
this problem graphically. Let us graph the constraints stated as linear inequalities:
5x + y ≤ 100 ... (1)
x + y ≤ 60 ... (2)
x ≥ 0 ... (3)
y ≥ 0 ... (4)
The graph of this system (shaded region) consists of the points common to all half planes
determined by the inequalities (1) to (4). Each point in this region represents a feasible choice
open to the dealer for investing in tables and chairs. The region, therefore, is called the
feasible region for the problem. Every point of this region is called a feasible solution to the
problem. Thus, we have,
Feasible region: The common region determined by all the constraints including non-
negative constraints x, y ≥ 0 of a linear programming problem is called the feasible region (or
solution region) for the problem. In the figure above, the region OABC (shaded) is the

feasible region for the problem. The region other than feasible region is called an infeasible
region.
Feasible solutions: Points within and on the boundary of the feasible region represent
solutions of the constraints. In the diagram, every point within and on the boundary of the
feasible region OABC represents feasible solution to the problem. For example, the point (10,
50) is a feasible solution of the problem and so are the points (0, 60), (20, 0) etc. Any point
outside the feasible region is called an infeasible solution. For example, the point (25, 40) is
an infeasible solution of the problem.
Optimal (feasible) solution: Any point in the feasible region that gives the optimal value
(maximum or minimum) of the objective function is called an optimal solution. Now, we see
that every point in the feasible region OABC satisfies all the constraints as given in (1) to (4),
and since there are infinitely many points, it is not evident how we should go about finding a
point that gives a maximum value of the objective function Z = 250x + 75y.
To handle this situation, we use the following theorems which are fundamental in solving
linear programming problems.
Theorem 1 Let R be the feasible region (convex polygon) for a linear programming problem
and let Z = ax + by be the objective function. When Z has an optimal value (maximum or
minimum), where the variables x and y are subject to constraints described by linear
inequalities, this optimal value must occur at a corner point* (vertex) of the feasible region (A
corner point of a feasible region is a point in the region which is the intersection of two
boundary lines.)
Theorem 2: Let R be the feasible region for a linear programming problem, and let Z = ax +
by be the objective function. If R is bounded**, then the objective function Z has both a
maximum and a minimum value on R and each of these occurs at a corner point (vertex) of
R. (A feasible region of a system of linear inequalities is said to be bounded if it can be
enclosed within a circle. Otherwise, it is called unbounded. Unbounded means that the
feasible region does extend indefinitely in any direction.)

Note that If R is unbounded, then a maximum or a minimum value of the objective function
may not exist. However, if it exists, it must occur at a corner point of R. (By Theorem 1).
In the above example, the corner points (vertices) of the bounded (feasible) region are: O, A,
B and C and it is easy to find their coordinates as (0, 0), (20, 0), (10, 50) and (0, 60)
respectively.
Let us now compute the values of Z at these points.

We have
We observe that the maximum profit to the dealer results from the investment strategy (10,
50), i.e. buying 10 tables and 50 chairs.
This method of solving linear programming problem is referred as Corner Point Method.
The method comprises of the following steps:
1. Find the feasible region of the linear programming problem and determine its
corner points (vertices) either by inspection or by solving the two equations of
the lines intersecting at that point. In our example points OCA were found by inspection while
point B was found by solving for the point where the two equations intersect 5x + y = 100 (1)
and x + y = 60 (2)
2. Evaluate the objective function Z = ax + by at each corner point.

2.4 SIMULATION
Many of the problems in decision sciences are solved by standard mathematical tools. But
application of these mathematical models is based on certain assumptions. In real business
scenario at times it becomes quite difficult to satisfy these assumptions. Under these
circumstances managers use the technique of simulation. It is a very popular technique.
Simulation means imitation. In simulation, a given system is copied and artificial environment
is created to examine the behaviour of the system under specified conditions. It is a
descriptive and not an optimizing technique. It does not generate solutions but enables
managers to test their alternative solutions on a model that is an imitation of the original
situation. Where optimization models cannot be easily applied because of non-fulfillment of
assumptions, cost or other practical difficulties; simulation models offer a convenient,
inexpensive and quick way to evaluate alternative systems.
The technique of simulation has very extensive areas of application. It is used to train
managers to handle real life business challenges and solve very complex business problems.
Through simulation, business engineers are taught to evaluate aerodynamic properties of
aeroplanes by placing their models in wind tunnels where air is blown to examine its impact
on the model. “Decision simulators” have been developed and are in use in the corporate
world through which managers test various alternative decisions and their impact on corporate
profitability. Another field where simulation is widely applied is designing. Alternative
system designs are developed and evaluated to measure system performance. Some of the
well-known areas of simulation application are waiting lines, production designing, capacity
planning, manpower and machine scheduling, location, layout and typically complex
problems of inventory and maintenance.
Process of simulation
The process of simulation involves the following steps:
1. Definition of the problem: For simulating a problem it must be well defined, its
purpose and objectives should be clearly specified. The objectives, controllable and
uncontrollable variables should also be defined. For example a wholesaler in apples
wishes to maximize his revenues by a simulation model. His objective is maximization
of revenue, controllable variable is quantity of apples procured and uncontrollable
variable is market demand for apples. In the absence of clarity of objective, the cost
and efforts put in developing a simulation model will be wasted.
2. Construction of a simulation model: A simulation model imitates a real situation. It can

be a physical model or a mathematical model or a mixture of both. We shall discuss
only mathematical models. Data collection is an important step of a mathematical
model and its evaluation. Type of data and their volume depends upon the nature and

scope of the problem. The simulation model must be developed in a manner so that
evaluation of various alternatives can be carried out.
3. Experimentation with the model constructed: After the development of the simulation
model it is run. The starting value of the variables is an important determinant of the
model as it may bias the results. Afterwards the number of runs is decided. If the
model is deterministic, with all its parameter known and constant, then only a single
run would be enough. But if the simulation is stochastic in nature, with the parameters
subject to random variation, then it should be run a large number of times to obtain a
better idea of the model performance with real system performance.
4. Evaluating results of the simulation: A simulation can be evaluated in a number of

ways. Some experts view it as statistical inference system. Others compare it with past
performance of similar system. But the true evaluation is its performance in real
system after the implementation of the model results. This validation is usually done
by analysts’ familiar with real life system.
Advantages and disadvantages of simulation

The most commonly known advantages of simulation are as follows:
1. These days many computer software packages are available. Their application has
made simulation very easy and simple.
2. It is capable of analyzing large and complex real-world situations that cannot be done
through other quantitative models.
3. Simulation answers what if questions. Managers would like to know in advance what
options are available and which one is better.
4. Simulation does not disrupt the real system because simulation experiments are
undertaken with the model and not on the system.
5. “Time compression” is possible with simulation. Years of experience, say in the field
of advertising, in the real system can be compressed into minutes or seconds.
6. Many quantitative techniques are based on certain probability distributions such as

Poisson or normal. In the absence of these distributions these techniques cannot be
used. But simulation is quite flexible and can use any probability distribution that the
user defines.
7. Simulation models are also used for training purposes under a wide range of situations.
The main disadvantages of simulation are:
1. Development of a simulation model for a complex situation may be very expensive,

needs considerable patience and efforts and time consuming. A complicated model
may take months or years to build.
2. Simulation is based on trial and error approach hence unable to provide optimal
solutions as one can obtain from linear programming. Moreover, it produces different
solutions in repeated runs.
3. Since simulation is based on random numbers it may be less accurate than

mathematical model. If direct mathematical model is available for solving a particular
problem, it is better to use that.
4. Simulation models are not standardized. For every situation the model is unique.
Therefore, different individuals may develop different models for the same situation.
AREAS OF SIMULATION APPLICATION

Simulation owes its popularity to its application in a large number of business areas where
other mathematical models fail. The detail of areas of its application is quite exhaustive but
we name a few of them. They are production planning, engineering, financial analysis,
research and development, information systems, materials, and personnel. Some of the most
common applications of simulation are explained here.
Waiting Lines Models

Simple waiting line problems can be solved by the known mathematical models of waiting
lines. For complex waiting line systems, it is not possible to apply the known formulas, and
simulation is often the only popular technique used to solve them. For example, for the coffee
home in Delhi’s Connaught Place with multiple waiting lines and multiple servers, simulation
may be the only model available to determine the appropriate number of servers required to
meet customer demand.
Inventory Management
Organizations face vast challenges in procurement and management of large number of items.
Questions of order size, safety stock, storage and in process inventory need to be answered.
Since the available inventory models assume constant demand but in practice it varies from
time to time; under such a situation the established inventory models are unable to solve
inventory problems but simulation is capable of solving such problems. Further, when
companies wish to use JIT inventory model, simulation can guide them about its benefits
without disrupting the current inventory practices.

