MATHEMATICS & STATISTICS

MBA
First Semester

Bharathidasan University
Centre for Distance and Online Education
Chairman:
Dr. M. Selvam
Vice-Chancellor
Bharathidasan University
Tiruchirappalli-620 024
Tamil Nadu
Co-Chairman:
Dr. G. Gopinath
Registrar
Bharathidasan University
Tiruchirappalli-620 024
Tamil Nadu
Course Co-Ordinator:
Dr. A. Edward William Benjamin
Director-Centre for Distance and Online Education
Bharathidasan University
Tiruchirappalli-620 024
Tamil Nadu
The Syllabus is Revised from 2021-22 onwards

Reviewer
Dr. O.M. Haja Mohideen, Head & dept of business Admn., Khadhir Mohideen college, Adiarampattinam.
Drt. M. Lakshmi bala, Asst prof. & Head, Dept. of business of Admin., K.N. Govt. arts college for women,
Thanjavur - 613 007

Authors
J. S. Chandan, (Units: 1, 2.6.2, 2.6.4, 3.1-3.4, 3.6, 3.8-3.11, unit-5)
C.R. Kothari, (Units: 2.1-2.3, 2.6.1, 2.6.3, 2.8-2.12, 3.7, unit-4, appendix)
G.S. Monga, (Units: 2.4, 2.6, 2.7, 3.5)
Vikas® Publishing House, (Units: 2.5)

"The copyright shall be vested with Bharathidasan University"

All rights reserved. No part of this publication which is material protected by this copyright notice
may be reproduced or transmitted or utilized or stored in any form or by any means now known or
hereinafter invented, electronic, digital or mechanical, including photocopying, scanning, recording
or by any information storage or retrieval system, without prior written permission from the Publisher.

Information contained in this book has been published by VIKAS® Publishing House Pvt. Ltd. and has
been obtained by its Authors from sources believed to be reliable and are correct to the best of their
knowledge. However, the Publisher, its Authors shall in no event be liable for any errors, omissions
or damages arising out of use of this information and specifically disclaim any implied warranties or
merchantability or fitness for any particular use.

Vikas® is the registered trademark of Vikas® Publishing House Pvt. Ltd.

VIKAS® PUBLISHING HOUSE PVT
LTDE-28, Sector-8, Noida -
201301 (UP)
Phone: 0120-4078900 • Fax: 0120-4078999
Regd. Office: A-27, 2nd Floor, Mohan Co-operative Industrial Estate, New Delhi 1100 44
Website: www.vikaspublishing.com • Email: [email protected]
COURSE INTRODUCTION

Statistics is considered as a mathematical science pertaining to the collection, analysis, interpretation or explanation,
and presentation of data and can be categorized as Inferential Statistics or Descriptive Statistics. Statistical
analysis is important for taking decisions and is widely used by academic institutions, natural and social sciences
departments, governments and business organizations. The word statistics is derived from the Latin word status
which means a political state or government. The subject of statistics is primarily concerned with making decisions
about various disciplines of markets and employment, such as stock market trends, unemployment rates in various
sectors of industries, demographic shifts, interest rates, inflation rates over the years, and so on. These conclusions
help to formulate specific policies and attitudes with respect to diverse areas of interest. Statistics, then, is a
science that deals with numbers or figures describing the state of affairs of various situations with which we are
generally or specifically concerned. Its objective is to summarize such data so that the summary gives us an
indication of some characteristics of a population or phenomenon that we wish to study. To ensure that conclusions
are meaningful, it is necessary to subject data to scientific analysis so that rational decisions can be made. Hence,
the field of statistics is concerned with proper collection of data, organizing this data into a manageable and
presentable form, and analysing and interpreting the data into useful conclusions.
Organized data can be presented in the form of tables, diagrams or graphs. This presentation in an orderly
manner facilitates the understanding and analysis of data. The basic purpose of data analysis is to make it useful
for drawing conclusions. This analysis may simply be a critical observation of data to draw meaningful conclusions
about it or it may involve highly complex and sophisticated mathematical techniques. Interpretation means drawing
conclusions from the data, which then form the basis of decision-making.
The objectives of this book are to:
• Introduce students to frequently used mathematical and basic statistical methods in business decision-
making.
• Explain how to assess these techniques and use them when appropriate.
• Formulate a suitable statistical problem, obtain a solution and interpret important features of this solution in
a business decision-making process.
The learning material is presented in a structured format so that it is easy to grasp. Each unit begins with an
outline of the Learning Objectives followed by Introduction to the topic of the unit. The detailed content is then
presented in a simple language, interspersed with Check Your Progress questions to enable the student to test his
understanding as and when he goes through each unit. Let Us Sum Up provided at the end of each unit helps in
quick recollection. Possible Questions is also provided for further practice.
The book is divided into five units.
UNIT 1: Concepts of Probability
UNIT 2: Probability Distributions
UNIT 3: Sampling and Hypothesis Testing
UNIT 4: Correlation and Regression Analysis
UNIT 5: Time Series Analysis
 MATHEMATICS & STATISTICS

DETAILED SYLLABUS

Unit–1: Concept and Definitions of Probability; Bayes’ Theorem and its Application Activity
Unit–2: Random Variable; Expectation and Variance; Continuous and Discrete Probability Functions;
Discrete Probability Distributions; Continuous Probability Distributions; Normal Distribution.
Unit–3: Sampling Procedures–Random and Non-Random Methods; Sample Size Determination, Sampling
Distribution and Standard Error; Hypothesis Testing; T-Distribution Testing ; Chi Square Test for
Independence of Attributes Activity
Unit–4: Correlation Analysis; Methods of Studying Simple Correlation; Properties of Correlation
Coefficient; Rank Correlation; Regression Analysis.
Unit–5: Components of Time Series Analysis; Forecasting with Decomposed Components of Time Series.
 CONTENTS
UNIT 1 CONCEPTS OF PROBABILITY 1-21
1.1 Learning Objectives
1.2 Introduction
1.3 Concept and Definitions of Probability
1.3.1 Sample Space
1.3.2 Events
1.3.3 Addition and Multiplication Theorems on Probability
1.3.4 Independent Events
1.3.5 Conditional Probability
1.4 Bayes’ Theorem and its Application Activity
1.5 Let Us Sum Up
1.6 Further Reading
1.7 Answers to Check Your Progress
1.8 Possible Questions
1.9 Learning Outcomes
UNIT 2 PROBABILITY DISTRIBUTIONS 23-66
2.1 Learning Objectives
2.2 Introduction
2.3 Random Variable
2.3.1 Techniques of Assigning Probabilities
2.3.2 Mean of Random Variable or the Expected Value of Random Variable
2.3.3 Variance and Standard Deviation of Random Variable
2.4 Expectation and Variance
2.4.1 Expected Value of a Function of X; 2.4.2 Variance
2.4.3 Expectation (Conditional); 2.4.4 Expectation (Iterated)
2.4.5 Expectation: Continuous Variables
2.5 Continuous and Discrete Probability Functions
2.6 Discrete Probability Distributions
2.6.1 Binomial Distribution—Properties
2.6.2 Applications of Binomial Distribution
2.6.3 Poisson Distribution—Properties
2.6.4 Applications of Poisson Distribution
2.7 Continuous Probability Distributions
2.8 Normal Distribution
2.8.1 Characteristics of Normal Distribution
2.8.2 Family of Normal Distributions
2.8.3 How to Measure the Area under the Normal Curve
2.9 Let Us Sum Up
2.10 Further Reading
2.11 Answers to Check Your Progress
2.12 Possible Questions
2.13 Learning Outcomes

UNIT 3 SAMPLING AND HYPOTHESIS TESTING 67-118

3.1 Learning Objectives
3.2 Introduction
3.3 Sampling Procedures–Random and Non-Random Methods
3.3.1 Sample Selection
3.3.2 Census and Sampling
 3.3.3 Sampling Techniques
3.3.4 Sampling and Non-Sampling Errors
3.4 Sample Size Determination, Sampling Distribution and Standard Error
3.4.1 Sampling Distribution of the Mean
3.4.2 Standard Error
3.5 Hypothesis Testing
3.5.1 Statistical Decision-Making
3.5.2 Type I and Type II Errors
3.5.3 Null and Alternative Hypotheses
3.5.4 Critical Region
3.5.5 Penalty
3.5.6 Standard Error
3.5.7 Testing of Hypothesis
3.5.8 Tests for a Sample Mean X
3.5.9 Test for Equality of Two Proportions
3.5.10 Large Sample Test for Equality of Two Means X1, X2
3.5.11 Small Sample Tests of Significance
3.5.12 Paired Observations
3.5.13 Test for a Given Population Variance
3.6 T-Distribution Testing
3.7 Chi Square Test for Independence of Attributes Activity
3.7.1 Degrees of Freedom
3.7.2 Conditions for the Application of Test
3.7.3 Areas of Application of Chi-Square Test
3.7.4 Steps in Finding the Value of Chi-square
3.7.5 Alternative Formula for Finding the Value of Chi-square in a (2 × 2) Table
3.7.6 Yates’ Correction
3.7.7 Chi-square as a Test of Population Variance
3.7.8 Additive Property of Chi-Square ( 2)
3.7.9 Important Characteristics of Chi-square ( 2) Test
3.8 Let Us Sum Up
3.9 Further Reading
3.10 Answers to Check Your Progress
3.11 Possible Questions
3.12 Learning Outcomes
UNIT 4 CORRELATION AND REGRESSION ANALYSIS 119-147
4.1 Learning Objectives
4.2 Introduction
4.3 Correlation Analysis
4.4 Methods of Studying Simple Correlation
4.5 Properties of Correlation Coefficient
4.5.1 Method of Least Squares
4.5.2 Simple Regression Coefficients
4.5.3 Karl Pearson’s Coefficient
4.5.4 Other Measures
4.6 Rank Correlation
4.7 Regression Analysis
4.8 Let Us Sum Up
4.9 Further Reading
4.10 Answers to Check Your Progress
4.11 Possible Questions
4.12 Learning Outcomes
UNIT 5 TIME SERIES ANALYSIS 149-170
5.1 Learning Objectives
5.2 Introduction
5.3 Components of Time Series Analysis
5.4 Forecasting with Decomposed Components of Time Series
5.4.1 Trend Analysis
5.4.2 Smoothing Techniques
5.4.3 Measuring Cyclical Effect
5.5 Let Us Sum Up
5.6 Further Reading
5.7 Answers to Check Your Progress
5.8 Possible Questions
5.9 Learning Outcomes
Appendix 171-181
Concepts of Probability Unit-1

UNIT 1 CONCEPTS OF
PROBABILITY
UNIT STRUCTURE
1.1 Learning Objectives
1.2 Introduction
1.3 Concept and Definitions of Probability
1.3.1 Sample Space
1.3.2 Events
1.3.3 Addition and Multiplication Theorems on Probability
1.3.4 Independent Events
1.3.5 Conditional Probability
1.4 Bayes’ Theorem and its Application Activity
1.5 Let Us Sum Up
1.6 Further Reading
1.7 Answers to Check Your Progress
1.8 Possible Questions
1.9 Learning Outcomes

1.1 LEARNING OBJECTIVES

After going through this unit, you will be able to:

• Explain sample space
• Describe events and their role in probability
• Explain the axiomatic approach to probability
• Solve the probability problems using addition and multiplication rules
• Define independent events
• Explain conditional probability
• Describe Bayes’ theorem

1.2 INTRODUCTION

In this unit, you will learn different theories of probability and will be able to explain
understand why probability is considered the most important tool in statistical
calculations. The subject of probability in itself is cumbersome; hence, only the
basic concepts will be discussed in this unit. The word probability or chance is very
commonly used in day-to-day conversation, and terms such as possible or probable
or likely, all have similar meanings.
Probability can be defined as a measure of the likelihood that a particular
event will occur. It is a numerical measure with a value between 0 and 1 of such a Probability: A measure
likelihood where the probability of 0 indicates that the given event cannot occur of the likelihood that a
and the probability of 1 assures certainty of such an occurrence. particular event will
occur.

Statistics for Management 1

Unit-1 Concepts of Probability

In this unit, you will also learn why all these uncertainties require knowledge
of probability so that calculated risks can be taken. Since the outcomes of most
decisions cannot be accurately predicted because of the impact of many
uncontrollable and unpredictable variables, it is necessary that all the known risks
be scientifically evaluated. Probability theory, sometimes referred to as the science
of uncertainty, is very helpful in such evaluations. It helps the decision-maker with
only limited information to analyse the risks and select the strategy of minimum
risk.
This unit will also discuss the laws of addition and multiplication. The law of
addition states that when two events are mutually exclusive, the probability that
either of the events will occur is the sum of their separate probabilities. The law of
multiplication is applicable when two events occur at the same time.
This unit will also describe the probability distributions and Bayes’ theorem.
Bayes’ theorem is based on the philosophy of science and tries to clarify the
relationship between theory and evidence.

1.3 CONCEPT AND DEFINITIONS OF PROBABILITY

The probability theory helps a decision-maker to analyse a situation and decide

accordingly. The following are few examples of such situations:
• What is the chance that sales will increase if the price of the product is
decreased?
• What is the likelihood that a new machine will increase productivity?
• How likely is it that a given project will be completed in time?
• What are the possibilities that a competitor will introduce a cheaper substitute
in the market?
Probability theory is also called the theory of chance and can be mathematically
derived using the standard formulas. A probability is expressed as a real number,
p [0, 1] and the probability number is expressed as a percentage (0 per cent to 100
per cent) and not as a decimal. For example, a probability of 0.55 is expressed as 55
per cent. When we say that the probability is 100 per cent, it means that the event is
certain while the 0 per cent probability means that the event is impossible. We can
also express probability of an outcome in the ratio format. For example, we have
two probabilities, i.e., ‘chance of winning’ (1/4) and ‘chance of not winning’
(3/4), then using the mathematical formula of odds, we can say,
‘chance of winning’ : ‘chance of not winning’ = 1/4 : 3/4 = 1 : 3 or 1/3
We are using the probability in vague terms when we predict something for
future. For example, we might say it will probably rain tomorrow or it will probably
be a holiday the day after. This is subjective probability to the person predicting,
but implies that the person believes the probability is greater than 50 per cent.
Different types of probability theories are:
(i) Axiomatic Probability Theory
(ii) Classical Theory of Probability
(iii) Empirical Probability Theory

2 Statistics for Management

Concepts of Probability Unit-1

(i) Axiomatic Probability Theory

The axiomatic probability theory is the most general approach to probability, and is
used for more difficult problems in probability. We start with a set of axioms, which
serve to define a probability space. These axioms are not immediately intuitive and
are developed using the classical probability theory.
(ii) Classical Theory of Probability
The classical theory of probability is based on the number of favourable outcomes
and the number of total outcomes. The probability is expressed as a ratio of these
two numbers. The term ‘favourable’ is not the subjective value given to the outcomes,
but is rather the classical terminology used to indicate that an outcome belongs to a
given event of interest.
Classical definition of probability: If the number of outcomes belonging to
an event E is NE , and the total number of outcomes is N, then the probability of
NE
event E is defined as pE = .
N
For example, a standard pack of cards (without jokers) has 52 cards. If we randomly
draw a card from the pack, we can imagine about each card as a possible outcome.
Therefore, there are 52 total outcomes. Calculating all the outcome events and their
probabilities, we have the following possibilities:
• Out of the 52 cards, there are 13 clubs. Therefore, if the event of interest is
drawing a club, there are 13 favourable outcomes, and the probability of this
13 1
event becomes: = .
52 4
• There are 4 kings (one of each suit). The probability of drawing a king is:
4 1
= .
52 13
• What is the probability of drawing a king or a club? This example is slightly
more complicated. We cannot simply add together the number of outcomes
for each event separately (4 + 13 = 17) as this inadvertently counts one of the
16
outcomes twice (the king of clubs). The correct answer is: from
52
13 4 1
+ – .
52 52 52
We have this from the probability equation, P(club) + P(king) – P(king of
clubs).
• Classical probability has limitations, because this definition of probability
implicitly defines all outcomes to be equiprobable and this can be only used
for conditions such as drawing cards, rolling dice, or pulling balls from urns.
We cannot calculate the probability where the outcomes are unequal
probabilities.

Statistics for Management 3

Unit-1 Concepts of Probability

It is not that the classical theory of probability is not useful because of the above
described limitations. We can use this as an important guiding factor to calculate
the probability of uncertain situations as mentioned above and to calculate the
axiomatic approach to probability.
Frequency of occurrence
This approach to probability is widely used to a wide range of scientific disciplines.
It is based on the idea that the underlying probability of an event can be measured
by repeated trials.
Probability as a measure of frequency: Let nA be the number of times event
A occurs after n trials. We define the probability of event A as,
nA
P ( A) = Lim
n n
It is not possible to conduct an infinite number of trials. However, it usually suffices
to conduct a large number of trials, where the standard of large depends on the
probability being measured and how accurate a measurement we need.
Definition of probability
nA
The sequence in the limit that will converge to the same result every time, or
n
that it will converge at all. To understand this, let us consider an experiment consisting
of flipping a coin an infinite number of times. We want that the probability of heads
must come up. The result may appear as the following sequence:
HTHHTTHHHHTTTTHHHHHHHHTTTTTTTTHHHHHHHHHHHHH
HHHTTTTTTTTTTTTTTTT...
This shows that each run of k heads and k tails are being followed by another run of
nA 1
the same probability. For this example, the sequence oscillates between and
n 3
2
3
which does not converge. These sequences may be unlikely, and can be right.
The definition given above does not express convergence in the required way, but it
shows some kind of convergence in probability. The problem of formulating exactly
can be considered using axiomatic probability theory.
(iii) Empirical Probability Theory
The empirical approach to determine probabilities relies on data from actual
experiments to determine approximate probabilities instead of the assumption of
equal likeliness. Probabilities in these experiments are defined as the ratio of the
frequency of the possibility of an event, f(E), to the number of trials in the experiment,
n, written symbolically as P(E) = f(E)/n. For example, while flipping a coin, the
empirical probability of heads is the number of heads divided by the total number
of flips.
The relationship between these empirical probabilities and the theoretical
probabilities is suggested by the Law of Large Numbers. The law states that as the
number of trials of an experiment increases, the empirical probability approaches
the theoretical probability. Hence, if we roll a die a number of times, each number
would come up approximately 1/6 of the time. The study of empirical probabilities
is known as statistics.

4 Statistics for Management

Concepts of Probability Unit-1

1.3.1 Sample Space

A sample space is the collection of all possible events or outcomes of an experiment.
For example, there are two possible outcomes of a toss of a fair coin: a head and a
tail. Then, the sample space for this experiment denoted by S would be: Sample space: The
S = [H, T] collection of all possible
So that the probability of the sample space equals 1, or events or outcomes of
an experiment.
P[S] = P[H,T] =1
This is so because in the toss of the coin, either a head or a tail, must occur. Similarly,
when we roll a die, any of the six faces can come as a result of the roll since there
are a total of six faces. Hence, the sample space is S = [1, 2, 3, 4, 5, 6], and P[S] =
1, since one of the six faces must occur.
1.3.2 Events
An event is an outcome or a set of outcomes of an activity or a result of a trial. For
example, getting two heads in the trial of tossing three fair coins simultaneously
would be an event. Event: An outcome or a
set of outcomes of an
• Elementary event: An elementary event, also known as a simple event, is a activity or a result of a
single possible outcome of an experiment. For example, if we toss a fair coin, trial.
then the event of a head coming up is an elementary event. If the symbol for
an elementary event is (E), then the probability of the event (E) is written as
P[E].
• Joint event: A joint event, also known as a compound event, has two or more
elementary events in it. For example, drawing a black ace from a pack of
cards would be a joint event, since it contains two elementary events of black
and ace.
• Simple probability: It refers to a phenomenon where only a simple or
elementary event occurs. For example, assume that event (E), the drawing of
a diamond card from a pack of 52 cards, is a simple event. Since there are 13
diamond cards in the pack and each card is equally likely to be drawn, the
probability of event (E) or P[E] = 13/52 or 1/4.
• Joint probability: It refers to the phenomenon of occurrence of two or more
simple events. For example, assume that event (E) is a joint event (or
compound event) of drawing a black ace from a pack of cards. There are two
simple events involved in the compound event, which are, the card being
black and the card being an ace. Hence, P[Black ace] or P[E] = 2/52 since
there are two black aces in the pack.
• Marginal probability: It is the probability of one event, ignoring any
information about the other event. Marginal probability is obtained by
summing (or integrating, more generally) the joint probability over the ignored
event. The marginal probability of A is written as P(A), and the marginal
probability of B is written as P(B).
• Complement of an event: The complement of any event A is the collection of
outcomes that are not contained in A. This complement of A is denoted as A

Statistics for Management 5

Unit-1 Concepts of Probability

(A prime). This means that the outcomes contained in A and the outcomes
contained in A must equal the total sample space. Therefore,
P[A] + P[A ] = P[S] = 1
or, P[A] = 1 – P[A ]
For example, if a passenger airliner has 300 seats and it is nearly full, but not totally
full, then event A would be the number of occupied seats and A would be the
number of unoccupied seats. Suppose there are 287 seats occupied by passengers
and only 13 seats are empty. Typically, the stewardess will count the number of
empty seats which are only 13 and report that 287 people are aboard. This is much
simpler than counting 287 occupied seats. Accordingly, in such a situation, knowing
event A is much more efficient than knowing event A.
• Mutually exclusive events: Two events are said to be mutually exclusive, if
both events cannot occur at the same time as outcome of a single experiment.
For example, if we toss a coin, then either event head or event tail would
occur, but not both. Hence, these are mutually exclusive events.
Venn diagrams
We can visualize the concept of events, their relationships and sample space using
Venn diagrams. The sample space is represented by a rectangular region and the
events and the relationships among these events are represented by circular regions
within the rectangle.
For example, two mutually exclusive events A and B are represented in the
Venn diagram as follows:

A B = S

Event P[A B] is represented in the Venn diagram as follows:

A B

Event [AB] is represented as follows:

A B

[AB]

6 Statistics for Management

Concepts of Probability Unit-1

Union of three events

The process of combining two events to form the union can be extended to three
events so that P[A B C] would be the union of events A, B, and C. This union
can be represented in a Venn diagram as follows:

A B

Example 1.1: A sample of 50 students is taken and a survey is made on the reading
habits of the sample selected. The survey results are shown as follows:

Event Number of students Magazines they read

[A] 20 Time
[B] 15 Newsweek
[C] 10 Filmfare
[AB] 8 Time and Newsweek
[AC] 6 Time and Filmfare
[BC] 4 Newsweek and Filmfare
[ABC] 2 Time and Newsweek and Filmfare

Find out the probability that a student picked up at random from this sample of 50
students does not read any of these 3 magazines.
Solution: The problem can be solved by a Venn diagram as follows:

A B

8 6 5

2
4 2
2

21
C

Since there are 21 students who do not read any of the three magazines, the probability
that a student picked up at random among this sample of 50 students who does not
read any of these three magazines is 21/50.
The problem can also be solved by the formula for probability for union of
three events, given as follows:

Statistics for Management 7

Unit-1 Concepts of Probability

P[A B C] = P[A] + P[B] + P[C] – P[AB] – P[AC] – P[BC] + P[ABC]

= 20/50 + 15/50 + 10/50 – 8/50 – 6/50 – 4/50 + 2/50
= 29/50
The above is the probability that a student picked up at random among the sample
of 50 reads either Time or Newsweek or Filmfare or any combination of the two or
all three. Hence the probability that such a student does not read any of these three
magazines is 21/50 which is [1 – 29/50].

1.3.3 Addition and Multiplication Theorems on Probability

Law of Addition
When two events are mutually exclusive, then the probability that either of the
events will occur is the sum of their separate probabilities. For example, if you roll
a single die, then the probability that it will come up with a face 5 or face 6, where
event A refers to face 5 and event B refers to face 6 and both events being mutually
exclusive events, is given by:
P[A or B] = P[A] + P[B]
or, P[5 or 6] = P[5] + P[6]
= 1/6 +1/6
= 2/6 = 1/3
P [A or B] is written as P[ A B] and is known as P [A union B].
However, if events A and B are not mutually exclusive, then the probability of
occurrence of either event A or event B or both is equal to the probability that event
A occurs plus the probability that event B occurs minus the probability that events
common to both A and B occur.
Symbolically, it can be written as:
P[ A B] = P[A] + P[B] – P[A and B]

P[A and B] can also be written as P[ A B], known as P [A intersection B] or simply

P[AB].

8 Statistics for Management

Concepts of Probability Unit-1

Events [A and B] consist of all those events which are contained in both A and
B simultaneously. For example, in an experiment of taking cards out of a pack of 52
playing cards, assume that:
Event A = An ace is drawn.
Event B = A spade is drawn.
Event [AB] = An ace of spade is drawn.
Hence, P[A B] = P[A] + P[B] – P[AB]
= 4/52 + 13/52 – 1/52
= 16/52
= 4/13
This is so, because there are 4 aces, 13 cards of spades, including 1 ace of spades
out of a total of 52 cards in the pack. The logic behind subtracting P[AB] is that the
ace of spades is counted twice—once in event A (4 aces) and once again in event B
(13 cards of spade including the ace).
Another example for P[ A B], where event A and event B are not mutually
exclusive is as follows:
Suppose a survey of 100 persons revealed that 50 persons read India Today
and 30 persons read Time magazine and 10 of these 100 persons read both India
Today and Time. Then:
Event [A] = 50
Event [B] = 30
Event [AB] = 10
Since event [AB] of 10 is included twice, both in event A as well as in event B, event
[AB] must be subtracted once in order to determine the event [ A B] which means
that a person reads India Today or Time or both. Hence,
P[ A B] = P [A] + P [B] – P [AB]
= 50/100 + 30/100 –10/100
= 70/100 = 0.7
Law of Multiplication
Multiplication rule is applied when it is necessary to compute the probability if
both events A and B will occur at the same time. The multiplication rule is different
if the two events are independent as against the two events being not independent.
If events A and B are independent events, then the probability that they both
will occur is the product of their separate probabilities. This is a strict condition so
that events A and B are independent if and only if,
P [AB] = P[A] × P[B]
or = P[A] P[B]

Statistics for Management 9

Unit-1 Concepts of Probability

For example, if we toss a coin twice, then the probability that the first toss results in
a head and the second toss results in a tail is given by,
P [HT] = P[H] × P[T]
= 1/2 × 1/2 = 1/4
However, if events A and B are not independent, meaning that the probability of
occurrence of an event is dependent or conditional upon the occurrence or non-
occurrence of the other event, then the probability that they will both occur is given
by,
P[AB] = P[A] × P[B/Given outcome of A]
This relationship is written as,
P[AB] = P[A] × P[B/A] = P[A] P[B/A]
Where P[B/A] means the probability of event B on the condition that event A has
occurred. As an example, assume that a bowl has 6 black balls and 4 white balls. A
ball is drawn at random from the bowl. Then a second ball is drawn without
replacement of the first ball back in the bowl. The probability of the second ball
being black or white would depend upon the result of the first draw as to whether
the first ball was black or white. The probability that both these balls are black is
given by,
P [Two black balls] = P [Black on 1st draw] × P [Black on 2nd draw/Black
on 1st draw]
= 6/10 × 5/9 = 30/90 = 1/3
This is so, because there are 6 black balls out of a total of 10, but if the first ball
drawn is black then we are left with 5 black balls out of a total of 9 balls.
1.3.4 Independent Events
Two events A and B are said to be independent events, if the occurrence of one
event is not influenced at all by the occurrence of the other. For example, if two fair
coins are tossed, then the result of one toss is totally independent of the result of the
other toss. The probability that a head will be the outcome of any one toss will
always be 1/2, irrespective of whatever the outcome is of the other toss. Hence,
these two events are independent.
Let us assume that one fair coin is tossed 10 times and it happens that the first
nine tosses resulted in heads. What is the probability that the outcome of the tenth
toss will also be a head? There is always a psychological tendency to think that a
tail would be more likely in the tenth toss since the first nine tosses resulted in
heads. However, since the events of tossing a coin 10 times are all independent
events, the earlier outcomes have no influence whatsoever on the result of the tenth
toss. Hence, the probability that the outcome will be a head on the tenth toss is still
1/2.
On the other hand, consider drawing two cards from a pack of 52 playing
cards. The probability that the second card will be an ace would depend upon whether
the first card was an ace or not. Hence, these two events are not independent
events.

10 Statistics for Management

Concepts of Probability Unit-1

1.3.5 Conditional Probability

In many situations, a manager may know the outcome of an event that has already
occurred and may want to know the chances of a second event occurring based
upon the knowledge of the outcome of the earlier event. We are interested in finding
out as to how additional information obtained as a result of the knowledge about
the outcome of an event affects the probability of the occurrence of the second
event. For example, let us assume that a new brand of toothpaste is being introduced
in the market. Based on the study of competitive markets, the manufacturer has
some idea about the chances of its success. Now, he introduces the product in a few
selected stores in a few selected areas before marketing it nationally. A highly positive
response from the test-market area will improve his confidence about the success
of his brand nationally. Accordingly, the manufacturer’s assessment of high
probability of sales for his brand would be conditional upon the positive response
from the test-market.
Let there be two events A and B. Then the probability of occurrence of event
A is given that event B has occured. The notation is given by,
P[ AB]
P[ A / B] =
P[B]
Where P[A/B] is interpreted as the probability of event A on the condition that event
B has occurred and P [AB] is the joint probability of event A and event B, and P[B]
is not equal to zero.
As an example, let us suppose that we roll a die and we know that the number
that came up is larger than 4. We want to find out the probability that the outcome
is an even number given that it is larger than 4.
Let Event A = Even
and Event B = Larger than 4

P[Even and larger than 4]

Then, P[Even / Larger than 4] = P[Larger than 4]

P[ AB]
or P[A/ B] = = (1/ 6)/(2 / 6) = 1/ 2
P[B]

But for independent events, P[AB] = P[A] P[B]. Thus, substituting this relationship
in the formula for conditional probability, we get,
P[ AB] P[ A]P[B]
P[ A / B] = = = P[ A]
P[B] P[B]
This means that P[A] will remain the same no matter what the outcome of event B
is. For example, if we want to find out the probability of a head on the second toss
of a fair coin, given that the outcome of the first toss was a head, this probability
would still be 1/2 because the two events are independent events and the outcome
of the first toss does not affect the outcome of the second toss.

Statistics for Management 11

Unit-1 Concepts of Probability

1.4 BAYES’ THEOREM AND ITS APPLICATION

ACTIVITY
Reverend Thomas Bayes (1702–1761), introduced his theorem on probability which
is concerned with a method for estimating the probability of causes which are
responsible for the outcome of an observed effect. Being a religious preacher himself
as well as a mathematician, his motivation for the theorem came from his desire to
prove the existence of God by looking at the evidence of the world that God created.
He was interested in drawing conclusions about the causes by observing the
consequences. The theorem contributes to the statistical decision theory in revising
prior probabilities of outcomes of events based upon the observation and analysis
of additional information.
Bayes’ theorem makes use of conditional probability formula where the
condition can be described in terms of the additional information which would
result in the revised probability of the outcome of an event.
Example 1.2: Suppose there are 50 students in our statistics class out of which 20
are male and 30 are female. Out of the 30 females, 20 are Indian students and 10 are
foreign students. Out of the 20 male students, 15 are Indians and 5 are foreigners, so
that out of all the 50 students, 35 are Indians and 15 are foreigners. This data can be
presented in a tabular form as follows:

Indian Foreigner Total

Male 15 5 20
Female 20 10 30
Total 35 15 50

Solution: Based upon this information, the probability that a student picked up at
random will be female is 30/50 or 0.6, since there are 30 females in the total class of
50 students. Now suppose that we are given additional information that the person
picked up at random is Indian, then what is the probability that this person is a
female? This additional information will result in revised probability or posterior
probability in the sense that it is assigned to the outcome of the event after this
additional information is made available.
Since we are interested in the revised probability of picking a female student
at random provided that we know that the student is Indian. Let A1 be the event
female, A2 be the event male and B be the event Indian. Then based upon our
knowledge of conditional probability, the Bayes’ theorem can be stated as follows,
P( A1) P(B / A1)
P( A1 / B) =
P( A1 ) P(B / A1 ) + P( A2 )( P(B / A2 )

In the example discussed above, there are 2 basic events which are A1 (female) and
A2 (male). However, if there are n basic events, A1, A2, ......An, then Bayes’ theorem
can be generalized as,

12 Statistics for Management

Concepts of Probability Unit-1

P( A )P(B / A )
P( A1 / B) = P( A )P(B / A ) + P( A )(P1 (B / A )1+ ... + P( A )P(B / A )
1 1 2 2 n n

Solving the case of 2 events we have,

(30 / 50)(20 / 30)

P( A1 / B) = = 20 / 35 = 4 / 7 = 0.57
(30 / 50)(20 / 30) + (20 / 50)(15 / 20)
This example shows that while the prior probability of picking up a female student
is 0.6, the posterior probability becomes 0.57 after the additional information that
the student is an American is incorporated in the problem.
Another example of application of Bayes’ theorem is as follows:
Example 1.3: A businessman wants to construct a hotel in New Delhi. He generally
builds three types of hotels. These are hotels with 50 rooms, 100 rooms and 150
rooms, depending upon the demand for rooms, which is a function of the area in
which the hotel is located, and the traffic flow. The demand can be categorized as
low, medium or high. Depending upon these various demands, the businessman has
made some preliminary assessment of his net profits and possible losses (in thousands
of dollars) for these various types of hotels. These payoffs are shown in the following
table.
States of Nature
Demand for Rooms
Low (A1) Medium (A2) High (A3)
0.2 0.5 0.3 Demand Probability
Number of Rooms R1=(50) 25 35 50
R2=(100) –10 40 70
R3=(150) –30 20 100

Solution: The businessman has also assigned ‘prior probabilities’ to the demand
structure or rooms. These probabilities reflect the initial judgement of the
businessman based upon his intuition and his degree of belief regarding the outcomes
of the states of nature.

Demand for Rooms Probability of Demand

Low (A1) 0.2
Medium (A2) 0.5
High (A3) 0.3

Based upon these values, the expected payoffs for various rooms can be computed
as follows,
EV (50) = ( 25 × 0.2) + (35 × 0.5) + (50 × 0.3) = 37.50
EV (100) = (–10 × 0.2) + (40 × 0.5) + (70 × 0.3) = 39.00
EV (150) = (–30 × 0.2) + (20 × 0.5) + (100 × 0.3) = 34.00
This gives us the maximum payoff of $39,000 for building a 100 rooms hotel.