Production and Manufacturing
There are large number of areas of production process and conversion of inputs where
simulation is frequently applied. The most common areas are scheduling, sequencing,
bottlenecks and assembly line balancing, layout and location, machine breakdown and
maintenance of machines. Some of these areas involve multiple variables that have different
probabilities of occurrence. Instead of mathematical models, simulation is a better technique
in such situations.
Capital Investment and Budgeting

Companies procure investment from a large number of sources such as shares, bonds, banks
and other agencies that carry different costs and observe different constraints. Similarly,
capital budgeting problems involve inward and outward cash flows relating to many sources.
Simulation is a very appropriate technique used by finance managers to deal with problems of
investment and capital budgeting.
Logistics
As logistics problems involve a large number of random variables, such as distance, suitable
mode of transportation under different circumstances and their costs, delivery schedule;
simulation can be used to tackle these complex problems and suggest suitable solution under
different situations. It is also used to handle complex problems of the combination of different
modes of transportation such as road and rail or shipping and road.
Public Services
Public services include water supply, fire fighting operations, medical facilities, public
transport system and law and order machinery. The size of these services is enormous and
their operations are complex. Because of their complexity and involvement of a large number
of random variables no other technique is found suitable except simulation.
Management of Environment
Civil society and governments have shown a great concern towards environment in the recent
years. Some of the plants needing environmental clearance in India are power plants, huge
steel plants, mining projects, nuclear power plants, waste disposal systems. Often permission
is not granted for many industries and plants without evaluating their impact on environment.
Simulation models have been developed to determine the impact of such a variety of large
projects on environment.
MONTE CARLO SIMULATION

Simulation can be of many types, but we shall focus on the probabilistic simulation using the
Monte Carlo method. Monte Carlo technique is based on the selection of random numbers
which range from 00 – 99, and these random numbers are used for trial run. It is a
mathematical process of simulation.

Steps in Monte Carlo Simulation
To understand the various steps of Monte Carlo simulation model, let us take an example. A
motor-cycle dealer wishes to simulate demand for motor-cycles for 10 days. He has the
following data:
Daily demand of frequency Probability Cumulative Interval of

Motor-cycles of demand of demand probability random-numbers
x P(x)
0 1 0.05 0.05 00 - 04
1 5 0.25 0.30 05 - 29
2 8 0.40 0.70 30 - 69
3 2 0.10 0.80 70 - 79
4 2 0.10 0.90 80 - 89
5 1 0.05 0.95 90 - 94
6 1 0.05 1.00 95 - 99
Total 20 1.00
Ten random numbers are: 09, 52, 71, 38, 69, 80, 92, 44, 67, 47.
1. Decide random variables. If the firm wishes to develop a simulation model for the sale
of motor-cycles, daily demand of the number of motor cycles (in the range of 0 to 6)
may be considered a random variable and denoted by x. Price may be considered
another random variable denoted by y.
2. Build a probability distribution for each random variable decided in step 1. On the
basis of actual sales of last 20 days’ demand frequency is noted and from this
frequency distribution, a probability distribution [P(x)] of demand can be developed.
3. Generate random numbers. Random numbers are such numbers each of which has
equal chance of selection at random. Instead of generating random numbers
personally, managers select them from the table of random numbers that are
conveniently available in computer software/books. Random numbers represent
random variables.
4. Establish an interval of random numbers for each variable. For creating this interval
cumulative probability distribution is computed. If the cumulative probability for zero
demand is 0.05, its range of random number will be 00 – 04; for cumulative
probability of 0.30, its range of random number will be 05 – 29 and so on.

5. Simulate a series of trials. There is no specific recommended number of runs as they
vary in different situations. The benefit of running a series of trials is that one can
achieve a reasonably stable result. The run of the above example is shown below:
Days Random Interval of Simulated

Numbers Random Numbers Demand
1 09 05 – 29 1
2 52 30 – 69 2
3 71 70 – 79 3
4 38 30 – 69 2
5 69 30 – 69 2
6 80 80 – 89 4
7 92 90 – 94 5
8 44 30 – 69 2
9 67 30 – 69 2
10 47 30 – 69 2
25
=2. 5
Average of daily demand = 10
If the manager so wishes he can undertake another trial run by selecting a new set of
ten random numbers, either using the old probability distribution or building a new
probability distribution on the basis of the first run.
EXERCISES
Q. 1 Explain the Monte Carlo Technique of simulation and how random variables are used
in a Monte Carlo process.
Q. 2 Why is simulation called a technique of last resort?
Q. 3 What are the advantages and disadvantages of simulation?
Q. 4 What are the main areas of application of simulation technique?
Example
Frontier bakery keeps stock of a popular brand of cake. Daily demand based on past
experience is as given below:
Daily 0 15 25 35 45 50
demand
Probability 0.01 0.15 0.20 0.50 0.12 0.02
Consider the following sequence of random numbers: 48, 78, 09, 51, 56, 77, 15, 14, 68, 09.
Using the sequence, simulate the demand for next 10 days.

Find out the stock situation if the owner of the bakery decides to make 35 cakes every day.
Also estimate the daily average demand for the cakes on the basis of simulated data.
Solution:
Random number coding: demand distribution

Daily Probability Cumulative Random number
deman Probability Intervals
d
0 0.01 0.01 00-00
15 0.15 0.16 01-15
25 0.20 0.36 16-35
35 0.50 0.86 36-85
45 0.12 0.98 86-97
50 0.02 1.00 98-99
The given random number 48 corresponds to 36-85 matching the demand for 35 cakes.
Calculation of demand corresponding to random numbers

Random Daily Stock at the
numbers demand End of every
Day if 35 cakes
Are orderd
48 35 0
78 35 0
09 15 20
51 35 20
56 35 20
77 35 20
15 15 40
14 15 60
68 35 60
09 15 80
Total demand 270
Thus, average daily demand is 270/10=27 cakes.

2.5 ASSIGNMENT AND TRANSPORTATION
The assignment model is a special form of a linear programming model. In the assignment
model, the supply at each source and the demand at each destination are each limited to one
unit. The following example from the text will be used to demonstrate the assignment model
and its special solution method. The Football Association of Zambia (FAZ) has four football
games on a particular match day. The Association wants to assign a team team of officials to
the four games in a way that will minimize the total distance traveled by the officials. The
distances in Kilometers for each team of officials to each game location are shown in the table
below:
Table 1
Game site
Officials Ndola Lusaka Solwezi Kitwe
A 210 90 180 160
B 100 70 130 200
C 175 105 140 170
D 80 65 105 120
The supply is always one team of officials, and the demand is for only one team of officials at
each game. The table is in the proper form for the assignment. The first step in the assignment
method of solution is to develop an opportunity cost table. We accomplish this by first
subtracting the minimum value in each row from every value in the row. These computations
are referred to as row reductions. In other words, the best course of action is determined for
each row, and the penalty or “lost opportunity” is developed for all other row values. The row
reductions for this example are shown in the table below.

Game site
A 120 0 90 70
B 30 0 60 130
C 70 0 35 65
D 15 0 40 55
Next, the minimum value in each column is subtracted from all column values. These
computations are called column reductions and are shown in the table below. It represents the
completed opportunity cost table for our example. Assignments can be made in this table
wherever a zero is present. For example, team A can be assigned to Lusaka. An optimal
solution results when each of the four teams can be uniquely assigned to a different game.
Game site
A 105 0 55 15
B 15 0 25 75
C 55 0 0 10
D 0 0 5 0
Notice in the table above that the assignment of team A to Lusaka means that no other team
can be assigned to that game. Once this assignment is made, the zero in row B is infeasible,
which indicates that there is not a unique optimal assignment for team B. Therefore, the table
above does not contain an optimal solution. A test to determine if four unique assignments
exist in the table is to draw the minimum number of horizontal or vertical lines necessary to
cross out all zeros through the rows and columns of the table. For example, the table below
shows that three lines are required to cross out all zeros.