Statistics for Management 13

Unit-1 Concepts of Probability

Now, the hotelier must decide whether to gather additional information

regarding the states of nature, so that these states can be predicted more accurately
than the preliminary assessment. The basis of such a decision would be the cost of
obtaining additional information. If this cost is less than the increase in maximum
expected profit, then such additional information is justified.
Suppose that the businessman asks a consultant to study the market and predict
the states of nature more accurately. This study is going to cost the businessman
$10,000. This cost would be justified if the maximum expected profit with the new
states of nature is at least $10,000 more than the expected payoff with the prior
probabilities. The consultant made some studies and came up with the estimates of
low demand (X1), medium demand (X2), and high demand (X3) with a degree of
reliability in these estimates. This degree of reliability is expressed as conditional
probability which is the probability that the consultant’s estimate of low demand
will be correct and the demand will be actually low. Similarly, there will be a
conditional probability of the consultant’s estimate of medium demand, when the
demand is actually low and so on. These conditional probabilities are expressed in
Table 1.1.
Table 1.1 Conditional Probabilities

X1 X2 X3
States of (A1) 0.5 0.3 0.2
Nature of (A2) 0.2 0.6 0.2
(Demand) (A3) 0.1 0.3 0.6

The values in the preceding table are conditional probabilities and are interpreted
as follows:
The upper north-west value of 0.5 is the probability that the consultant’s
prediction will be for low demand (X1) when the demand is actually low. Similarly,
the probability is 0.3 that the consultant’s estimate will be for medium demand (X2)
when in fact the demand is low and so on. In other words, P(X1/ A1) = 0.5 and P(X2/
A1) = 0.3. Similarly, P(X1 / A2) = 0.2 and P(X2 / A2) = 0.6 and so on.
Our objective is to obtain posteriors which are computed by taking the
additional information into consideration. One way to reach this objective is to first
compute the joint probability which is the product of prior probability and conditional
probability for each state of nature. Joint probabilities as computed is given as,

State Prior Joint Probabilities

of Nature Probability P(A1X1) P(A1X2) P(A1X3)

A1 0.2 0.2 × 0.5 = 0.10 0.2 × 0.3 = 0.06 0.2 × 0.2 = 0.04
A2 0.5 0.5 × 0.2 = 0.10 0.5 × 0.6 = 0.30 0.5 × 0.2 = 0.10
A3 0.3 0.3 × 0.1 = 0.03 0.3 × 0.3 = 0.09 0.3 × 0.6 = 0.18
Total Marginal Probabilities. = 0.23 = 0.45 = 0.32

14 Statistics for Management

Concepts of Probability Unit-1

Now, the posterior probabilities for each state of nature Ai are calculated as follows:

Joint probability of Ai and Xj

P(Ai / Xj )
Marginal probability of Xj

By using this formula, the joint probabilities are converted into posterior probabilities
and the computed table for these posterior probabilities is given as,
States of Nature Posterior Probabilities

P(A1/X1) P(A1/X2) P(A1/X3)

A1 0.1/0.23 = 0.435 0.06/0.45 = 0.133 0.04/0.32 = 0.125

A2 0.1/0.23 = 0.435 0.30/0.45 = 0.667 0.1/0.32 = 0.312
A3 0.03/0.23 = 0.130 0.09/0.45 = 0.200 0.18/0.32 = 0.563

Total = 1.0 = 1.0 = 1.0

Now, we have to compute the expected payoffs for each course of action with the
new posterior probabilities assigned to each state of nature. The net profits for each
course of action for a given state of nature is the same as before and is restated
below. These net profits are expressed in thousands of dollars.

Low (A1) Medium (A2) High (A3)

Number of Rooms (R1) 25 35 50
(R2) – 10 40 70
(R3) – 30 20 100

Let Oij be the monetary outcome of course of action (i) when (j) is the corresponding
state of nature, so that in the above case Oi1 will be the outcome of course of action
R1 and state of nature A1, which in our case is $25,000. Similarly, Oi2 will be the
outcome of action R2 and state of nature A2, which in our case is –$10,000, and so
on. The expected value EV (in thousands of dollars) is calculated on the basis of the
actual state of nature that prevails as well as the estimate of the state of nature as
provided by the consultant. These expected values are calculated as follows,
Course of action = Ri
Estimate of consultant = Xi
Actual state of nature = Ai
Where, i = 1, 2, 3
Then,
(A) Course of action = R1 = Build 50 rooms hotel
A
R P i O
EV 1
=
X1 X 1 i1
= 0.435(25) + 0.435 (–10) + 0.130 (–30)
= 10.875 – 4.35 – 3.9 = 2.625
A
R1 P i O
EV
=
X2 X 2 i1

Statistics for Management 15

Unit-1 Concepts of Probability

= 0.133(25) + 0.667 (–10) + 0.200 (–30)

= 3.325 – 6.67 – 6.0 = –9.345
A
R1 P i O
EV
=
X3 X 3 i1
= 0.125(25) + 0.312(–10) + 0.563(–30)
= 3.125 – 3.12 – 16.89
= –16.885
(B) Course of action = R2 = Build 100 rooms hotel
A
R P i O
EV 2
=
X1 X 1 i2
= 0.435(35) + 0.435 (40) + 0.130 (20)
= 15.225 + 17.4 + 2.6 = 35.225
A
R2 P i O
EV
=
X2 X 1 i2
= 0.133(35) + 0.667 (40) + 0.200 (20)
= 4.655 + 26.68 + 4.0 = 35.335
A
R2 P i O
EV
=
X3 X 3 i2
= 0.125(35) + 0.312(40) + 0.563(20)
= 4.375 + 12.48 + 11.26 = 28.115
(C) Course of action = R3 = Build 150 rooms hotel
A
R P i O
EV 3
=
X1 X 1 i3
= 0.435(50) + 0.435(70) + 0.130 (100)
= 21.75 + 30.45 + 13 = 65.2
R3 Ai
EV P O
X2 = X 2 i3
= 0.133(50) + 0.667 (70) + 0.200 (100)
= 6.65 + 46.69 + 20 = 73.34
A
R3 P i O
EV
=
X3 X 3 i3
= 0.125(50) + 0.312(70) + 0.563(100)
= 6.25 + 21.84 + 56.3 = 84.39
The expected values in thousands of dollars, as calculated above, are presented
below in a tabular form.

16 Statistics for Management

Concepts of Probability Unit-1
Expected Posterior Payoffs

Outcome EV (R1/Xi) EV (R2/Xi) EV (R3/Xi)

X1 2.625 35.225 65.2
X2 –9.345 35.335 73.34
X3 –16.885 28.115 84.39

This table can now be analysed in the following manner.

If the outcome is X1, it is desirable to build 150 rooms hotel, since the expected
payoff for this course of action is maximum of $65,200. Similarly, if the outcome is
X2, the course of action should again be R3 since the maximum payoff is $73,34.
Finally, if the outcome is X3, the maximum payoff is $84,390 for course of action
R 3.
Accordingly, given these conditions and the payoff, it would be advisable to
build a 150 rooms hotel.

Statistics for Management 17

Unit-1 Concepts of Probability

1.5 LET US SUM UP

• Statistical data is used to draw certain conclusions with the effective use of
probability.
• Various terms and concepts can be applied in the decision-making process
and also in inferring the occurrence of several events in the business
environment of an organization.
• The outcomes of most decisions cannot be accurately predicted because of
the impact of many uncontrollable and unpredictable variables. So, it is
necessary to scientifically evaluate the known risks. Probability theory, often
known as the science of uncertainty, is helpful in such evaluations.
• It also helps the decision-maker to analyse risks with the help of limited
information and select a strategy of minimum risk.
• Probability distribution refers to the listing of all possible outcomes of an
experiment together with their probabilities.
• Bayes’ theorem contributes to the statistical decision theory in revising prior
probabilities of outcomes of events.

1.6 FURTHER READING

Elhance, D.N. Fundamental of Statistics. Allahabad: Kitab Mahal.
Gupta, S.C. Fundamentals of Business Statistics. New Delhi: Sultan Chand & Sons.

1.7 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress-1

1. The classical theory of probability is based on the number of favourable
outcomes and the number of total outcomes. The probability is expressed as
a ratio of these two numbers. The term ‘favourable’ is not the subjective
value given to the outcomes, but is rather the classical terminology used to
indicate that an outcome belongs to a given event of interest.
2. The term ‘simple probability’ refers to the phenomenon where only a simple
or elementary event occurs. For example, assume that event (E), the drawing
of a diamond card from a pack of 52 cards, is a simple event. Since there are
13 diamond cards in the pack and each card is equally likely to be drawn, the
probability of event (E) or P[E] = 13/52 or 1/4.
The term ‘joint probability’ refers to the phenomenon of occurrence of two
or more simple events. For example, assume that event (E) is a joint event (or
compound event) of drawing a black ace from a pack of cards. There are two
simple events involved in the compound event, i.e., the card being black and
the card being an ace. Hence, P[Black ace] or P[E] = 2/52 since there are two
black aces in the pack.

18 Statistics for Management

Concepts of Probability Unit-1

3. Two events are said to be mutually exclusive, if both events cannot occur at
the same time as the outcome of a single experiment. For example, if we toss
a coin, then either event head or event tail would occur, but not both. Hence,
these are mutually exclusive events.
Check Your Progress-2
1. The addition rule states that when two events are mutually exclusive, then
the probability that either of the events will occur is the sum of their separate
probabilities. For example, if you roll a single die then the probability that it
will come up with a face 5 or face 6, where event A refers to face 5 and event
B refers to face 6, both events being mutually exclusive events, is given by,
P[A or B] = P[A] + P[B]
Or, P[5 or 6] = P[5] + P[6]
= 1/6 +1/6
= 2/6 = 1/3
P [A or B] is written as P[A B] and is known as P [A union B].
2. The multiplication rule is applied when it is necessary to compute the
probability in case two events occur at the same time.
3. The Bayes’ theorem on probability gives a method for estimating the
probability of causes which are responsible for the outcome of an observed
effect. The theorem contributes to the statistical decision theory in revising
prior probabilities of outcomes of events based upon the observation and
analysis of additional information.

1.8 POSSIBLE QUESTIONS

Short-Answer Questions
1. Explain the concept of probability.
2. What are the different theories of probability? Explain briefly.
3. What are mutually exclusive events?
4. What do you mean by simple probability?
5. Explain the axiomatic approach to probability.
6. Explain the concept of multiplication rule.
7. What is Bayes’ theorem? What is its importance in statistical calculations?
Long-Answer Questions
1. A family plans to have two children. What is the probability that both children
will be boys? (List all the possibilities and then select the one which would
be two boys.)
2. A wheel of fortune has numbers 1 to 40 painted on it, each number being at
equal distance from the other so that when the wheel is rotated, there is the
same chance that the pointer will point at any of these numbers. Tickets have
been issued to contestants numbering 1 to 40. The number at which the wheel

Statistics for Management 19

Unit-1 Concepts of Probability

stops after being rotated would be the winning number. What is the probability
that,
(i) Ticket number 29 wins
(ii) A person who bought 5 tickets numbered 18 to 22 (inclusive), wins the
prize
3. Two fair dice are rolled. Comment on the probability of getting:
(i) A sum of 10 or more.
(ii) A pair of which atleast one number is 3.
(iii) A sum of 8, 9, or 10.
(iv) One number less than 4.
4. An urn contains 12 white balls and 8 red balls. Two balls are to be selected in
succession, at random and without replacement. What is the probability that,
(i) Both balls are white.
(ii) The first ball is white and the second ball is red.
(iii) One white ball and one red ball are selected.
Would the probabilities change if the first ball after being identified is put
back in the urn before the second ball is selected?
5. 200 students from the college were surveyed to find out if they were taking
any of the Management, Marketing or Finance courses. It was found that 80
of them were taking Management courses, 70 of them were taking Marketing
courses and 50 of them were taking Finance courses. It was also found that
30 of them were taking Management and Marketing courses, 30 of them
were taking Management and Finance courses and 25 of them were taking
Marketing and Finance courses. It was further determined that 20 of these
students were taking courses in all the three areas. What is the probability
that a particular student is not taking any course in any of these areas?
6. A family plans to have three children. List all the possible combinations and
find the probability that all the three children will be boys.
7. A movie house is filled with 700 people and 60 per cent of these people are
females. 70 per cent of these people are seated in the no smoking area including
300 females. What is the probability that a person picked up at random in the
movie house is:
(i) A male.
(ii) A female smoker.
(iii) A male or a non-smoker.
(iv) A smoker if we knew that the person is a male.
(v) Are the events sex and smoking statistically independent?
8. If a fair dice is rolled once, state the probability of getting:
(i) An odd number
(ii) A number greater than 3
9. In a computer course, the probability that a student will get an A is 0.09. The
probability that he will get a B grade is 0.15 and the probability that he will

20 Statistics for Management

Concepts of Probability Unit-1

get a C grade is 0.45. What is the probability that the student will get either a
D or an F grade?
10. An urn contains 12 white balls and 8 red balls. Two balls are to be selected in
succession, at random and without replacement. What is the probability that:
(i) Both balls are white.
(ii) The first ball is white and the second ball is red.
(iii) One white ball and one red ball are selected.
(iv) Would the probabilities change if the first ball after being identified is
put back in the urn before the second ball is selected?
11. In a statistics class, the probability that a student picked up at random comes
from a two parent family is 0.65, and the probability that he will fail the exam
is 0.20. What is the probability that such a randomly selected student will be
a low achiever given that he comes from a two parent family?
12. The following is a breakdown of faculty members in various ranks at the
college.
Rank Number of Males Number of Females
Professor 20 12
Assoc. Professor 18 20
Asst. Professor 25 30
What is the probability that a faculty member selected at random is:
(i) A female.
(ii) A female professor.
(iii) A female given that the person is a professor.
(iv) A female or a professor.
(v) A professor or an assistant professor.
(vi) Are the events of being a male and being an associate professor
statistically independent events?

1.9 LEARNING OUTCOMES

• Learn sample space
• Events and their role in probability
• The axiomatic approach to probability
• The probability problems using addition and multiplication rules
• Learn independent events
• Learn conditional probability
• bescrie Bayes’ theorem

Statistics for Management 21

Probability Distributions Unit-2

UNIT 2 PROBABILITY
DISTRIBUTIONS
UNIT STRUCTURE
2.1 Learning Objectives
2.2 Introduction
2.3 Random Variable
2.3.1 Techniques of Assigning Probabilities
2.3.2 Mean of Random Variable or the Expected Value of Random Variable
2.3.3 Variance and Standard Deviation of Random Variable
2.4 Expectation and Variance
2.4.1 Expected Value of a Function of X; 2.4.2 Variance
2.4.3 Expectation (Conditional); 2.4.4 Expectation (Iterated)
2.4.5 Expectation: Continuous Variables
2.5 Continuous and Discrete Probability Functions
2.6 Discrete Probability Distributions
2.6.1 Binomial Distribution—Properties
2.6.2 Applications of Binomial Distribution
2.6.3 Poisson Distribution—Properties
2.6.4 Applications of Poisson Distribution
2.7 Continuous Probability Distributions
2.8 Normal Distribution
2.8.1 Characteristics of Normal Distribution
2.8.2 Family of Normal Distributions
2.8.3 How to Measure the Area under the Normal Curve
2.9 Let Us Sum Up
2.10 Further Reading
2.11 Answers to Check Your Progress
2.12 Possible Questions
2.13 Learning Outcomes

2.1 LEARNING OBJECTIVES

After going through this unit, you will be able to:

• Explain the expectation property of a random variable
• Explain the functionality of a random variable
• List the probability density functions
• List the probability distribution functions

2.2 INTRODUCTION

Randomness means each possible entity has the same chance of being considered.
A random variable may be qualitative or quantitative in nature. In this unit, you will
study probability distribution, which means listing of all possible outcomes of an
experiment together with their probabilities. It may be discrete or continuous. You
will also study the concept of expectation. An expected value is the sum of each
possible outcome and the probability of occurrence of outcome. Expectation may
Statistics for Management 23
Unit-2 Probability Distributions

be conditional or iterated. You will study about the moment generating function
(MGF), which generates the moments for the probability distribution of a random
variable.

2.3 RANDOM VARIABLE

A random variable is a variable that takes on different values as a result of the

outcomes of a random experiment. In other words, a function which assigns
numerical values to each element of the set of events that may occur (i.e., every
element in the sample space) is termed a random variable. The value of a random
variable is the general outcome of the random experiment. One should always make
a distinction between the random variable and the values that it can take on. All
these can be illustrated by a few examples shown in Table 2.1.
Table 2.1 Random Variable

Random Variable Values of the Description of the Values of

Random Variable the Random Variable
X 0, 1, 2, 3, 4 Possible number of heads in
four tosses of a fair coin
Y 1, 2, 3, 4, 5, 6 Possible outcomes in a
single throw of a die
Z 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 Possible outcomes from
throwing a pair of dice
M 0, 1, 2, 3, ............................ S Possible sales of newspapers
by a newspaper boy,
S representing his stock

All these above stated random variable assignments cover every possible
outcome and each numerical value represents a unique set of outcomes. A random
variable can be either discrete or continuous. If a random variable is allowed to
take on only a limited number of values, it is a discrete random variable but if it is
allowed to assume any value within a given range, it is a continuous random variable.
Random variables presented in the above table are examples of discrete random
variables. We can have continuous random variables if they can take on any value
within a range of values, for example, within 2 and 5, in that case we write the
values of a random variable x as follows:
2 x 5

2.3.1 Techniques of Assigning Probabilities

We can assign probability values to the random variables. Since the assignment of
probabilities is not an easy task, we should observe certain rules in this context as
given below:
(i) A probability cannot be less than 0 or greater than 1, i.e., 0 pr 1, where pr
represents probability.
(ii) The sum of all the probabilities assigned to each value of the random variable
must be exactly 1.

24 Statistics for Management

Probability Distributions Unit-2

There are three techniques of assignment of probabilities to the values of the random
variable:
(a) Subjective probability assignment. It is the technique of assigning
probabilities on the basis of personal judgement. Such assignment may differ
from individual to individual and depends upon the expertise of the person
assigning the probabilities. It cannot be termed as a rational way of assigning
probabilities but is used when the objective methods cannot be used for one
reason or the other.
(b) A-priori probability assignment. It is the technique under which the
probability is assigned by calculating the ratio of the number of ways in
which a given outcome can occur to the total number of possible outcomes.
The basic underlying assumption in using this procedure is that every
possible outcome is likely to occur equally. But at times, the use of this
technique gives ridiculous conclusions. For example, we have to assign
probability to the event that a person of age 35 will live upto age 36. There
are two possible outcomes, he lives or he dies. If the probability assigned
in accordance with a-priori probability assignment is half then the same
may not represent reality. In such a situation, probability can be assigned
by some other techniques.
(c) Empirical probability assignment. It is an objective method of assigning
probabilities and is used by decision-makers. Using this technique the
probability is assigned by calculating the relative frequency of occurrence of
a given event over an infinite number of occurrences. However, in practice
only a finite (perhaps very large) number of cases are observed and relative
frequency of the event is calculated. The probability assignment through this
technique may as well be unrealistic, if future conditions do not happen to be
a reflection of the past.
Thus, what constitutes the ‘best’ method of probability assignment can only
be judged in the light of what seems best to depict reality. It depends upon the
nature of the problem and also on the circumstances under which the problem
is being studied.
2.3.2 Mean of Random Variable or the Expected Value of Random
Variable
Mean of random variable is the sum of the values of the random variable weighted
by the probability that the random variable will take on the value. In other words,
it is the sum of the product of the different values of the random variable and
their respective probabilities. Symbolically, we write the mean of a random variable,
say X, as X . The Expected value of the random variable is the average value that
would occur if we have to average an infinite number of outcomes of the random
variable. In other words, it is the average value of the random variable in the
long run. The expected value of a random variable is calculated by weighting
each value of a random variable by its probability and summing over all
values. The symbol for the expected value of a random variable is E (X).

Statistics for Management 25

Unit-2 Probability Distributions
Mathematically, we can write the mean and the expected value of a random
variable, X, as follows:
n
X Xi . pr. Xi
i 1

and,
n
E X Xi .pr. X i
i 1
Thus, the mean and the expected value of a random variable are conceptually
and numerically the sam_e but usually denoted by different symbols and as such
the two symbols, viz., X and E (X) are completely interchangeable. We can,
therefore, express the two as follows:
n
E X Xi . pr. Xi X
i 1

Where Xi is the ith value X can take.

Sum of Random variables

If we are given the means or the expected values of different random variables, say
X, Y, and Z to obtain the mean of the random variable (X + Y + Z), then it can be
obtained as under:
E X Y Z E X E Y E Z X Y Z

2.3.3 Variance and Standard Deviation of Random Variable

The mean or the expected value of a random variable may not be adequate
enough at times to study the problem as to how a random variable actually
behaves and we may as well be interested in knowing something about how the
values of random variables are dispersed about the mean. In other words, we
want to measure the dispersion of random variable (X) about its expected value,
i.e., E(X). The variance and the standard deviation provide measures of this
dispersion.
The variance of a random variable is defined as the sum of the squared
deviations of the values of the random variable from the expected value weighted
by their probability. Mathematically, we can write it as follows:
n
2 2
Var X X
Xi E X . pr. X i
i1

Alternatively, it can also be written as,

2 2
Var X X X i 2 pr. X i E X

Where, E (X) is the expected value of random variable.

Xi is the ith value of random variable.
pr. (Xi) is the probability of the ith value.

26 Statistics for Management

Probability Distributions Unit-2
The standard deviation of random variable is the square root of the
variance of random variable and is denoted as,
2 =
X X

The variance of a constant time random variable is the constant squared

times the variance of random variable. This can be symbolically written as,
Var (cX ) = c2 Var (X)
The variance of a sum of independent random variables equals the sum of
the variances.
Thus,
Var (X + Y + Z ) = Var(X) + Var(Y ) + Var(Z)
If X, Y and Z are independent of each other.
The following examples will illustrate the method of calculation of these
measures of a random variable.
Example 2.1: Calculate the mean, variance and standard deviation of random
variable sales from the following information provided by a sales manager of a
certain business unit for a new product:
Monthly Sales (in units) Probability
50 0.10
100 0.30
150 0.30
200 0.15
250 0.10
300 0.05
Solution:
The given information may be developed as shown in the following table for
calculating the mean, variance and the standard deviation for random variable
sales:
Monthly Sales Probability (Xi) pr (Xi) (Xi – E (X))2 Xi – E (X)2
(in units)1 Xi pr (Xi) pr (Xi)
X1 50 0.10 5.00 (50 – 150)2 1000.00
= 10000
X2 100 0.30 30.00 (100 – 150)2 750.00
= 2500
X3 150 0.30 45.00 (150 – 150)2 0.00
=0
X4 200 0.15 30.00 (200 – 150)2 375.00
= 2500
X5 250 0.10 25.00 (250 – 150)2 1000.00
= 10000
X6 300 0.5 15.00 (300 – 150)2 1125.00
= 22500
(Xi) pr (Xi) [Xi – E (X)2]
= 150.00 pr (Xi) = 4250.00

Statistics for Management 27

Unit-2 Probability Distributions
_
Mean of random variable sales = X
or E (X) = (Xi).pr(Xi) = 150
Variance of random variable sales,
n
2 2
or X Xi E X . pr X i 4250
i 1

Standard deviation of random variable sales,

2
or X X 4250 65.2 approx.
The mean value calculated above indicates that in the long run the average sales
will be 150 units per month. The variance and the standard deviations measure
the variation or dispersion of random variable values about the mean or the
expected value of random variable.
Example 2.2: Given are the mean values of four different random variables viz.,
A, B, C, and D.
A 20, B 40, C 10, D 5
Find the mean value of the random variable (A + B + C + D).
Solution:
E(A + B + C + D) = E(A) + E(B) + E(C) + E(D)
= A B C D
= 20 + 40 + 10 + 5
= 75
Hence, the mean value of random variable (A + B + C + D) is 75.
Example 2.3: If X represents the number of heads that appear when one coin is
tossed and Y the number of heads that appear when two coins are tossed, compare
the variances of the random variables X and Y. The probability distributions of X
and Y are as follows:
Xi pr (Xi) Yi pr (Yi)
0 1/2 0 1/4
1 1/2
1 1/2 2 1/4
Total = 1 Total = 1
Solution:
2 2
2
1 1 1 1
Variance of X X = 0 . 1 .
2 2 2 2
1 1 1
= 0.25
8 8 4
2 1 1
Variance of Y 2 = 0 1 . (1 1)2
Y 4 2
1 1 1
+ ( 2 1) .
2
= +0+ = 0.50
4 4 4

28 Statistics for Management

Probability Distributions Unit-2

The variance of the number of heads for two coins is double the variance
of the number of heads for one coin.

2.4 EXPECTATION AND VARIANCE

The expected value (or mean) of X is the weighted average of the possible values
that X can take. Here, X is a discrete random variable and each value is being weighted
according to the probability of the possibility of the occurrence of the event. The
expected value of X is usually written as E(X) or .
E(X) = × P(X = x)
Hence, the expected value is the sum of:
Each of the possible outcomes + The probability of the outcome occurring
Therefore, expectation is the outcome you expect of an experiment.
Let us consider the following example,
What is the expected value when we roll a fair die?
There are six possible outcomes 1, 2, 3, 4, 5, 6. Each one of these has a probability
of 1/6 of occurring. Let X be the outcome of the experiment.
Then,
P(X =1) = 1/6 (this shows that the probability that the outcome of the
experiment is 1 is 1/6)
P(X = 2) = 1/6 (the probability that you throw a 2 is 1/6)
P(X = 3) = 1/6 (the probability that you throw a 3 is 1/6)
P(X = 4) = 1/6 (the probability that you throw a 4 is 1/6)
P(X = 5) = 1/6 (the probability that you throw a 5 is 1/6)
P(X = 6) = 1/6 (the probability that you throw a 6 is 1/6)
E(X) = 1×P(X = 1) + 2×P(X = 2) + 3×P(X = 3) + 4×P(X=4) + 5×P(X=5) +
6 × P (X = 6)

Statistics for Management 29

Unit-2 Probability Distributions

Therefore,
E(X) = 1/6 + 2/6 + 3/6 + 4/6 + 5/6 + 6/6 = 7/2 or 3.5
Hence, the expectation is 3.5, which is also the halfway between the possible values
the die can take, and so this is what you should have expected.
2.4.1 Expected Value of a Function of X
To find E[ f(X) ], where f(X) is a function of X, we use the following formula:
E[ f(X) ] = f(x)P(X = x)
Let us consider the above example of die, and calculate E(X2)
Using the notation above, f(x) = x2
f(1) = 1, f(2) = 4, f(3) = 9, f(4) = 16, f(5) = 25, f(6) = 36
P(X = 1) = 1/6, P(X = 2) = 1/6, etc.
Hence, E(X2) = 1/6 + 4/6 + 9/6 + 16/6 + 25/6 + 36/6 = 91/6 = 15.167
The expected value of a constant is just the constant, as for example E(1) = 1.
Multiplying a random variable by a constant multiplies the expected value by that
constant.
Therefore, E[2X] = 2E[X]
An important formula, where a and b are constants, is:
E[aX + b] = aE[X] + b
Hence, we can say that the expectation is a linear operator.
2.4.2 Variance
The variance of a random variable tells us something about the spread of the possible
values of the variable. For a discrete random variable X, the variance of X is written
as Var(X).
Var(X) = E[(X – )2]

Where, is the expected value E(X).

This can also be written as:
Var(X) = E(X2) – 2
The standard deviation of X is the square root of Var(X).
Note: The variance does not behave in the same way as expectation, when we
multiply and add constants to random variables.
Var[aX + b] = a2Var(X)
Because, Var[aX + b] = E[ (aX + b)2 ] – (E [aX + b])2
= E[ a2X2 + 2abX + b2] – (aE(X) + b)2
= a2E(X2) + 2abE(X) + b2 – a2E2(X) – 2abE(X) – b2
= a2E(X2) – a2E2(X) = a2Var(X)
30 Statistics for Management
Probability Distributions Unit-2

2.4.3 Expectation (Conditional)

The expectation of a random variable X with probability density function (PDF)
p(x) is theoretically defined as:

E[ X ] = z xp( x)dx
If we consider two random variables X and Y (not necessarily independent), then
their combined behaviour is described by their joint probability density function
p(x, y) and is defined as:

p{x X < x + dx, y Y < y + dy} = p(x, y).dx.dy

The marginal probability density of X is defined as,

pX (x) = z p(x, y)dy

For any fixed value y of Y, the distribution of X is the conditional distribution of X,
where Y = y , and it is denoted by p(x , y).
2.4.4 Expectation (Iterated)
The expectation of the random variable is expressed as:
E[X] = E[E[X | Y]]
This expression is known as the ‘Theorem of iterated expectation’ or ‘Theorem
of double expectation’. Symbolically, it can be expressed as:
(i) For the discrete case:
E[X] = E[X | Y = y].P{Y = y}
y

(ii) For the continuous case:

E[ X ] = z +
E[ X |Y = y]. f ( y).dy

2.4.5 Expectation: Continuous Variables

If x is a continuous random variable we define that,

E(x) = x P(x)dx = u
–

The expectation of a function h(x) is,

Eh(x) = h(x) P(x)dx

–

The rth moment about the mean is,

E(x – )r = (x – μ)r P(x)dx

–

Statistics for Management 31

Unit-2 Probability Distributions

Example 2.4: A newspaper boy earns 100 a day if there is suspense in the news.
He loses 10 a day if it is an eventless newspaper. What is the expectation of his
earnings if the probability of suspense news is 0.4?
Solution: E(x) = p1x1 + p2x2
= 0.4 × 100 – 0.6 × 10
= 40 – 6 = 34
Example 2.5: A player tossing three coins, earns 10 for 3 heads, 6 for 2 heads
and 1 for 1 head. He loses 25, if 3 tails appear. Find his expectation.
1 1 1 1
Solution: p(HHH) = . . = = p say 3 heads
1
2 2 2 8
1 1 1 3
p(HHT) = 3 C2 . . . = = p 2 say 2 heads, 1 tail
2 2 2 8
1 1 1 3
p(HTT) = 3 C . . . = = p say 1 head, 2 tails
1 3
2 2 2 8
1.1.1 3
p(TTT) = = = p 4 say 3 tails
2 2 2 8
E(x) = p1x1 + p2x2 + p3x3 – p4x4
1 3 3 1
= 10 + 6+ 2– 25
8 8 8 8
9
= = 1.125
8
Example 2.6: A and B roll a die. Whoever gets a 6, first wins 550. Find their
individual expectations if A makes a start. What will be the answer if B makes a
start?
1
Solution: A may not get 6 in 1st trial, p1 =
6
A may not get in 1st, B may not get in 2nd and A may get in 3rd,
2
5 5 1 5 1
p3 = = and so on.
6 6 6 6 6
2
5 5 1 5 1
A’s winning chance = = +...........
6 6 6 6 6
Where p(x) is the density function x,

1 1 6
= 2 =
6 5 11
1–
6

32 Statistics for Management

Probability Distributions Unit-2
2
5
Geometric progression with common ratio
6

6 5
B’s winning chance = 1 – =
11 11
6
A wins 550 with probability p =
11
5
A gets nothing if he loses with probability q =
11
Expectation of A = p . x + q . 0
6 5
= 550 + 0 = 300
11 11
5
Similarly B wins with p =
11
6
B gets nothing if he loses with q =
11
5 6
Expectation of B = 550 + 0 = 250
11 11
If B starts first his expectation would be 300 and A’s would be 250.
Example 2.7: Calculate the standard deviation (S.D.) when x takes the values 0, 1,
2 and 9 with probability, 0.4, 0 – 0.1.
Solution: x takes the values 0, 1, 2, 9 with probability 0.4, 0.2, 0.3, 0.1.
= E(x) = (x
i i
p ) = 0 × 0.4 + 12 × 0.2 + 3 × 0.3 + 9 × 0.1 = 2.0
E(x2) = xi 2pi = 02 × 0.4 + 12 × 0.2 + 3 × 0.3 + 92 × 0.1 = 11.0
V(x) = E(x2) – 2
= 11 – 2 = 9
S.D.(x) = 9 =3
Example 2.8: The purchase of some shares can give a profit of 400 with probability
1/100 and 300 with probability 1/20. Comment on a fair price of the share.
1 1
Solution: Expected value E(x) = x p = 400 + 300 = 19
i i
100 20
Example 2.9: Find the variance and standard deviation (S.D.) of the following
probability distribution:

x1 1 2 3 4
p1 0.1 0.3 0.2 0.4

Statistics for Management 33

Unit-2 Probability Distributions

Solution: E(x) = p xi i= 0.1 × 1 + 0.3 × 2 + 0.2 × 3 + 0.4 × 4 = 2.9

Variance (x) = V(x) = E(x2) – x 2
= p xi 2i – x 2 = 0.1 12 + 0.3 22 + 0.2 32 + 0.4 42 – 2.92
= 0.1 + 1.2 + 1.8 + 6.4 – 8.41 = 1.09
S.D.(x) = V (x) = 1.09 = 1.04
Example 2.10: Prove that the variance of a constant is zero.
Solution: If k is a constant, it has no variability.
V(k) = 0
If the constant k is attached to a variable x then,
V(kx) = k2V(x)
V(2x) = 4V(x)
V(2 + 3x) = V(3x) = 9 V(x)
Example 2.11: A box contains 2n tickets of which nCi tickets bear the number
i(i = 1, 2, 3, .... , n). If a set of m tickets is drawn, find the expected value of the sum
of their numbers.
Solution: To find E(A), where A = x1 + x2 + ...+ xm. Here xi can take values 0, 1, 2, ... ,
n with probabilities,
n n n
C C C
0 1 n
, , ....
2n 2n 2n
n n n n
C0 C1 C2 Cn
E(x ) = p x = .0 + .1+ .2 +.... + .n
i i
i
2n 2n 2n 2n
1
= [n C + 2n C + .... + nnC ]
n 1 2 n
2
n
[1+n–1 C
+ n –1C 2 + ....+
n–1
= C ] (By taking n common)
n 1 n –1
2
n n
= (1 +1)n–1 = [Expand (1 + 1)n – 1 and check]
2mn mn2
E(x ) =
E(A) = i
i=1 2

2.5 CONTINUOUS AND DISCRETE PROBABILITY

FUNCTIONS

Probability density may be defined as a measure of existence. Logic can be analysed

with the help of a graph. In graphical representation, a probability density function
gives information about the existence of samples at various locations and the entire
graph area can be considered as a sample space.

34 Statistics for Management

Probability Distributions Unit-2

A mathematical description of samples in regions can be given under a

coordinate system. The regions and the existence of samples at various locations
can be described in this coordinate system by mathematical functions. For example,
to locate a point in a sample space it is expressed as a pair of (x,y) coordinates and
a mathematical function describes a location for each sample.
The function that shows the location of each sample is called a ‘density
function’. By showing locations of each sample, density of the samples can be
understood. If we assume that the odds of selecting any one sample is the same as
for any other, then the PDF (probability density function), p(x,y), gives the probability
of a point in sample space to lie between two given limits of the variables.
Thus, in case of a discrete variable x, the probability function, PDF(x), is the
probability of occurrence of x, variable of continuous nature, PDF(x) shows
probability density of a variable x. Thus, probability of a value between x and x+dx
is PDF(x) × dx.
A ‘cumulative density function’, CDF(x), shows the probability that the said
variable assumes a value <= x .
The mean or the average value is available in majority of cases. This mean is
taken as sum of products x × PDF(x). In case of continuous variables, mean is given
by an integral of X × PDF(x), integrated in the range.
Let f be a non-negative function mapped R R then the probability
distribution has density f probability in the interval [a, b] is given as,
b

a
f (x) dx in case of any two numbers a and b
Total integral of f must be 1. It is converse is also true which tells that a
function f with total integral is equal to 1 then, f represents the probability density
for some probability distribution.
A probability density function is in fact, a refined version of a histogram with
very small or infinitesimal interval. Due to this, the curve is smooth. If sampling is
done, taking many values of a random variable which is of continuous nature, a
histogram is produced that depicts a histogram showing probability density, in a
very narrow output range. This can be termed a probability density function if and
only if it is non-negative and the area under the graph is 1. Putting mathematically
with logical connective showing ‘conjunction’, it is given as:

f (x) 0 x f (x) dx = 1
All distributions do not show density function. It is said to have a density
function f(x) only if its CDF, denoted as F(x), is continuous.
This is expressed mathematically as:
d
F ( x) = f ( x)
dx
Link between discrete and continuous distributions
A probability density function describes the association of variable in a distribution
that is continuous: taking a set of two state variables in the interval [a, b].
Discrete random variables may be represented in probability distribution using
a delta function. If we consider a binary discrete random variable by taking two
Statistics for Management 35
Unit-2 Probability Distributions

distinct values, say 1 and 2 which are equally likely, the probability density of such
a variable is given by:
1
f (t) = ( (t + 1) + (t 1))
2
We may generalize this as follows: If a discrete variable assumes ‘n’ different
values in the set of real numbers, then the associated probability density function is
given by:
n

f (t) = Pi (t xi )
i =1

Where i = 1, 2,3....., n, and x1, ......,. xn stand for discrete values for the variables
and SPi (where i = 1, 2, …., n) are probabilities associated with these values.
The method is used for knowing the characteristics of the mean, its variance
and kurtosis.
The method is used to show mathematically the characteristic of Brownian
movement and deciding on its initial configuration.
Probability functions associated with multiple variables
In case of random variables, of continuous nature, SXi, where i = 1, 2,3, 4, ….., n,
one can define a PDF for the whole set. This is a term coined as joint PDF which is
defined as a function with n variables. This is known by a different name MDF
which means marginal density function.
Independence
Random variables X1, …,…, Xn of continuous nature are independent of each other
if and only if fX1, ...., Xn (x1, ..., xn) = fX1(x1) ... fXn(xn).
Corollary
If a JPDF (joint probability distribution function) of a vector of n random variables
is shown as a product of n functions of one variable fX1, ...., Xn (x1, ..., xn) = f1(x1) ·
fn(xn), then all these variables are independent of each other. MPDF (marginal
probability density function) for each is expressed as:
fi (xi )
fXi (xi ) =
fi (x) dx

For example, this illustrates the definition of a MDPF (multidimensional probability

density functions) as stated above. This condition is considered for a function with
a set of two variables. This condition is considered in case of a function of a set of
two variables. Let us call R a 2-dimensional random vector of coordinates (X,Y):
the probability to obtain R in the quarter plane of positive x and y is

Pr( X > 0,Y > 0) = fX ,Y (x, y) dxdy

0 0

36 Statistics for Management

Probability Distributions Unit-2

Sums of independent random variables

Let there be two random variables, u and v. Here, each of the variables has a PDF;
then the sum of these two PDFs, is taken as convolution of these two and is the
convolution of their separate density functions. This is given mathematically, as
below:

fu + v (x) = fu( y) fv (x y) dy

2.6 DISCRETE PROBABILITY DISTRIBUTIONS

When a random variable x takes discrete values x1, x2,.... , xn with probabilities p1,
p2,.. ,pn, we have a discrete probability distribution of X.
The function p(x) for which X = x1, x2,..., xn takes values p1, p2, ... ,pn, is the
probability function of X.
The variable is discrete because it does not assume all values. Its properties
are:
p(xi) = Probability that X assumes the value x.
= Prob (x = xi) = pi
p(x) 0, p(x) = 1
For example, four coins are tossed and the number of heads X noted. X can take
value 0, 1, 2, 3, 4 heads.
4
1 1
p(X = 0) = =
2 16
3
p(X = 1) = 4 C 1 1 4
1
=
2 2 16

2 2
p(X = 2) = 4 C 2 1 1
=
6
2 16
2
Statistics for Management 37
Unit-2 Probability Distributions
3
p(X = 3) = 4 C 1 1 4
3 =
2 2 16
4 0
p(X = 4) = 4 C 4 1 1 = 1
2 16
2
6
16

5
16

4
16

3
16

2
16

1
16

O
0 1 2 3 4

4 6 4 1
p(x) = 1 + 4 + + + = 1
x=2 16 16 16 16 16
This is a discrete probability distribution.
Example 2.12: If a discrete variable X has the following probability function, then,
find (i) a (ii) p(X 3) (iii) p(X 3).
Solution: x1 p(xi)
0 0
1 a
2 2a
3 2a2
4 4a2
5 2a
Since p(x) = 1 , 0 + a + 2a + 2a2 + 4a2 + 2a = 1
6a2 + 5a – 1 = 0, so that (6a – 1) (a + 1) = 0
1
a= or a = –1 (not admissible)
6
1 5
2 2
For a = , p(X 3) = 0 + a + 2a + 2a = 2a + 3a =
6 9
4
2
p(X 3) = 4a + 2a =
9

38 Statistics for Management

Probability Distributions Unit-2

Discrete Distributions
There are several discrete distributions. Some other discrete distributions are
described below.
Uniform or rectangular distribution
Each possible value of the random variable x has the same probability in the uniform
distribution. If x takes vaues x1, x2 ...... ,xk, then
1
p(xi, k) =
k
The numbers on a die follow the uniform distribution,
1
p(xi, 6) = (Here, x = 1, 2, 3, 4, 5, 6)
6

Bernoulli trials
In a Bernoulli experiment, an even E either happens or does not happen (E ).
Examples are, getting a head on tossing a coin, getting a six on rolling a die, and so
on.
The Bernoulli random variable is written,
X = 1 if E occurs
= 0 if E occurs
Since there are two possible values it is a case of a discrete variable where,
Probability of success = p = p(E)
Profitability of failure = 1 – p = q = p(E )
We can write,
For k = 1, f(k) = p
For k = 0, f(k) = q
For k = 0 or 1, f(k) = pkq1–k
Negative binomial
In this distribution, the variance is larger than the mean.
Suppose, the probability of success p in a series of independent Bernoulli
trials remains constant.
Suppose the rth success occurs after x failures in x + r trials.
1. The probability of the success of the last trial is p.
2. The number of remaining trials is x + r – 1 in which there should be r – 1
successes. The probability of r – 1 successes is given by,
x + r –1
Cr –1 p r –1 q x
The combined pobability of cases (1) and (2) happening together is,
x + r –1
p(x) = px C r –1 pr –1 qx x = 0, 1, 2,....

Statistics for Management 39

Unit-2 Probability Distributions

This is the Negative Binomial distribution. We can write it in an alternative

form,
p(x) = – r Cx pr ( q)x x = 0, 1, 2,....
This can be summed up as follows:
In an infinite series of Bernoulli trials the probability that x + r trials will be
required to get r successes is the negative binomial,
x + r –1
p(x) = Cr –1 p r –1 q x r 0
If r = 1, it becomes the geometric distribution.
If p 0, , rp = m a constant, then the negative binomial tends to be the
Poisson distribution.
Geometric distribution
Suppose the probability of success p in a series of independent trials remains constant.
Suppose, the first success occurs after x failures, i.e., there are x failures
preceding the first success. The probability of this event will be given by p(x) = qxp
(x = 0, 1, 2, ).
This is the geometric distribution and can be derived from the negative
binomial. If we put r = 1 in the negative binomial distribution:
x + r –1
p(x) = Cr –1 p r –1 q x
We get the geometric distribution,
p(x) = x C0 p1 qx = pqx
p
x p
p(x) = q p= =1
n=0 1 q

p
E(x) = Mean =
q
p
Variance =
q2
x
1
Mode =
2
Example 2.13: Find the expectation of the number of failures preceding the first
success in an infinite series of independent trials with constant probability p of
success.
Solution: The probability of success in,
1st trial = p (Success at once)
2nd trial = qp (One failure then success, and so on)
3rd trial = q2p (Two failures then success, and so on)

40 Statistics for Management

Probability Distributions Unit-2

The expected number of failures preceding the success,

E(x) = 0 . p + 1. pq + 2p2p + ............
= pq(1 + 2q + 3q2 + ......... )
1 1 q
pq = qp =
=
(1 – q) 2
p 2
p
Since p = 1 – q.
Hypergeometic distribution
From a finite population of size N, a sample of size n is drawn without replacement.
Let there be N1 successes out of N.
The number of failures is N2 = N – N1
The disribution of the random variable X, which is the number of successes
obtained in the above case, is called the hypergeometic distribution.
N
CN C 1

x n x
p(x) = N (X = 0, 1, 2,....., n)
Cn
Here, x is the number of successes in the sample and n – x is the number of
failures in the sample.
It can be shown that,
N1
Mean : E(X) = n
N
N n nN nN 2
Variance : Var (X) = 1
– 1

N 1 N N
Example 2.14: There are 20 lottery tickets with three prizes. Find the probability
that out of 5 tickets purchased exactly two prizes are won.
Solution: We have N1 = 3, N2 = N – N1 = 17, x = 2, n = 5.
3
C2 17C3
p(2) = 20
C5
3
C0 17C 5
The probability fo no pize p(0) = 20
C5
3
C1 17 C4
The probability of exactly 1 prize p(1) = 20
C5
Example 2.15: Examine the nature of the distibution when r balls are drawn, one at
a time without replacement, from a bag containing m white and n black balls.
Solution: It is the hypergeometric distribution. It corresponds to the probability
that x balls will be white out of r balls so drawn and is given by,
x
C x nC r x
p(x) = m+ n
Cr

Statistics for Management 41

Unit-2 Probability Distributions

Multinomial
There are k possible outcomes of trials, viz., x1, x2, ..., xk with probabilities p1, p2, ...,
pk, n independent trials are performed. The multinomial distibution gives the
probability that out of these n trials, x1 occurs n1 times, x2 occurs n2 times, and so on.
n!
p1n pn2 ..... pnk ,
1 2
This is given by n !n !....n !
1 2 k

where, ni = n
i=1

2.6.1 Binomial Distribution—Properties

Binomial distribution (or the Binomial probability distribution) is a widely used
probability distribution concerned with a discrete random variable and as such is an
example of a discrete probability distribution. The binomial distribution describes
discrete data resulting from what is often called the Bernoulli process. The tossing
of a fair coin a fixed number of times is a Bernoulli process and the outcome of such
tosses can be represented by the binomial distribution. The name of Swiss
Bernoulli process: The
tossing of a fair coin a mathematician Jacob Bernoulli is associated with this distribution. This distribution
fixed number of times. applies in situations where there are repeated trials of any experiment for which
only one of two mutually exclusive outcomes (often denoted as ‘success’ and
‘failure’) can result on each trial.
Bernoulli process
Binomial distribution is considered appropriate in a Bernoulli process which has
the following characteristics:
(a) Dichotomy. This means that each trial has only two mutually exclusive possible
outcomes, e.g., ‘Success’ or ‘failure’, ‘Yes’ or ‘No’, ‘Heads’ or ‘Tails’ and the
like.
(b) Stability. This means that the probability of the outcome of any trial is known
(or given) and remains fixed over time, i.e., remains the same for all the
trials.
(c) Independence. This means that the trials are statistically independent, i.e., to
say the happening of an outcome or the event in any particular trial is
independent of its happening in any other trial or trials.
Probability function of binomial distribution
The random variable, say X, in the binomial distribution is the number of ‘successes’
in n trials. The probability function of the binomial distribution is written as under:
f (X = r) = nCr prqn–r
r = 0, 1, 2…n
where, n = Numbers of trials
p = Probability of success in a single trial
q = (1 – p) = Probability of ‘failure’ in a single trial
r = Number of successes in ‘n’ trials
42 Statistics for Management
Probability Distributions Unit-2

Parameters of binomial distribution

Binomial distribution depends upon the values of p and n which in fact are its
parameters. Knowledge of p truly defines the probability of X since n is known by
definition of the problem. The probability of the happening of exactly r events in n
trials can be found out using the above stated binomial function.
The value of p also determines the general appearance of the binomial
distribution, if shown graphically. In this context, the usual generalizations are:
(a) When p is small (say 0.1), the binomial distribution is skewed to the right,
i.e., the graph takes the form shown in the following figure.
Probability

0 1 2 3 4 5 6 7 8
No. of Successes
(b) When p is equal to 0.5, the binomial distribution is symmetrical and the graph
takes the form as shown in the following figure.
Probability

0 1 2 3 4 5 6 7 8
No. of Successes
(c) When p is larger than 0.5, the binomial distribution is skewed to the left and
the graph takes the form as shown in the following figure.
Probability

0 1 2 3 4 5 6 7 8
No. of Successes
But if ‘p’ stays constant and ‘n’ increases, then as ‘n’ increases the vertical
lines become not only numerous but also tend to bunch up together to form a bell
shape, i.e., the binomial distribution tends to become symmetrical and the graph
takes the shape as shown in the following figure.
Probability

0, 1, 2, ..........
No. of Successes

Statistics for Management 43

Unit-2 Probability Distributions

Important measures of binomial distribution

The ex_pected value of random variable [i.e., E(X)] or mean of random variable
(i.e., X ) of the binomial distribution is equal to n.p and the variance of random
variable is equal to n. p. q or n. p. (1 – p). Accordingly the standard deviation of
binomial distribution is equal to n. p.q. The other important measures relating
to binomial distribution are as under:
1 2p
Skewness =
n. p.q

1 6 p 6q2
Kurtosis = 3
n. p.q

When to use binomial distribution

The use of binomial distribution is most appropriate in situations fulfilling the
conditions outlined above. Two such situations, for example, can be described as
follows.
(a) When we have to find the probability of 6 heads in 10 throws of a fair coin.
(b) When we have to find the probability that 3 out of 10 items produced by a
machine, which produces 8 per cent defective items on an average, will be
defective.
Example 2.16: A fair coin is thrown 10 times. The random variable X is the number
of head(s) coming upwards. Using the binomial probability function, find the
probabilities of all possible values which X can take and then verify that binomial
distribution has a mean: X = n.p. and variance: 2 = n.p.q.
Solution: Since the coin is fair and so, when thrown, can come either with head
1 1
upward or tail upward. Hence, p (head) and q (no head) . The required
2 2
probability function is,
f(X = r) = nCr prqn–r
r = 0, 1, 2…10
The following table of binomial probability distribution is constructed using this
function.
_ _ _
Xi (Number Probability pri Xi pri (Xi – X) (Xi – X)2 (Xi – X)2.pi
of Heads)
0 10
C0 p0 q10 = 1/1024 0/1024 –5 25 25/1024
1 10
C1 p1 q9 = 10/1024 10/1024 –4 16 160/1024
2 10
C2 p2 q8 = 45/1024 90/1024 –3 9 405/1024
3 10
C3 p3 q7 = 120/1024 360/1024 –2 4 480/1024
4 10
C4 p4 q6 = 210/1024 840/1024 –1 1 210/1024
10
5 C5 p5 q5 = 252/1024 1260/1024 0 0 0/1024
10
6 C6 p6 q4 = 210/1024 1260/1024 1 1 210/1024
10
7 C7 p7 q3 = 120/1024 840/1024 2 4 480/1024

44 Statistics for Management

Probability Distributions Unit-2
10 8 2
8 C8 p q = 45/1024 360/1024 3 9 405/1024
10 9 1
9 C9 p q = 10/1024 90/1024 4 16 160/1024
10
10 C10 p10 q0 = 1/1024 10/1024 5 25 25/1024
_
X
_ = 5120/1024 Variance_= 2 =
X=5 (Xi– X)2.pri =
2560/1024 = 2.5

1
The mean of the binomial distribution1 is given by n. p. = 10 × = 5 and the
2
1 1
variance of this distribution is equal to n. p. q. = 10 × × = 2.5.
2 2
These values are exactly the same as we have found them in the above table.
Hence, these values stand verified with the calculated values of the two
measures as shown in the table.
2.6.2 Applications of Binomial Distribution
Binomial distribution is one of the simplest and most frequently used discrete
probability distributions and is very useful in many practical situations involving
either/or types of events. It has certain distinct properties which are enumerated as
follows:
1. It describes the distribution of probabilities where there are only two mutually
exclusive outcomes for each trial of an experiment. For example, when tossing
a coin, there are only two possible outcomes, namely a head and a tail, which
are both mutually exclusive events. Similarly, when checking the quality of a
product, we can see that either the item is good or it is defective. These two
possible outcomes are denoted as success or failure. Success is simply the
outcome in which we are interested. Similarly, if we are interested in the
probability of a head in the toss of a coin, then the outcome tail will be
considered a failure. The probability of success is symbolized by p and the
probability of failure is symbolized by q, which is also (1 – p).
2. Each trial is independent of other trials. This means that the outcome of any
trial is independent of the outcome of any other trial.
3. The probability of success p remains constant from trial to trial. Similarly,
the probability of failure q or (1 – p) remains constant over all observations.
4. The process is performed under the same conditions for a fixed and finite
number of trials, say (n).
The concept of binomial distribution may be clear with the following case:
Suppose we toss a coin five times. The outcome head is designated as success
and outcome tail as failure, with probability of success and failure being respectively

1. The value of the binomial probability function for various values of n and p are also available in
tables (known as binomial tables) which can be used for the purpose to ease calculation work. The
tables are of considerable help particularly when n is large. (See Appendix)

Statistics for Management 45

Unit-2 Probability Distributions

(p) and (q). Suppose that the following was the sequence of outcomes of these
tosses.
H, T, H, T, T
This means that there are two successes and three failures in the given order
as above. The probability of this sequence of outcomes can be found by multiplication
rule of joint probability of mutually exclusive events and is given as:

p q p q q = p2q3
However, if we are not concerned with any particular sequence of the outcome,
but only in the outcome of two successes in any order out of the five tosses, then it
can be shown that there are ten different ways of obtaining two heads out of five
tosses. By applying the addition rule of probability, we can see that the probability
of getting any sequence with two heads and three tails would be ten times the
probability of the single sequence obtained above.
Accordingly, the probability of any two heads in 5 tosses would be,
10 × p2q3
In our case, we have,
p = 0.5
q = (1 – p) = 0.5
x = number of successes desired = 2
n = number of trials undertaken = 5
(n – x) = number of failures
Hence, the probability of two heads out of five tosses is,

10(0.5)2 (0.5)3 = 0.3125

In our case, we have the number of successes (x) = 2 and the number of trials
tosses (n) = 5 and there are only ten possible combinations in which two heads can
occur. However, as the number of trials increases, it becomes more and more difficult
to list all the possible sequences. In such cases, the counting method, is required to
be used. The number of sequences for any number of tosses or trials (n) and any
number of successes (x) is given by the formula,
n!
n
Cx =
x!(n x)!
n
The symbol Cx is also simply written as
n .
x
Hence, the general formula of calculating the probability of x successes out
of n trials is given by,
n
P(x) = pxq(n x)

x
This expression is known as the binomial formula.

46 Statistics for Management

Probability Distributions Unit-2

Based upon this formula, the probability in our example of two heads in five
tosses can be calculated as follows,

5
P(2) = (.5) 2 (.5)3
2

5!
= (.5)2 (.5)3
2!(5 – 2)!

5 4 3 2 1 2 3
= (.5) (.5)
2 1 3 2 1

= 10(.5)2 (.5)3
= 0.3125
Example 2.17: If a new drug is found to be effective 40% of the time, then what is
the probability that in a random sample of 4 patients, it will be effective on 2 of
them?
Solution: Let us define effective as success and non-effective as failure.
Then,
p = 0.4 (since the drug is effective 40% of the time)
q = (1 – p) = (1 – 0.4) = 0.6
x=2
n=4
Now,
n
P(x) = ( p) x (q)n x
x

4
= (.4) 2 (.6) 2
2
4!
= (.4)2 (.6)2
2!(4 – 2)!
= 6 × .16 × .36
= 0.3456
2.6.3 Poisson Distribution—Properties
Poisson distribution is also a discrete probability distribution with which is associated
the name of a Frenchman, Simeon Denis Poisson who developed this distribution.
This distribution is frequently used in the context of operations research and for this
reason has a great significance for management people. This distribution plays an
important role in Queuing theory, Inventory control problems and also in Risk
models.

Statistics for Management 47

Unit-2 Probability Distributions
Unlike binomial distribution, Poisson distribution cannot be deducted on
purely theoretical grounds based on the conditions of the experiment. In fact, it
must be based on experience, i.e., on the empirical results of past experiments relating
to the problem under study. Poisson distribution is appropriate especially when the
probability of happening of an event is very small (so that q or (1– p) is almost
equal to unity) and n is very large such that the average of series (viz., n. p.) is a
finite number. Experience has shown that this distribution is good for calculating
the probabilities associated with X occurrences in a given time period or specified
area.
The random variable of interest in Poisson distribution is the number of
occurrences of a given event during a given interval (interval may be time, distance,
area, etc.). We use capital X to represent the discrete random variable and lower
case x to represent a specific value that capital X can take. The probability function
of this distribution is generally written as under:
x
e
f ( Xi = x) =
x!
x = 0, 1, 2…
Where, = Average number of occurrences per specified interval.2 In other
words, it is the mean of the distribution.
e = 2.7183 being the basis of natural logarithms.
x = Number of occurrences of a given event.

Poisson process
The distribution applies in case of Poisson process which has the following
characteristics:
• Concerning a given random variable, the mean relating to a given interval
can be estimated on the basis of past data concerning the variable under
study.
• If we divide the given interval into very very small intervals we will find:
(a) The probability that exactly one event will happen during the very very
small interval is a very small number and is constant for every other
very small interval.
(b) The probability that two or more events will happen within a very small
interval is so small that we can assign it a zero value.
(c) The event that happens in a given very small interval is independent,
when the very small interval falls during a given interval.
(d) The number of events in any small interval is not dependent on the
number of events in any other small interval.

2. For Binomial distribution we had stated that mean = n. p.

Mean for Poisson distribution (or ) = n. p.
p = /n

Hence, mean = .n
n

48 Statistics for Management

Probability Distributions Unit-2

Parameter and important measures of Poisson distribution

Poisson distribution depends upon the value of , the average number of
occurrences per specified interval which is its only parameter. The probability of
exactly x occurrences can be found out using Poisson probability function stated
above.3 The expected value or the mean of Poisson random variable is and its
variance is also .4 The standard deviation of Poisson distribution is, λ .
Underlying the Poisson model is the assumption that if there are on the
average occurrences per interval t, then there are on the average k occurrences
per interval kt. For example, if the number of arrivals at a service counted in a
given hour, has a Poisson distribution with = 4, then y, the number of arrivals
at a service counter in a given 6 hour day, has the Poisson distribution = 24, i.e.,
6×4.

When to use Poisson distribution

The use of Poisson distribution is resorted to those cases when we do not know
the value of ‘n’ or when ‘n’ cannot be estimated with any degree of accuracy. In
fact, in certain cases it does not make any sense in asking the value of ‘n’. For
example, the goals scored by one team in a football match are given, it cannot
be stated how many goals could not be scored. Similarly, if one watches carefully
one may find out how many times the lightning flashed but it is not possible to
state how many times it did not flash. It is in such cases we use Poisson distribution.
The number of deaths per day in a district in one year due to a disease, the
number of scooters passing through a road per minute during a certain part of the
day for a few months, the number of printing mistakes per page in a book
containing many pages, are few other examples where Poisson probability
distribution is generally used.
Example 2.18: Suppose that a manufactured product has 2 defects per unit of
product inspected. Use Poisson distribution and calculate the probabilities of
finding a product without any defect, with 3 defects and with four defects.
Solution: If the product has 2 defects per unit of product inspected. Hence, = 2.
Poisson probability function is as follows:
x
.e
f ( Xi = x) =
x!
x = 0, 1, 2,…
Using the above probability function, we find the required probabilities as under:
20.e 2
P(without any defects, i.e., x = 0) =
0

x
3. There are tables which gives the e– values. These tables also give the e– values for
x
x = 0, 1, 2, . . . for a given and thus facilitate the calculation.
4. Variance of the Binomial distribution is n. p. q. and the variance of Poisson distribution is .
Therefore, = n. p. q. Since q is almost equal to unity and as pointed out earlier n. p.= in Poisson
distribution. Hence, variance of Poisson distribution is also .

Statistics for Management 49

Unit-2 Probability Distributions

1. 0.13534
= 0.13534
1
2
23.e 2 2 2 0.13534
P(with 3 defects, i.e., x = 3) = =
3 3 2 1
0.54136
= 0.18045
3
2
24.e 2 2 2 2 0.13534
P(with 4 defects, i.e., x = 4) = =
4 4 3 2 1
0.27068
= 0.09023
3
Example 2.19: How would you use a Poisson distribution to find approximately
the probability of exactly 5 successes in 100 trials the probability of success in
each trial being p = 0.1?
Solution: In the question we have been given,
n = 100 and p = 0.1
= n.p = 100 × 0.1 = 10
To find the required probability, we can use Poisson probability function as an
approximation to Binomial probability function as shown below:

( n. p) .e (n. p)
x x
.e
f ( Xi = x ) = =
x! x!

105.e 10
100000 0.00005 5.00000
or P(5)7 =
5 5 4 3 2 1 5 4 3 2 1
1
= = 0.042
24
2.6.4 Applications of Poisson Distribution
Poisson distribution is another theoretical discrete distribution, which is useful for
modelling certain real situations. It differs from the binomial distribution in the
sense that in the binomial distribution we must be able to count the number of
successes and the number of failures, while in Poisson distribution, all we want to
know is the average number of successes in a given unit of time or space. In many
situations, it is not possible to count the number of failures even though we can
know the number of successes. For example, in the case of patients coming to the
hospital for emergency treatment, we can always count the number of patients
arriving in any given hour. If the number of patients arriving is considered as the
number of successes, then we cannot know the number of failures because it is not
possible to count the number of patients not coming for emergency treatment in
that hour. Accordingly, it is not possible to determine the total number of possible
outcomes (successes and failures), and hence, binomial distribution cannot be applied
as a decision-making tool. In such a situation, we can use Poisson distribution, if
we know the average number of patients arriving for emergency treatment per hour.

50 Statistics for Management

Probability Distributions Unit-2

It is assumed that such arrival of patients is a random phenomenon and hence, the
exact number of patients arriving in any hour is not predictable. Other examples of
Poisson distribution are telephone calls going through a switchboard system, the
number of cars passing through India Gate, the number of customers coming to a
bank for service, and so on. All these arrivals can be described by a discrete random
variable that takes on integer values (0, 1, 2, 3, ...).
Characteristics of the Poisson distribution
A physical situation must possess certain characteristics before it can be characterized
by poisson distribution. Some of these characteristics are:
(a) In a very small time interval between t and (t ), (where is infinitesimally
small), the probability that exactly one event will occur is a very small number,
(the event is a rare event) and is constant for every such small time period.
(b) The probability that two or more events will occur in this small time period
(t to t ) is so small that it can be assigned a value of zero.
(c) The events must be random and independent of each other. The occurrence
of one event cannot influence the chances of another event occurring, nor can
the occurrence of any one event be predicted in advance.
With the characteristics as described here, we need to know the average of
events per unit of time. The symbol for this average is (lambda) and it could be
the average number of cars passing under a bridge in any given hour or it could be
the average number of machine breakdowns per month, or it could be the average
number of customers arriving at a bank per day, and so on.
The probability that exactly (x) events will occur in a given time is given as
follows:
x
e
P(x) =
x!
where is the average number of occurrences per unit of time and e is the
base of the natural logarithms and is equal to 2.71828...... .
Example 2.20: Assume that on an average 3 persons enter a bank for service every
10 minutes. What is the probability that exactly 5 customers will enter the bank in
a given 10-minute period, assuming that the process can be described by Poisson
distribution.
Solution:
x –
e
P(x) =
x!

(3)5 (2.71828)–3
=
5!

(243)(.0498)
= = 0.1008
120

Statistics for Management 51

Unit-2 Probability Distributions
Thus, the probability of 5 customers arriving in a bank in any given 10-minute
period is 0.1008. Other computations can be made the same way to show the
probability of arrival of 0, 1, 2, 3, 4 ... customers.
Poisson distribution is found to be particularly useful in waiting line or queuing
type of problems where, by knowing the rate of arrival of units at a service station
and the rate at which these units are served, the number of service stations as well
as the average waiting time for service for each arriving unit can be reasonably
determined.
Example 2.21: Customers arrive at a photocopying machine at an average rate of
two every 10 minutes. The number of arrivals is distributed according to Poisson
distribution. What is the probability that:
(a) There will be no arrivals during any period of ten minutes.
(b) There will be exactly one arrival during this time period.
(c) There will be more than two arrivals during this time period.
x
e
Solution: We know that with = 2, P(x) = for x = 0,1, 2....
x!
(2)0 (2.71828) 2

(a) When x = 0, P(0) = = 0.1353

0!
(2) (2.71828) 2
1

(b) When x = 1, P(1) = = 0.2707

1!
2 2
(2) (2.71828)
(c) When x = 2, P(2) = = 0.2707
2!
Then, the probability of more than two arrivals in a 10-minute period is:
P(x > 2) = 1 – [P(0) + P(1) + P(2)]
= 1 – [0.1353 + 0.2707 + 0.2707]
= 1 – 0.6767 = 0.3233

2.7 CONTINUOUS PROBABILITY DISTRIBUTIONS

Continuous variate: A
random variate can take
any value in the given When a random variate can take any value in the given interval a x b, it is a
interval a x b. continuous variate and its distribution is a Continuous Probability Distribution.

52 Statistics for Management

Probability Distributions Unit-2
Theoretical distributions are often continuous. They are useful in practice
because they are convenient to handle mathematically. They can serve as good
approximations to discrete distributions.
The range of the variate may be finite or infinite.
A continuous random variable can take all values in a given interval.
A continuous probability distribution is represented by a smooth curve.
The total area under the curve for a probability distribution is necessarily
unity. The curve is always above the x axis because the area under the curve for any
interval represents probability and probabilities cannot be negative.
If X is a continous variable, the probability of X falling in an interval with end
points z , z may be written p(z X z ).
1 2 1 2
This probability corresponds to the shaded area under the curve.

z1 z2

A function is a probability density function if,

–
p( x)dx = 1, p( x)>0, – < x < , i.e., the area under the curve p(x) is 1
and the probability of x lying between two values a, b, i.e., p(a < x < b) is positive.
The most prominent example of a continuous probability function is the normal
distribution.
Cumulative Probability Function (CPF)
The cumulative probability function (CPF) shows the probability that x takes a
value less than or equal to, say, z and corresponds to the area under the curve up
to z:
z
p(x z) = –
p(x)dx

This is denoted by F(x).

2.8 NORMAL DISTRIBUTION

Among all the probability distributions, the normal probability distribution is by far
the most important and frequently used continuous probability distribution. This is
so because this distribution fits well in many types of problems. This distribution is
of special significance in inferential statistics since it describes probabilistically the
link between a statistic and a parameter (i.e., between the sample results and the
population from which the sample is drawn). The name Karl Gauss, 18th century
mathematician–astronomer, is associated with this distribution and in honour of his
contribution, this distribution is often known as the Gaussian distribution.

Statistics for Management 53

Unit-2 Probability Distributions
Normal distribution can be theoretically derived as the limiting form of many discrete
distributions. For instance, if in the binomial expansion of (p + q)n, the value of
1
‘n’is infinity and p = q = , then a perfectly smooth symmetrical curve would be
2
obtained. Even if the values of p and q are not equal but if the value of the exponent
‘n’ happens to be very very large, we get a curve of normal probability smooth and
symmetrical. Such curves are called normal probability curves (or at times known
as normal curves of error) and such curves represent the normal distributions.5
The probability function in case of normal probability distribution6 is given as:
1 x 2
1 2
f ( x) = e
. 2
Where, = The mean of the distribution
2
= Variance of the distribution
2
The normal distribution is thus defined by two parameters viz., and . This
distribution can be represented graphically as in Figure 2.1.