Game site
A 105 0 55 15
B 15 0 25 75
C 55 0 0 10
D 0 0 5 0
The three lines indicate that there are only three unique assignments, whereas four are
required for an optimal solution. (Note that even if the three lines could have been drawn
differently, the subsequent solution method would not be affected.) Next, subtract the
minimum value that is not crossed out from all other values not crossed out. Then add this
minimum value to those cells where two lines intersect. The minimum value not crossed out
in the table above is 15. The second iteration for this model with the appropriate changes is
shown below.
Game site
A 90 0 40 0
B 0 0 10 60
C 55 15 0 10
D 0 15 5 0
No matter how the lines are drawn in the table above, at least four are required to cross out all
the zeros. This indicates that four unique assignments can be made and that an optimal
solution has been reached. Now let us make the assignments from the table above. First, team
A can be assigned to either the Lusaka or Kitwe game. We will assign team A to Lusaka first.
This means that team A cannot be assigned to any other game, and no other team can be
assigned to Lusaka. Therefore, row A and the Atlanta column can be eliminated. Next, team B
is assigned to Ndola. (Team B cannot be assigned to Lusaka, which has already been
eliminated.) The third assignment is of team C to the Solwezi game. This leaves team D for

the Kitwe game. These assignments and their respective distances (from Table 1) are
summarized as follows.
Assignment Distance
Team A : Lusaka 90
Team B : Ndola 100
Team C : Solwezi 140
Team D : Kitwe 120
450 kilometres
Now let us go back and make the initial assignment of team A to Kitwe (the alternative
assignment we did not initially make). This will result in the following set of assignments.
Assignment Distance
Team A : Kitwe 160
Team B : Lusaka 70
Team C : Solwezi 140
Team D : Ndola 80
450 Kilometres
These two assignments represent multiple optimal solutions for our example problem. Both
assignments will result in the officials traveling a minimum total distance of 450 kilometres.

2.6 QUEUING THEORY
We are all familiar with queues – but few of us enjoy them. Managers can shorten queues but
only by using more servers, and this means higher costs. Models of queues look for ways to
reduce the time for which customers wait and still give acceptable costs. In practice, not all
queues involve people and there are many queues of inanimate – and even intangible –
objects.
Features of queues
Queues form when customers want some kind of service, but find that the server is already
busy. Then they have the choice of either joining a queue and waiting to be served, or leaving
and possibly going to a competitor with a server who is not busy. Here we consider the
situation where they decide to join the queue. As you know, you are likely to meet a queue
whenever you buy a train ticket, get money from a bank, go to a supermarket, wait for traffic
lights to change and in many other circumstances.
However, not all queues involve people – for example, there are queues of programs waiting
to be processed on a computer, telephone calls waiting to use a satellite, items moving along
an assembly line, faulty equipment waiting to be repaired, ships queuing for a berth,
aeroplanes queuing to land and so on. By convention a ‘customer’ is anyone or anything
wanting a service, and a ‘server’ is the person or thing providing that service. Whatever
parties are involved; all queues have common features – the figure below shows a basic
queuing system.
A single server queuing system

There are many different configurations of queues, with variations including:
 the number of servers may vary from one to a very large number
 customers may arrive singly or in batches – for example, when a bus-load of people
arrive at a restaurant
 arrivals may be at random, following some probability distribution, or organized

through an appointment system
 customers may form a single queue or a separate queue for each server
 customers may be served individually or in batches – for example, at an airport

customers are served a plane-load at a time
 servers may be in parallel – where each does the same job – or in series – where each
gives part of the service and then passes the customer on to the next stage
 service time may be constant or variable
 customers may be served in order of arrival or in some other order – for example,
hospitals admit patients in order of urgency.
The quality of a service is judged – at least in part – by the time that customers have to wait.
You might assume any process in which customers have to wait a long time is inherently
inefficient – more efficient operations would have smoother flows of work and move
customers and servers quickly on to other tasks. Specifically, you know from experience that
to stand in a long queue of people is irritating (at best) and encourages customers to move to
other services and organizations. So an initial aim might be to have short waiting times.
However, there can be genuine reasons for having long queues – particularly when demand is
variable, service is restricted or organizations are trying to reduce the cost of servers. For
instance, you probably have to queue in morning rush-hour traffic (because of variable
demand on the road network), to wait for specialized medical treatment (because there is a
limited supply of servers) and at a supermarket checkout (because it is too expensive to
provide enough check-outs to eliminate queues).

The length of a queue depends on three factors:
 the rate at which customers arrive
 the time taken to serve each customer
 the number of servers available.
Call centers, for example, try to reduce the time that callers wait by reducing the time servers
spend talking to each one. Usually, though, the easiest thing that managers can adjust is the
number of servers, and in any particular circumstances having a lot of servers gives the
shortest queues. Unfortunately, servers cost money and these short queues come with high
costs. So managers look for a balance that seems to satisfy all parties – combining reasonable
queue length with acceptable costs. The point of balance differs according to circumstances.
When you visit a doctor’s surgery you often have a long wait.
This is because the doctor’s time is considered expensive while patients’ time is cheap. To
make sure that doctors do not waste their valuable time waiting for patients, they make
appointments close together and patients are expected to wait. On the other hand, in petrol
stations the cost of servers (petrol pumps) is low and customers can nearly always drive to a
competitor when there is a queue. Then it is better to have a large number of servers with low
utilization, ensuring that customers only have a short wait in any queue.
Single-server queues
The simplest type of queue has:
 a single server dealing with a queue of customers
 random arrival of customers joining a queue
 all customers waiting to be served in first-come-first-served order
 random service time.

It has been discussed that a Poisson distribution describes random occurrences, so we can use
this to describe customer arrivals. When the average number of customers arriving in unit
time is λ the probability of r arrivals in unit time is given by the Poisson distribution:
Service time is also random – but now the data is continuous so we cannot use a discrete
Poisson distribution. Instead, we use a closely related continuous distribution called a negative
exponential distribution. It has the useful feature that the probability of service being
completed within some specified time, T, is:
And the probability that service is not completed by time T is 1 minus this, or:
When an average of 10 customers are served an hour the probability that service takes less
than 5 minutes is:
The probability that it takes more than 10 minutes is:
Now we have descriptions of the random arrival of customers in terms of λ (the mean arrival
rate) and of random service times in terms of μ (the mean service rate). If the mean arrival rate
is greater than the mean service rate, the queue will never settle down to a steady state but will
continue to grow indefinitely. So any analysis of queues must assume a steady state where μ is
greater than λ.
We can derive some standard results for a single-server queue – which are called the
operating characteristics.
Imagine a queue where the mean arrival rate is 2 customers an hour and the mean service rate
is 4 customers an hour. On average a customer arrives every 30 minutes, and takes 15 minutes
to serve, so the server is busy for half the time. Extending this reasoning, we can suggest that
a server is generally busy for a proportion of time λ / μ, and this is described as the utilization
factor. This is also the proportion of time that a customer is either being served or waiting –
technically described as being ‘in the system’. Looking at it in other ways, it is also the
probability that an arriving customer has to wait, and is the average number of customers
being served at any time.
utilization factor = λ / μ
The probability that there is no customer in the system is:
P0 = 1 − the probability there is a customer in the system
=1−λ/μ
This is the probability that a new customer is served without any wait.
The probability that there are n customers in the system is:
We can use this result – to calculate some other characteristics of the queue. To start with, the
average number of customers in the system is:
L=
The average number of customers in the queue is the average number in the system minus the
average number being served:

Lq =
The average time a customer has to spend in the system is:
The average time spent in the queue is:
Wq

UNIT 3: Applying business planning techniques
Learning Outcomes:
 Apply project planning and control

 Describe Material Requirements Planning (MRP) systems
 Analyze Inventory management
 Analyze Aggregate planning
Material Requirements Planning (MRP) systems
Material Requirements Planning is a time phased priority-planning technique that calculates

material requirements and schedules supply to meet demand across all products and parts in
one or more plants.
Information Technology plays a major role in designing and implementing Material

Requirements Planning systems and processes as it provides information about manufacturing
needs (linked with customer demand) as well as information about inventory levels. MRP
techniques focus on optimizing inventory. MRP techniques are used to explode bills of
material, to calculate net material requirements and plan future production.
MRPII stands for Manufacturing Resource Planning and represents an extension of MRP.
MRPII points to computer based planning and scheduling designed to improve management’s
control of manufacturing and its support functions. MRPII maps an extension of MRP to
capture all manufacturing requirements including materials, human resources, scheduling, etc.
An MRP system is intended to simultaneously meet three objectives:
 Ensure materials and products are available for production and delivery to customers.
 Maintain the lowest possible level of inventory.
 Plan manufacturing activities, delivery schedules and purchasing activities
What is Material Requirements Planning (MRP)?
The globalization of the economy and the liberalization of the trade markets have formulated
new conditions in the market place which are characterized by instability and intensive
competition in the business environment. Competition is continuously increasing with respect
to price, quality and selection, service and promptness of delivery. Removal of barriers,
international cooperation, technological innovations cause competition to intensify. In terms
of manufacturing emphasis is placed on reducing cost while improving quality. In addition,
other factors such as timely delivery of the product become critical (this is captured by
emphasis in Just in Time or JIT in short) techniques.
A key question to a MRP process is the number of times a company replenishes (or turns
around) inventory within a year. There are accounts of inventory annual turnover ratios of
greater than 100, mainly reported by Japanese companies. One can readily realize that a high
inventory ratio is likely to be conducive to lowering production cost since less capital is tied
up to unused inventory.
MRP systems use four pieces of information to determine what material should be ordered
and when:
 the master production schedule, which describes when each product is scheduled to be
manufactured;
 bill of materials, which lists exactly the parts or materials required to make each product;
 production cycle times and material needs at each stage of the production cycle time; and,
 Supplier lead times.
The master schedule and bill of materials indicate what materials should be ordered; the
master schedule, production cycle times and supplier lead times then jointly determine when
orders should be placed.
The Master Production Schedule includes quantities of products to be produced at a given

time period. Quantities are included both at aggregate and detailed levels. Aggregate may
refer to monthly production and detailed may refer to weekly or daily production. The master
production schedule takes the form of a table in which rows represent products and columns
represent time components. Entries of the table map to units of each product to be produced at
a given time period.
Bill of Materials gives information about the product structure, i.e., parts and raw material
units necessary to manufacture one unit of the product of interest.
MRP was pioneered in the 1970’s with the work of Orlicky. Later evolved or became part of
integrated to Manufacturing Resource Planning systems (or MRPII). MRPII is a computer
based planning and scheduling system designed to improve management’s control of
manufacturing and its support functions.
In today’s corporate environment MRPII is often termed as ERP (or Enterprise Resource
Planning).
MRPII represents a group of software programs designed to tie together disparate company
functions to create more efficient operations in areas such as assembly or delivery of products
or services.
Thus MRP has evolved to become a component of a MRPII system. Technically, MRPII
extends MRP and links it with the company’s information resources such as human resource
information system, financial management, accounting, sales, etc. Such extension is typical
according to modern trends in business management and modeling and made possible by
advances in information technology. On the other hand, the need to integrate is well
established in management thinking and practice. Since the pioneering work of Anthony
during the sixties, management decision-making processes are viewed from extending from
strategic planning, to management control and to operational control. MRP systems lay in-
between management control and operational control processes. However, as detailed
production data are linked with overall organizational information resources it becomes clear
that MRP and MRPII system implementations play a significant role in company’s corporate
advantage.
Problems with MRP systems:
The major problem with MRP systems is the integrity of the data. If there are any errors in the
inventory data, the bill of materials (commonly referred to as 'BOM’) data, or the master
production schedule, then the output data will also be incorrect.
Most vendors of this type of system recommend at least 99% data integrity for the system to
give useful results.
Another major problem with MRP systems is the requirement that the user specify how long it
will take a factory to make a product from its component parts (assuming they are all
available). Additionally, the system design also assumes that this "lead-time" in
manufacturing will be the same each time the item is made, without regard to quantity being
made, or other items being made simultaneously in the factory.
Inventory Management
A business can run smoothly its operating activities only when appropriate amount of
inventory is maintained. Inventory affects all operating activities like manufacturing,
warehousing, sales etc. The amount of opening inventory and closing inventory should be
sufficient enough so that the other business activities are not adversely affected. Thus,
inventory plays an important role in operations management.
Meaning & Types of Inventory
Inventory is an asset that is owned by a business that has the express purpose of being sold to
a customer. Inventory refers to the stock pile of the product a firm is offering for sale and the
components that make up the product. In other words, the inventory is used to represent the
aggregate of those items of tangible assets which are –
 Held for sale in ordinary course of the business.

 In process of production for such sale.
 To be currently consumed in the production of goods or services to be available for sale.
The inventory may be classified into three categories:
 Raw material and supplies: It refers to the unfinished items which go in the production
process.
 Work in Progress: It refers to the semi-finished goods which are not 100% complete but
some work has been done on them.
 Finished goods: It refers to the goods on which 100% work has been done and which are
ready for sale.

Meaning of Inventory Management
Inventory management is the practice overseeing and controlling of the ordering, storage and
use of components that a company uses in the production of the items it sells. A component of
supply chain management, inventory management supervises the flow of goods from
manufacturers to warehouses and from these facilities to point of sale. Inventory control
means efficient management of capital invested in raw materials and supplies, work- in –
progress and finished goods.
Significance of holding inventory
Inventory is considered to be one of the most important assets of a business. Its management
needs to be proactive, accurate and efficient. Inventory is essential for every organization to
ensure smooth running of the production process, to reduce the ordering cost of inventory, to
take advantage of quantity discount, avoid opportunity loss on sales, to utilize and optimize
the plant capacity and to reduce the overall price. Thus, it can be said that inventory is
inevitable and has to be maintained in appropriate quantity. However, the concept of Just In
Time (JIT) is becoming popular which is an inventory strategy companies employ to increase
efficiency and decrease waste by receiving goods only as they are needed in the production
process, thereby reducing inventory costs. This method requires producers to forecast demand
accurately.
Objectives of Inventory Management
The objective of inventory management is to maintain inventory at an appropriate level to

avoid excess or shortage of inventory. Inventory management systems reduce the cost of
carrying inventory and ensure that the supply of raw material and finished goods remains
continuous throughout the business operations. The objectives specifically may be divided
into two categories mentioned below:
A. Operating objectives: They are related to the operating activities of the

business like purchase, production, sales etc.
a. To ensure continuous supply of materials.
b. To ensure uninterrupted production process.
c. To minimize the risks and losses incurred due to shortage of inventory.
d. To ensure better customer services.
e. Avoiding of stock out danger.
B. Financial Objectives:
a. To minimize the capital investment in the inventory.
b. To minimize inventory costs.
c. Economy in purchase.

Apart from the above objectives, inventory management also emphasize to bring
down the adverse impacts of holding excess inventory. Holding excess inventory
lead to the following consequences:
 Unnecessary investment of funds and reduction in profit.

 Increase in holding costs.
 Loss of liquidity.
 Deterioration in inventory.