Fig. 2.1 Curve Representing Normal Distribution

2.8.1 Characteristics of Normal Distribution

The characteristics of the normal distribution or that of a normal curve are, as given
below:
1. It is a symmetric distribution.7

5. Quite often, mathematicians use the normal approximation of the binomial distribution whenever
‘n’ is equal to or greater than 30 and np and nq each are greater than 5.
6. Equation of the normal curve in its simplest form is,
x2
2
2
y y0 .e
Where, y = The computed height of an ordinate at a distance of X from the mean.
y0 = The height of the maximum ordinate at the mean. It is a constant in the equation
and is worked out as under:
Ni
y0
2
7. A symmetric distribution is one which has no skewness. As such it has the following statistical
properties:
(a) Mean=Mode=Median (i.e., X=Z=M)
(b) (Upper Quantile – Median)=(Median – Lower Quantile) (i.e., Q3–M = M–Q1)
(c) Mean Deviation=0.7979(Standard Deviation)
Q Q
3 1 0.6745 (Standard Deviation)
(d)
2

54 Statistics for Management

Probability Distributions Unit-2
2. The mean defines where the peak of the curve occurs. In other words, the
ordinate at the mean is the highest ordinate. The height of the ordinate at a distance
of one standard deviation from the mean is 60.653% of the height of the mean
ordinate and similarly the height of other ordinates at various standard deviations
( s ) from mean happens to be a fixed relationship with the height of the mean
ordinate.
3. The curve is asymptotic to the base line which means that it continues to approach
but never touches the horizontal axis.
4. The variance ( 2) defines the spread of the curve.
5. Area enclosed between mean ordinate and an ordinate at a distance of one standard
deviation from the mean is always 34.134% of the total area of the curve. It means
that the area enclosed between two ordinates at one sigma (SD) distance from
the mean on either side would always be 68.268% of the total area. This can be
shown as follows:

(34.134% + 34.134%) = 68.268%

Area of the total
curve between ± 1( )

X or X
–3 –2 – + +2 +3

Similarly, the other area relationships are as follows:

Between Area Covered to Total Area of the
Normal Curve 8
µ±1 S.D. 68.27%
µ±2 S.D. 95.45%
µ±3 S.D. 99.73%
µ ± 1.96 S.D. 95%
µ ± 2.578 S.D. 99%
µ ± 0.6745 S.D. 50%
6. The normal distribution has only one mode since the curve has a single peak. In
other words, it is always a unimodal distribution.
7. The maximum ordinate divides the graph of normal curve into two equal parts.
8. In addition to all the above stated characteristics the curve has the following
properties:
_
(a) µ = x
(b) µ2= 2 = Variance
4
(c) µ =3
4
(d) Moment Coefficient of Kurtosis = 3

8. This also means that in a normal distribution, the probability of area lying between various
limits are as follows:
Limits Probability of area lying within the stated limits
µ ± 1 S.D. 0.6827
µ ± 2 S.D. 0.9545
µ ± 3 S.D. 0.9973 (This means that almost all cases
lie within µ ± 3 S.D. limits)

Statistics for Management 55

Unit-2 Probability Distributions

2.8.2 Family of Normal Distributions

We can have several normal probability distributions but each particular normal
distribution is being defined by its two parameters viz., the mean (µ) and the standard
deviation ( ). There is, thus, not a single normal curve but rather a family of normal
curves. We can exhibit some of these as under:
Normal curves with identical means but different standard deviations:
Curve having small standard
deviation say ( = 1)
Curve having large standard
deviation say ( = 5)
Curve having very large standard
deviation say ( = 10)
in a normal
distribution

Normal curves with identical standard deviation but each with different means:

Normal curves each with different standard deviations and different means:

2.8.3 How to Measure the Area under the Normal Curve

We have stated above some of the area relationships involving certain intervals of
standard deviations (plus and minus) from the means that are true in case of a
normal curve. But what should be done in all other cases? We can make use of the
statistical tables constructed by mathematicians for the purpose. Using these tables
we can find the area (or probability, taking the entire area of the curve as equal to 1)
that the normally distributed random variable will lie within certain distances from
the mean. These distances are defined in terms of standard deviations. While using
the tables showing the area under the normal curve we talk in terms of standard
variate (symbolically Z ) which really means standard deviations without units of
measurement and this ‘Z’ is worked out as under:
X
Z

56 Statistics for Management

Probability Distributions Unit-2

Where, Z = The standard variate (or number of standard deviations from X to the
mean of the distribution).
X = Value of the random variable under consideration.
µ = Mean of the distribution of the random variable.
= Standard deviation of the distribution.
The table showing the area under the normal curve (often termed as the standard
normal probability distribution table) is organized in terms of standard variate (or
Z) values. It gives the values for only half the area under the normal curve, beginning
with Z = 0 at the mean. Since the normal distribution is perfectly symmetrical the
values true for one half of the curve are also true for the other half. We now illustrate
the use of such a table for working out certain problems.
Example 2.22: A banker claims that the life of a regular savings account opened
with his bank averages 18 months with a standard deviation of 6.45 months. Answer
the following: (a) What is the probability that there will still be money in 22 months
in a savings account opened with the said bank by a depositor? (b) What is the
probability that the account will have been closed before two years?
Solution: (a) For finding the required probability we are interested in the area of
the portion of the normal curve as shaded and shown below:

σ = 6.45

μ = 18
z=0 X = 22

Let us calculate Z as under:

X 22 18
Z= 0.62
6.45
The value from the table showing the area under the normal curve for Z = 0.62 is
0.2324. This means that the area of the curve between µ = 18 and X = 22 is 0.2324.
Hence, the area of the shaded portion of the curve is (0.5) – (0.2324) = 0.2676 since
the area of the entire right hand portion of the curve always happens to be 0.5. Thus
the probability that there will still be money in 22 months in a savings account is
0.2676.
(b) For finding the required probability we are interested in the area of the portion
of the normal curve as shaded and shown in figure:

σ = 6.45

μ = 18
X = 24
z=0

Statistics for Management 57

Unit-2 Probability Distributions

For the purpose we calculate,

24 18
Z 0.93
6.45

The value from the concerning table, when Z = 0.93, is 0.3238 which refers to the
area of the curve between µ = 18 and X = 24. The area of the entire left hand portion
of the curve is 0.5 as usual.
Hence, the area of the shaded portion is (0.5) + (0.3238) = 0.8238 which is the
required probability that the account will have been closed before two years, i.e.,
before 24 months.
Example 2.23: Regarding a certain normal distribution concerning the income of
the individuals we are given that mean=500 rupees and standard deviation =100
rupees. Find the probability that an individual selected at random will belong to
income group,
(a) 550 to 650 (b) 420 to 570
Solution: (a) For finding the required probability we are interested in the area of
the portion of the normal curve as shaded and shown below:

= 100

= 500 X = 650
z = 0 X = 550
For finding the area of the curve between X = 550 to 650, let us do the following
calculations:
550 500 50 = 0.50
Z= =
100 100
Corresponding to which the area between µ = 500 and X = 550 in the curve as per
table is equal to 0.1915 and,
650 500 150
Z 1.5
100 100
Corresponding to which, the area between µ = 500 and X = 650 in the curve, as per
table, is equal to 0.4332.
Hence, the area of the curve that lies between X = 550 and X = 650 is,
(0.4332) – (0.1915) = 0.2417
This is the required probability that an individual selected at random will belong to
income group of 550 to 650.
(b) For finding the required probability we are interested in the area of the portion
of the normal curve as shaded and shown below:
To find the area of the shaded portion we make the following calculations:

58 Statistics for Management

Probability Distributions Unit-2

= 100

= 100
z=0
X = 420 X = 570
570 500
Z = = 0.70
100
Corresponding to which the area between µ = 500 and X = 570 in the curve as per
table is equal to 0.2580.
420 500
and Z= = 0.80
100
Corresponding to which the area between µ = 500 and X = 420 in the curve as per
table is equal to 0.2881.
Hence, the required area in the curve between X = 420 and X = 570 is,
(0.2580) + (0.2881) = 0.5461
This is the required probability that an individual selected at random will belong to
income group of 420 to 570.
1''
Example 2.24: A certain company manufactures 1 all-purpose rope using imported
2
hemp. The manager of the company knows that the average load-bearing capacity
of the rope is 200 lbs. Assuming that normal distribution applies, find the standard
1''
deviation of load-bearing capacity for the 1 rope if it is given that the rope has a
2
0.1210 probability of breaking with 68 lbs. or less pull.
Solution: Given information can be depicted in a normal curve as shown below:
Probability of this
area (0.5) – (0.1210) = 0.3790

σ = ? (to be found out)

Probability of this area

(68 lbs. or less)
as given is 0.1210

μ = 200
X = 68 z=0

Statistics for Management 59

Unit-2 Probability Distributions

If the probability of the area falling within µ = 200 and X = 68 is 0.3790 as stated
above, the corresponding value of Z as per the table9 showing the area of the normal
curve is – 1.17 (minus sign indicates that we are in the left portion of the curve).
Now to find , we can write,
X
Z

68 200
or 1.17 =

or 1.17 = 132
or = 112.8 lbs. approx.
Thus, the required standard deviation is 112.8 lbs. approximately.
Example 2.25: In a normal _distribution, 31 per cent items are below 45 and 8 per
cent are above 64. Find the X and of this distribution.
Solution: We can depict the given information in a normal curve as shown below:

X X

X
X

If the probability of the area falling within µ and X = 45 is 0.19 as stated above, the
corresponding value of Z from the table showing the area of the normal curve is –
0.50. Since, we are in the left portion of the curve, we can express this as under,
45
0.50 = (1)
Similarly, if the probability of the area falling within µ and X = 64 is 0.42, as stated
above, the corresponding value of Z from the area table is, +1.41. Since, we are in
the right portion of the curve we can express this as under,
64
1.41 (2)

If we solve equations (1) and (2) above to obtain the value of µ or X , we have,
– 0.5 = 45 – µ (3)
1.41 = 64 – µ (4)
By subtracting equation (4) from equation (3) we have,
– 1.91 = –19
= 10

9. The table is to be read in the reverse order for finding Z value (See Appendix).

60 Statistics for Management

Probability Distributions Unit-2

Putting = 10 in equation (3) we have,

– 5 = 45 –
= 50
_
Hence, X (or µ)=50 and =10 for the concerning normal distribution.

2.9 LET US SUM UP

• Statistical data is used to draw certain conclusions with the effective use of
probability. The various terms and concepts can be applied in the decision-
making process and also in inferring the occurrence of several events in the
business environment of an organization.
• The outcomes of most decisions cannot be accurately predicted because of
the impact of many uncontrollable and unpredictable variables, so it is
necessary to scientifically evaluate the known risks.

Statistics for Management 61

 Unit-2 Probability Distributions

• The expected value is the weighted average of the possible values, a variable
can take. The variance of a random variable explains the spread of the possible
values of the variable.
• Probability distribution refers to the listing of all possible outcomes of an
experiment together with their probabilities. It may be discrete or continuous
depending upon the nature of the variable being considered.
• Binomial distribution is probably the best known of discrete distributions.
• The normal distribution, or Z-distribution, is often used to approximate the
binomial distribution. However, if the sample size is very large, Poisson
distribution is a philosophically more correct alternative to binomial
distribution than normal distribution.
• One of the main differences between Poisson distribution and the binomial
distribution is that in using binomial distribution, all eligible phenomena are
studied, whereas in Poisson only the cases with a particular outcome are
studied.
• Amongst all, normal probability distribution is by far the most important and
frequently used distribution because it fits well with many types of problems.

2.10 FURTHER READING

Chance, William A. Statistical Methods for Decision Making. Illinois: Richard D
Irwin, 1969.
Chandan, J.S. Statistics for Business and Economics. New Delhi: Vikas Publishing
House Pvt. Ltd., 1998.
Chandan, J.S., Jagjit Singh and K.K. Khanna. Business Statistics. New Delhi: Vikas
Publishing House Pvt. Ltd., 1995.
Meyer, Paul L. Introductory Probability and Statistical Applications. Massachusetts:
Addison-Wesley Publishing Co., 1970.
Neter, John, and William Wasserman. Fundamental Statistics for Business and
Economics. New York: Allyn and Bacon, 1956.
Yamane, Taro. Statistics—An Introductory Analysis. New York: Harper & Row,
1964.
M.Selvam 2014 Maritime Logistics and Documentation , Lakshmi Publications
, New Delhi
M.Selvam 2013 Quality Management , Lakshmi Publications , New Delhi
Besterfield .H (2018) Total Quality Management

2.11 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress–1

1. A random variable is a variable that takes on different values as a result of the
outcomes of a random experiment.

62 Statistics for Management

Probability Distributions Unit-2

2. A random variable can be either discrete or continuous. If a random variable

is allowed to take on only a limited number of values, it is a discrete random
variable but if it is allowed to assume any value within a given range, it is a
continuous random variable.
3. A-priori probability assignment is the technique under which the probability
is assigned by calculating the ratio of the number of ways in which a given
outcome can occur to the total number of possible outcomes. The basic
underlying assumption in using this procedure is that every possible outcome
is likely to occur equally.
4. The variance of random variable is defined as the sum of the squared deviations
of the values of random variable from the expected value weighted by their
probability.
Check Your Progress–2
1. The expected value (or mean) of X is the weighted average of the possible
values that X can take. Here, X is a discrete random variable and each value is
being weighted according to the probability of the possibility of the occurrence
of the event. The expected value of X is usually written as E(X) or m.
2. The variance of a random variable tells us something about the spread of the
possible values of the variable. For a discrete random variable X, the variance
of X is written as Var(X).
Var(X) = E[(X – m)2]
Where, m is the expected value E(X).
3. The function that shows the location of each sample is called a ‘density
function’. By showing the location of each sample, density of the samples
can be understood.
Check Your Progress–3
1. The parameters of binomial distribution are p and n, where p specifies the
probability of success in a single trial and n specifies the number of trials.
2. We need to use binomial distribution under the following circumstances:
(a) When we have to find the probability of heads in 10 throws of a fair
coin.
(b) When we have to find the probability that 3 out of 10 items produced by
a machine, which produces 8% defective items on an average, will be
defective.
3. Poisson distribution is a discrete probability distribution that is frequently
used in the context of Operations Research. Unlike binomial distribution,
Poisson distribution cannot be deduced on purely theoretical grounds based
on the conditions of the experiment. In fact, it must be based on experience,
i.e., on the empirical results of past experiments relating to the problem under
study.

Statistics for Management 63

Unit-2 Probability Distributions

Check Your Progress–4

1. Normal distribution is the most important and frequently used continuous
probability distribution among all the probability distributions. This is so
because this distribution well fits in many types of problems. This distribution
is of special significance in inferential statistics since it describes
probabilistically the link between a statistic and a parameter.
2. The characteristics of a normal distribution are:
(a) It is a symmetric distribution.
(b) The curve is asymptotic to the base line, which means that it continues
to approach but never touches the horizontal axis.
(c) The variance ( 2) defines the spread of the curve.
(d) The normal distribution has only one mode since the curve has a single
peak. In other words, it is always a unimodal distribution.
3. When n is large approaching to infinity and p is small approaching to zero,
Poisson distribution is considered as an approximation of binomial
distribution.

2.12 POSSIBLE QUESTIONS

Short-Answer Questions

1. Explain the concept of expectation of a random variable.

2. Explain briefly the terms ‘expectation’ and ‘expected value’.
3. Define variance.
4. Why is random variable considered important in statistics?
5. Explain briefly the probability distribution and its types.
6. Describe binomial distribution and its measures.
7. Poisson distribution can be an approximation of binomial distribution. Explain.
8. When is the Poisson distribution used?
9. Write the formula for measuring the area under the curve.
10. Explain the circumstances when the normal probability distribution can be
used.
Long-Answer Questions

1. Explain expectation with the help of examples.

2. If it rains a person can earn 100 per day by hiring out his vehicle. If it is fair
weather he loses 10. The probability of raining is 0.4. Find his expectation.
3. There are 3 tickets numbered 0, 2, 3. One ticket is selected and replaced;
another ticket is selected and replaced. A third ticket is selected and replaced
once again. If X stands for the sum of the 3 ticket numbers, construct the
probability distribution of X and find m and 2 .

64 Statistics for Management

Probability Distributions Unit-2
4. A car insurance policy offers 1000 after an accident the probability of whose
occurrence is 0.04. If the expected gain is to be zero what should be the
premium?
5. In a queue the probability of the number of people joining per minute given
below.
Number of persons 0 1 2 3
Joining the queue
Probability 0.4 0.3 0.2 0.1
6. From an urn containing 20 black and 30 red balls, 6 balls are drawn at random.
Find the probability that no red ball is selected.
7. Examine the nature of the distribution when r balls are drawn, one at a time,
without replacement, from a bag containing m white and n black balls.
8. 20 per cent of bolts in a factory are defective. Deduce the probability
distribution of the number of defectives in a sample of 5.
9. (a) Explain the meaning of Bernoulli process pointing out its main
characteristics.
(b) Give a few examples narrating some situations wherein binomial
pr. distribution can be used.
10. State the distinctive features of the Binomial, Poisson and Normal probability
distributions. When does a Binomial distribution tend to become a Normal
and a Poisson distribution? Explain.
11. Explain the circumstances when the following probability distributions are
used:
(a) Binomial distribution
(b) Poisson distribution
(c) Normal distribution
12. Certain articles were produced of which 0.5 per cent are defective, are packed
in cartons, each containing 130 articles. When proportion of cartons are free
from defective articles? What proportion of cartons contain 2 or more
defectives?
(Given e–0.5=0.6065).
13. The following mistakes per page were observed in a book:
No. of Mistakes No. of Times the Mistake
Per Page Occurred
0 211
1 90
2 19
3 5
4 0
Total 325
Fit a Poisson distribution to the data given above and test the goodness of
fit.
Statistics for Management 65
Unit-2 Probability Distributions

14. In a distribution exactly normal, 7 per cent of the items are under 35 and 89
per cent are under 63. What are the mean and standard deviation of the
distribution?
15. Fit a normal distribution to the following data:
Height in inches Frequency
60–62 5
63–65 18
66–68 42
69–71 27
72–74 8

2.13 LEARNING OUTCOMES

• The expectation property of a random variable

• The functionality of a random variable
• The probability density functions
• The probability distribution functions

66 Statistics for Management

Sampling and Hypothesis Testing Unit-3

UNIT 3 SAMPLING AND

HYPOTHESIS TESTING
UNIT STRUCTURE
3.1 Learning Objectives
3.2 Introduction
3.3 Sampling Procedures—Random and Non-Random Methods
3.3.1 Sample Selection; 3.3.2 Census and Sampling
3.3.3 Sampling Techniques; 3.3.4 Sampling and Non-Sampling Errors
3.4 Sample Size Determination, Sampling Distribution and Standard Error
3.4.1 Sampling Distribution of the Mean; 3.4.2 Standard Error
3.5 Hypothesis Testing
3.5.1 Statistical Decision-Making; 3.5.2 Type I and Type II Errors
3.5.3 Null and Alternative Hypotheses; 3.5.4 Critical Region
3.5.5 Penalty; 3.5.6 Standard Error; 3.5.7 Testing of Hypothesis
3.5.8 Tests for a Sample Mean X
3.5.9 Test for Equality of Two Proportions
3.5.10 Large Sample Test for Equality of Two Means X1, X2
3.5.11 Small Sample Tests of Significance
3.5.12 Paired Observations
3.5.13 Test for a Given Population Variance
3.6 T-Distribution Testing
3.7 Chi-Square Test for Independence of Attributes Activity
3.7.1 Degrees of Freedom
3.7.2 Conditions for the Application of Test
3.7.3 Areas of Application of Chi-Square Test
3.7.4 Steps in Finding the Value of Chi-Square
3.7.5 Alternative Formula for Finding the Value of Chi-Square in a (2 × 2) Table
3.7.6 Yates’ Correction
3.7.7 Chi-Square as a Test of Population Variance
3.7.8 Additive Property of Chi-Square ( 2)
3.7.9 Important Characteristics of Chi-Square ( 2) Test
3.8 Let Us Sum Up
3.9 Further Reading
3.10 Answers to Check Your Progress
3.11 Possible Questions
3.12 Learning Outcomes

3.1 LEARNING OBJECTIVES

After going through this unit, you will be able to:

• Explain random and non-random sampling methods
• Explain the methods of sample size determination
• Explain hypothesis testing
• Explain paired t-test
• List the areas of application of chi-square test
Statistics for Management 67
Unit-3 Sampling and Hypothesis Testing

3.2 INTRODUCTION
Sampling is the process of selecting a sample from the population. This unit
introduces you to sampling procedures, where you will learn about random and
non-random sampling methods. You will also learn the different methods of sample
size determination.
Hypothesis is an assumption that is tested to find its logical or empirical
consequences. A hypothesis should be clear and accurate. Various concepts such as
null and alternate hypotheses help to verify the testability of an assumption.
In this unit, you will also learn about the chi-square test and t-test. Any
statistical hypothesis test in which the test statistic has a chi-square distribution
when the null hypothesis is true is termed as a chi-square test.

3.3 SAMPLING PROCEDURES—RANDOM AND NON-

RANDOM METHODS
A sample is a portion of the total population that is considered for study and analysis.
For instance, if we want to study the income pattern of professors at City University
of New York and there are 10,000 professors, then we may take a random sample of
Sample: A portion of
the total population that
only 1,000 professors out of this entire population. Then this number of 1,000
is considered for study professors constitutes a sample. The summary measure that describes a characteristic
and analysis. such as average income of this sample is known as a statistic.
Sampling is the process of selecting a sample from the population. It is
technically and economically not feasible to take the entire population for analysis.
So we must take a representative sample out of this population for the purpose of
such analysis. A sample is part of the whole, selected in such a manner as to be
representing the whole.
Random sample
Random sample is a collection of items selected from the population in such a
manner that each item in the population has exactly the same chance of being selected,
so that the sample taken from the population would be truly representative of the
population. The degree of randomness of selection would depend upon the process
of selecting the items from the sample. A true random sample would be free from
all biases whatsoever. For example, if we want to take a random sample of five
students from a class of twenty-five students, then each one of these twenty-five
students should have the same chance of being selected in the sample. One way to
do this would be writing the names of all students on separate but small pieces of
paper, folding each piece of this paper in a similar manner, putting each folded
piece into a container, mixing them thoroughly and drawing out five pieces of paper
from this container.
Sampling without replacement
The sample as taken in the previous example is known as sampling without
replacement, because each person can only be selected once so that once a piece of
paper is taken out of the container, it is kept aside so that the person whose name
appears on this piece of paper has no chance of being selected again.

68 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Sampling with replacement

There are certain situations in which the piece of paper once selected and taken into
consideration is put back into the container in such a manner that the same person
has the same chance of being selected again as any other person. For example, if we
are randomly selecting five persons for award of prizes so that each person is eligible
for any and all prizes, then once the slip of paper is drawn out of the container and
the prize is awarded to the person whose name appears on the paper, the same piece
of paper is put back into the container and the same person has the same chance of
winning the second prize as anybody else.
3.3.1 Sample Selection
The third step in the primary data collection process is selecting an adequate sample.
It is necessary to take a representative sample from the population, since it is
extremely costly, time-consuming and cumbersome to do a complete census. Then,
depending upon the conclusions drawn from the study of the characteristics of such
a sample, we can draw inferences about the similar characteristics of the population.
If the sample is truly representative of the population, then the characteristics of the
sample can be considered to be the same as those of the entire population. For
example, the taste of soup in the entire pot of soup can be determined by tasting one
spoonful from the pot if the soup is well stirred. Similarly, a small amount of blood
sample taken from a patient can determine whether the patient’s sugar level is normal
or not. This is so because the small sample of blood is truly representative of the
entire blood supply in the body.
Sampling is necessary because of the following reasons: First, as discussed
earlier, it is not technically or economically feasible to take the entire population
into consideration. Second, due to dynamic changes in business, industrial and social
environment, it is necessary to make quick decisions based upon the analysis of
information. Managers seldom have the time to collect and process data for the
entire population. Thus, a sample is necessary to save time. The time element has
further importance in that if the data collection takes a long time, then the values of
some characteristics may change over the period of time so that data may no longer
be up to date, thus defeating the very purpose of data analysis. Third, samples, if
representative, may yield more accurate results than the total census. This is due to
the fact that samples can be more accurately supervised and data can be more
carefully selected. Additionally, because of the smaller size of the samples, the
routine errors that are introduced in the sampling process can be kept at a minimum.
Fourth, the quality of some products must be tested by destroying the products. For
example, in testing cars for their ability to withstand accidents at various speeds,
the environment of accidents must be simulated. Thus, a sample of cars must be
selected and subjected to accidents by remote control. Naturally, the entire population
of cars cannot be subjected to these accident tests and hence, a sample must be
selected.
One important aspect to be considered is the size of the sample. The sampling
size—which is the number of sampling units selected from the population for
investigation—must be optimum. If the sample size is too small, it may not
appropriately represent the population or the universe as it is known, thus leading

Statistics for Management 69

Unit-3 Sampling and Hypothesis Testing

to incorrect inferences. Too large a sample would be costly in terms of time and
money. The optimum sample size should fulfil the requirements of efficiency,
representativeness, reliability and flexibility. What is an optimum sample size is
also open to question. Some experts have suggested that 5 per cent of the population
properly selected would constitute an adequate sample, while others have suggested
as high as 10 per cent depending upon the size of the population under study.
However, proper selection and representation of the sample is more important than
size itself. The following considerations may be taken into account in deciding
about the sample size:
(a) The larger the size of the population, the larger should be the sample size.
(b) If the resources available do not put a heavy constraint on the sample size,
a larger sample would be desirable.
(c) If the samples are selected by scientific methods, a larger sample size would
ensure greater degree of accuracy in conclusions.
(d) A smaller sample could adequately represent the population, if the population
consists of mostly homogeneous units. A heterogeneous universe would
require a larger sample.
3.3.2 Census and Sampling
Under the census or complete enumeration survey method, data is collected for
each and every unit (e.g., person, consumer, employee, household, organization) of
the population or universe which are the complete set of entities and which are of
interest in any particular situation. In spite of the benefits of such an all-inclusive
approach, it is infeasible in most of the situations. Besides, the time and resource
constraints of the researcher, infinite or huge population, the incidental destruction
of the population unit during the evaluation process (as in the case of bullets,
explosives, etc.) and cases of data obsolescence (by the time census ends) do not
permit this mode of data collection.
Sampling is simply a process of learning about the population on the basis of
a sample drawn from it. Thus, in any sampling technique, instead of every unit of
the universe, only a part of the universe is studied and conclusions are drawn on
that basis for the entire population. The process of sampling involves selection of a
sample based on a set of rules, collection of information and making an inference
about the population. It should be clear to the researcher that a sample is studied not
for its own sake, but the basic objective of its study is to draw inference about the
population. In other words, sampling is a tool which helps us know the characteristics
of the universe or the population by examining only a small part of it. The values
obtained from the study of a sample, such as the average and dispersion are known
as ‘statistics’ and the corresponding such values for the population are called
‘parameters’.
Although diversity is a universal quality of mass data, every population has
characteristic properties with limited variation. The following two laws of statistics
are very important in this regard.
1. The law of statistical regularity states that a moderately large number of items
chosen at random from a large group are almost sure on the average to possess

70 Statistics for Management

Sampling and Hypothesis Testing Unit-3

the characteristics of the large group. By random selection, we mean a

selection where each and every item of the population has an equal chance of
being selected.
2. The law of inertia of large numbers states that, other things being equal,
larger the size of the sample, more accurate the results are likely to be.
Hence, a sound sampling procedure should result in a representative, adequate and
homogeneous sample while ensuring that the selection of items should occur
independently of one another.

3.3.3 Sampling Techniques

The various methods of sampling can be grouped under two broad categories:
Probability (or random) sampling and Non-probability (or non-random) sampling.
Probability sampling methods are those in which every item in the universe
has a known chance, or probability of being chosen for the sample. Thus, the sample
selection process is objective (independent of the person making the study) and
hence, random. It is worth noting that randomness is a property of the sampling
procedure instead of an individual sample. As such, randomness can enter processed
sampling in a number of ways and hence, random samples may be of many types.
These methods include: (a) Simple Random Sampling, (b) Stratified Random
Sampling, (c) Systematic Sampling and (d) Cluster Sampling.
Non-probability sampling methods are those which do not provide every
item in the universe with a known chance of being included in the sample. The
selection process is, at least, partially subjective (dependent on the person making
the study). The most important difference between random and non-random
sampling is that whereas the pattern of sampling variability can be ascertained in
case of random sampling, there is no way of knowing the pattern of variability in
the non-random sampling process. The non-probability methods include:
(a) Judgement Sampling, (b) Quota Sampling and (c) Convenience Sampling.
Figure 3.1 depicts the broad classification and sub-classification of various
methods of sampling.
Sampling Methods

Non-probability Probability
Samples Samples

Judgement Quota Convenience

Sampling Sampling Sampling

Simple Random Stratified Systematic Cluster

Sampling Sampling Sampling Sampling

Fig. 3.1 Methods of Sampling

Statistics for Management 71

Unit-3 Sampling and Hypothesis Testing

1. Non-probability sampling methods

(i) Judgement sampling. In this method of sampling, the choice of sample items
depends exclusively on the judgement of the investigator. The sample here is
based on the opinion of the researcher, whose discretion will clinch the sample.
Though the principles of sampling theory are not applicable to judgement
sampling, it is sometimes found to be useful. When we want to study some
unknown traits of a population, some of whose characteristics are known, we
may then stratify the population according to these known properties and
select sampling units from each stratum on the basis of judgement. Naturally,
the success of this method depends upon the excellence in judgement.
(ii) Convenience sampling. A convenience sample is obtained by selecting
convenient population units. It is also called a chunk, which refers to that
fraction of the population being investigated, which is selected neither by
probability nor by judgement but by convenience. A sample obtained from
readily available lists such as telephone directories is a convenience sample
and not a random sample, even if the sample is drawn at random from such
lists. In spite of the biased nature of such a procedure, convenience sampling
is often used for pilot studies.
(iii) Quota sampling. It is a type of judgement sampling and is perhaps the
most commonly used sampling technique in non-probability category. In a
quota sample, quotas (or minimum targets) are set up according to some
specified characteristics, such as age, income group, religious or political
affiliations, and so on. Within the quota, the selection of the sample items
depends on personal judgement. Because of the risk of personal prejudice
entering the sample selection process, quota sampling is not widely used in
practical works.
It is worth noting that similarity between quota sampling and stratified
random sampling is confined to dividing the population into different strata.
The process of selection of items from each of these strata in the case of
stratified random sampling is random, while it is not so in the case of quota
sampling. Quota sampling is often used in public opinion studies.
2. Probability sampling methods
The following are the probability sampling methods:
(i) Simple random sampling
(ii) Stratified random sampling
(iii) Systematic sampling
(iv) Multistage or cluster sampling
(i) Simple random sampling
Simple random sampling refers to that sampling technique in which each and every
unit of the population has an equal chance of being selected in the sample. One
should not mistake the term ‘arbitrary’ for ‘random’. To ensure randomness, one
may adopt either the lottery method or consult the table of random numbers,
preferably the latter. Being a random method, it is independent of personal bias
72 Statistics for Management
Sampling and Hypothesis Testing Unit-3

creeping into the analysis besides enhancing the representativeness of the sample.
Furthermore, it is easy to assess the accuracy of the sampling estimates because
sampling errors follow the principles of chance. However, a completely catalogued
universe is a prerequisite for this method. The sample size requirements would be
usually larger under random sampling than under stratified random sampling, to
ensure statistical reliability. It may escalate the cost of collecting data as the cases
selected by random sampling tend to be too widely dispersed geographically.
(ii) Stratified random sampling
In stratified random sampling method, the universe to be sampled is subdivided
(stratified) into groups which are mutually exclusive, but collectively exhaustive
based on a variable known to be correlated with the variable of interest. Then, a
simple random sample is chosen independently from each group. This method differs
from simple random sampling in that, in the latter the sample items are chosen at
random from the entire universe. In stratified random sampling, the sampling is
designed in such a way that a designated number of items is chosen from each
stratum. If the ratio of items between various strata in the population matches with
the ratio of corresponding items between various strata in the sample, it is called
proportionate stratified sampling; otherwise, it is known as disproportionate stratified
sampling. Ideally, we should assign greater representation to a stratum with a larger
dispersion and smaller representation to one with small variation. Hence, it results
in a more representative sample than simple random sampling.
(iii) Systematic sampling
Systematic sampling is also known as quasi-random sampling method because once
the initial starting point is determined, the remainder of the items selected for the
sample are predetermined by the sampling interval. A systematic sample is formed
by selecting one unit at random and then selecting additional units at evenly spaced
intervals until the sample has been formed. This method is popularly used in those
cases where a complete list of the population from which sample is to be drawn is
available. The list may be prepared in alphabetical, geographical, numerical or
some other order. The items are serially numbered. The first item is selected at
random generally by following the lottery method. Subsequent items are selected
by taking every Kth item from the list, where ‘K’ stands for the sampling interval or
the sampling ratio, i.e., the ratio of the population size to the size of the sample.
Symbolically,
K = N / n , where K = Sampling interval; N = Universe size; n = Sample size. In
case K is a fractional value, it is rounded off to the nearest integer.

(iv) Cluster sampling

Under cluster sampling method, the random selection is made of primary,
intermediate and final (or the ultimate) units from a given population or stratum.
There are several stages in which the sampling process is carried out. At first, the
stage units are sampled by some suitable method such as simple random sampling.
Then, a sample of second stage units is selected from each of the selected first stage
units, by applying some suitable method which may or may not be the same method

Statistics for Management 73

Unit-3 Sampling and Hypothesis Testing

employed for the first stage units. For example, in a survey of 10,000 households in
AP, we may choose a few districts in the first stage, a few towns/villages/mandals
in the second stage and select a number of households from each town/village/
mandal selected in the previous stage. This method is quite flexible and is particularly
useful in surveys of underdeveloped areas, where no frame is generally sufficiently
detailed and accurate for subdivision of the material into reasonably small sampling
units. However, a multistage sample is, in general, less accurate than a sample
containing the same number of final stage units which have been selected by some
suitable single stage process.
3.3.4 Sampling and Non-Sampling Errors
The basic objective of a sample is to draw inferences about the population from
which such sample is drawn. This means that sampling is a technique which helps
us in understanding the parameters or the characteristics of the universe or the
population by examining only a small part of it. Therefore, it is necessary that the
sampling technique be a reliable one. The randomness of the sample is especially
important because of the principle of statistical regularity, which states that a sample
taken at random from a population is likely to possess almost the same characteristics
as those of the population. However, in the total process of statistical analysis,
some errors are bound to be introduced. These errors may be the sampling errors or
the non-sampling errors. Sampling errors arise due to drawing faulty inferences
about the population based upon the results of the samples. In other words, it is the
difference between the results that are obtained by the sample study and the results
that would have been obtained if the entire population was taken for such a study,
provided that the same methodology and manner was applied in studying both the
sample as well as the population. For example, if a sample study indicates that 25
per cent of the adult population of a city does not smoke and the study of the entire
adult population of the city indicates that 30 per cent are non-smokers, then this
difference would be considered as the sampling error. This sampling error would be
smallest if the sample size is large relative to the population, and vice versa.
Non-sampling errors, on the other hand, are introduced due to technically faulty
observations or during the processing of data. These errors could also arise due to
defective methods of data collection and incomplete coverage of the population,
because some units of the population are not available for study, inaccurate
information provided by the participants in the sample and errors occurring during
editing, tabulating and mathematical manipulation of data. These are the errors
which can arise even when the entire population is taken under study.
Both the sampling as well as the non-sampling errors must be reduced to a minimum
in order to get as representative a sample of the population as possible.

3.4 SAMPLE SIZE DETERMINATION, SAMPLING

DISTRIBUTION AND STANDARD ERROR

One of the major objectives of statistical analysis is to know the ‘true’ values of
different parameters of the population. Since it is not possible due to time, cost and
other constraints to take the entire population for consideration, random samples
74 Statistics for Management
Sampling and Hypothesis Testing Unit-3

are taken from the population. These samples are analysed properly and they lead
to generalizations that are valid for the entire population. The process of relating
the sample results to the population is referred to as, ‘Statistical Inference’ or
‘Inferential Statistics’.
Statistical inference:
In general, a single sample is taken and its mean X is considered to represent The process of relating
the population mean. However, in order to use the sample mean to estimate the the sample results to the
population mean, we should examine every possible sample (and its mean, etc.) population.
that could have occurred, because a single sample may not be representative enough.
If it was possible to take all the possible samples of the same size, then the distribution
of the results of these samples would be referred to as, ‘sampling distribution’. The
distribution of the means of these samples would be referred to as, ‘sampling
distribution of the means’.
The relationship between the sample means and the population mean can
best be illustrated by Example 3.1.
Example 3.1: Suppose a babysitter has 5 children under her supervision with average
age of 6 years. However, individually, the age of each child be as follows:
X1 = 2
X2 = 4
X3 = 6
X4 = 8
X5 = 10
Now these 5 children would constitute our entire population, so that N = 5.
Solution:
X
The population mean μ
N

2 + 4 + 6 + 8 +10
= = 30 / 5 = 6
5
and the standard deviation is given by the formula:

( μ)2
σ
N

Now, let us calculate the standard deviation.