UNIT 4: Demonstrating the use of Probability in decision making
Learning Outcomes:
 Apply addition and multiplicative laws of probability

 Use probability trees and conditional probabilities
 Apply Bayes’ rule
 Apply discrete probability distributions
 Apply continuous probability distributions
1.1 PROBABILITY THEORY

The concept of probability
Likelihood and chance are expressions used in everyday life to denote a level of uncertainty.
Probability is the mathematical term used to describe the likelihood of an event happening.
Introducing probability
Probability is a measure of likelihood and can be stated as a percentage, a ratio, or more
usually as a number from 0 to 1.
Consider the following
Probability = 0 = impossibility. Something with a probability of 0 will not happen
Probability = 1 = certainty. Something with a probability of 1 is certain to happen.
Probability = 1/2 = a 50% chance of something happening.
Probability = 1/4 = a 1 in 4 chance of something happening.
Expressing probabilities
In statistics, probabilities are more commonly expressed as proportions than as percentages.
(a) When expressed as a proportion, the probability is a value between 0 and 1.
(b) The sum of the probabilities of all possible outcomes is 1
Notation of probabilities
It is usual to use the letter p to refer to the probability that something will happen
We may also use the letter q to refer to the probability that something will not happen
An outcome may therefore have a probability of p that it will happen and q that it will not
happen, and q = (1-p). For example, if a student has a 70% chance of passing an examination,
the probability of passing is p = 0.7 and the probability of failure is q = 0.3.

Assessing probabilities
It is useful to consider how probability can be quantified. A businessman might estimate that
if the selling price of a product is raised by K2,000, there would be a 90% probability that
demand would fall by 30%. But how would he have reached his estimate of 90% probability?
There are several ways of assessing probabilities
 Probabilities may be measurable with mathematical certainty
- If a card is taken from a deck of 52 cards, there is a 0.25 probability that it will be
a card in the spades suit.
- There is a 1 in 13 probability that the card will be an Ace.
Probabilities that can be measured with mathematical certainty are called theoretical
probabilities, or a ‘priori’ probabilities.
 Probabilities may be measurable from an analysis of past experience, on the
assumption that what has happened in the past will continue to happen in the future.
 In some cases, a manager may estimate a probability on the basis of his experience or
judgement.
 Probabilities can be estimated from research or surveys, such as market research
surveys.
Theoretical and empirical probabilities
Probabilities that cannot be measured with mathematical certainty, but are based on past
experience or research are known as empirical probabilities.
Probabilities that are based on judgement or guesswork are called subjective probabilities.
Probabilities that can be measured with mathematical certainty are called theoretical
probabilities, or a priori probabilities. The formula for a theoretical probability is:
Probability of an event = Number of ways that the event can occur
Total number of possible outcomes
Example 1
What is the probability of throwing an even number from a six-sided fair die?
Number of ways that the event can occur: There are three even numbers on a 6-sided die: 2,
4, 6
Total number of possible outcomes: On a 6-sided die there are six possible numbers which
could result, being:
1, 2, 3, 4, 5, 6.
Therefore, the probability of rolling an even number is: (3/6) = (1/2) = 0.5 = 50%
Exercise
There are 2 red balls, 14 blue balls, 35 black balls, and 8 white balls in a box.

If a ball is chosen at random from the bag, what is the probability that it:
(a) Will be a black ball
(b) Will be a white ball
(c) Will not be a red ball
(d) Will be neither black nor white?
PROBABILITY LAWS OF ADDITION AND MULTIPLICATION
Probability rules are important to understand when dealing with probabilities of more than one
event or occurrence of an event.
The simple addition rule
The simple addition rule or law of probability applies when there are two mutually exclusive
events, A and B. The probability that A or B will occur is the sum of the probability that A
will occur and the probability that B will occur.
P(A or B) = P(A) + P(B)
Mutually exclusive outcomes are outcomes that cannot occur all at the same time.
Example
A major conference of accountancy is held in Johannesburg. 1200 delegates attended the
conference. Out of these 60 Came from Zambia a 24 from Botswana. What is the probability
that a delegate is chosen random from the audience will have come from either Zambia or
Botswana?
Let A = Delegate from Zambia
B= Delegate from Botswana
P(A) = 60/1200 = 0.05
P(B) = 24/1200 = 0.02
Coming from Zambia and coming from Botswana are mutually exclusive events. A delegate
cannot have come from both countries. So the probability that a randomly selected delegate
will have come from either Zambia or Botswana = P(A or B) = 0.05 + 0.02 = 0.07
The simple multiplication rule
The simple multiplication rule of probability applies when there are two independent events,
A and B. The probability that both A and B will occur is the probability that A will occur
multiplied by the probability that B will occur.
P(A and B) = P(A) * P(B)
Independent events are events where the outcome of one event in no way affects the outcome
of the other events
Example

A construction company has tendered two contracts. The probability of getting a government
contract to build a school is 0.7, while the probability of getting the contract to build a clinic is
0.3. The probability of getting either contract is independent of the probability of getting the
other contract. What is the probability that the company will get the contracts to build both the
school and the clinic?
Let A = contract to build a school B= contract to build a clinic
P(A) = 0.7 P(B) = 0.3
P(A and B) = 0.7 x 0.3 = 0.21
The general rule of addition

The general rule of addition is an adjustment to the simple addition rule. It applies when
outcomes are not mutually exclusive, so that both outcomes can happen at the same. However,
there is no positive connection between the two possible outcomes. This means that if A
happens, this does not affect whether (or not) B will happen.
The general rule of addition for two events, A and B, which are not mutually exclusive, is as
follows.
P(A or B) = P(A) + P(B) – P(A and B)
Example
A firm of accountants has 50 employees. 30 are qualified accountants and 10 are women.
There is no connection between the gender of an employee and whether he or she is an
accountant.
What is the probability that an employee of the company, selected at random, will be either a
qualified accountant or female?
An employee maybe both a qualified accountant and female, so the general rule of addition
applies in this case.
Let A = being an accountant B = being female
P(A) = 30/50 = 0.6 P(B) = 10/50 = 0.20
Probability of being either a qualified accountant or female = 0.60 + 0.20 – (0.60*0.20) = 0.80
– 0.12 = 0.68
The general rule of multiplication
The general rule of multiplication applies when there are independent events A and B, where
the outcome of one event does not affect the outcome of the other event. However, the
outcome of the second event is conditional or dependent on the outcome of the first event.
Dependent or conditional events are events where the outcome of one event depends on or is
conditional on the outcome of the others.
The rule is for calculating the probability that both events will happen.
P(A and B) = P(A) x P(B/A) = P(B) x P(A/B)
P(B/A) means the probability that B will happen given that A has happened.
P(A/B) means the probability that A will happen given that B has happened.
Example
A bag contains 4 blue balls and 6 white balls. Two balls will be drawn at random from the
bag, one after the other. The first ball will not be replaced in the bag after it has been drawn
out. What is the probability that:
(A) The first ball drawn will be a blue ball (only one ball is drawn)
(B) The second ball drawn will also be blue given that the first is blue.
sWhen A and B are independent events, then P(B/A) = P(B) since by definition, the
occurrence B (and therefore P(B)) does not depend upon the occurrence of A. Similarly,
P(A/B) = P(A)
Probability trees
Probability trees are a useful pictorial representation of a probability problem. They show all
of the possible outcomes.
When there are complex situations where a large number of outcomes may occur, and
especially when there are conditional probabilities, it may help to construct a probability tree
to show all the possible outcomes and their associated probabilities. Tree diagrams are useful
for laying out all the possible outcomes or alternatives when dealing with independent events
involving uncertainty in business. It provides a systematic and logical way of solving decision
problems.
Constructing a probability tree
There are two stages in preparing a probability tree:
 Drawing the tree itself to show all the possibilities or possible outcomes
 Putting in the numbers (the probabilities and outcome values)
It is conventional to draw probability trees from left to right, so that all the possible final
outcomes are shown on the right hand side of the tree.
When there are several different possible outcomes, each possible outcome is shown on a
separate branch (a line) that is drawn from left to right on the tree.
BAYES THEOREM
Bayes theorem is a formula which reverses conditional probabilities. It can be used to estimate
the probability of a prior outcome, given a final outcome that is conditional on it. Bayes
theorem reverses conditional probability to find a conditional probability P(A/B), given its
inverse P(B/A).
Exercise
A hospital might conduct test to check whether patients have a disease. The tests are correct
most of the time, but are incorrect some of the time. A test finds that a patient has the disease.
What is the probability that the patient does have the disease, given that the test is positive?
What is the probability that the test result is incorrect, and the patient does not have the
disease?
Similarly if the test shows that a patient does not have the disease, what is the probability that
the test is incorrect, and the patient does have the disease?
The theorem is used to solve problems involving pairs of conditional probabilities of the
general rule of multiplication.
Since P (B) X P (A/B) = P (A) X P (B/A)
It follows that:
P (A/B) =
Exercise
In the accounting department of a company, 60 % of bookkeepers have been given formal
training in accountancy and 4% have not. Employees who have not been trained may make a
mistake, and it has been discovered that they make a mistake 50% of the time.
A supervisor checks work of an employee selected at random and finds that the task has been
performed correctly. What is the probability that the task was performed by an employee who
has not been trained?
Exercise
The following example is more complex.
Mopani and konkola copper mines have to be tested for the emission of pollutants. It is known
that 25% of all mines emit excessive pollutants. When tested, 99% of all copper mines that
emits excessive pollutants will fail the test, but 17% of the mines that do not emit excessive of
pollutants will also fail the test.
What is the probability that a mine that actually emits excessive amounts of pollutants?
TAKE HOME QUESTION

Chipata General Hospital records indicate that 20% of chipata residents have malaria, and the
laboratory test accurately discloses the malaria parasites 95% of the time. The tests also
indicate positive among people who do not have malaria: this happens on 10% of occasions.
Determine the probability that a person with malaria actually tests positive.

Bayes theorem can be used to work out the probability of a prior event, given a conditional
outcome. It re works the general rules of multiplication so that:
P (A/B) =
1.2 Probability distributions

Introduction: to nature of a probability of a probability distribution
A probability distribution is a form of frequency distribution, where possible outcomes are not
expressed as a number of frequencies, but as a probability that each particular outcome will
occur. In other words, a probability distribution is the spread of possible outcomes and the
associated probability that each outcome will happen. The total of the probabilities within a
probability distribution equals 1 (100%)
Coin toss
EXAMPLE
A fair coin is tossed three times. Construct a probability distribution for the number of times
the coin will be expected to come down heads.
Constructing a probability distribution from empirical evidence

A probability distribution can be constructed using empirical evidence, starting with a
frequency distribution table of actual recorded observations.
Relative frequency
EXAMPLE
Motor vehicles are tested for defective by taking samples of six completed cars. Records for
the most recent 200 samples taken for testing shows that the following number of defectives
were found:
Number of defectives, x: 0123456

Frequency (number of samples) 150 30 14 4 2 0
This frequency distribution can be converted into a probability distribution.
The above can be transformed into a probability distribution by dividing the frequency of each
by the total frequency in this example, 200. This is the relative frequency of distribution.

Note that probabilities in a probability distribution must always add up to one.
In the real world, it has been found that many probability distributions have a similar shape or
pattern, and this can help with making estimates of probabilities of future events or
outcomes .this will be explained in more detail later.
Discrete probability distributions
A discrete probability distribution is a probability distribution in which all the possible
outcomes are identifiable as separate (discrete) values and there is a probability for each of
these specific outcomes. For example the number of times an outcome may occur may be 0, 1,
2, and 3 and so on, and probabilities are given to each possible outcome.
...................................
A probability distribution presents all the possible outcomes from a number of events and
their associated probabilities .the total of possibilities in a probability distribution is 1.
In a discrete probability distribution, there are specific possible outcomes, and a probability
that each of these will occur. For example there are probabilities that an event will happen 0
times, 1 time, 2 times, 3 times and so on.
Binomial Distribution
Consider an experiment of tossing a coin 3 times. You will be able to write the sample space
and be able to write the probability of any interest of the required random variable, such as;
find the probability that exactly one head appears on the second toss. All you need to do is to
write the sample space and the event which will give you the required probability. Soppose
now that you are asked to toss a coin 20 times, and being asked to find the probability that at
least two heads. This method of writing the sample space becomes tedious and all we have to
do is use the binomial dsitribution where we need to know that the experiment has two
possible outcomes, a head called the ‘success’ or no head which is ‘failure’. So we define a
random variable x as the number of heads that can appear as ; X ={0,1,2,3,....,20} .
In genral, if a single trial can have only two possible mutually exclusive and exhaustive
results, either ‘success’ with the probabilty p or ‘failure’ with probability 1-p( = q), then in a
series of n trials, the probability of x successes is given by
p( x )=P ( X=x )= ( nx ) p (1− p) =(nx ) p q
x n− x x n− x
p( X =x ) is the term in p x of the binomial expansion of (q+ p )n and so this distribution is

called the Binomial distribution. It is applicable in many situations and gives the probability
distribution of the discrete random variable X where X is the number of successes in n trials
when the following criteria apply:
There are a fixed number of trials. Think of trials as repetitions of an experiment. The letter n
denotes the number of trials.

There are only 2 possible outcomes, called "success" and, "failure" for each trial. The letter p
denotes the probability of a success on one trial and q denotes the probability of a failure on
one trial. p+q=1.
The n trials are independent and are repeated using identical conditions. Because the n trials
are independent, the outcome of one trial does not affect the outcome of any other trial.
Another way of saying this is that for each individual trial, the probability, p, of a success and
probability, q, of a failure remain the same. For example, randomly guessing at a true - false
statistics question has only two outcomes. If a success is guessing correctly, then a failure is
guessing incorrectly. Suppose Joe always guesses correctly on any statistics true - false
question with probability p=0.6. Then, q=0.4 .This means that for every true - false statistics
question Joe answers, his probability of success (p=0.6) and his probability of failure (q=0.4)
remain the same.
The outcomes of a binomial experiment fit a binomial probability distribution. The random
variable X= the number of successes obtained in the n independent trials.
The mean, μ, and variance, σ2, for the binomial probability distribution is μ=np and σ2=npq.
The standard deviation, σ, is then σ=\ npq .
Example 1
At ABC University, the withdrawal rate from an elementary physics course is 30% for any
given term. This implies that, for any given term, 70% of the students stay in the class for the
entire term. A "success" could be defined as an individual who withdrew. The random
variable is X = the number of students who withdraw from the elementary physics course per
term.
Example 2
Suppose you play a game that you can only either win or lose. The probability that you win
any game is 55% and the probability that you lose is 45%. If you play the game 20 times,
what is the probability that you win 15 of the 20 games? Here, if you define X = the number of
wins, then X takes on the values X = 0, 1, 2, 3, ..., 20. The probability of a success is p=0.55.
The probability of a failure is q=0.45. The number of trials is n=20. The probability question
can be stated mathematically as P(X =15), using the general model given above as:
p( x )=P ( X=x )= ()
n x
x
n
()
p (1− p) n− x= p x qn− x
x
Where p=0.55, q=0.45 , n = 20 and x= 15. You substitute these values in this model and your
answer is the required probability.
Example 3
A fair coin is flipped 15 times. What is the probability of getting more than 10 heads? Let X =
the number of heads in 15 flips of the fair coin. X takes on the values x = 0, 1, 2, 3, ..., 15.
Since the coin is fair, p = 0.5 and q = 0.5. The number of trials is n = 15. The probability
question can be stated mathematically as P(X >10) .
Example 4

Approximately 70% of statistics students do their homework in time for it to be collected and
graded. In a statistics class of 50 students, what is the probability that at least 40 will do their
homework on time?
Example 5
It has been stated that about 41% of adult workers have a high school diploma but do not
pursue any further education. If 20 adult workers are randomly selected, find the probability
that at most 12 of them have a high school diploma but do not pursue any further education.
How many adult workers do you expect to have a high school diploma but do not pursue any
further education?
Let X = the number of workers who have a high school diploma but do not pursue any further
education.
X takes on the values 0, 1, 2, ..., 20 where n = 20 and p = 0.41. q = 1 - 0.41 = 0.59. X ~
B(20,0.41)
Find P(X ≤12). P(X ≤12) =0.9738. (calculator or computer)
The result is P(X ≤12) =0.9738.
The probability at most 12 workers have a high school diploma but do not pursue any further
education is 0.9738
The graph of X ~ B(20,0.41) is:
The y-axis contains the probability of X, where X = the number of workers who have only a
high school diploma.
The number of adult workers that you expect to have a high school diploma but not pursue
any further education is the mean, μ = np
The formula for the variance is σ2=npq. The standard deviation is σ = √ npq
σ=√(20)(0.41)(0.59)=2.20 ..
Exercise
1. The probability that a person chosen at random is left handed is 0.1. What is the
probability that in a group of ten people there is one, and only one, who is left handed?
What is the most likely number of left handed people in a group of ten?