X (X– )2
2 6 16
4 6 4
6 6 0
8 6 4
10 6 16

Total = (X– )2 = 40

Statistics for Management 75

Unit-3 Sampling and Hypothesis Testing

Then,

40
σ 8 2.83
5
Now, let us assume the sample size, n = 2, and take all the possible samples of size
2, from this population. There are 10 such possible samples. These are as follows,
along with their means.
X1, X2 (2, 4) X1 = 3
X1, X3 (2, 6) X2 = 4
X1, X4 (2, 8) X3 = 5
X1, X5 (2,10) X4 = 6
X2, X3 (4, 6) X5 = 5
X2, X4 (4, 8) X6 = 6
X2, X5 (4, 10) X7 = 7
X3, X4 (6, 8) X8 = 7
X3, X5 (6, 10) X9 = 8
X4, X5 (8, 10) X10 = 9
Now, if only the first sample was taken, the average of the sample would be 3.
Similarly, the average of the last sample would be 9. Both these samples are totally
unrepresentative of the population. However, if a grand mean X of the distribution
of these sample means is taken, then,
10

Xi
X = i=1
10

3+ 4 +5+ 6+5+6 +7 +7 +8+9

= 60 / 10 = 6
10
This grand mean has the same value as the mean of the population. Let us
organize this distribution of sample means into a frequency distribution and
probability distribution.
Sample mean Freq. Rel.freq. Prob.
3 1 1/10 .1
4 1 1/10 .1
5 2 2/10 .2
6 2 2/10 .2
7 2 2/10 .2
8 1 1/10 .1
9 1 1/10 .1
1.00

76 Statistics for Management

Sampling and Hypothesis Testing Unit-3

This probability distribution of the sample means is referred to as ‘sampling

distribution of the mean.’
3.4.1 Sampling Distribution of the Mean
The sampling distribution of the mean can thus be defined as, ‘A probability
distribution of all possible sample means of a given size, selected from a population’.
Accordingly, the sampling distribution of the means of the ages of children as
tabulated in Example 3.1, has three predictable patterns. These are as follows:
(i) The mean of the sampling distribution and the mean of the population are
equal. This can be shown as follows:
Sample mean ( X ) Prob. P( X )
3 .1
4 .1
5 .2
6 .2
7 .2
8 .1
9 .1
1.00
Then,
= XP( X ) = (3 × .l) + (4 × .l) + (5×.2) + (6 × .2) + (7 × .2) + (8 × .l) + 9 ×.l) = 6
This value is the same as the mean of the original population.
(ii) The spread of the sample means in the distribution is smaller than in the
population values. For example, the spread in the distribution of sample means
above is from 3–9, while the spread in the population was from 2–10.
(iii) The shape of the sampling distribution of the means tends to be, ‘Bell-
shaped’ and approximates the normal probability distribution, even when
the population is not normally distributed. This last property leads us to the
‘Central Limit Theorem’.
Central limit theorem
Central Limit Theorem states that, ‘Regardless of the shape of the population, the
distribution of the sample means approaches the normal probability distribution as
the sample size increases.’
The question now is how large should the sample size be in order for the
distribution of sample means to approximate the normal distribution for any type of
population. In practice, the sample sizes of 30 or larger are considered adequate for
this purpose. This should be noted however, that the sampling: distribution would
be normally distributed, if the original population is normally distributed, no matter
what the sample size.

Statistics for Management 77

Unit-3 Sampling and Hypothesis Testing

As we can see from our sampling distribution of the means, the grand mean
X of the sample means or – × equals , the population mean. However, realistically
speaking, it is not possible to take all the possible samples of size n from the
population. In practice only one sample is taken, but the discussion on the sampling
distribution is concerned with the proximity of ‘a’ sample mean to the population
mean.
It can be seen that the possible values of sample means tend towards the
population mean, and according to Central Limit Theorem, the distribution of sample
means tend to be normal for a sample size of n being larger than 30. Hence, we can
draw conclusions based upon our knowledge about the characteristics of the normal
distribution.
For example, in the case of sampling distribution of the means, if we know
the grand mean –×of this distribution, which is equal to p, and the standard deviation
of this distribution, known as ‘Standard error of free mean’ and denoted by –× ,
then we know from the normal distribution that there is a 68.26 per cent chance that
a sample selected at random from a population, will have a mean that lies within
one standard error of the mean of the population mean. Similarly, this chance
increases to 95.44 per cent, that the sample mean will lie within two standard errors
of the mean ( –× ) of the population mean. Hence, knowing the properties of the
sampling distribution tells us as to how close the sample mean will be to the true
population mean.
3.4.2 Standard Error
Standard error of the mean ( –×)
Standard error of the mean ( –×)is a measure of dispersion of the distribution of
sample means and is similar to the standard deviation in a frequency distribution
and it measures the likely deviation of a sample mean from the grand mean of the
sampling distribution.
If all sample means are given, then ( –×)can be calculated as follows:

( – ) where N = Number of sample means

=
N
Thus we can calculate –× for Example 3.1 of the sampling distribution of the
ages of 5 children as follows:
X (μ ) (X μ )2
3 6 9
4 6 4
5 6 1
6 6 0
7 6 1
8 6 4
9 6 9
(X μ )2 = 28

78 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Then,
( – )
=
N

28
=
7
= 4=2
However, since it is not possible to take all possible samples from the
population, we must use alternate methods to compute –× .
The standard error of the mean can be computed from the following formula,
if the population is finite and we know the population mean. Hence,
σ
σ = (N n)
n (N 1)

Where,
= Population standard deviation
N = Population size
n = Sample size
This formula can be made simpler to use by the fact that we generally deal
with very large populations, which can be considered infinite, so that if the population
size A’ is very large and sample size n is small, as for example in the case of items
tested from assembly line operations, then,
(N – n)
would approach 1.
(N – 1)

Hence,
=
n
(N – n)
The factor (N – n)
is also known as the ‘finite correction factor’, and should be
used when the population size is finite.
As this formula suggests, –× decreases as the sample size (w) increases,
meaning that the general dispersion among the sample means decreases, meaning
further that any single sample mean will become closer to the population mean, as
the value of ( –×) decreases. Additionally, since according to the property of the
normal curve, there is a 68.26 per cent chance of the population mean being within
one –× of the sample mean, a smaller value of –× will make this range shorter;
thus making the population mean closer to the sample mean (refer Example 3.2).
Example 3.2: The IQ scores of college students are normally distributed with the
mean of 120 and standard deviation of 10.
(a) What is the probability that the IQ score of any one student chosen at random
is between 120 and 125?
(b) If a random sample of 25 students is taken, what is the probability that the
mean of this sample will be between 120 and 125?

Statistics for Management 79

Unit-3 Sampling and Hypothesis Testing
Solution:
(a) Using the standardized normal distribution formula,

125
= 120
= 10

(X – )
Z=

125 – 120
Z= = 5 / 10 = .5
10
The area for Z = .5 is 19.15.
This means that there is a 19.15 per cent chance that a student picked up at
random will have an IQ score between 120 and 125.
(b) With the sample of 25 students, it is expected that the sample mean will be
much closer to the population mean, hence it is highly likely that the sample
mean would be between 120 and 125.
The formula to be used in the case of standardized normal distribution for
sampling distribution of the means is given by,
X–
Z=

where,

=
n
Hence,

125
=120
=10

X–
Z=

80 Statistics for Management

Sampling and Hypothesis Testing Unit-3

10
where, = = 10 = =2
n 25 5

Then,
125 – 120
Z= = 5 / 2 = 2.5
2
The area for Z = 2.5 is 49.38.
This shows that there is 49.38 per cent chance that the sample mean will be
between 120 and 125. As the sample size increases further, this chance will also
increase. It can be noted that the probability of a sample mean being between 120
and 125 is much higher than the probability of an individual student having an IQ
between 120 and 125.

3.5 HYPOTHESIS TESTING

A hypothesis is an approximate assumption that a researcher wants to test for its

logical or empirical consequences. Hypothesis refers to a provisional idea whose
merit needs evaluation, but has no specific meaning. Though it is often referred as
a convenient mathematical approach for simplifying cumbersome calculation. Setting
up and testing hypothesis is an integral art of statistical inference. Hypotheses are
often statements about population parameters like variance and expected value.
During the course of hypothesis testing, some inference about population like the
mean and proportion are made. Any useful hypothesis will enable predictions by
reasoning including deductive reasoning. According to Karl Popper, a hypothesis
must be falsifiable and that a proposition or theory cannot be called scientific if it
does not admit the possibility of being shown false. Hypothesis might predict the
outcome of an experiment in a lab setting the observation of a phenomenon in
nature. Thus, hypothesis is an explanation of a phenomenon or proposal suggesting
a possible correlation between multiple phenomena. Hypothesis: A phen-
omenon or proposal
The characteristics of hypothesis are: suggesting a possible
correlation between
• Clear and accurate: Hypothesis should be clear and accurate so as to draw multiple phenomena.
a consistent conclusion.

Statistics for Management 81

Unit-3 Sampling and Hypothesis Testing

• Statement of relationship between variables: If a hypothesis is relational,

it should state the relationship between different variables.
• Testability: A hypothesis should be open to testing so that other deductions
can be made from it and can be confirmed or disproved by observation. The
researcher should do some prior study to make the hypothesis a testable one.
• Specific with limited scope: A hypothesis, which is specific, with limited
scope, is easily testable than a hypothesis with limitless scope. Therefore, a
researcher should pay more time to do research on such kind of hypothesis.
• Simplicity: A hypothesis should be stated in the most simple and clear terms
to make it understandable.
• Consistency: A hypothesis should be reliable and consistent with established
and known facts.
• Time limit: A hypothesis should be capable of being tested within a reasonable
time. In other words, it can be said that the excellence of a hypothesis is
judged by the time taken to collect the data needed for the test.
• Empirical reference: A hypothesis should explain or support all the sufficient
facts needed to understand what the problem is all about.
A hypothesis is a statement or assumption concerning a population. For the purpose
of decision-making, a hypothesis has to be verified and then accepted or rejected.
This is done with the help of observations. We test a sample and make a decision on
the basis of the result obtained. Decision-making plays a significant role in different
areas such as marketing, industry and management.
3.5.1 Statistical Decision-Making
Testing a statistical hypothesis on the basis of a sample enables us to decide whether
the hypothesis should be accepted or rejected. The sample data enables us to accept
or reject the hypothesis. Since the sample data give incomplete information about
the population, the result of the test need not be considered to be final or
unchallengeable. The procedure on which the basis of sample results, enables to
decide whether a hypothesis is to be accepted or rejected. This is called Hypothesis
Testing or Test of Significance.
Note 1: A test provides evidence, if any, against a hypothesis, usually called a null
hypothesis. The test cannot prove the hypothesis to be correct. It can give some
evidence against it.
The test of hypothesis is a procedure to decide whether to accept or reject a
hypothesis.
Note 2: The acceptance of a hypothesis implies, if there is no evidence from the
sample that we should believe otherwise.
The rejection of a hypothesis leads us to conclude that it is false. This way of putting
the problem is convenient because of the uncertainty inherent in the problem. In
view of this we must always briefly state a hypothesis that we hope to reject.
A hypothesis stated in the hope of being rejected is called a null hypothesis and is
denoted by H0.

82 Statistics for Management

Sampling and Hypothesis Testing Unit-3

If H0 is rejected, it may lead to the acceptance of an alternative hypothesis denoted

by H1.
For example, a new fragrance soap is introduced in the market. The null hypothesis
H0, which may be rejected, is that the new soap will not be better than the existing
soap.
Similarly, a dice is suspected to be rolled. Roll the dice a number of times to test.
The Null Hypothesis H0: p = 1/6 for showing six.
The Alternative hypothesis H1: p 1/6.
For example, skulls found at an ancient site may all belong to race X or race Y on the
basis of their diameters. We may test the hypothesis, that the mean is of the
population from which the present skulls came. We have the hypotheses.
H0 : = x
, H1 : = y

Here, we should not insist on calling either hypothesis null and the other alternative
since the reverse could also be true.
3.5.2 Type I and Type II Errors
There are two types of errors in statistical hypothesis, which are as follows:
• Type I error: In this type of error, you may reject a null hypothesis when
it is true. It means rejection of a hypothesis, which should have been
accepted. It is denoted by (alpha) and is also known alpha error.
• Type II error: In this type of error, you are supposed to accept a null
hypothesis when it is not true. It means accepting a hypothesis, which
should have been rejected. It is denoted by (beta) and is also known as
beta error.
Type I error can be controlled by fixing it at a lower level. For example, if
you fix it at 2%, then the maximum probability to commit Type I error is
0.02. But, reducing Type I error has a disadvantage when the sample size is
fixed, as it increases the chances of Type II error. In other words, it can be
said that both types of errors cannot be reduced simultaneously. The only
solution of this problem is to set an appropriate level by considering the costs
and penalties attached to them or to strike a proper balance between both
types of errors.
In a hypothesis test, a Type I error occurs when the null hypothesis is rejected when
it is in fact true; that is, H0 is wrongly rejected. For example, in a clinical trial of a
new drug, the null hypothesis might be that the new drug is no better, on average,
than the current drug; that is H0: there is no difference between the two drugs on
average. A Type I error would occur if we concluded that the two drugs produced
different effects, when in fact there was no difference between them.
In a hypothesis test, a Type II error occurs when the null hypothesis H0, is not
rejected when it is in fact false. For example, in a clinical trial of a new drug, the
null hypothesis might be that the new drug is no better, on average, than the current
drug; that is H0: there is no difference between the two drugs on average. A Type II

Statistics for Management 83

Unit-3 Sampling and Hypothesis Testing

error would occur if it were concluded that the two drugs produced the same effect,
that is, there is no difference between the two drugs on average, when in fact they
produced different ones.
In how many ways can we commit errors?
We reject a hypothesis when it may be true. This is type I error.
We accept a hypothesis when it may be false. This is type II error.
The other true situations are desirable:
We accept a hypothesis when it is true. We reject a hypothesis when it is false.

Accept H0 Reject H0

H0 Accept True H0 Reject True H0

Desirable Type I Error
True

H1 Accept False H0 Reject False H0

Type II Error Desirable
False

The level of significance implies the probability of Type I error. A 5 per cent level
implies that the probability of committing a Type I error is 0.05. A 1 per cent level
implies 0.01 probability of committing Type I error.
Lowering the significance level and hence the probability of Type I error is good
but unfortunately, it would lead to the undesirable situation of committing Type II
error.
To sum up:
• Type I Error: Rejecting H0when H 0is true.
• Type II Error: Accepting H0when H 0is false.
Note. The probability of making a Type I error is the level of significance of a
statistical test. It is denoted by
Where, = Prob. (Rejecting H 0 / H0 true)
1– = Prob. (Accepting H 0 / H0 true)
The probability of making a Type II error is denoted by .
Where, = Prob. (Accepting H 0 / H0 false)
1– = Prob. (Rejecting H / H false) = Prob. (The test correctly rejects H when
0 0 0
H0 is false.
1– is called the power of the test. It depends on the level of significance , sample size
n and the parameter value.
3.5.3 Null and Alternative Hypotheses
Hypothesis is usually considered as the principal instrument in research.

84 Statistics for Management

Sampling and Hypothesis Testing Unit-3

In the context of statistical analysis, while comparing any two methods, the
following concepts or assumptions are taken into consideration:
• Null hypothesis: While comparing two different methods in terms of their
superiority, wherein the assumption is that both the methods are equally good
is called null hypothesis. It is also known as statistical hypothesis and is
symbolized as H0.
• Alternate hypothesis: While comparing two different methods, regarding
their superiority, wherein, stating a particular method to be good or bad
as compared to the other is called alternate hypothesis. It is symbolized as
H1.
Comparison of null hypothesis with alternate hypothesis
Following are the points of comparison between null hypothesis and alternate
hypothesis:
• Null hypothesis is always specific, while alternate hypothesis gives an
approximate value.
• The rejection of null hypothesis involves great risk, which is not the case of
alternate hypothesis.
Null hypothesis is more frequently used in statistics than alternate hypothesis because
it is specific and is not based on probabilities.
The hypothesis to be tested is called the null hypothesis and is denoted by H0.This
is to be tested against other possible states of nature called alternative hypothesis.
The alternative is usually denoted by H1.
The null hypothesis implies that there is no difference between the statistic and the
population parameter. To test whether there is any difference between the sample
mean X and the population , we write the null hypothesis.
H0: X = µ
The alternative hypothesis would be,
H1: µ
This means > µ or < µ. This is called a two-tailed hypothesis.
The alternative hypothesis H1: > µ is right tailed.
The alternative hypothesis H1: < µ is left tailed.
These are one sided or one-tailed alternatives.
Note 1: The alternative hypothesis H1 implies all such values of the parameter,
which are not specified by the null hypothesis H0.
Note 2: Testing a statistical hypothesis is a rule, which leads to a decision to accept
or reject a hypothesis.
A one-tailed test requires rejection of the null hypothesis when the sample statistic
is greater than the population value or less than the population value at a certain
level of significance.
1. We may want to test if the sample mean exceeds the population mean .
Then the null hypothesis is,

Statistics for Management 85

 Unit-3 Sampling and Hypothesis Testing

H0: > µ
2. In the other case the null hypothesis could be,
H0: < µ
Each of these two situations leads to a one-tailed test and has to be dealt with in the
same manner as the two-tailed test. Here, the critical rejection is on one side only,
right for > µ and left for < µ. Both the Figures 3.2 and 3.3 here show a 5 per cent
level of test of significance.
For example, a minister in a certain government has an average life of 11 months
without being involved in a scam. A new party claims to provide ministers with an
average life of more than 11 months without scam. We would like to test if, on the
average, the new ministers last longer than 11 months. We may write the null
hypothesis H0: = 11 and alternative hypothesis H1: > 11.

Fig. 3.2 H : >

0

Fig. 3.3 H 0: <

3.5.4 Critical Region

The Critical Region (CR), or Rejection Region (RR), is a set of values for testing
statistic for which the null hypothesis is rejected in a hypothesis test. It means, the
sample space for the test statistic is partitioned into two regions; one region as the
critical region will lead us to reject the null hypothesis H0, the other not. So, if the
observed value of the test statistic is a member of the critical region, we conclude
that ‘reject H0’; if it is not a member of the critical region then we conclude that ‘do
not reject H0’.
We shall consider test problems arising out of Type I Error.
The level of significance of a test is the maximum probability with which we are
willing to take a risk of Type I error.
If we take a 5 per cent significance level (p = 0.05), we are 95 per cent confident
(p = 0.95) that a right decision has been made.
A 1 per cent significance level (p = 0.01), makes us 99 per cent confident (p = 0.99)
about the correctness of the decision.

86 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Critical region is the area of the sampling distribution in which the test statistic
must fall for the null hypothesis to be rejected.
We can say that the critical region corresponds to the range of values of the statistic,
which according to the test requires the hypothesis to be rejected.
• Two-tailed and one-tailed tests: A two-tailed test rejects the null hypothesis
if the sample mean is either more or less than the hypothesized value of the
mean of the population. It is considered to be apt when null hypothesis is of
some specific value whereas alternate hypothesis is not equal to the value of
null hypothesis. In a two-tailed curve there are two rejection regions, also
called critical regions.

Acceptance and rejection

regions in case of a two-tailed
test
(With 5% significance level)
Rejection Rejection
region Acceptance region (Accept H0 region
if the sample mean X falls in
this region)
LIMIT
LIMIT

0.475 of 0.475 of
area area

{Both taken together equals

0.95 or 95% of area}

Z = –1.96 2H 0= Z = 1.96

Reject H0 if the sample

mean ( X ) falls in either
of these two regions

• Conditions for the occurrence of one-tailed test: When the population mean
is either lower or higher than some hypothesized value, one-tailed test is
considered to be appropriate where the rejection is only on the left tail of the
curve. This is known as left-tailed test.

Statistics for Management 87

Unit-3 Sampling and Hypothesis Testing

For example, what will happen if the acceptance region is made larger? will
decrease. It will be more easily possible to accept H0 when H0 is false (Type II
error), i.e., it will lower the probability by making a Type I error, but raise that of ,
Type II error.
Note: , are probabilities of making an error; 1 – , l – are probabilities of
making correct decisions.

Can we say α + β = 1?
No. Each is concerned with a different type of error. But both are not independent
of each other.
3.5.5 Penalty
Usually Type II error is considered the worse of the two though, it is mainly the
circumstances of a case that decide the answer to this question.

88 Statistics for Management

Sampling and Hypothesis Testing Unit-3

If Type I error means accepting the hypothesis that a guilty person is innocent
and if Type II error means accepting the hypothesis that an innocent person is guilty,
then Type II error would be dangerous. The penalties and costs associated with an
error determine the balance or trade-off between Type I and Type II errors.
Usually, Type I error is shown as the shaded area, say 5 per cent of a normal
curve which is supposed to represent the data. If the sample statistic, say the sample
mean, falls in the shaded area, the hypothesis is rejected at 5 per cent level of
significance.
3.5.6 Standard Error
The concept of Standard Error (SE) of statistic, is used to test the precision of a
sample and provides the confidence limits for the corresponding population
parameter.
The statistic may be the sample arithmetic mean , the sample proportion p,
etc.
The SE of any such statistic is the standard deviation of the sampling
distribution of the statistic. Given below is SE in common use.
2 2
1 + 2
SE( X1 X2) =
n1 n2
SE( p p )= P1Q1 P2Q2
+
1 2
n1 n2

SE of difference between two means X1, X 2 or two proportions p1, p2 sample sizes
n1, n2 can be stated as,

2 2
1 + 2
SE( X1 X2) =
n1 n2
SE( p p )= P1Q1 P2Q2
+
1 2
n1 n2

Where, n is the number of observations.

X is the sample mean.
µ is the population mean.
s is the sample standard deviation.
σ is the population standard deviation.
p is the sample proportion, q = 1 – p.
P is the population proportion, Q = l – P.
3.5.7 Testing of Hypothesis
Statistical decisions have to be made in the presence of uncertainty. In testing of
hypothesis, the choice is between H0 and H1. In estimation, there are several choices
available. The design of experiments requires one to choose between the nature and
the extent of observations. All these have to be done in the presence of uncertainty.

Statistics for Management 89

Unit-3 Sampling and Hypothesis Testing

A decision function D(x), assigns to every possible outcome a unique action. This
may result in loss—positive or negative—depending on an unknown parameter w.
So, the loss function is L(w, D), which depends on the outcome x is a random
variable. Its expected value is called the risk function.

3.5.8 Tests for a Sample Mean X

We have to test the null hypothesis that the population mean has a specified value µ,
i.e., H0: X = µ. For large n, if H0 is true then,

X
z is approximately nominal. The theoretical region for z depending on
SE( X )
the desired level of significance can be calculated.
Example 3.3: A factory produces items, each weighing 5 kg with variance 4. Can a
random sample of size 900 with mean weight 4.45 kg be justified as having been
taken from this factory?
Solution:
n = 900
X = 4.45
µ =5
= 4=2
X X 4.45 5
z = = = 8.25
SE( X ) / n 2 / 30

We have z > 3. The null hypothesis is rejected. The sample may not be regarded as
originally from the factory at 0.27% level of significance (corresponding to 99.73%
acceptance region).
3.5.9 Test for Equality of Two Proportions
If P1, P2 are proportions of some characteristic of two samples of sizes n1, n2, drawn
from populations with proportions P1, P2, then we have H0: P1 = P2 vs H1:P1 ¹ P2.
Case (I): If H0 is true, then let P1 = P2 = p
Where, p can be found from the data:
n1P1 n2 P2
p
n1 n2
q 1 p

p is the mean of the two proportions.

1 1
SE(P P2 ) pq
1
n1 n2

90 Statistics for Management

Sampling and Hypothesis Testing Unit-3

P1 P2
z , P is approximately normal (0,1)
SE(P1 P2 )

We write z ~ N(0, 1)
The usual rules for rejection or acceptance are applicable here.
Case (II): If it is assumed that the proportion under question is not the same in the
two populations from which the samples are drawn and that P1, P2 are the true
proportions, we write,

P1q1 P2 q2
SE(P1 P2 ) = +
n1 n2

We can also write the confidence interval for P1 – P2.

For two independent samples of sizes n1, n2 selected from two binomial populations,
the 100 (1 – ) per cent confidence limits for P – P are,
1 2

P1q1 P2q2
(P1 P2 ) z /2
n1 n2

The 90 per cent confidence limits would be [with a = 0.1, 100 (1 – a) = 0.90]

(P1 P1q1 P2q2

P2 ) 1.645
n1 n2

Example 3.4: Out of 5000 interviewees, 2400 are in favour of a proposal, and out
of another set of 2000 interviewees, 1200 are in favour. Is the difference significant?
2400 1200
Where, P= = 0.48 P = = 0.6
1 2
5000 2000
2400 1200
Solution: Given, P = = 0.48 P = = 0.6
1 2
5000 2000
n1 = 5000 n2 = 2000

0.48 0.52 0.6 0.4

SE = + = 0.013 (using Case (II))
5000 2000

P1 P2 0.12
z 9.2 3
SE 0.013
The difference is highly significant at 0.27% level.

3.5.10 Large Sample Test for Equality of Two Means X1 , X 2

Suppose two samples of sizes n1 and n2 are drawn from populations having means
, and standard deviations , .
1 2 1 2

Statistics for Management 91

Unit-3 Sampling and Hypothesis Testing

To test the equality of means X1, X 2 we write,

H0 : 1 2
H1 : 1 2

If we assume H0 is true then,

z X1 X2
2 2
1 2 , approximately normally distributed with mean 0, and S.D. = 1.
n1 n2
We write z ~ N (0, 1)
As usual, if | z | > 2 we reject H0 at 4.55 per cent level of significance, and so on.
Example 3.5: Two groups of sizes 121 and 81 are subjected to tests. Their means
are found to be 84 and 81 and standard deviations 10 and 12. Test for the significance
of difference between the groups.
Solution: X1 = 84 X 2 = 81 n1 = 121 n2 = 81
= 10 = 12
1 2

z X1 X2 84 81
2 2
z
1 2 , = 1.86 < 1.96
n1 n2 121 81
The difference is not significant at the 5 per cent level of significance.
3.5.11 Small Sample Tests of Significance
The sampling distribution of many statistics for large samples is approximately
normal. For small samples with n < 30, the normal distribution, as shown in Example
3.5, can be used only if the sample is from a normal population with known .
If is not known, we can use student’s t distribution instead of the normal. We then
replace by sample standard deviation with some modification as given below.
Let x1, x2, ..., xn be a random sample of size n drawn from a normal population with
mean and S.D. . Then,

x
t .
s/ n 1
Here, t follows the student’s t distribution with n – 1 degrees of freedom.
Note: For small samples of n < 30, the term n 1 , in SE = s / n 1 , corrects the
bias, resulting from the use of sample standard deviation as an estimator of .
Also,
s2 n 1 n 1
or s S
S2 n n

92 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Procedure: Small Samples

To test the null hypothesis H0 : , against the alternative hypothesis H1 : .
X
Calculate t and compare it with the table value with n – 1 degrees of
SE( X )
freedom (d.f.) at level of significance a per cent.
If this value > table value, reject H0.
If this value < table value, accept H0.
(Significance level idea same as for large samples)
We can also find the 95 per cent (or any other) confidence limits for .
For the two-tailed test (use the same rules as for large samples; substitute t for z) the
95 per cent confidence limits are,
X t s/ n 1
Rejection Region. At a per cent level for two-tailed test, if | t | > ta/2 reject.
For one-tailed test, (right) if t > ta reject.
(left) if t > –ta reject.
At 5 per cent level the three cases are,
If | t | > t0.025 reject two-tailed.
If t > t0.05 reject one-tailed right.
If t £ t0.05 reject one-tailed left.
For proportions, the same procedure is to be followed.
Example 3.6: A firm produces tubes of diameter 2 cm. A sample of 10 tubes is
found to have a diameter of 2.01 cm and variance 0.004. Is the difference significant?
Given t0.05,9= 2.26
Solution:
X
t=
s/ n 1
2.01 2
=
0.004/(10 1)
0.01
=
0.021
= 0.48
Since, |t| < 2.26, the difference is not significant at 5 per cent level.
3.5.12 Paired Observations
t-Test for the Difference of Means
Let (x1, y1), (x2, y2,), ...,(xn, yn), be the pairs of values for the same subjects, e.g., sales
data before (x) and after an advertisement campaign (y).
Performance of candidates before (x) and after training (y)

Statistics for Management 93

Unit-3 Sampling and Hypothesis Testing
We have to test the significance of the difference between x, y values.
For each pair (xi-, yi), find di = xi- – yi.
H 0: 1= ,2i.e., no difference before and after and H : 1 1
.We 2
find the mean d of d values and use the statistics:
d
t
S/ n
2
(d d)
S
n 1
Example 3.7: Eleven students were given a test and their marks noted. After training,
their marks in a second test were noted. Do the marks indicate any benefit from
training?
Solution:

Student 1 2 3 4 5 6 7 8 9 10 11

x 23 20 19 21 18 20 18 17 23 16 19

y 24 19 22 18 20 22 20 20 23 20 17

di –1 1 –3 3 –2 –2 –2 –3 0 –4 2

11
d= = 1
11
2
(d d )
s= = 2.49
n 1
df = 11 1 = 10
d 1
t= = = 0.121
2.49 / 11 2.49 11

The difference is not significant.

3.5.13 Test for a Given Population Variance
In the test for given population variance, the variance is the square of the standard
deviation, whatever you say about a variance can be, for all practical purposes,
extended to a population standard deviation.
To test the hypothesis that a sample x1, x2, …xn of size n has a specified variance
0.

2 2
Null hypothesis H0 : 0
or
0
2 2
H1 : 0

94 Statistics for Management

Sampling and Hypothesis Testing Unit-3

ns2 (x – x )2
2
Test statistic 2 2
0 0

2
If is greater than the table value, we reject the null hypothesis.

3.6 T-DISTRIBUTION TESTING

Sir William S. Gosset (pen name Student) developed a significance test and through
it made significant contribution in the theory of sampling applicable in case of
small samples. When population variance is not known, the test is commonly known
as Student’s t-test and is based on the t distribution.
Like the normal distribution, t distribution is also symmetrical but happens to
be flatter than the normal distribution. Moreover, there is a different t distribution
for every possible sample size. As the sample size gets larger, the shape of the t
distribution loses its flatness and becomes approximately equal to the normal
distribution. In fact, for sample sizes more than 30, the t distribution is so close to
the normal distribution that we will use the normal to approximate the t distribution.
Thus, when n is small, the t distribution is far from normal, but when n is infinite, it
is identical with normal distribution.
For applying t-test in context of small samples, the t value is calculated first
and, then the calculated value is compared with the table value of t at certain level
of significance for given degrees of freedom. If the calculated value of t exceeds the
table value (say t0.05), we infer that the difference is significant at 5 per cent level,
but if calculated value is t0, is less than its concerning table value, the difference is
not treated as significant.
The t-test is used when two conditions are fullfiled:
(i) The sample size is less than 30, i.e., when n 30.
(ii) The population standard deviation ( p) must be unknown.
In using the t-test, we assume the following:
(i) The population is normal or approximately normal.
(ii) The observations are independent and the samples are randomly drawn
samples.
(iii) There is no measurement error.
(iv) In the case of two samples, population variances are regarded as equal if
equality of the two population means is to be tested.
The following formulae are commonly used to calculate the t value:
(i) To test the significance of the mean of a random sample

|X |
t=
S | SEx X
where, X = Mean of the sample

Statistics for Management 95

Unit-3 Sampling and Hypothesis Testing

µ = Mean of the universe

SEx = S.E. of mean in case of small sample and is worked out as follows:

(x i x )2
n
SEx = s
=
n n
and the degrees of freedom = (n – 1)
The above stated formula for t can as well be stated as under:
|x |
t=
SEx
|x |
=
(x x )2
n 1
n
|x |
= n
(x x )2n
1
If we want to work out the probable or fiducial limits of population mean (µ) in
case of small samples, we can use either of the following:
(a) Probable limits with 95 per cent confidence level:

=X SEx (t0.05 )
(b) Probable limits with 99 per cent confidence level:
= X SEx (t0.01 )
At other confidence levels, the limits can be worked out in a similar manner, taking
the concerning table value of t just as we have taken t0.05 in (a) and t0.01 in (b) above.
(ii) To test the difference between the means of the two samples
| X1 X 2 |
t=
SE x x
1 2

where, X1 = Mean of the sample 1

X2 = Mean of the sample 2
SEx 1 x2 = Standard error of difference between two sample means and is
worked out as follows:

2 2
(X x )1 + (X 2 i x 2)
=
1i
SEx1 x2
n1 + n2 2
1 1
+
n1 n2

96 Statistics for Management

Sampling and Hypothesis Testing Unit-3

and the degrees of freedom = (n1 + n2 – 2).

When the actual means are in fraction, then use of assumed means is convenient. In
such a case, the standard deviation of difference, i.e.,

( x 1i + x 1)2 + ( x 2i x 2)2
n1 + n2 2

can be worked out by the following short-cut formula:

( x 1i A 1)2 + (x A )2 1 n (x 1 A )2 n (x A2i )2
= 2i 1i 2 2 2

n1 + n2 2

where, A1 = Assumed mean of sample 1

A2 = Assumed mean of sample 2
X1 = True mean of sample 1
X2 = True mean of sample 2
(iii) To test the significance of an observed correlation coefficient
r
t= n 2
1 r2
Here, t is based on (n – 2) degrees of freedom.
(iv) In context of the ‘difference test’.
Difference test is applied in the case of paired data and in this context t is
calculated as follows:
xDitt 0 xDiff 0
t= = n
6Diff n 6Diff

where, XDiff or
D = Mean of the differences of sample items.
0 = The value zero on the hypothesis that there is no difference
Diff.
= Standard deviation of difference and is worked out as

D XDiff ) (n 2
1)
or

D 2 ( D )2 n
(n 1)
D = Differences
n = Number of pairs in two samples and is based on (n –1)
degrees of freedom.