2. A shopkeeper found that 5% of the eggs received from a central distributing agency
were stale on delivery, and reduced his prices by 5%. A housewife requiring ten fresh
eggs was advised to purchase a dozen, the shopkeeper claiming that it was more likely
than not that all the eggs in the dozen would be fresh, and furthermore that there was
only a one in ten chance of two eggs in the dozen being stale. Are the shopkeeper’s
claims valid?
What is the probability that, of the dozen eggs purchased,

(a) one egg is stale
(b) not more than two eggs are stale?
Poisson distribution
Consider the number of telephone calls arriving at a small telephone exchange in 100
consective time intervals of five minutes. What kind of distribution would we expect in this
situation? We may think that this is a binomial distribution, since the number of calls is a
discrete variable. However, a little further thought reveals certain problems, such as what
number corresponds to the binomial parameter n and what meaning can we give to p, the
probability of a “ success”, in this cas the occurance of a telephone call? In this particular
situation there is one statistic which we can calculate and this is the mean number of telephone
calls , m , for a five –minute time interval, for example;
Consider the distribution of phone calls arriving at an exchange given below:
Number of time fx
intervals, f
Number of
calls
0 71 0
1 23 23
2 4 8
3 2 6
4 or more 0 0
100 37
m=
∑ fx =37 =0 .37
Therefore, ∑ f 100
For this module we consider the discrete distribution , with m , the mean number of calls for a
five minutes time interval follows a poisson distribution which gives the probability of x
events occuring in a given interval as:

mx e− λ
P( x )=P ( X=x )=
x!
Where the mean number of events in the same interval is m . Besides being applicable to
events randomly distributed in time such as telephone calls. It is also applicable to events
randomly distributed in space. For it to be applicable the following conditions must be
satisfied:
Two or more events cannot occur simultaneously;
The events are independent;
The mean number of events in a given interval is constant.
m is the parameter of the distribution.
Example
The number of calls per minute passing a certain point on a road is poisson distributed with
mean 4.
Find (i) the probability that 4 cars pass in a minute
(ii) the probability that 8 cars pass in two minutes
Solutions
The mean number of cars in a minute is 4, meaning m=4
mx e− λ
P( x )=P ( X=x )=
Using the formula x ! , we have, x = 4 and substituting in this
4 4 e−4
p( X =4 )= =0 .195
formula we have 4!
88 e−8
p( X =8 )= =0. 140
The mean number of cars in two minutes is 8, so we have 8!
The mean and variance of a poisson distribution
These can be deduced from the mean and variance of the Binomial distribution in the limit
when n→ ∞ , p→0 but np → m remains constant. For a binomial distribution μ→ np .
Therefore for a poisson distribution , μ= m
2
For a Binomial distribution σ =npq . As n→ ∞ , p→0 and q=1 . Writing np → m and
q=1 gives, σ 2= m
The regorous proof of these results will not given in this module as it is more mathematical.
However, we have found an important property of the poisson distribution which is that its
mean is equal to its variance. This gives a simple test of whether or not a distribution is
approximately poissonian.
Exercise
1. The number of emergency admissions each day to a hospital is found to have a possion
distribution with mean 2.
(a) Evaluate the probability that on a particular day there will be no emergency
admissions.
(b) At the beginning of one day the hospital has five beds available for
emergencies. Calculate the probability that this will be an insufficient number for the day.
(c ) Calculate the probability that there will be exactly three admissions altogether
on two consective days.

The Normal Distribution
A mathematical model which describes approximately the shape of many frequency
distribution, such as the one in fig.. below; is the normal distribution.
For a continuous random variable X, with mean μ and standard deviation σ , the normal
distibution curve is given as the one given above. This curve has the following properties;
It is symmetrical about x= μ
It has one maximum at x= μ
As x →±∞ , f (x )→0
The area under the curve is 1.
Thus the Normal distribution may be a suitable model for a variate which is
Continuous
Unimodal
Has a symmetrical frequency distribution,
Has class frequencies which fall away rapidly as the variate moves away from the mean
2
A convenient way of denoting that X is Normally distributed with mean μ and variance σ
2
is N ( μ , σ ) . μ and σ are parameters of the distribution. Their values determine the shape
of the distribution.
Standardization
f(x) is a probability density function. It does not give the probability that X takes a certain
value but that X lies in a certain range: the probability X lies between x and x +δx is f ( x )δx .
x2
∫ f ( x )dx
The probability that X lies between x 1 and x 2 is thus given by x1
which is the area
under the curve shown below:

Unfortunately, evaluation of this integral is very difficulty nfor all values of μ and σ , since
they can take an infinite number of values. In order to overcome this challenge of intergartion,
we standardize the given variate. A standard deviate Z is defined such that when X = x the
value of Z is given by:
x −μ
Z=
σ
Which is the number of standard deviations by which x departs from μ . Using the Normal
distibution table we can find the probability that x lies between x 1 and x 2 we can write
Area between x1 and x2 = p( x 1 < Z < x 2 ) you can find this probability using the
Normal distribution table.
Example
The weight of a population of women are normally distributed with mean 60 kg and s.d. 5 kg.
What is the probability that the weight of a woman chosen at random is:
(a) less than 61 kg
(b) between 61kg and 65 kg
(c ) greater than 70 kg
Solutions
2
We have μ=60 and σ =5 if X is the weight of a woman, the X ≈N (60 , 5 ) and Z, where
(61−60)
z= =0 . 2
(a) x = 61 , 5
From the Normal distribution table, we have
p( Z<0 .2 )=0 .5793

Therfore the probability that a woman chosen at random weighs less tha 61k is 0.5793.
(65−60)
x=65 , z= =1
(b) we find the 5 ,
The probability that the weight of a woman lies between 61 and 65 kg is found as:
p(0 . 2<z<1)= p( z <1 )− p( z ,<0 .2 )=0 . 262
See figure below;
Therefore, the probability that a woman weighs between 61 and 65 kgs is 0.262.
(70−60)
x=70 , z= =2
(c ) 5
Therfore, p( z >2)=1− p( z <2)=0 .0228

That is the probability that the woman weighs more than 70kg is 0.0228.
If Z≈N (0 , 1) , find
(a) p( Z< 1) (b) p( Z> 2) (c ) p( Z<−2. 92 ) (d) p( Z<0 .543 )
If Z≈N (0 , 1) , find
(a) p(1 .3<Z <2 .3 ) (b) p(−1. 2<Z<0 .9 ) (c) p(−0 . 074<Z<2. 155 )
3. For a Normal population with mean 10, standard deviation 5, find the
probability thata member chosen at random gives a value of the variates which is
(a) < 11 (b) > 11 (c ) >5 (d) <5 (e) between 5 and 11
4. IQ scores are N(100,225)> Find the probability that a personchosen at randomhas an
IQ sore of;
(a) more than 140 (b) less than 90 (c ) between 120 and 1

UNIT 5: Applying estimation techniques and hypothesis tests to aid the
decision-making process
Learning Outcomes:

 Expain different estimation techniques
 Perform hypothesis testing
 Use Chi-Square test
 Apply Analysis of variance (Anova)
ESTIMATION TECHNIQUES WITH SAMPLES

Samples are used in statistics to make estimates which might be of value to decision makers.
Typically, the samples are used to estimate the mean of a population or a proportion in a
population. Samples are not 100% exact, and there will be some error in the sample estimate.
For example, a sample estimate of the mean of a population will not be exactly the same as
the true population mean, but might be a little over or under the population mean.
This unit is concerned mainly with estimating a population mean or proportion from a sample,
allowing for the existence of sampling error. The usual method of estimating is to establish a
confidence interval, from the sample results. A confidence interval is a range within which the
true population mean or proportion lies, at a given level of confidence.
Statistical estimation techniques are concerned mainly with making estimates based on sample
statistics. The techniques described in this unit are only valid when:
(a) The samples are taken at random
(b) The samples are quite large (30 or more)
The sampling distribution of the sample means
If we were to take a large number of samples and calculate the means of all the samples, the
average of the sample means should be the same as the population mean, but the values of the
sample means would be (approximately) normally distributed around the population mean.
This conclusion is based on the so called central limit theorem.
The standard distribution of the sample means around the population mean is called the
standard error of the mean (se).
Formula for the standard error of the mean or standard deviation of the sample mean
A very small standard error indicates that the sample means are a good approximation of the
population mean, so the smaller the standard error, the better.
However, what if we only have one sample from which to estimate the population mean, or
just a few samples?