Statistics for Management 97

Unit-3 Sampling and Hypothesis Testing

The following examples would illustrate the application of t-test using the above
stated formulae.
Example 3.8:
A sample of 10 measurements of the diameter of a sphere, gave a mean X = 4.38
inches and a standard deviation, = 0.06 inches. Find (a) 95% and (b) 99%
confidence limits for the actual diameter.
Solution:
On the basis of the given data the standard error of mean
0.06
= s= 0.06 = = 0.02
n 1 10 1 3

Assuming the sample mean 4.38 inches to be the population mean, the required
limits are as follows:
(i) 95% confidence limits = X SEx (t0.05 ) with degrees of freedom
= 4.38 ± .02(2.262)
= 4.38 ± .04524
i.e., 4.335 to 4.425
(ii) 99% confidence limits = X SEx (t0.01 ) with 9 degrees of freedom
= 4.38 ± .02(3.25) = 4.38 ± .0650
i.e., 4.3150 to 4.4450.
Example 3.9:
The specimen of copper wires drawn from a large lot have the following breaking
strength (in kg. wt.):
578, 572, 570, 568, 572, 578, 570, 572, 596, 544

Test whether the mean breaking strength of the lot may be taken to be 578 kg. wt.
Solution:
We take the hypothesis that there is no difference between the mean height of the
sample and the given height of universe. In other words we can write,
H0 : = X , H0 : X . Then on the basis of the sample data, the mean and standard
deviation has been worked out as follows:
2
S. No. X (X – X ) (X1 – X )
1 578 6 36
2 572 0 0
3 570 –2 4
4 568 –4 16
5 572 0 0

98 Statistics for Management

Sampling and Hypothesis Testing Unit-3

6 578 6 36
7 570 –2 4
8 572 0 0
9 596 24 576
10 544 –28 784
n = 10 X = 5720 (X – X )2 = 1456
i i

x 5720
X= =
n 10
= 572
(x x )s2
s =
n 1
1456 1456
= =
10 1 9
=12.72
12.72
SE x = s
=
n 10
12.72
= = 4.03
3.16
|x | | 572 578 |
t= =
SEx 4.03
= 1.488
Degrees of freedom = n –1 = 9
At 5% level of significance for 9 degrees of freedom, the table value of t = 2.262.
For a two-tailed test.
The calculated value of t is less than its table value and hence the difference is
insignificant. The mean breaking strength of the lot may be taken to be 578 Kg. wt.
with 95% confidence level.
Example 3.10:
Sample of sales in similar shops in two towns are taken for a new product with the
following results:
Mean sales Variance ‘Size of sample
Town A 57 5.3 5
Town B 61 4.8 7
Is there any evidence of difference in sales in the two towns?
Solution:
We take the hypothesis that there is no difference between the two sample means
concerning sales in the two towns. In other words, H0 : X1 = X 2 , H0 : X1 X 2 . Then,
we work out the concerning t value as follows:
Statistics for Management 99
Unit-3 Sampling and Hypothesis Testing

|X1 X2 |
t=
SEx1 x 2

where, x1 = Mean of the sample concerning Town A

x2 = Mean of the sample concerning Town B
SEx x = Standard Error of the difference between two means
1 2

(x 1i x )12 + (x x )2 2 1 1
SEx1 x2 = 2i
+
n1 + n2 2 n1 n2

Hence,

| 57 61 | 4
t= =
1.421 1.421
= 2.82
Degrees of freedom = (n1 + n2 – 2) = (5 + 7 – 2) = 10
Table value of t at 5% level of significance for 10 degrees of freedom is 2.228, for
a two-tailed test.
The calculated value of t is greater than its table value. Hence, the hypothesis is
wrong and the difference is significant.
Example 3.11:
The sales data of an item in six shops before and after a special promotional campaign
is:
Shops A B C D E F
Before the
promotional
campaign 53 28 31 48 50 42
After the campaign 58 29 30 55 56 45
Can the campaign be judged to be a success? Test at 5% level of significance.
Solution:
We take the hypothesis that the campaign does not bring any improvement in sales.
We can thus write:
In order to judge this, we apply the ‘difference test’. For this purpose we calculate
the mean and standard deviation of differences in two sample items as follows:

Shops Sales before Sales after Difference = D (D – D ) (D – D )2

campaign campaign (i.e., increase or
XBi XAi decrease after the
campaign)

A 53 58 +5 +1.5 2.25
B 28 29 +1 –2.5 6.25
C 31 30 –1 –4.5 20.25

100 Statistics for Management

Sampling and Hypothesis Testing Unit-3
D 48 55 +7 +3.5 12.25
E 50 56 +6 +2.5 6.25
F 42 45 +3 –0.5 0.25

n=6 D = 21 (D – D )2 = 47.50
D 21
Mean of difference or X = = = 3.5
Diff
n 6
Standard deviation of difference

(D D)2 47.50
Diff = = = 3.08
n 1 6 1
X 0
t = Diff = n
Diff

= 1.14 × 2.45 = 2.793

Degrees of freedom = (n – 1) = (6 – 1) = 5
Table value of t at 5% level of significance for 5 degrees of freedom = 2.015 for
one-tailed test.
Since, the calculated value of t is greater than its table value, the difference is
significant. Thus, the hypothesis is wrong and the special promotional campaign
can be taken as a success.
Example 3.12:
Memory capacity of 9 students was tested before and after training. From the
following scores, state whether the training was effective or not.

Student 1 2 3 4 5 6 7 8 9
Before (XBi) 10 15 9 3 7 12 16 17 4
After (XAi) 12 17 8 5 6 11 18 20 3

Solution:
We take the hypothesis that training was not effective. We can write,
H0 : xA = X B , H0 : X > X B . We apply the difference test for which purpose first of all
we calculate the mean and standard deviation of difference as follows:

Students Before XBi After XAi Difference = D D2

1 10 12 2 4
2 15 17 2 4
3 9 8 –1 1
4 3 5 2 4
5 7 6 –1 1
6 12 11 –1 1

Statistics for Management 101

Unit-3 Sampling and Hypothesis Testing

7 16 18 2 4
8 17 20 3 9
9 4 3 –1 1

n=9 D=7 D2 = 29

D 7
D= = = 0.78
n 9
D2 (D)2 n 29 (0.78)2 9
Diff = = = 1.71
n 1 9 1
0.78
t= = 1.369
1 71
Degrees of freedom = (n – 1) = (9 – 1) = 8
Table value of t at 5% level of significance for 8 degrees of freedom
= 1.860 for one-tailed test.
Since the calculated value of t is less than its table value, the difference is insignificant
and the hypothesis is true. Hence, it can be inferred that the training was not effective.
Example 3.13:
It was found that the coefficient of correlation between two variables calculated
from a sample of 25 items was 0.37. Test the significance of r at 5% level with the
help of t-test.
Solution:
To test the significance of r through t-test, we use the following formula for
calculating t value:
r
t= n 2
1 r2
0.37
= 25 2
1 (0.37)2
=1.903
Degrees of freedom = (n–2) = (25–2) = 23
The table value of at 5% level of significance for 23 degrees of freedom is 2.069
for a two-tailed test.
The calculated value of t is less than its table value, hence r is insignificant.
Example 3.14:
A group of seven week old chickens reared on high protein diet weighs 12, 15, 11,
16, 14, 14 and 16 ounces; a second group of five chickens similarly treated except
that they receive a low protein diet weighs 8, 10, 14, 10 and 13 ounces. Test at 5%
level whether there is significant evidence that additional protein has increased the
weight of chickens. (Use assumed mean (or A1) = 10 for the sample of 7 and assumed
mean (or A2) = 8 for the sample of 5 chickens in your calculation).

102 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Solution:
We take the hypothesis that additional protein has not increased the weight of the
chickens. We can write, Ho: X1 > X2 Ho : X1 > X2.
Applying t-test, we work out the value of t for measuring the significance of two
sample means as follows:

X X2
t = SE1
x x
1 2

Calculation can be done as follows:

(X1i – A1) (X1i – A1)2 X (X – A ) (X – A ) 2
X1 2 2i 2 2i 2

A1 = 10 A2= 8

12 2 4 8 0 0
15 5 25 10 2 4
11 1 1 14 6 36
16 6 36 10 2 4
14 4 16 13 5 25
14 4 16
16 6 36

n1=7 (X 1i– A )1 (X 1i– A )12 n2=5 (X 2i– A )2 (X 2i– A )22

=28 =134 =15 = 69
X1 = A1 + (x1i A1 ) / n1
28
=10 + = 14
7
(x2i A2 )
Similarly, X 2 = A 1+
n2
15
=8+ = 11
5
Hence,

(X A )12 + ( X A )2 2 n ( X 1 A1 )2 n (X A 2)2 1 1
SE X1 X2 = 1i 2i 1 2 2
+
n1 + n2 2 n1 n2
134 + 69 7 (14 10)2 5(11 8)2 1 1
= +
7+5 2 7 5
= (2.14) (.59) = 1.2626
We now calculate the value under t,
X X 2 14 11
t= 1 = = 2.376
SEX X 1.26261 2

Statistics for Management 103

Unit-3 Sampling and Hypothesis Testing

Degree of freedom = (n1 + n2 – 2) = (7 + 5 – 2) = 10

The table value of t at 5% level of significance for 10 degrees of freedom = 1.812
for one-tailed test.
The calculated value of t is higher than its table value and hence the difference is
significant, which means the hypothesis is wrong. It can therefore be concluded
that additional protein has increased the weight of chickens.

3.7 CHI-SQUARE TEST FOR INDEPENDENCE OF

ATTRIBUTES ACTIVITY

Chi-square test is a non-parametric test of statistical significance for bivariate tabular

analysis (also known as cross-breaks). Any appropriate test of statistical significance
Chi-square test: A non-
lets you know the degree of confidence you can have in accepting or rejecting a
parametric test of hypothesis. Typically, the Chi-square test is any statistical hypothesis test, in which
statistical significance the test statistics has a chi-square distribution when the null hypothesis is true. It is
for bivariate tabular performed on different samples (of people) who are different enough in some
analysis. characteristic or aspect of their behaviour that we can generalize from the samples
selected. The population from which our samples are drawn should also be different
in the behaviour or characteristic. Amongst the several tests used in statistics for
judging the significance of the sampling data, Chi-square test, developed by Prof.
Fisher, is considered as an important test. Chi-square, symbolically written as 2
(pronounced as Ki-square), is a statistical measure with the help of which, it is
possible to assess the significance of the difference between the observed frequencies
and the expected frequencies obtained from some hypothetical universe. Chi-square
tests enable us to test whether more than two population proportions can be
considered equal. In order that Chi-square test may be applicable, both the frequencies
must be grouped in the same way and the theoretical distribution must be adjusted
to give the same total frequency which is equal to that of observed frequencies. c2 is
calculated with the help of the following formula:

2 ( f0 fe )2
fe

104 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Where, f0 means the observed frequency; and

fe means the expected frequency.
2
Whether or not a calculated value of is significant, it can be ascertained by
2
looking at the tabulated values of (given at the end of this book in appendix
part) for given degrees of freedom at a certain level of confidence (generally a
5% level is taken). If the calculated value of 2 exceeds the table value, the
difference between the observed and the expected frequency is taken as
significant but if the table value is more than the calculated value of 2, then the
difference between the observed and expected frequencies is considered as
insignificant, i.e., considered to have arisen as a result of chance and as such can
be ignored.
3.7.1 Degrees of Freedom
As already stated in the earlier unit, the number of independent constraints
determines the number of degrees of freedom (or df). If there are 10 frequency
classes and there is one independent constraint, then there are (10 – 1) = 9 degrees
of freedom. Thus, if n is the number of groups and one constraint is placed by
making the total of observed and expected frequencies equal, df = (n – 1); when
two constraints are placed by making the totals as well as the arithmetic means
equal then df = (n – 2), and so on. In the case of a contingency table (i.e., a table
with two columns and more than two rows or a table with two rows but more
than two columns or a table with more than two rows and more than two columns)
or in the case of a 2 × 2 table the degrees of freedom is worked out as follows:
df = (c – 1)(r – 1)
Where, c = Number of columns
r = Number of rows
3.7.2 Conditions for the Application of Test
The following conditions should be satisfied before the test can be applied:
(i) Observations recorded and used are collected on a random basis.
(ii) All the members (or items) in the sample must be independent.
(iii) No group should contain very few items, say less than 10. In cases where
the frequencies are less than 10, regrouping is done by combining the
frequencies of adjoining groups so that the new frequencies become greater
than 10. Some statisticians take this number as 5, but 10 is regarded as
better by most statisticians.
(iv) The overall number of items (i.e., N) must be reasonably large. It should
at least be 50, howsoever small the number of groups may be.
(v) The constraints must be linear. Constraints which involve linear equations
in the cell frequencies of a contingency table (i.e., equations containing
no squares or higher powers of the frequencies) are known as linear
constraints.
3.7.3 Areas of Application of Chi-Square Test
Chi-square test is applicable in large number of problems. The test is, in fact, a
technique through the use of which it is possible for us to (a) Test the goodness

Statistics for Management 105

Unit-3 Sampling and Hypothesis Testing

of fit; (b) Test the homogeneity of a number of frequency distributions; and (c)
Test the significance of association between two attributes. In other words, Chi-
square test is a test of independence, goodness of fit and homogeneity. At times
Chi-square test is used as a test of population variance also.
As a test of goodness of fit, 2 test enables us to see how well the distribution of
observe data fits the assumed theoretical distribution such as Binomial
distribution, Poisson distribution or the Normal distribution.
As a test of independence, 2 test helps explain whether or not two
attributes are associated. For instance, we may be interested in knowing whether
a new medicine is effective in controlling fever or not and 2 test will help us in
deciding this issue. In such a situation, we proceed on the null hypothesis that the
two attributes (viz., new medicine and control of fever) are independent. Which
means that new medicine is not effective in controlling fever. It may, however,
be stated here that 2 is not a measure of the degree of relationship or the form of
relationship between two attributes but it simply is a technique of judging the
significance of such association or relationship between two attributes.
As a test of homogeneity, 2 test helps us in stating whether different
samples come from the same universe. Through this test, we can also explain
whether the results worked out on the basis of sample/samples are in conformity
with well defined hypothesis or the results fail to support the given hypothesis.
As such the test can be taken as an important decision-making technique.
As a test of population variance. Chi-square is also used to test the
significance of population variance through confidence intervals, specially in
case of small samples.
3.7.4 Steps in Finding the Value of Chi-Square
The various steps involved are as follows:
(i) First of all calculate the expected frequencies.
(ii) Obtain the difference between observed and expected frequencies and
find out the squares of these differences, i.e., calculate (f 0 – f )e2.
(iii) Divide the quantity ( f0 – f ) 2 obtained, as stated above by the corresponding
e

( f0 fe )2
expected frequency to get .
fe
( f0 fe )2 ( f0 fe )2
(iv) Then find summation of values or what we call
fe fe
2
This is the required value.
The 2 value obtained as such should be compared with the relevant table value
of 2 and inference may be drawn as stated above.
The following examples illustrate the use of Chi-square test.
Example 3.15: A dice is thrown 132 times with the following results:

Number Turned Up 1 2 3 4 5 6
Frequency 16 20 25 14 29 28
Test the hypothesis that the dice is unbiased.

106 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Solution: Let us take the hypothesis that the dice is unbiased. If that is so, the
probability of obtaining any one of the six numbers is 1/6 and as such the
1
expected frequency of any one number coming upward is 132 22. Now, we
6
can write the observed frequencies along with expected frequencies and work out
the value of 2 as follows:

No. Turned Observed Expected ( f0 – f e) (f0 – f e)2 ( f0 fe )2

Up Frequency Frequency fe
(or f0) (or fe )
1 16 22 –6 36 36/22
2 20 22 –2 4 4/22
3 25 22 3 9 9/22
4 14 22 –8 64 64/22
5 29 22 7 49 49/22
6 28 22 6 36 36/22
( f0 fe )2
=9
fe

Hence, the calculated value of 2 = 9.

Degrees of freedom in the given problem is (n – 1) = (6 – 1) = 5.
The table value of 2 for 5 degrees of freedom at 5% level of significance is
11.071. If we compare the calculated and table values of 2 we find that
calculated value is less than the table value and as such could have arisen due
to fluctuations of sampling. The result thus supports the hypothesis and it can be
concluded that the dice is unbiased.
Example 3.16:
Find the value of 2 for the following information:
Class Observed A B C D E
Frequency 8 29 44 15 4
Theoretical (or
Expected) Frequency 7 24 38 24 7
Solution:
Since some of the frequencies are less than 10, we shall first regroup the given
data as follows and then work out the value of 2:
Class Observed Frequency Expected Frequency (f0 – fe) ( f0 fe )2
(f0) (fe) fe

A and B (8+29) = 37 (7+24) = 31 6 36/31

C 44 38 6 36/38
D and E (15+4) = 19 (24+7) = 31 –12 144/31

Statistics for Management 107

Unit-3 Sampling and Hypothesis Testing

2 ( f0 fe )2
= 6.76 approx.
fe

Example 3.17:
Two research workers classified some people in income groups on the basis of
sampling studies. Their results are as follows:
Income Groups
Investigators Poor Middle Rich
Total
A 160 30 10 200
B 140 120 40 300
Total 300 150 50 500
Show that the sampling technique of at least one research worker is defective.
Solution:
Let us take the hypothesis that the sampling techniques adopted by the research
workers are similar (i.e., there is no difference between the techniques adopted
by the research workers). This being so, the expectation of A investigator
classifying the people in,
200 300
(i) Poor income group 120
500

200 150
(ii) Middle income group 60
500

200 50
(iii) Rich income group 20
500
Similarly, the expectation of B investigator classifying the people in
300 300
(i) Poor income group 180
500

300 150
(ii) Middle income group 90
500

300 50
(iii) Rich income group 30
500
2
We can now calculate value of as follows:

Groups Observed Expected (f0 – fe) (f0 – fe)2/fe

Frequency Frequency
(f0) (fe)
Investigator A
Classifies people as poor 160 120 40 1600/120 = 13.33

108 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Classifies people as middle class 30 60 –30 900/60 = 15.00

Classifies people as rich 10 20 –10 100/20 = 5.00
Investigator B
Classifies people as poor 140 180 –40 1600/180 = 8.88
Classifies people as middle class 120 90 30 900/90 = 10.00
Classifies people as rich 40 30 10 100/30 = 3.33

2 ( f0 fe )2
= 55.54
fe

Degrees of freedom = (c – 1)(r – 1)

= (3 – 1)(2 – 1) = 2
2
The table value of for two degrees of freedom at 5% level of significance is
5.991. The calculated value of 2 is much higher than this table value which
means that the calculated value cannot be said to have arisen just because of
chance. It is significant. Hence, the hypothesis does not hold good. This means
that the sampling techniques adopted by the two investigators differ and are not
similar. Naturally, then the technique of one must be superior than that of the
other.
3.7.5 Alternative Formula for Finding the Value of Chi-Square in a
(2 × 2) Table
There is an alternative method of calculating the value of 2 in the case of a
(2 × 2) table. Let us write the cell frequencies and marginal totals in case of a
(2 × 2) table as follows:
a b (a + b)
c d (c + d)
(a + c) (b + d) N
2
Then the formula for calculating the value of will be stated as follows:

2
(ad bc)2 N
=
(a c)(b d )(a b)(c d )

Where, N means the total frequency, ad means the larger cross product, bc means
the smaller cross product and (a + c), (b + d), (a + b) and (c + d) are the
marginal totals. The alternative formula is rarely used in finding out the value of
Chi-square as it is not applicable uniformly in all cases but can be used only in
a (2 × 2) contingency table.
3.7.6 Yates’ Correction
F. Yates has suggested a correction in 2 value calculated in connection with a (2
× 2) table particularly when cell frequencies are small (since no cell frequency
should be less than 5 in any case, though 10 is better as stated earlier) and 2 is
just on the significance level. The correction suggested by Yates is popularly

Statistics for Management 109

Unit-3 Sampling and Hypothesis Testing

known as Yates’ correction. It involves the reduction of the deviation of observed,

from expected frequencies which of course reduces the value of 2. The rule for
correction is to adjust the observed frequency in each cell of a (2 × 2) table in
such a way as to reduce the deviation of the observed from the expected
frequency for that cell by 0.5, and this adjustment is made in all the cells without
disturbing the marginal totals. The formula for finding the value of 2 after
applying Yates’ correction is written as follows:

2
N.(ad bc 0.5 N )2
(corrected) =
(a b)(c d )(a c)(b d )
In case we use the usual formula for calculating the value of Chi-square
2 ( f0 fe )2 then Yates’ correction can be applied as follows:
viz.,
fe
2 2
f01 fe1 0.5 f02 fe2 0.5
2
(corrected) =
fe1 fe2
It may again be emphasized that Yates’ correction is made only in case of
(2 × 2) table and that too when cell frequencies are small.
3.7.7 Chi-Square as a Test of Population Variance
2
is used, at times, to test the significance of population variance ( p)2 through
confidence intervals. This, in other words, means that we can use 2 test to judge
if a random sample has been drawn from a normal population with mean ( ) and
with specified variance ( p )2. In such a situation, the test statistic for a null
hypothesis will be as follows:
2 ( Xi X s )2 n( s )2
= with (n–1) degrees of freedom.
( p)
2
( p )2
By comparing the calculated value (with the help of the above formula) with the
table value of 2 for (n–1) df at a certain level of significance, we may accept or
reject the null hypothesis. If the calculated value is equal or less than the table
value, the null hypothesis is to be accepted but if the calculated value is greater
than the table value, the hypothesis is rejected. This can be made clear by an
example.
Example 3.18:
Weight of 10 students is as follows:
Sl. No. 1 2 3 4 5 6 7 8 9 10
Weight in kg. 38 40 45 53 47 43 55 48 52 49
Can we say that the variance of the distribution of weights of all students from
which the above sample of 10 students was drawn is equal to 20 square kg? Test
this at 5% and 1% level of significance.

110 Statistics for Management

Sampling and Hypothesis Testing Unit-3

Solution:
First, we should work out the standard deviation of the sample ( s).
Calculation of the sample standard deviation:
2
Sl. No. Xi Xi Xs ( Xi Xs)
Weight in kg
1 38 – 9 81
2 40 – 7 49
3 45 – 2 04
4 53 + 6 36
5 47 + 0 00
6 43 – 4 16
7 55 + 8 64
8 48 + 1 01
9 52 + 5 25
10 49 + 2 04
2
n = 10 Xi = 470 ( Xi Xs) = 280
Xi 470
Xs = 47 kg
n 10

( Xi X s )2 280
s
= 28 5.3 kg
n 10
)s2 = 28
Taking the null hypothesis as H : ( )2 = ( )2
0 p s
2
n( s ) 10 28 280
The test statistic 2
= 14
( p )2 20 20
Degrees of freedom in this case is (n – 1) = 10 – 1 = 9
At 5% level of significance, the table value of 2 = 16.92, and at 1% level of
significance it is 21.67 for 9 df, and both these values are greater than the
calculated value of 2 which is 14. Hence, we accept the null hypothesis and
conclude that the variance of the given distribution can be taken as 20 square kg
at 5% as well as at 1% level of significance.
3.7.8 Additive Property of Chi-Square( 2)
An important property of 2 is its additive nature. This means that several values
of 2 can be added together and if the degrees of freedom are also added, this
number gives the degrees of freedom of the total value of 2. Thus, if a number of
2
values have been obtained from a number of samples of similar data, then
because of the additive nature of 2, we can combine the various values of 2 by just
simply adding them. Such addition of various values of 2 gives one value of 2
which helps in forming a better idea about the significance of the problem under
consideration. The following example illustrates the additive property of the 2.

Statistics for Management 111

Unit-3 Sampling and Hypothesis Testing
2
Example 3.19: The following values of are obtained from different
investigations carried to examine the effectiveness of a recently invented
medicine for checking malaria.
2
Investigation df
1 2.5 1
2 3.2 1
3 4.1 1
4 3.7 1
5 4.5 1
What conclusion would you draw about the effectiveness of the new medicine on
the basis of the five investigations taken together?
Solution: By adding all the values of 2, we obtain a value equal to 18.0. Also by
adding the various d.f. as given in the question, we obtain a figure 5. We can now
state that the value of 2 for 5 degrees of freedom (when all the five investigations
are taken together) is 18.0.
Let us take the hypothesis that the new medicine is not effective. The table value
of 2 for 5 degrees of freedom at 5% level of significance is 11.070. But our
calculated value is higher than this table value which means that the difference
is significant and is not due to chance. As such the hypothesis is wrong and it can
be concluded that the new medicine is effective in checking malaria.
3.7.9 Important Characteristics of Chi-Square ( 2) Test
(i) Chi-square test is based on frequencies and not on parameters like mean and
standard deviation.
(ii) This test is used for testing the hypothesis and is not useful for estimation.
(iii) This test possesses the additive property.
(iv) This test can also be applied to a complex contingency table with several
classes and as such is a very useful test in research work.
(v) This test is an important non-parametric (or a distribution free) test as no
rigid assumptions are necessary in regard to the type of population and no
need of the parameter values. It involves less mathematical details.
2
A word of caution in using Test
Chi-square test is no doubt a most frequently used test but its correct application is
equally an uphill task. It should be borne in mind that the test is to be applied only
when the individual observations of sample are independent, which means that the
occurrence of one individual observation (event) has no effect upon the occurrence
of any other observation (event) in the sample under consideration. The researcher,
while applying this test, must remain careful about all these things and must
thoroughly understand the rationale of this important test before using it and drawing
inferences concerning his hypothesis.

112 Statistics for Management

Sampling and Hypothesis Testing Unit-3

3.8 LET US SUM UP

• Sampling is the process of selecting a sample from the population. It is

technically and economically not feasible to take the entire population for
analysis.
• Random sample is a collection of items selected from the population in such
a manner that each item in the population has exactly the same chance of
being selected, so that the sample taken from the population would be truly
representative of the population.
• The various methods of sampling can be grouped under two broad categories:
Probability (or random) sampling and Non-probability (or non-random)
sampling.
• Non-probability sampling methods are those which do not provide every item
in the universe with a known chance of being included in the sample.
• Sampling errors arise due to drawing faulty inferences about the population
based upon the results of the samples.
• Non-sampling errors are introduced due to technically faulty observations or
during the processing of data.

Statistics for Management 113

Unit-3 Sampling and Hypothesis Testing

• A hypothesis is a statement or assumption concerning a population. For the

purpose of decision-making, a hypothesis has to be verified and then accepted
or rejected.
• Testing a statistical hypothesis on the basis of a sample enables us to decide
whether the hypothesis should be accepted or rejected.
• The Chi-square test is any statistical hypothesis test, in which the test statistics
has a chi-square distribution when the null hypothesis is true.

3.9 FURTHER READING

Chance, William A. Statistical Methods for Decision Making. Illinois: Richard D
Irwin, 1969.
Chandan, J.S. Statistics for Business and Economics. New Delhi: Vikas Publishing
House Pvt. Ltd., 1998.
Chandan, J.S., Jagjit Singh and K.K. Khanna. Business Statistics. New Delhi: Vikas
Publishing House Pvt. Ltd., 1995.
Kothari, C.R. Quantitative Techniques. New Delhi: Vikas Publishing House Pvt.
Ltd., 1984.
Levin, Richard I. Statistics for Management. New Jersey: Prentice-Hall, 1991.
Meyer, Paul L. Introductory Probability and Statistical Applications. Massachusetts:
Addison-Wesley Publishing Co., 1970.
Neter, John, and William Wasserman. Fundamental Statistics for Business and
Economics. New York: Allyn and Bacon, 1956.
Yamane, Taro. Statistics—An Introductory Analysis. New York: Harper & Row,
1964.

3.10 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress-1

1. Sampling is necessary because of the following reasons: First, as discussed
earlier, it is not technically or economically feasible to take the entire
population into consideration. Second, due to dynamic changes in business,
industrial and social environment, it is necessary to make quick decisions
based upon the analysis of information.
2. The non-probability methods include: (a) Judgement Sampling, (b) Quota
Sampling and (c) Convenience Sampling.
3. The process of relating the sample results to population is referred to as,
‘Statistical Inference’ or ‘Inferential Statistics’.
Check Your Progress-2
1. While comparing two different methods in terms of their superiority, wherein,
the assumption is that both the methods are equally good is called null
hypothesis. It is also known as statistical hypothesis and is symbolized as
H0.

114 Statistics for Management

Sampling and Hypothesis Testing Unit-3

2. The Critical Region (CR), or Rejection Region (RR), is a set of values for
testing statistic for which the null hypothesis is rejected in a hypothesis test.
3. The t-test is used when two conditions are fullfiled,
(i) The sample size is less than 30, i.e., when n 30.
(ii) The population standard deviation ( p) must be unknown.
Check Your Progress-3
1. Chi-square, symbolically written as 2 (pronounced as Ki-square), is a
statistical measure with the help of which, it is possible to assess the
significance of the difference between the observed frequencies and the
expected frequencies obtained from some hypothetical universe. Chi-square
tests enable us to test whether more than two population proportions can be
considered equal.
2. As a test of goodness of fit, 2 test enables us to see how well the distribution
of observed data fits the assumed theoretical distribution such as binomial
distribution, Poisson distribution or the normal distribution.
3. (i) Chi-square test is based on frequencies and not on parameters like mean
and standard deviation.
(ii) This test is used for testing the hypothesis and is not useful for estimation.

3.11 POSSIBLE QUESTIONS

Short-Answer Questions
1. Define sampling.
2. Explain briefly the different types of sampling.
3. What is sampling distribution? Give examples.
4. Define population in statistical terms.
5. What is a hypothesis?
6. Explain the importance of statistical decision-making.
7. Define null and alternate hypotheses.
8. Describe the various types of errors that occur in statistical hypotheses.
9. Describe standard error.
10. What do you mean by the level of significance?
11. What is critical region?
12. Describe the one-tailed test.
13. What is the importance of a two-tailed test in statistics?
14. Write the importance of a small sample test.
15. Explain t-test and Chi-square test.

Statistics for Management 115

Unit-3 Sampling and Hypothesis Testing

Long-Answer Questions
1. Differentiate between probability samples and non-probability samples. Under
what circumstances would non-probability types of samples be more useful
in statistical analyses?
2. How does sampling with replacement differ from sampling without
replacement? Give some examples of situations where sampling has to be
done without replacement.
3. Explain in detail the situations that would require:
(a) Judgement sampling
(b) Quota sampling
(c) Stratified sampling
4. Differentiate between sampling errors and non-sampling errors. Under what
circumstances would each type of error occur? What steps can be taken to
minimize the impact of such errors upon statistical analyses?
5. Your college has a total population of 5000 students. It is desired to estimate
the proportion of students who use drugs.
(a) What type of sampling would be necessary to reach a meaningful
conclusion regarding the drug use habits of all students?
(b) What type of sampling would you select so that the sample is most
representative of the population?
(c) Drug use being a sensitive issue, what type of questions would you
include in your questionnaire? What type of questions would you avoid?
Give reasons.
6. The lottery method of sample selection is still the most often used method.
Discuss this method in detail and give reasons as to why a sample selected by
the lottery method would be representative of the population.
7. You are the chairman of the Department of Business Administration and you
have been asked to make a report on the current status of students who
graduated with B.S. in Business during the two years of 1989 and 1990.
Records have indicated that a total of 425 students graduated from the
department during these two years. The report is to include information
regarding sex of the student, grade point average at the time of graduation,
whether the students completed MBA degree or started a job after B.S. degree,
current employment position and current annual salary. Prepare a proposal
for this survey and include in this proposal:
(a) Objectives of the survey
(b) Type of sampling technique
(c) Size of the sample
8. At New Delhi airport, there is a green channel and a red channel. Passengers
without any custom duty articles can go through the green channel. Some
passengers are stopped for a random check. What type of random sampling
would be appropriate in such situations? Would judgement sampling be more
appropriate? Give reasons.

116 Statistics for Management

Sampling and Hypothesis Testing Unit-3
9. The average length of stay in a hospital after open heart surgery is believed to
be about 10 days. A random sample of 10 patients who had open heart surgery
at North Shore Hospital in the previous month showed the number of days
spent by each patient in the hospital as follows:
10, 8, 7, 12, 16, 9, 8, 10, 12, 13
Calculate the point estimate of the average number of days that a patient
spends in the hospital after open heart surgery.
10. A rate analyst for a telephone company wished to study residential monthly
telephone costs in Brooklyn. He took a random sample of 10 households
scattered all over Brooklyn and checked their telephone bills. With the data
so obtained, he calculated the mean and the standard deviation, which are
recorded below:
x = $47.30
s = $8.20
Construct a 90 per cent confidence interval for the true average monthly
telephone bill for all households in Brooklyn.
11. It is desired to estimate the proportion of people who make purchases of
gasoline at the local Mobil gas station by credit cards. A random sample of
100 gasoline purchasers was taken and it was found that 30 per cent of the
purchasers paid by credit cards. Establish a 95 per cent confidence interval
estimate of the true proportion of gasoline purchases made by credit cards.
12. It is desired to estimate the proportion of junior executives who change their
first job within the first five years. This proportion is to be estimated within 3
per cent of error and .95 degree of confidence is to be used. A study conducted
several years ago revealed that 30 per cent of such junior executives changed
their first job within 5 years.
13. It is believed that 75 per cent of American people would like to see the sale of
such assault weapons as AK-47 banned as war against crime. A congressional
representative would like to take a survey before presenting a bill in Congress.
At a confidence level of 95 per cent, what should be the sample size for each
of the following margins of error?
(a) 10 per cent margin of error
(b) 5 per cent margin of error
(c) 2 per cent margin of error
(d) In general how does the decrease in the margin of error affect the sample
size?
14. A die was rolled 400 times. It showed a 6 coming up 80 times. Can you say
the die is unbiased? p = 80/400=1/5 P=1/6
15. A company supplied 500 units of an item. The number of defectives was
found to be 42 as against the company’s conviction that 6 per cent items
could be defective. Examine the tenability of the company claim.
16. The average score of two groups A B were found to be 25 and 22 with S.D. 4
and 5.5 respectively. Test for the equality of the two group scores. Given
n1 = n2 = 400.

Statistics for Management 117

Unit-3 Sampling and Hypothesis Testing

17. 800 ore pieces from a mine were found to contain an average of 74.5 gm of
gold. From a nearby mine, 1600 pieces had 75 gm gold. Test the equality of
the averages from the two mines, each having an S.D. of 2.4.
18. To test the goodness of a coin, it is tossed 5 times. It is considered a bad coin
if more than 4 heads show up. (a) What is the probability of Type I error?
(b) If the probability of a head is 0.2, what is the probability of Type II error?
19. 10 persons randomly selected are found to have heights 63, 63, 66, 67, 68,
69, 70, 71, 71, 71 inches. Discuss the suggestion that the mean height in the
population is 66 inches.
20. 360 persons out of 600 are found to suffer from pollution induced bronchitis
in one city. In another, 400 out of 500 are found to suffer from bronchitis. Is
there any significant difference in the incidence of bronchitis?

3.12 LEARNING OUTCOMES

• Random and non-random sampling methods

• The methods of sample size determination
• Hypothesis testing
• Learn paired t-test
• The areas of application of chi-square test

118 Statistics for Management

Correlation and Regression Analysis Unit-4

UNIT 4 CORRELATION AND

REGRESSION ANALYSIS
UNIT STRUCTURE
4.1 Learning Objectives
4.2 Introduction
4.3 Correlation Analysis
4.4 Methods of Studying Simple Correlation
4.5 Properties of Correlation Coefficient
4.5.1 Method of Least Squares
4.5.2 Simple Regression Coefficients
4.5.3 Karl Pearson’s Coefficient
4.5.4 Other Measures
4.6 Rank Correlation
4.7 Regression Analysis
4.8 Let Us Sum Up
4.9 Further Reading
4.10 Answers to Check Your Progress
4.11 Possible Questions
4.12 Learning Outcomes

4.1 LEARNING OBJECTIVES

After going through this unit, you will be able to:

• Explain the types of correlations
• Explain the various methods of studying simple correlation
• Describe the properties of coefficient
• Discuss regression analysis in detail

4.2 INTRODUCTION

In this unit, you will learn about correlation analysis. This technique looks at indirect
relationships and establishes the variables that are most closely associated with a
given data or mindset. It is the process of finding how accurately the line fits using
the observations. Correlation analysis can be referred to as the statistical tool used
to describe the degree to which one variable is related to another. The relationship,
if any, is assumed to be a linear one. In fact, the word ‘correlation’ refers to the
relationship or the interdependence between two variables. There are various Correlation analysis:
The statistical tool used
phenomena that are related to each other. The theory by means of which quantitative to describe the degree to
connections between two sets of phenomena are determined is called the ‘Theory which one variable is
of Correlation’. On the basis of this theory, you can study the comparative changes related to another.
occurring in two related phenomena and their cause–effect relation can also be
examined. Thus, correlation is concerned with the relationship between two related
and quantifiable variables and can be positive or negative.
Statistics for Management 119
Unit-4 Correlation and Regression Analysis

In this unit, you will also learn about regression analysis, which is the
mathematical process of using observations to find the line of best fit through the
data in order to make estimates and predictions about the behaviour of variables.
This technique is used to determine the statistical relationship between two or more
variables and to make prediction of one variable on the basis of one or more other
variables.

4.3 CORRELATION ANALYSIS

Correlation analysis is the statistical tool generally used to describe the degree to
which one variable is related to another. The relationship, if any, is usually assumed
to be a linear one. This analysis is used quite frequently in conjunction with regression
analysis to measure how well the regression line explains the variations of the
dependent variable. In fact, the word correlation refers to the relationship or
interdependence between two variables. There are various phenomena which are
related to each other. For instance, when demand of a certain commodity increases,
its price goes up and when its demand decreases, its price comes down. Similarly,
with age the height of children, with height the weight of children, with money the
supply and the general level of prices go up. Such sort of relationships can as well
be noticed for several other phenomena. The theory by means of which quantitative
connections between two sets of phenomena are determined is called the ‘Theory
of Correlation’.
On the basis of the theory of correlation, one can study the comparative
changes occurring in two related phenomena and their cause–effect relation can
be examined. It should, however, be borne in mind that relationships like ‘black
cat causes bad luck’, ‘filled up pitchers result in good fortune’ and similar other
beliefs of the people cannot be explained by the theory of correlation, since they
are all imaginary and are incapable of being justified mathematically. Thus,
correlation is concerned with relationship between two related and quantifiable
variables. If two quantities vary in sympathy, so that a movement (an increase or
decrease) in one tends to be accompanied by a movement in the same or opposite
direction in the other and the greater the change in one, the greater is the change
in the other, the quantities are said to be correlated. This type of relationship is
known as correlation or what is sometimes called, in statistics, as covariation.
For correlation, it is essential that the two phenomena should have cause–
effect relationship. If such relationship does not exist then one should not talk of
correlation. For example, if the height of the students as well as the height of the
trees increases, then one should not call it a case of correlation because the two
phenomena, viz., the height of students and the height of trees are not even casually
related. However, the relationship between the price of a commodity and its
demand, the price of a commodity and its supply, the rate of interest and savings,
etc. are examples of correlation, since in all such cases the change in one
phenomenon is explained by a change in another phenomenon.
It is appropriate here to mention that correlation in case of phenomena
pertaining to natural sciences can be reduced to absolute mathematical term, e.g.,

120 Statistics for Management

Correlation and Regression Analysis Unit-4

heat always increases with light. However, in phenomena pertaining to social

sciences, it is often difficult to establish any absolute relationship between two
phenomena. Hence, in social sciences, we must take the fact of correlation being
established if in a large number of cases, two variables always tend to move in the
same or opposite direction.
Correlation can either be positive or it can be negative. Whether correlation
is positive or negative would depend upon the direction in which the variables are
moving. If both variables are changing in the same direction, then correlation is
said to be positive, but when the variations in the two variables take place in opposite
direction, the correlation is termed as negative. This can be explained as follows:
Changes in Independent Changes in Dependent Nature of
Variable Variable Correlation

Increase (+) Increase (+) Positive (+)

Decrease (–) Decrease (–) Positive (+)

Increase (+) Decrease (–) Negative (–)

Decrease (–) Increase (+) Negative (–)

Statisticians have developed two measures for describing the correlation

between two variables, viz., the coefficient of determination and the coefficient of
correlation. These two methods are explained in detail in the following sections.

4.4 METHODS OF STUDYING SIMPLE

CORRELATION

Coefficient of Determination
The coefficient of determination (symbolically indicated as r2, though some people
would prefer to put it as R2) is a measure of the degree of linear association or
correlation between two variables, say X and Y, one of which happens to be an
independent variable and the other being a dependent variable. This coefficient is
based on the following two types of variations:
( )
2
(i) The variation of the Y values around the fitted regression line, viz., Y ˆ
Y ,
technically known as the unexplained variation.
(ii) The variation of the Y values around their own mean, viz., (Y Y ) ,
2

technically known as the total variation.

If we subtract the unexplained variation from the total variation, we obtain
what is known as the explained variation, i.e., the variation explained by the line of
regression. Thus, Explained Variation = (Total variation) – (Unexplained variation)

(Y Y ) (Y Yˆ )
2 2
=
2
Yˆ Y

Statistics for Management 121

Unit-4 Correlation and Regression Analysis

The Total and Explained as well as Unexplained variations can be shown as

given in Figure. 4.1.
Y-axis

100

80 Y
Explained Variation
( i.e.,Y Y ) at a
specific point
60
Y Mean line of Y
X
40

20

X- axis
0 20 40 60 X 80 100 120
Income (’00 Rs)
Fig. 4.1 Diagram Showing Total, Explained and Unexplained Variations

Coefficients of determination is that fraction of the total variation of Y which is

explained by the regression line. In other words, coefficient of determination is the
ratio of explained variation to total variation in the Y variable related to the X variable.
Coefficient of determination can be algebraically stated as,
Explained variation
r2 = Total variation
2
Yˆ Y
= 2
Y Y
Alternatively, r2 can also be stated as,
Explained variation
r2 = 1 –
Total variation

2
Yˆ Y
=1– 2
Y Y

Interpreting r2
The coefficient of determination can have a value ranging from 0 – 1. The value of
1 can occur only if the unexplained variation is 0, which simply means that all the
data points in the Scatter diagram fall exactly on the regression line. For a 0 value to
occur, (Y Y )2 = (Y Ŷ )2 , which simply means that X tells us nothing about Y and
hence there is no regression relationship between X and Y variables. Values between
0 and 1 indicate the ‘Goodness of fit’ of the regression line to the sample data. The
higher the value of r2, the better the fit. In other words, the value of r2 will lie

122 Statistics for Management

Correlation and Regression Analysis Unit-4
2
somewhere between 0 and 1. If r has a 0 value then it indicates no correlation, but
if it has a value equal to 1 then it indicates that there is perfect correlation and as
such the regression line is a perfect estimator. However, in most cases, the value of
r2 will lie somewhere between these two extremes of 1 and 0. One should remember
that r2 close to 1 indicates a strong correlation between X and Y, while an r2 near 0
means there is little correlation between these two variables. r2 value can as well be
interpreted by looking at the amount of the variation in Y, the dependant variable,
that is explained by the regression line. Supposing, we get a value of r2 = 0.925 then
this would mean that the variations in independent variable (say X) would explain
92.5 per cent of the variation in the dependent variable (say Y). If r2 is close to 1 then
it indicates that the regression equation explains most of the variations in the
dependent variable (see Example 4.1).
Example 4.1: Calculate the coefficient of determination (r2) using the provided
data. Calculate and analyse the result.
Observations 1 2 3 4 5 6 7 8 9 10
Income (X) (‘00 ) 41 65 50 57 96 94 110 30 79 65
Consumption
Expenditure (Y) (‘00 ) 44 60 39 51 80 68 84 34 55 48
Solution:
r2 can be worked out as follows:
2
Unexplained variation Y Yˆ
2
Since, r =1 =1 2
Total variation
Y Y
2 2
As, Y Y Y2 Y2 nY , we can write,
2
Y Yˆ
r2 = 1
Y2 nY 2
Calculating and putting the various values, we have the following equation:
260.54 260.54
r2 = 1 1 0.897
2 2526.10
34223 10 56.3

Analysis of result: The regression equation used to calculate the value of the
coefficient of determination (r2) from the sample data shows that, about 90 per cent
of the variations in consumption expenditure can be explained. In other words, it
means that the variations in income explain about 90 per cent of variations in
consumption expenditure.

Observation 1 2 3 4 5 6 7 8 9 10
Income (X) (’00 ) 41 65 50 57 96 94 110 30 79 65
Consumption
Expenditure (Y) (’00 ) 44 60 39 51 80 68 84 34 55 48

Statistics for Management 123

Unit-4 Correlation and Regression Analysis

4.5 PROPERTIES OF CORRELATION COEFFICIENT

The coefficient of correlation, symbolically denoted by ‘r’, is another important
measure to describe how well one variable is explained by another. It measures the
degree of relationship between the two casually related variables. The value of this
coefficient can never be more than +1 or less than –1. Thus, +1 and –1 are the limits
of this coefficient. For a unit change in independent variable, if there happens to be
a constant change in the dependent variable in the same direction, then the value of
the coefficient will be +1 indicative of the perfect positive correlation; but if such a
change occurs in the opposite direction, the value of the coefficient will be –1,
indicating the perfect negative correlation. In practical life, the possibility of obtaining
either a perfect positive or perfect negative correlation is very remote particularly
in respect of phenomena concerning social sciences. If the coefficient of correlation
has a zero value then it means that there exists no correlation between the variables
under study.
There are several methods of finding the coefficient of correlation, but the following
ones are considered important:
(i) Coefficient of correlation by the method of least squares
(ii) Coefficient of correlation using simple regression coefficients
(iii) Coefficient of correlation through product moment method or Karl Pearson’s
coefficient of correlation
Whichever of these three methods we adopt, we get the same value of r.
4.5.1 Method of Least Squares
Under this method, first, the estimating equation is obtained using the least square
method of simple regression analysis. The equation is worked out as,
Yˆ a bX i
2
Total variation Y Y
2
Unexplained variation Y Yˆ
2
Explained variation Yˆ Y

Then, by applying the following formulae, we can find the value of the coefficient
of correlation as,
Explained variation
r= r2
Total variation
Unexplained variation
= 1
Total variation
2
Y Yˆ
1
= 2
Y Y

This clearly shows that the coefficient of correlation happens to be the square root
of the coefficient of determination.

124 Statistics for Management

Correlation and Regression Analysis Unit-4

Short-cut formula for finding the value of ‘r’ by the method of least squares
can be repeated and readily written as,

a Y b XY nY 2
r=
Y2 nY 2
Where, a = Y-intercept
b = Slope of the estimating equation
X = Values of the independent variable
Y_ = Values of dependent variable
Y = Mean of the observed values of Y
n = Number of items in the sample
(i.e., pairs of observed data)
The plus (+) or the minus (–) sign of the coefficient of correlation worked out
by the method of least squares, is related to the sign of ‘b’ in the estimating
equation, viz., Yˆ a bXi . If ‘b’ has a minus sign, the sign of ‘r’ will also be
minus, but if ‘b’ has a plus sign, then the sign of ‘r’ will also be plus. The value
of ‘r’ indicates the degree along with the direction of the relationship between the
two variables X and Y.

4.5.2 Simple Regression Coefficients

Under this method, the estimating equation of Y and the estimating equation of X is
worked out using the method of least squares. From these estimating equations we
find the regression coefficient of X on Y, i.e., the slope of the estimating equation of
X
X (symbolically written as b XY) and this happens to be equal to r and similarly,
Y

we find the regression coefficient of Y on X, i.e., the slope of the estimating equation
Y
of Y (symbolically written as b ) and this happens to be equal to r . For finding
YX
X

‘r’, the square root of the product of these two regression coefficients are worked
out as1
r= bXY .bYX

X Y
= r .r
Y X

= r2 = r

1. The short-cut formulae to workout bXY and bYX :

XY nXY
bXY =
2
Y nY 2
XY nXY
and bYX =
X 2
nX2

Statistics for Management 125

Unit-4 Correlation and Regression Analysis

As stated earlier, the sign of ‘r’ will depend upon the sign of the regression
coefficients. If they have minus sign, then ‘r’ will take minus sign but the sign of ‘r’
will be plus if regression coefficients have plus sign.
4.5.3 Karl Pearson’s Coefficient
Karl Pearson’s method is the most widely used method of measuring the relationship
between two variables. This coefficient is based on the following assumptions:
(i) There is a linear relationship between the two variables, which means that a
straight line would be obtained if the observed data is plotted on a graph.
(ii) The two variables are casually related, which means that one of the variables
is independent and the other one is dependent.
(iii) A large number of independent causes operate on both the variables so as to
produce a normal distribution.
According to Karl Pearson, ‘r’ can be worked out as,
XY
r = n
X Y
_
Where, X = (X – X_)
Y = (Y – Y )
X = Standard deviation of
X2
X series and is equal to
n
Y = Standard deviation of
Y2
Y series and is equal to
n
n = Number of pairs of X and Y observed
A short-cut formula, known as the Product Moment Formula, can be derived
from the above stated formula as,
XY
r = n
X Y

XY
= 2
X Y2
n n
XY
n= 2
X Y2
The above formulae are based on obtaining true means (viz. X and Y ) first and
then doing all other calculations. This happens to be a tedious task, particularly
if the true means are in fractions. To avoid difficult calculations, we make use of

126 Statistics for Management

Correlation and Regression Analysis Unit-4

the assumed means in taking out deviations and doing the related calculations. In
such a situation, we can use the following formula for finding the value of ‘r’:2
a. In case of ungrouped data:
dX.dY dX dY
n n n
r =
dX 2 dX 2
dY 2 dY 2
.
n n n n

dX dY
dX .dY
n
= 2
dX 2
dY
dX 2 dY 2
n n

Where, dX = (X – XA) XA = Assumed average of X

dY = (Y – YA) YA = Assumed average of Y
dX2 = (X – XA)2
dY2 = (Y – YA)2
dX . dY = (X – XA) (Y – YA)
n = Number of pairs of observations of X and Y
b. In case of grouped data:
fdX .dY fdX fdY
n n . n
r =
fdX 2 fdX 2
fdY 2 fdY 2
– –
n n n n

2. In case we take assumed mean to be zero for X variable as for Y variable, then our formula will be
as,

nXY nX Y
n
r =
X2 X 2
Y2 Y 2

n n n n

XY
XY
n
or r =
X2 2 Y2 2
X Y
n n

XY nXY
r= 2
X nX 2
Y2 nY 2

Statistics for Management 127

Unit-4 Correlation and Regression Analysis

fdX . fdY
fdX .dY
n
or r =
2 2
fdX fdY
fdX 2 fdY 2
n n

Where, fdX.dY =0 f (X – XA) (Y – YA)

fdX = f (X – XA)
fdY = f (Y – YA)
fdY2 = f (Y – YA)2
fdX2 = f (X – XA)2
n = Number of pairs of observations of X and Y

Probable Error (P.E.) of the coefficient of correlation

Probable Error (P.E.) of r is very useful in interpreting the value of r and is worked
out for Karl Pearson’s coefficient of correlation as,
1 r2
P.E. = 0.6745
n
If r is less than its P.E., it is not at all significant. If r is more than P.E., there
is correlation. If r is more than six times its P.E. and greater than ± 0.5, then it
is considered significant. Let us consider Example 4.2.
Example 4.2: From the following data calculate ‘r’ between X and Y applying the
following three methods:
(i) The method of least squares
(ii) The method based on regression coefficients
(iii) The product moment method of Karl Pearson
Verify the obtained result of any one method with that of another.

X 1 2 3 4 5 6 7 8 9
Y 9 8 10 12 11 13 14 16 15
Solution:
Let us develop the following table for calculating the value of ‘r’:

X Y X2 Y2 XY
1 9 1 81 9
2 8 4 64 16
3 10 9 100 30
4 12 16 144 48

128 Statistics for Management

Correlation and Regression Analysis Unit-4

5 11 25 121 55
6 13 36 169 78
7 14 49 196 98
8 16 64 256 128
9 15 81 225 135
n=9
X = 45 Y = 108 X2 = 285 Y2 = 1356 XY = 597
_ _
X = 5; Y = 12
(i) Coefficient of correlation by the method of least squares is worked out as
follows:
First find out the estimating equation,
Yˆ = a + bXi

XY n XY
Where, b= 2
X2 nX

597 9 5 12 597 540 57

= = = 0.95
285 9 25 285 225 60
_ _
and a = Y – bX
= 12 – 0.95(5) = 12 – 4.75 = 7.25

Hence, Yˆ = 7.25 + 0.95Xi

Now ‘r’ can be worked out as follows by the method of least squares,

Unexplained variation
r= 1
Total variation

2 2
Y YYˆˆ Yˆˆ Y
= 1 2 = 2
Y Y Y Y

2
a Y b XY nY
= 2
Y2 nY
This is as per short-cut formula,

7.25(108) + 0.95(597) 9(12)

2

r=
9(12)
2
1356

783 567.15 1296

=
1356 1296

Statistics for Management 129

Unit-4 Correlation and Regression Analysis

54.15
= = 0.9025 = 0.95
60
(ii) Coefficient of correlation by the method based on regression coefficients is
worked out as,
Regression coefficients of Y on X,
XY n XY
i.e., bYX = 2
2
X nX
597 9 5 12 597 540 57
= 285 9 5 285 225 60
Regression coefficient of X on Y,
XY n XY
i.e., bXY = 2
2
Y nY
597 9 5 12 597 540 57
= 2
1356 9 12 1356 1296 60
Hence, r = bYX . bXY

57 57 = 57 = 0.95
=
60 60 60
(iii) Coefficient of correlation by the product moment method of Karl Pearson is
worked out as,
XY n XY
r= 2 2
X2 nX Y2 nY
597 9 5 12
=
285 9 5 2 1356 9 12 2

597 540 57 57
= = = 0.95
285 225 1356 1296 60 60 60
Hence, we get the value of r = 0.95. We get the same value by applying the other
two methods also. Therefore, whichever method we apply, the results will be the
same.
4.5.4 Other Measures
Two other measures are often talked about along with the coefficients of
determinations and that of correlation. These are as follows:
(i) Coefficient of nondetermination: Instead of using coefficient of
determination, sometimes coefficient of nondetermination is used. Coefficient
of nondetermination (denoted by k2) is the ratio of unexplained variation to
total variation in the Y variable related to the X variable. Algebrically,
2
Y Yˆ
Unexplained variation
k2 = = 2
Total variation Y Y

130 Statistics for Management

Correlation and Regression Analysis Unit-4

Concerning the data of Example 4.1, coefficient of nondetermination will be

calculated as follows:
260.54
k2 0.103
2526.10
The value of k2 shows that about 10 per cent of the variation in consumption
expenditure remains unexplained by the regression equation we had worked out,
viz., Yˆ 14.000 + 0.616Xi. In simple terms, this means that variable other than
X is responsible for 10 per cent of the variations in the dependent variable Y in
the given case.
Coefficient of nondetermination can as well be worked out as,
k2 = 1 – r2
Accordingly for Example 4.1, it will be equal to 1–0.897 = 0.103.
Note: Always remember that r2 + k2 = 1.

(ii) Coefficient of alienation: Based on k2, we can work out one more measure,
namely the coefficient of alienation, symbolically written as ‘k’. Thus,
coefficient of alienation, i.e., ‘k’ = k 2 .
Unlike r + k2 = 1, the sum of ‘r’ and ‘k’ will not be equal to 1 unless one of the
two coefficients is 1 and in this case the remaining coefficients must be zero. In
all other cases, ‘r’ + ‘k’ > 1. Coefficient of alienation is not a popular measure
from a practical point of view and is used very rarely.

4.6 RANK CORRELATION

If observations on two variables are given in the form of ranks and not as numerical
values, it is possible to compute what is known as rank correlation between the two
series.
The rank correlation, written as , is a descriptive index of agreement
between ranks over individuals. It is the same as the ordinary coefficient of
correlation computed on ranks, but its formula is simpler. Rank correlation: A
descriptive index of
6 Di2 agreement between
1
n(n2 1) ranks over individuals.

Here, n is the number of observations and Di , the positive difference between

ranks associated with the individuals i.
Like r, the rank correlation lies between –1 and +1. Consider Examples 4.3
and 4.4 for better understanding

Statistics for Management 131

Unit-4 Correlation and Regression Analysis

Example 4.3: The ranks given by two judges to 10 individuals are as follows:
Rank given by
Individual Judge I Judge II D D2
x y = x–y
1 1 7 6 36
2 2 5 3 9
3 7 8 1 1
4 9 10 1 1
5 8 9 1 1
6 6 4 2 4
7 4 1 3 9
8 3 6 3 9
9 10 3 7 49
10 5 2 3 9
D2 = 128
Solution:
The rank correlation is given by,
6 D2 6 128
ρ =1 = 1 = 1 0.776 = 0.224
n3 n 103 10
The value of = 0.224 shows that the agreement between the judges is not high.
Example 4.4: Consider Example 4.3 and compute r and compare.
Solution:
The simple coefficient of correlation r for the previous data is calculated as follows:
x y x2 y2 xy
1 7 1 49 7
2 5 4 25 10
7 8 49 64 56
9 10 81 100 90
8 9 64 81 72
6 4 36 16 24
4 1 16 1 4
3 6 9 36 18
10 3 100 9 30
5 2 25 4 10
x = 55 y = 55 x2 = 385 y2 = 385 xy = 321
55 55
321 10
10 10 18.5 18.5 = 0.224
r = 2 2 = =
55 55 82.5 82.5 82.5
385 10 385 10
10 10
This shows that the Spearman for any two sets of ranks is the same as the
Pearson r for the set of ranks. However, it is much easier to compute .

132 Statistics for Management

Correlation and Regression Analysis Unit-4
Often, the ranks are not given. Instead, the numerical values of observations are
given. In such a case, we must attach the ranks to these values to calculate .
Example 4.5: Show by means of diagrams various cases of scatter expressing
correlation between x, y.
Solution:
(i) (ii) Y
Y

O X
Positive slope
Direct linear relationship
O X High scatter r low, positive
Negative slope
Inverse linear relationship
High scatter r low, negative
(iii) Y
Y
(iv)

O X O X
Slope = 0 No relationship
r=0 r~–0
(v) Y
(vi) Y

O X
Direct
curvilinear relationship
O X
Inverse
curvilinear relationship
(vii)( Y
v
i
i
)

O X
Perfect relationship
But, r = 0 because of
non-linear relation

Statistics for Management 133

Unit-4 Correlation and Regression Analysis

Correlation analysis helps us in determining the degree to which two or more

variables are related to each other.
When there are only two variables, we can determine the degree to which one
variable is linearly related to the other. Regression analysis helps in determining the
pattern of relationship between one or more independent variables and a dependent
variable. This is done by an equation estimated with the help of data.

4.7 REGRESSION ANALYSIS

The term ‘regression’ was first used in 1877 by Sir Francis Galton who made a
study that showed that the height of children born to tall parents will tend to move
back or ‘regress’ toward the mean height of the population. He designated the word
regression as the name of the process of predicting one variable from another variable.
He coined the term multiple regression to describe the process by which several
variables are used to predict another. Thus, when there is a well-established
relationship between variables, it is possible to make use of this relationship in
making estimates and to forecast the value of one variable (the unknown or the
dependent variable) on the basis of the other variable/s (the known or the independent
variable/s). A banker, for example, could predict deposits on the basis of per capita
income in the trading area of bank. A marketing manager, may plan his advertising
expenditures on the basis of the expected effect on total sales revenue of a change in
the level of advertising expenditure. Similarly, a hospital superintendent could project
his need for beds on the basis of total population. Such predictions may be made by
using regression analysis. An investigator may employ regression analysis to test
his theory having the cause and effect relationship. All these explain that regression
analysis is an extremely useful tool especially in problems of business and industry
involving predictions.
Assumptions in regression analysis
While making use of the regression techniques for making predictions, the following
are always assumed:
(i) There is an actual relationship between the dependent and independent
variables.
(ii) The values of the dependent variable are random but the values of the
independent variable are fixed quantities without error and are chosen by the
experimentor.
(iii) There is a clear indication of direction of the relationship. This means that
dependent variable is a function of independent variable. (For example, when
we say that advertising has an effect on sales, then we are saying that sales
has an effect on advertising).
(iv) The conditions (that existed when the relationship between the dependent
and independent variable was estimated by the regression) are the same when
the regression model is being used. In other words, it simply means that the
relationship has not changed since the regression equation was computed.

134 Statistics for Management

Correlation and Regression Analysis Unit-4

(v) The analysis is to be used to predict values within the range (and not for
values outside the range) for which it is valid.
Simple Linear Regression Model
In case of simple linear regression analysis, a single variable is used to predict
another variable on the assumption of linear relationship (i.e., relationship of the
type defined by Y = a + bX) between the given variables. The variable to be predicted
is called the dependent variable and the variable on which the prediction is based is
called the independent variable.
Simple linear regression model3 (or the Regression Line) is stated as,
Yi = a + bXi + ei
Where, Yi = The dependent variable
Xi = The independent variable
ei = Unpredictable random element (usually called
residual or error term)
(i) a represents the Y-intercept, i.e., the intercept specifies the value of the
dependent variable when the independent variable has a value of zero.
(However, this term has practical meaning only if a zero value for the
independent variable is possible).
(ii) b is a constant, indicating the slope of the regression line. Slope of the line
indicates the amount of change in the value of the dependent variable for a
unit change in the independent variable.
If the two constants (viz., a and b) are known, the accuracy of our prediction of Y
(denoted by Yˆ and read as Y-hat) depends on the magnitude of the values of ei . If in
the model, all the ei tend to have very large values then the estimates will not be
very good, but if these values are relatively small, then the predicted values ( Yˆ )
will tend to be close to the true values (Yi).
Estimating the intercept and slope of the regression model (or estimating the
regression equation)
The two constants or the parameters viz., ‘a’ and ‘b’ in the regression model for the
entire population or universe are generally unknown and as such are estimated from
sample information. The following are the two methods used for estimation:
(i) Scatter diagram method
(ii) Least squares method

ˆ
3. Usually, the estimate of Y denoted by Y is written as,
Yˆ a bXi
on the assumption that the random disturbance to the system averages out or has an expected
value of zero (i.e., e = 0) for any single observation. This regression model is known as the
Regression line of Y on X from which the value of Y can be estimated for the given value of X.

Statistics for Management 135

Unit-4 Correlation and Regression Analysis

1. Scatter diagram method

This method makes use of the Scatter diagram also known as Dot diagram. Scatter
diagram4 is a diagram representing two series with the known variable, i.e.,
independent variable plotted on the X-axis and the variable to be estimated, i.e.,
dependent variable to be plotted on the Y-axis on a graph paper (see Figure 4.2) to
get the following information illustrated in Table 4.1:
Table 4.1 Table Derived from Scatter Diagram

Income Consumption Expenditure

X Y
(Hundreds of Rupees) (Hundreds of Rupees)
41 44
65 60
50 39
57 51
96 80
94 68
110 84
30 34
79 55
65 48

y-axis
Consumption Expenditure (’00 Rs)

120
100
80
60
40
20
x-axis
0 20 40 60 80 100 120

Fig. 4.2 Scatter Diagram

The scatter diagram by itself is not sufficient for predicting values of the dependent
variable. Some formal expression of the relationship between the two variables is
necessary for predictive purposes. For the purpose, one may simply take a ruler and
draw a straight line through the points in the scatter diagram and this way can

4.
(2) (3) (4) (5)
(1)
Five possible forms, which Scatter diagram may assume has been depicted in the above five
diagrams. Diagram (1) is indicative of perfect positive relationship. Diagram (2) shows perfect
negative relationship. Diagram (3) shows no relationship, Diagram (4) shows positive
relationship and Diagram (5) shows negative relationship between the two variables under
consideration.

136 Statistics for Management

Correlation and Regression Analysis Unit-4

determine the intercept and the slope of the said line and then the line can be defined
as Yˆ = a + bXi , with the help of which we can predict Y for a given value of X.
However, there are shortcomings in this approach. For example, if five different
persons draw such a straight line in the same scatter diagram, it is possible that
there may be five different estimates of a and b, especially when the dots are more
dispersed in the diagram. Hence, the estimates cannot be worked out only through
this approach. A more systematic and statistical method is required to estimate the
constants of the predictive equation. The least squares method is used to draw the
best fit line.
2. Least square method
The least squares method of fitting a line (the line of best fit or the regression line)
through the scatter diagram is a method which minimizes the sum of the squared
vertical deviations from the fitted line. In other words, the line to be fitted will pass
through the points of the scatter diagram in such a way that the sum of the squares
of the vertical deviations of these points from the line will be a minimum.
The meaning of the least squares criterion can be easily understood through
Figure 4.3, where the earlier Figure 4.2 in scatter diagram has been reproduced
along with a line which represents the least squares line to fit the data.

Fig. 4.3 Scatter Diagram, Regression Line and Short Vertical Lines Representing ‘e’

In Figure 4.3, the vertical deviations of the individual points from the line are shown
as the short vertical lines joining the points to the least squares line. These deviations
will be denoted by the symbol ‘e’. The value of ‘e’ varies from one point to another.
In some cases it is positive, while in others it is negative. If the line drawn happens
to be the least squares line, then the values of ei is the least possible. It is because
of this feature the method is known as Least Squares Method.
Why we insist on minimizing the sum of squared deviations is a question that
needs explanation. If we denote the deviations from the actual value Y to the estimated
n

value Ŷ as (Y – Yˆ ) or ei, it is logical that we want the (Y – Yˆ ) or ei , to be as small

i=1
n
as possible. However, mere examining (Y – Yˆ ) or e , is inappropriate, since
i
i=1

Statistics for Management 137

Unit-4 Correlation and Regression Analysis

any ei can be positive or negative. Large positive values and large negative values
could cancel one another. However, large values of ei regardless of their sign,
n
indicate a poor prediction. Even if we ignore the signs while working out | ei | ,
i1

the difficulties may continue. Hence, the standard procedure is to eliminate the
effect of signs by squaring each observation. Squaring each term accomplishes
two purposes, viz., (i) it magnifies (or penalizes) the larger errors, and (ii) it cancels
the effect of the positive and negative values (since a negative error when squared
becomes positive). The choice of minimizing the squared sum of errors rather
than the sum of the absolute values implies that there are many small errors rather
than a few large errors. Hence, in obtaining the regression line, we follow the
approach that the sum of the squared deviations be minimum and on this basis
work out the values of its constants viz., ‘a’ and ‘b’ also known as the intercept
and the slope of the line. This is done with the help of the following two normal
equations:5
Y = na + b X
XY = a X + b X2
In these two equations, ‘a’ and ‘b’ are unknowns and all other values viz., X,
Y, X2, XY, are the sum of the products and cross products to be calculated
from the sample data, and ‘n’ means the number of observations in the sample.
Example 4.6 explains the Least squares method.
Example 4.6: Fit a regression line Yˆ = a + bXi by the method of Least squares to
the following sample information.

Observations 1 2 3 4 5 6 7 8 9 10
Income (X) (’00 ) 41 65 50 57 96 94 110 30 79 65
Consumption
Expenditure (Y) (’00 ) 44 60 39 51 80 68 84 34 55 48

Solution:
We are to fit a regression line Yˆ = a + bXi to the given data by the method of Least
Squares. Accordingly, we work out the ‘a’ and ‘b’ values with the help of the
normal equations as stated above and also for the purpose, work out X, Y,

5. If we proceed centering each variable, i.e., setting its origin at its mean, then the two equations
will be as under:
Y = na + b X
XY = a X + b X2
But since Y and X will be zero, the first equation and the first term of the second equation will
disappear and we shall simply have the following equations:
XY = b X2
b = XY/ X2
The value of ‘a’ can then be worked out as: _ _
a = Y – bX

138 Statistics for Management

Correlation and Regression Analysis Unit-4
2
XY, X values from the given sample information table on summations for
regression equation.
Summations for Regression Equation

Observations Income Consumption XY X2 Y2

X Expenditure
Y
(’00 ) (’00 )
1 41 44 1804 1681 1936
2 65 60 3900 4225 3600
3 50 39 1950 2500 1521
4 57 51 2907 3249 2601
5 96 80 7680 9216 6400
6 94 68 6392 8836 4624
7 110 84 9240 12100 7056
8 30 34 1020 900 1156
9 79 55 4345 6241 3025
10 65 48 3120 4225 2304
n = 10 X = 687 Y =563 XY = 42358 X2= 53173 Y2 = 34223
Putting the values in the required normal equations we have,
563 = 10a + 687b
42358 = 687a + 53173b
Solving these two equations for a and b we obtain,
a = 14.000 and b = 0.616
Hence, the equation for the required regression line is,
Yˆ = a + bXi
or, Yˆ = 14.000 + 0.616Xi
This equation is known as the regression equation of Y on X from which Y values
can be estimated for given values of X variable.6

Checking the accuracy of equation

After finding the regression line, one can check its accuracy also. The method to be
used for the purpose follows from the mathematical property of a line fitted by the
method of least squares, viz., the individual positive and negative errors must sum
to zero. In other words, using the estimating equation one must find out whether the
term Y Yˆ is zero and if this is so, then one can reasonably be sure that he has
not committed any mistake in determining the estimating equation.
6. It should be pointed out that the equation used to estimate the Y variable values from values of X
should not be used to estimate the values of X variable from given values of Y variable. Another
regression equation (known as the regression equation of X on Y of the type
X = a + bY) that reverses the two value should be used if it is desired to estimate X from value of
Y.

Statistics for Management 139

Unit-4 Correlation and Regression Analysis

The problem of prediction

When we talk about prediction or estimation, we usually imply that if the relationship
Yi = a + bXi + ei exists, then the regression equation, Yˆ a bXi provides a base for
making estimates of the value of Y which will be associated with particular values
of X. In Example 4.6, we worked out the regression equation for the income and
consumption data as,
Yˆ = 14.000 + 0.616Xi
On the basis of this equation, we can make a point estimate of Y for any given
value of X. Suppose we wish to estimate the consumption expenditure of individuals
with income of 10,000. We substitute X = 100 for the same in our equation and
get an estimate of consumption expenditure as,
Yˆ 14.000 0.616 100 75.60
Thus, the regression relationship indicates that individuals with 10,000 of
income may be expected to spend approximately 7,560 on consumption. However,
this is only an expected or an estimated value and it is possible that actual
consumption expenditure of the individual with that income may deviate from this
amount and if so, then our estimate will be an error, the likelihood of which will be
high if the estimate is applied to any one individual. The interval estimate method is
considered better and it states an interval in which the expected consumption
expenditure may fall. Remember that the wider the interval, the greater the level of
confidence we can have, but the width of the interval (or what is technically known
as the precision of the estimate) is associated with a specified level of confidence
and is dependent on the variability (consumption expenditure in our case) found in
the sample. This variability is measured by the standard deviation of the error term,
‘e’, and is popularly known as the standard error of the estimate.
Standard error of the estimate
Standard error of estimate is a measure developed by statisticians for measuring the
reliability of the estimating equation. Like the standard deviation, the Standard Error
(S.E.) of Yˆ measures the variability or scatter of the observed values of Y around
the regression line. Standard Error of Estimate (S.E. of Yˆ ) is worked out as,
ˆ (Y Yˆ)2 e2
S.E. of Y (or Se )
n 2 n 2
where, S.E. of Yˆ (or Se) = Standard error of the estimate
Y = Observed value of Y
Yˆ = Estimated value of Y
e = The error term = (Y– Yˆ )
n = Number of observations in the sample
Note: In the above formula, n – 2 is used instead of n because of the fact that two degrees of
freedom are lost in basing the estimate on the variability of the sample observations about the
line with two constants viz., ‘a’ and ‘b’ whose position is determined by those same sample
observations.