The standard deviation of the whole population is not known. To overcome this problem, the
standard deviation of a sample is taken as the best estimate of the standard deviation of the
whole population.
When we only have one sample, we can therefore estimate the se from the data in the
individual sample. The estimated se of the mean is:
Se = a/_/n
Whre n is the size of the sample ans a is the standard deviation of the population
Points to note in the formula:
(a) The size of the se depends on n, the size of the sample. The larger the sample sieze
chosen , the smaller the se will become. Large samples are more likely to be
represenataiv of the population as a whole.
(b) The se is related to the standard deviation, a , of the population. The larger the
dispersion of values in the population, he moer likely it will be that the meanof a
sample will vary from the true mean of the population.
When estimating with samples , the means of the samples are used as an estimator of thhe
population mean and a sampe mean and its se can be used to estimate the population
mean.
Some properties of a good estimator are as follows
(a). It must be unbiased. This means that the mean of the sampling distribution equals the
population mean.
(b). It must be efficient. The most efficient estimator is the one with the smallest variance
and S.D
(c). It must be consistent. This means that as the sample size increases, the value of the
statistic approaches the population value.
(d). It must be sufficient. This means that it uses all the relevant information.
Central limit theorem (CLT)
The central limit theorem states that, given a probability distribution with a mean u and
variance a2, the sampling distributon of the mean approaches a normal distribution with a
mean (u) and a variance a2 /N as N, the sample size increases.
The CLT is a very important statement in statistical analysis, it allows the presumption of
a normal curve for large samples. The sampling distribution of the can be approximated
closely with a normal distribution, with a S.D equal to the se of the sample means.
The mean or proportion of the sample will not be equal to the true population mean or
proportion. If we took a large number of samples, their mean (or proportion) would be
approximately normally distributed around the true population mean or proportion, with a
S.D equal to the se of the mean (or proportion).

We can estiate the se of the population mean (or proportion) from an individual sample.
The se of th mean = s/ _/n
The se of a proportion = _/pq/n
Estimating a population mean: confidence limits
Point estimates and interval estimates of a population mean
If a sample of n items is drawn and the mean is found and used to estimate the population
mean, the single estimate of u is known as a point estimate of the population mean. If you
take a sample and compute its mean, the sample mean will not be exactly the same as the
population mean. It will be abit higher or abit lower than the population mean.
An erro in the estimation process is more likely to occur when using a point estimate. This
sampling error is the difference between the population mean and sample mean. It would
therefore be more reasonable to state a confidence interval (C. I.) estimate of the
population mean. A confidence interval denotes:
(a). a range within which the population mean might be found and
(b). The level of confidence or probability that the population mean lies withi the stated
confidence interval.
Confidence interval (C. I.) for population mean (u), when population variance, (a2) is
known.
If x bar is the mean of a random sample of size n taken from either
(i) a NORMAL population with known variance α2
Or
(ii) ANY population with known variance α2 provided that n is large, (n ≥ 30 )
then, the 100%(1 – α) C. I is given by:
NOTE: When constructing confidence intervals, the interval that is central, or symmetric,
about the mean is required, unless otherwise stated.
Example.
A machine produces steel sheets whose weights are known to be normally distributed with
a standard deviation of 2.4 kg. A random sample of 36 sheets had a mean weight of 31.4
kg. Find a 99% confidence interval for the population mean.
Solution
Let X be the random variable (r.v) ‘the weight, in kg, of a steel sheet’.

Example.
On the basis of the results obtained from a random sample of 100 women from
Mporokoso district, the 95% confidence interval for the mean height of the women in the
district is found to be (177.22 cm, 179.18 cm). Find the value of X bar , the mean of the
sample, and the standard deviation α of the normal distribution from which the sample is
drawn. Calculate the 98% confidence interval for the mean height.
Solution

Confidence Interval Estimation of population proportion, p
So far we have discussed the estimation of the population mean, which is quantitative in
nature. This concept of estimation can be extended to qualitative data where the data is
available in proportion or percentage form. In this situation the parameter of interest is p
which is the proportion of times a certain desirable outcome occurs.
We have established through the Central Limit Theorem that when large samples of size n are
selected from a population having a proportion of desirable outcomes p then the sampling
distribution of proportions is normally distributed. Here, n should be considered large if both
nṗ and n(1 - ṗ) are at least 5. In that case:
Example
A survey of 500 people shopping at a shopping mall, selected at random, showed that 350 of
them used credit cards for their purchases and 150 of them used cash.
(a) Construct a 95% confidence interval estimate of the population of all persons at the mall
who use credit card for shopping.
(b) What would our confidence level be, if we make the assertion that the proportion of
shoppers at the mall who shop with a credit card is between 67% and 73%?
Solution
(a) There are 350 out of a total sample of 500 who pay by credit card. Hence, the sample
proportion of credit card shoppers is:

(b).
Using the Standard normal table, we see that the area under the curve for Z = 1.5 is 0.4332.
This area is on each side of the mean so that the total area is 0.8664. In other words, our
confidence level is 86.64% that the population proportion of shoppers using credit card is
between 67% and 73%.
Recommended Readings
Main Text
The main textbook recommended is Zambia Institute of Chartered Accountants (2011)

Decision Making Techniques Study Manual
Students can also refer to Morris, C (2008), Quantitative Approaches in Business Studies, 7th
edition, FT Prentice Hall,

BAF 204 Decision Making Techniques

Uploaded by

Copyright:

Available Formats

BAF 204 Decision Making Techniques

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BAF 204 Decision Making Techniques

Uploaded by

Copyright:

Available Formats

About this Decision Making Techniques Module

How this Decision Making Techniques Module is structured

The Module overview

The overview also provides guidance on:

The Module content

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

Module learning outcomes

Self-study time recommended is 4 hours per week

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

Assignments There shall be two assignments given for this module.

National Institute of Public Administration- Outreach Programmes Division

Activity Assessment Assignment Case study

Discussion Group Help Note it!

Outcomes Reading Reflection Study skills

Summary Terminology Time frame Tip

National Institute of Public Administration- Outreach Programmes Division

How long will it take?

About the study guide

Activities, self-help questions and assessments

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

After studying this unit, the learner will be able to:

 Describe types of business decisions

 Describe the decision making process

 Apply simple decision making techniques

1.1 Types of business decision

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

After studying the unit, you will be able to:

 Apply regression analysis

1.1 Correlation and regression analysis

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

The correlation coefficient and the coefficient of determination

National Institute of Public Administration- Outreach Programmes Division

The coefficient of determination r2

National Institute of Public Administration- Outreach Programmes Division

Where n = number of pairs of data

National Institute of Public Administration- Outreach Programmes Division

Where n = the number of pairs of data

National Institute of Public Administration- Outreach Programmes Division

USING EXCEL FOR LINEAR REGRESSION ANALYSIS (PRACTICAL IN

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division

Examples of seasonal variations

Summarizing the components

Methods of finding the trend

National Institute of Public Administration- Outreach Programmes Division

Over what time period should a moving average be taken?

National Institute of Public Administration- Outreach Programmes Division

Moving averages of an even number of results

Required :calculate a moving average of four periods

National Institute of Public Administration- Outreach Programmes Division

Finding the seasonal variations

National Institute of Public Administration- Outreach Programmes Division

National Institute of Public Administration- Outreach Programmes Division