140 Statistics for Management

Correlation and Regression Analysis Unit-4

The square of the Se, also known as the variance of the error term, is the basic
measure of reliability. The larger the variance, the more significant are the magnitudes
of the e’s and the less reliable is the regression analysis in predicting the data.
Interpreting the standard error of estimate and finding the confidence
limits for the estimate in large and small samples
The larger the S.E. of estimate (SEe), the greater happens to be the dispersion,
or scattering, of given observations around the regression line. However, if the
S.E. of estimate happens to be zero, then the estimating equation is a ‘perfect’
estimator (i.e., cent per cent correct estimator) of the dependent variable.
(i) In case of large samples, i.e., where n > 30 in a sample, it is assumed that the
observed points are normally distributed around the regression line and we may
find that,
• 68 per cent of all points lie within Yˆ 1 SEe limits.
• 95.5 per cent of all points lie within Yˆ 2 SEe limits.
• 99.7 per cent of all points lie within Yˆ 3 SEe limits.
This can be stated as,
a. The observed values of Y are normally distributed around each estimated
value of Yˆ and;

b. The variance of the distributions around each possible value of Yˆ is the

same.
(ii) In case of small samples, i.e., where n 30 in a sample the ‘t’ distribution
is used for finding the two limits more appropriately.
This is done as follows:
Upper limit = Yˆ + ‘t’ (SEe)
Lower limit = Yˆ – ‘t’ (SEe)
Where, Yˆ = The estimated value of Y for a given value of X.
SEe = The standard error of estimate.
‘t’ = Table value of ‘t’ for given degrees of freedom for a specified
confidence level.

Some other details concerning simple regression

Sometimes, the estimating equation of Y also known as the regression equation of Y
on X, is written as,

Yˆ Y = r
Y
Xi X
X

Y
r Xi X Y
or, Yˆ = X

Where, r = Coefficient of simple correlation between X and Y

Y = Standard deviation of Y
Statistics for Management 141
Unit-4 Correlation and Regression Analysis
= Standard deviation of X
_X
X = Mean of X
_
Y = Mean of Y
Yˆ = Value of Y to be estimated
Xi = Any given value of X for which Y is to be estimated
This is based on the formula we have used, i.e., Yˆ a bXi . The coefficient of
Xi is defined as,

Y
Coefficient of Xi = b = r
X

(Also known as regression coefficient of Y on X or slope of the regression line

of Y on X) or bYX.
2
XY n XY Y2 nY
= 2 2
Y2 nY 2 X2 nX X2 nX
XY n XY
= 2
X2 nX

Y
and a = r X Y
X

r Y
= Y since b
bX
X

Similarly, the estimating equation of X, also known as the regression equation of X

on Y, can be stated as,

Xˆ X
X
= r Y Y
Y

X
or Xˆ = r Y Y X
Y
and the
X XY n XY
Regression coefficient of X on Y (or b XY) r 2
2
Y Y nY
If we are given the two regression equations as stated above, along with the values
of ‘a’ and ‘b’ constants to solve the same for finding the value of X and Y, then the
values of X and Y so obtained, are the mean values of X (i.e., X ) and the mean
value of Y (i.e., Y ).
If we are given the two regression coefficients (viz., bXY and bYX), then we can
work out the value of coefficient of correlation by just taking the square root of the
product of the regression coefficients as shown,
r= bYX .bXY

142 Statistics for Management

Correlation and Regression Analysis Unit-4

Y X
= r .r
X Y

= r.r =r
The (±) sign of r will be determined on the basis of the sign of the given regression
coefficients. If regression coefficients have minus sign then r will be taken with
minus (–) sign and if regression coefficients have plus sign then r will be taken with
plus (+) sign, (Remember that both regression coefficients will necessarily have the
same sign, whether it is minus or plus, for their sign is governed by the sign of
coefficient of correlation.) To understand it better, see Examples 4.7 and 4.8.
Example 4.7: Given is the following information:
X Y
Mean 39.5 47.5
Standard Deviation 10.8 17.8
Simple correlation coefficient between X and Y is = + 0.42.
Find the estimating equation of Y and X.
Solution:
Estimating equation of Y can be worked out as,

(X )
Y
Yˆ Y r –X
= i
X

ˆ r Y
(X – X )+Y
or Y = i
X
17.8
= 0.42 (X i 39.5) + 47.5
10.8
= 0.69 X i 27.25 47.5
= 0.69Xi + 20.25
Similarly, the estimating equation of X can be worked out as
Xˆ
X
= r X
(Y i
Y )
Y

or Xˆ = r
X
(Y i
Y +X )
Y
10.8
or = 0.42 (Yi 47.5) + 39.5
17.8
= 0.26Yi – 12.35 + 39.5
= 0.26Yi + 27.15
Example 4.8: The following is the given data:
Variance of X = 9
Regression equations:
4X – 5Y + 33 = 0
20X – 9Y – 107 = 0

Statistics for Management 143

Unit-4 Correlation and Regression Analysis

Find: (i) Mean values of X and Y

(ii) Coefficient of Correlation between X and Y
(iii) Standard deviation of Y
Solution:
(i) For finding the mean values of X and Y, we solve the two given regression
equations for the values of X and Y as follows:
4X – 5Y + 33 = 0 ...(1)
20X – 9Y –107 = 0 ...(2)
If we multiply equation (1) by 5, we have the following equations:
20X – 25Y = –165 ...(3)
20X – 9Y = 107 ...(2)
– + –
– 16Y = –272
Subtracting equation (2) from (3)
or Y = 17
Putting this value of Y in Equation (1) we have,
4X = – 33 + 5(17)
33 85 52
or X= 13
4 4
_
Hence, X = 13 and Y = 17
(ii) For finding the coefficient of correlation, we first presume one of the two
given regression equations as the estimating equation of X. Let equation
4X – 5Y + 33 = 0 be the estimating equation of X, then we have,
5Yi 33
Xˆ
4 4
and
5
From this we can write bXY .
4
The other given equation is then taken as the estimating equation of Y and can
be written as,
20 Xi 107
Yˆ
9 9
20
and from this we can write bYX .
9
If the above equations are correct then r must be equal to,
r = 5 / 4 20 / 9 25 / 9 = 5/3 = 1.6
which is an impossible equation, since r can in no case be greater than 1. Hence,
we change our supposition about the estimating equations and by reversing it, we
re-write the estimating equations as,

144 Statistics for Management

Correlation and Regression Analysis Unit-4

9Yi 107
Xˆ
20 20
4 Xi 33
and Yˆ
5 5
Hence, r= 9 / 20 4 / 5
= 9 / 25
= 3/5
= 0.6
Since, regression coefficients have plus signs, we take r = + 0.6.
(iii) Standard deviation of Y can be calculated,
Variance of X = 9 Standard deviation of X = 3
4
bYX r Y
= 0.6 Y
0.2 Y
X 5 3
Hence, Y = 4
Alternatively, we can work it out as,

= 9 = 0.6
Y 1.8
b X =
XY
r
Y
20 3 Y

Hence, Y= 4

Statistics for Management 145

 Unit-4 Correlation and Regression Analysis

4.8 LET US SUM UP

• Correlation analysis is the statistical tool generally used to describe the degree
to which one variable is related to another.
• The coefficient of determination (symbolically indicated as r2, though some
people would prefer to put it as R2) is a measure of the degree of linear
association or correlation between two variables, say X and Y, one of which
happens to be the independent variable and the other being the dependent
variable.
• The coefficient of correlation, symbolically denoted by ‘r’, is an important
measure to describe how well one variable is explained by another. It measures
the degree of relationship between two casually-related variables.
• In case of simple linear regression analysis, a single variable is used to predict
another variable on the assumption of linear relationship (i.e., relationship of
the type defined by Y = a + bX) between the given variables.

4.7 FURTHER READING

Chance, William A. Statistical Methods for Decision Making. Illinois: Richard D
Irwin, 1969.
Chandan, J.S. 1998. Statistics for Business and Economics. New Delhi: Vikas
Publishing House.
Chandan, J.S., Jagjit Singh and K.K. Khanna. Business Statistics. New Delhi: Vikas
Publishing House, 1995.
Elhance, D.N. Fundamental of Statistics. Allahabad: Kitab Mahal, 2006.
Freud, J.E., and F.J. William. Elementary Business Statistics–The Modern Approach.
New Jersey: Prentice-Hall International, 1997.
Goon, A.M., M.K. Gupta, and B. Das Gupta. Fundamentals of Statistics. Vols. I &
II, Kolkata: The World Press Pvt. Ltd, 1983.
Gupta, S.C. Fundamentals of Business Statistics. Mumbai: Himalaya Publishing
House, 2008.
Neter, John, and William Wasserman. Fundamental Statistics for Business and
Economics. New York: Allyn and Bacon, 1956.
Yamane, Taro. Statistics—An Introductory Analysis. New York: Harper & Row,
1964.
M.Selvam 2014 Maritime Logistics and Documentation , Lakshmi Publications
, New Delhi
M.Selvam 2013 Quality Management , Lakshmi Publications , New Delhi
Besterfield .H (2018) Total Quality Management

146 Statistics for Management

Correlation and Regression Analysis Unit-4

4.10 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress-1

1. The theory by means of which quantitative connections between two sets of
phenomena are determined is called the ‘Theory of Correlation’.
2. The coefficient of determination (symbolically indicated as r2, though some
people would prefer to put it as R2) is a measure of the degree of linear
association or correlation between two variables, say X and Y, one of which
happens to be independent variable and the other being dependent variable.
3. The coefficient of correlation, symbolically denoted by ‘r’, is another
important measure to describe how well one variable is explained by another.
It measures the degree of relationship between the casually-related variables.
4. The variable to be predicted is called the dependent variable and the variable
on which the prediction is based is called the independent variable.
5. The following are the two methods used for estimation:
(i) Scatter diagram method
(ii) Least squares method

4.11 POSSIBLE QUESTIONS

Short-Answer Questions
1. What is the importance of correlation analysis?
2. How will you determine the coefficient of determination?
3. When is correlation positive and when is it negative?
4. What are the assumptions in Karl Pearson’s coefficient?
5. What is the relationship between coefficient of nondetermination and
coefficient of alienation?
6. List the basic precautions and limitations of regression and correlation
analyses.
7. Differentiate between scatter diagram and least square method.
Long-Answer Questions
1. Explain the method to calculate the coefficient of correlation using simple
regression coefficient.
2. Describe Karl Pearson’s method of measuring coefficient of correlation.
3. Explain Spearman’s rank correlation.
4. What is regression analysis? What are the assumptions in it?
5. Explain scatter diagram and the least square method in detail. Also, mention
how scatter diagram helps in studying correlation between two variables.

Statistics for Management 147

4.12 LEARNING OUTCOMES
• The types of correlations
• The various methods of studying simple correlation
• The properties of coefficient
• Learn regression analysis in detail
Time Series Analysis Unit-5

UNIT 5 TIME SERIES ANALYSIS

UNIT STRUCTURE
5.1 Learning Objectives
5.2 Introduction
5.3 Components of Time Series Analysis
5.4 Forecasting with Decomposed Components of Time Series
5.4.1 Trend Analysis
5.4.2 Smoothing Techniques
5.4.3 Measuring Cyclical Effect
5.5 Let Us Sum Up
5.6 Further Reading
5.7 Answers to Check Your Progress
5.8 Possible Questions
5.9 Learning Outcomes

5.1 LEARNING OBJECTIVES

After going through this unit, you will be able to:
• Classify the time series
• Analyse the components of time series
• Study the influence of time series analysis
• Explain the different methods of measuring trend
• List the different methods of measuring seasonal variations
• Explain smoothing techniques

5.2 INTRODUCTION
In this unit, you will learn how time series analysis differs from regression analysis.
We often see a number of charts on company drawing boards or in newspapers, where
we see lines goingup and down from left to right on a graph. The vertical axis represents
a variable such as productivity or crime data in the city, and the horizontal axis
represents the different periods of increasing time such asdays, weeks, months or years.
The analysis of the movements of such variables over periods of time is referred to as
time series analysis. Time series can then be defined as a set of numeric observations
of the dependent variable, measured at specific points in time in a chronological order,
usually at equal intervals, in order to determine the relationship of time to such
variables.
In this unit, you will also learn that one of the major elements of planning, and
specifically strategic planning of any organization is accurately forecasting the future
events that would have an impact on the operations of the organization. Previous
performances must be studied so as to forecast future activity. Even in our daily lives,
we plan our future events on the basis of a reasonable estimate of the future environment
that could affect our plans, whether it is forecasting rain on our picnic on Saturday, or
forecasting economic conditions for ten years. Textbook publishers must predict future
Statistics for Management 149
Unit-5 Time Series Analysis

sales of books to print enough copies for students; financial advisors must predict the
values of a variety of economic factors in order to advise clients regarding stocks,
bonds and other business opportunities. Similarly, hotel builders in a citymust project
the future influx of tourists, and so on. The quality of such forecasts is strongly related
to the relevant information that can be extracted and used from past data. In that respect,
time series can be used to determine patterns in the data of the past over a period of
time and extrapolate the data into the future.

5.3 COMPONENTS OF TIME SERIES ANALYSIS

The time series analysis method is quite accurate where future is expected to be
similar to past. The underlying assumption in time series is that the same factors
will continue to influence the future patterns of economic activity in a similar
manner as in the past. These techniques are fairly sophisticated and require
experts to use these methods.
The classical approach is to analyse a time series in terms of four distinct types
of variations or separate components that influence a time series.
1. Secular Trend or Simply Trend (T)
Trend is a general long-term movement in the time series value of the variable
(Y) over a fairly long period of time. The variable (Y) is the factor that we are
interested in evaluating for the future. It could be sales, population, crime rate,
and so on.
Trend is a common word, popularly used in day-to-day conversation, such
as population trends, inflation trends and birth rate. These variables are observed
over a long period of time and any changes related to time are noted and calculated
and a trend of these changes is established. There are many types of trends; the
series may be increasing at a slow rate or at a fast rate or these may be decreasing
at various rates. Some remain relatively constant and some reverse their trend
from growth to decline or from decline to growth over a period of time. These
changes occur as a result of the general tendency of the data to increase or
decrease as a result of some identifiable influences.
If a trend can be determined and the rate of change can be ascertained, then
tentative estimates on the same series values into the future can be made. However,
such forecasts are based upon the assumption that the conditions affecting the
steady growth or decline are reasonably expected to remain unchanged in the
future. A change in these conditions would affect the forecasts. As an example,
a time-series involving increase in population over time can be shown as,

150 Statistics for Management

Time Series Analysis Unit-5

2. Cyclical Fluctuations (C)

Cyclical fluctuations refer to regular swings or patterns that repeat over a long
period of time. The movements are considered cyclical only if they occur after
time intervals of more than one year. These are the changes that take place as a Cyclical fluctuations:
result of economic booms or depressions. These may be ups and downs, and are Regular swings or
patterns that repeat over
recurrent in nature and have a duration of several years—usually lasting for two
a long period of time.
to ten years. These movements also differ in intensity or amplitude and each
phase of movement changes gradually into the phase that follows it. Some
economists believe that the business cycle completes four phases every twelve to
fifteen years. These four phases are: prosperity, recession, depression and recovery.
However, there is no agreement on the nature or causes of these cycles.
Even though, measurement and prediction of cyclical variation is very
important for strategic planning, the reliability of such measurements is highly
questionable due to the following reasons:
(i) These cycles do not occur at regular intervals. In the twenty-five years from
1956–1981 in America, it is estimated that the peaks in the cyclical activity
of the overall economy occurred in August 1957, April 1960, December 1969,
November 1973 and January 1980.1 This shows that they differ widely in
timing, intensity and pattern, thus making reliable evaluation of trends very
difficult.
(ii) The cyclic variations are affected by many erratic, irregular and random forces
which cannot be isolated and identified separately, nor can their impact be
measured accurately.
The cyclic variation for revenues in an industry against time is shown graphically
as follows:

3. Seasonal Variation (S)

Seasonal variation involves patterns of change that repeat over a period of one
year or less. Then they repeat from year to year and they are brought about by
fixed events. For example, sales of consumer items increase prior to Diwali due
to gift giving tradition. The sale of automobiles are much higher during the last
three to four months of the year due to the introduction of new models. This data
may be measured monthly or quarterly.
Since these variations repeat during a period of twelve months, they can be
predicted fairly and accurately. Some factors that cause seasonal variations are as
follows:
1
Mark L. Berenson and David M. Levine, Basic Business Statistics (New Jersey: Prentice-Hall, 1983),
618.

Statistics for Management 151

Unit-5 Time Series Analysis

(i) Season and climate. Changes in the climate and weather conditions
have a profound effect on sales. For example, the sale of umbrellas in India is
always more during monsoons. Similarly, during winter, there is a greater demand
for woollen clothes and hot drinks, while during summer months there is an
increase in the sales of fans and air conditioners.
(ii) Customs and festivals. Customs and traditions affect the pattern of
seasonal spending. For example, Mother’s Day or Valentine’s Day in America see
increase in gift sales preceding these days. In India, festivals such as Baisakhi
and Diwali mean a big demand for sweets and candy. It is customary all over the
world to give presents to children when they graduate from high school or college.
Accordingly, the month of June, when most students graduate, is a time for the
increase of sale for presents befitting the young.
An accurate assessment of seasonal behaviour is an aid in business planning
and scheduling such as in the area of production, inventory control, personnel,
advertising, and so on. The seasonal fluctuations over four repeating quarters in
a given year for sale of a given item is illustrated as:

4. Irregular or Random Variation (I)

Irregular variations are accidental, random or simply due to chance factors. Thus,
they are wholly unpredictable. These fluctuations may be caused by such isolated
incidents as floods, famines, strikes or wars. Sudden changes in demand or a
breakthrough in technological development may be included in this category.
Accordingly, it is almost impossible to isolate and measure the value and the
impact of these erratic movements on forecasting models or techniques. This
phenomenon may be graphically shown as follows:

152 Statistics for Management

Time Series Analysis Unit-5

It is traditionally acknowledged that the value of the time series (Y) is a

function of the impact of variable trend (T), seasonal variation (S), cyclical variation
(C) and irregular fluctuation (I). These relationships may vary depending upon
assumptions and purposes. The effects of these four components might be additive,
multiplicative, or a combination thereof in a number of ways. However, the
traditional time series analysis model is characterized by multiplicative relationship,
so that:
Y = T × S× C× I
This model is appropriate for those situations where percentage changes
best represent the movement in the series and the components are not viewed as
absolute values but as relative values.
Another approach to define the relationship may be additive, so that:
Y=T+S+C+I
This model is useful when the variations in the time series are in absolute
values and can be separated and traced to each of these four parts and each part
can be measured independently.

5.4 FORECASTING WITH DECOMPOSED

COMPONENTS OF TIME SERIES

5.4.1 Trend Analysis

While chance variations are difficult to identify, separate, control or predict, a
more precise measurement of trend, cyclical effects and seasonal effects can be
made in order to make the forecasts more reliable. In this section, we discuss
techniques that would allow us to describe trend.
When a time series shows an upward or downward long-term linear trend,
then regression analysis can be used to estimate this trend and project the trends
into forecasting the future values of the variables involved. The equation for the
straight line used to describe the linear relationship between the independent
variable X and the dependent variable Y is:
Y = b0 + b1X
where, b0 = Intercept on the Y-axis and b1 = Slope of the straight line

Statistics for Management 153

Unit-5 Time Series Analysis

In time series analysis, the independent variable is time, so we will use the
symbol t in place of X and we will use the symbol Yt in place of Yc which we
have used previously.
Hence, the equation for linear trend is given as:
Yt = b0 + b1t
where, Yt = Forecast value of the time series in period t
b0 = Intercept of the trend line on Y-axis
b1 = Slope of the trend line
t = Time period
As discussed earlier, we can calculate the values of b0 and bl by the following
formulae:
n (ty) – ( t)( y)
b1 , and b0 y b1t
n( t 2 ) – ( t)2
where, y = Actual value of the time series in period time t
n = Number of periods
y
y = Average value of time series
n
t
t=
t = Average value of
n
Knowing these values, we can calculate the value of y.
Example 5.1:
A car fleet owner has 5 cars which have been in the fleet for several different years.
The manager wants to establish if there is a linear relationship between the age of
the car and the repairs in hundreds of dollars for a given year. This way, he can
predict the repair expenses for each year as the cars become older. The information
for the repair costs he had collected for the last year on these cars is as follows:
Car # Age (t) Repairs (Y)
1 1 4
2 3 6
3 3 7
4 5 7
5 6 9
The manager wants to predict the repair expenses for the next year for the
two cars that are 3 years old.
Solution:
The trend in repair costs suggests a linear relationship with the age of the
car, so that the linear regression equation is given as:
Yt b0 b1t

154 Statistics for Management

Time Series Analysis Unit-5

where n (ty) ( t)( y)

b1
n( t 2 ) ( t)2
and b0
y b1t
To calculate the various values, let us form a new table as follows:
Age of Car (t) Repair Cost (Y ) tY t2
1 4 4 1

3 6 18 9

3 7 21 9

5 7 35 25

6 9 54 36

Total 18 33 132 80
Knowing that n = 5, let us substitute these values to calculate the regression
coefficients b0 and b1.
5(132) (18)(33)
Then, b1
5(80) (18)2
660 – 594
=
400 – 324
66
= = 0.87
76
and b0 y b1t

y 33
where y= 6.6
n 5
t 18
and t= 3.6
n 5
Then, b0 6.6 0.87(3.6)
= 6.6 – 3.13
= 3.47
Hence, Yt 3.47 0.87t
The cars that are 3 years old now will be 4 years old next year, so that t =
4.
Hence, Y( 4) 3.47 0.87(4)

= 3.47 + 3.48
= 6.95

Statistics for Management 155

Unit-5 Time Series Analysis

Accordingly, the repair costs on each car that is 3 years old are expected to
be $695.00.

5.4.2 Smoothing Techniques

Smoothing techniques improve the forecasts of future trends provided that the
time series is fairly stable with no significant trend, cyclical or seasonal effect
and the objective is to smooth out the irregular component of the time series
through the averaging process. There are two techniques that are generally
employed for such smoothing.
1. Moving averages 2. Exponential smoothing
1. Moving averages. The concept of the moving averages is based on the idea
that any large irregular component of time series at any point in time will have
a less significant impact on the trend, if the observation at that point in time is
averaged with such values immediately before and after the observation under
consideration. For example, if we are interested in computing the three-period
moving average for any time period, then we will take the average of the value
in such time period, the value in the period immediately preceding it and the
value in the time period immediately following it. Let us illustrate this concept
with the help of an example.
Example 5.2: Let the following table represent the number of cars sold in the
first 6 weeks of the first two months of the year by a given dealer. Our objective
is to calculate the three-week moving average.
Week Sales
1 20
2 24
3 22
4 26
5 21
6 22
Solution: The moving average for the first three-week period is given as:
20 + 24 + 22 66
Moving average = 22
3 3
This moving average can then be used to forecast the sale of cars for week
4. Since the actual number of cars sold in week 4 is 26, we note that the error
in the forecast is (26 – 22) = 4.
The calculation for the moving average for the next three periods is done
by adding the value for week 4 and dropping the value for week 1, and taking
the average for weeks 2, 3 and 4. Hence,
24 + 22 + 26 72
Moving average = = = 24
3 3
Then, this is considered to be the forecast of sales for week 5. Since the
actual value of the sales for week 5 is 21, we have an error in our forecast of
(21 – 24) = – (3).
156 Statistics for Management
Time Series Analysis Unit-5

The next moving average for weeks 3 to 5, as a forecast for week 6 is

given as:
22 + 26 + 21 69
Moving average = = = 23
3 3
The error between the actual and the forecast value for week 6 is (22 – 23)
= – (1). (Since the actual value of the sales for week 7 is not given, there is no
need to forecast such values).
Our objective is to predict the trend and forecast the value of a given
variable in the future as accurately as possible so that the forecast is reasonably
free from random variations. To do that, we must have the sum of individual
errors, as discussed earlier, as little as possible. However, since errors are irregular
and random, it is expected that some errors would be positive in value and others
negative, so that the sum of these errors would be highly distorted and would be
closer to zero. This difficulty can be avoided by squaring each of the individual
forecast errors and then taking the average. Naturally, the minimum values of
these errors would also result in the minimum value of the ‘average of the sum
of squared errors’. This is shown as follows:
Week Time Series Value Moving Average Error Error Squared
1 20
2 24
3 22
4 26 22 4 16
5 21 24 –3 9
6 22 23 –1 1

Then the average of the sum of squared errors, also known as mean squared
error (MSE) is given as:
16 + 9 +1 26
MSE = = = 8.67
3 3
The value of MSE is an often-used measure of the accuracy of the forecasting
method, and the method which results in the least value of MSE is considered
more accurate than others. The value of MSE can be manipulated by varying the
number of data values to be included in the moving average. For example, if we
had calculated the value of MSE by taking 4 periods into consideration for
calculating the moving average, rather than 3, then the value of MSE would be
less. Accordingly, by using trial and error method, the number of data values
selected for use in forecasting would be such that the resulting MSE value would
be minimum.
2. Exponential smoothing. In the moving average method, each value contributes
equally towards the calculation of the moving average, irrespective of the number Exponential smooth-
of time periods taken into consideration. In most actual situations, this is not a ing: Each observation in
realistic assumption. Because of the dynamics of the environment over a period the moving average cal-
of time, it is more likely that the forecast for the next period would be closer to culation receives the
the most recent previous period than the more distant previous period, so that the same weight.

Statistics for Management 157

Unit-5 Time Series Analysis

more recent value should get more weight than the previous value, and so on.
The exponential smoothing technique uses the moving average with appropriate
weights assigned to the values taken into consideration in order to arrive at a
more accurate or smoothed forecast. It takes into consideration the decreasing
impact of the past time periods as we move further into the past time periods.
This decreasing impact as we move down into the time period is exponentially
distributed and hence, the name exponential smoothing.
In this method, the smoothed value for period t, which is the weighted
average of that period’s actual value and the smoothed average from the previous
period (t – 1), becomes the forecast for the next period (t + l). Then the exponential
smoothing model for time period (t + l) can be expressed as follows:
F(t 1) Yt (1 )Ft
where F(t + 1) = The forecast of the time series for period (t + 1)
Yt = Actual value of the time series in period t
= Smoothing factor (0 1)
Ft = Forecast of the time series for period t
The value of is selected by the decision-maker on the basis of the degree of
smoothing required. A small value of means a greater degree of smoothing. A
large value of means very little smoothing. When = 1, then there is no
smoothing at all so that the forecast for the next time period is exactly the same
as the actual value of times series in the current period. This can be seen by:
F(t 1) Yt (1 )Fr
when 1
F(t 1) Yt 0Ft Yt
The exponential smoothing approach is simple to use and once the value of
is selected, it requires only two pieces of information, namely Yt and Ft to
calculate F(t 1) .
To begin with the exponential smoothing process, we let Ft equal the actual
value of the time series in period t, which is Y1. Hence, the forecast for period
2 is written as:
F2 Y1 (1 ) F1
But since we have put F1=Y1 , hence,
F2Y1 (1 ) Y1
= Y1
Let us now apply exponential smoothing method to the problem of forecasting
car sales as discussed in the case of moving averages. The data once again is
given as follows:

158 Statistics for Management

Time Series Analysis Unit-5

Week Time Series Value (Yt)

1 20
2 24
3 22
4 26
5 21
6 22
Let 0.4
Since F2 is calculated earlier as equal to Y1 = 20, we can calculate the value
of F3 as follows:
F3 = 0 .4Y2 + (1 – 0.4) F2
Since F2 = Y1, we get
F3
0.4(24) 0.6(20) 9.6 12
= 21.6
Similar values can be calculated for subsequent periods, so that,
F4 = 0.4Y3 0.6F3
= 0.4(22) + 0.6(21.6)
= 8.8 + 12.96 = 21.76
F5 = 0.4Y4 + 0.6F4
= 0.4(26) + 0.6(21.76)
= 10.4 + 13.056 = 23.456
F6 = 0.4Y5 + 0.6F5
= 0.4(21) + 0.6(23.456)
= 8.4 + 14.07
= 22.47
and, F7 = 0.4Y6 + 0.6F6
= 0.4(22) + 0.6(22.47)
= 8.8 + 13.48
= 22.28
Now we can compare the exponential smoothing forecast value with the
actual values for the six time periods and calculate the forecast error.
Week Time Series Value Exponential Smoothing Error
(Yt ) Forecast Value (Ft ) (Yt – Ft)
1 20 – –
2 24 20.000 4.0
3 22 21.600 0.4
4 26 21.760 4.24
5 21 23.456 – 2.456
6 22 22.470 – 0.47

Statistics for Management 159

Unit-5 Time Series Analysis
(The value of F7 is not considered because the value of Y7 is not given).
Let us now calculate the value of MSE for this method with selected value
of = 0.4. From the previous table:
Forecast errors Squared Forecast Error
(Yt Ft ) (Yt Ft )
4 16
0.4 0.16
4.24 17.98
– 2.456 6.03
– 0.47 0.22
Total = 40.39
Then,
MSE = 40.39/5
= 8.08
The previous value of MSE was 8.67. Hence, the current approach is a
better one.
The choice of the value for is very significant. Let us look at the exponential
smoothing model again.
F(t 1) Yt (1 )Ft

Yt Ft Ft
Ft (Yt Ft )
where (Yt – Ft) is the forecast error during the time period t.
The accuracy of the forecast can be improved by carefully selecting the
value of . If the time series contains substantial random variability then a small
value of (known as smoothing factor or smoothing constant) is preferable. On
the other hand, a larger value of would be desirable for time series with
relatively little random variability (Yt – Ft).

5.4.3 Measuring Cyclical Effect

Cyclic variation, as we have discussed earlier, is a pattern that repeats over time
periods longer than one year. These variations are generally unpredictable in
relation to the time of occurrence, duration as well as amplitude. However, these
Cyclic variation: A
variations have to be separated and identified. The measure we use to identify
pattern that repeats over
time periods longer than cyclical variation is the percentage of trend and the procedure used is known as
one year. the residual trend.
As we have discussed earlier, there are four components of time series.
These are secular trend (T), seasonal variation (S), cyclical variation (C) and
irregular (or chance) variation (I). Since the time period considered for seasonal
variation is less than one year, it can be excluded from the study, because when
we look at time series consisting of annual data spread over many years, then
only the secular trend, cyclical variation and irregular variation are considered.

160 Statistics for Management

Time Series Analysis Unit-5

Since secular trend component can be described by the trend line (usually
calculated by line of regression), we can isolate cyclical and irregular components
from the trend. Furthermore, since irregular variation occurs by chance and cannot
be predicted or identified accurately, it can be reasonably assumed that most of
the variation in time series left unexplained by the trend component can be
explained by the cyclical component. In that respect, cyclical variation can be
considered as residual, once other causes of variation have been identified.
The measure of cyclic variation as percentage of trend is calculated as
follows:
1. Determine the trend line (usually by regression analysis).
2. Compute the trend value Yt for each time period (t) under consideration.
3. Calculate the ratio Y/Yt for each time period.
4. Multiply this ratio by 100 to get the percentage of trend, so that:
Y
100.
Percentage of trend =
Yt

Example 5.3:
The following is the data for energy consumption (measured in quadrillions of
BTU) in the United States from 1981–1986 as reported in the Statistical Abstracts
of the United States.
Year Time Period (t) Annual Energy
Consumption (Y)
1981 1 74.0
1982 2 70.8
1983 3 70.5
1984 4 74.1
1985 5 74.0
1986 6 73.9
Assuming a linear trend, calculate the percentage of trend for each year
(cyclical variation).
Solution:
First, we find the secular trend by the regression line method, which is
given by:
Yt b0 b1t

n (ty) ( t)( y)
where b1
n( t 2 ) ( t)2

and b0 y b1t

Statistics for Management 161

Unit-5 Time Series Analysis

Let us make a table for these values.

t Y tY t2
1 74.0 74.0 1
2 70.8 141.6 4
3 70.5 211.5 9
4 74.1 296.4 16
5 74.0 370.0 25
6 73.9 443.4 36
t = 21 Y 437.3 tY 1536.9 t2 91
Substituting these values we get,
6(1536.9) (21)(437.3)
b1
6(91) (21)2

9221.4 – 9183.3
=
546 – 441
38.1
= = 0.363
105
and b0 = y – b1t

y 437.3
where y 72.88
n 6
21
t 3.5
6
Hence, b0 72.88 0.363(3.5)
= 72.88 – 1.27
= 71.61
Then, Yt 71.61 0.363t
Calculating the value of Yt for each time period, we get the following table
for the percentage of trend (Y/Yt)100.
Time Period Energy Consumption Trend Percentage of Trend
(t) (Y) (Yt) (Y/Yt)100
1 74.0 71.97 102.82
2 70.8 72.34 97.87
3 70.5 72.70 96.97
4 74.1 73.06 101.42
5 74.0 73.43 100.77
6 73.9 73.79 100.15

162 Statistics for Management

Time Series Analysis Unit-5

The following graph shows the actual energy consumption (Y), trend line
(Yt) and the cyclical fluctuations above and below the trend line over the time
period (t) for 6 years.

Yt

Frequently, we draw a graph of cyclic variation as the percentage of trend.

This process eliminates the trend line and isolates the cyclical component of the
time series.
It must be understood that cyclical fluctuations are not accurately predictable,
and hence, we cannot predict the future cyclic variations based upon such past
cyclic variations.

The percentage of trend figures show that in 1981, the actual consumption
of energy was 102.82 per cent of expected consumption that year and in 1983,
the actual consumption was 96.97 per cent.

Statistics for Management 163

Unit-5 Time Series Analysis

g
s

164 Statistics for Management

Time Series Analysis Unit-5

5.5 LET US SUM UP

• Accurate forecasting is an essential element of planning of any organization
or policy. This requires studying previous performances in order to forecast
future activities.
• When a projection of the pattern of future economic activity is known and
the level of future business activity is understood, the desirability of an
alternative course of action and the selection of an optimum alternative can
be examined and forecast.
• The quality of such forecasts is strongly related to the relevant information
that can be extracted from past data.
• Hence, the time series analysis method helps in making accurate predictions
and also in situations where the future is expected to be similar to or at least
predictive from the past.

5.6 FURTHER READING

Chandan, J. S. Statistics for Business and Economics. New Delhi: Vikas Publishing
House Pvt. Ltd., 1998.
Gupta, S. C. Fundamentals of Statistics. New Delhi: Himalaya Publishing House,
2006.
Hooda, R. P. Statistics for Business and Economics. New Delhi: Macmillan India
Ltd., 2002.
Kothari, C. R. Quantitative Techniques. New Delhi: Vikas Publishing House Pvt.
Ltd., 1984.
Monga, G. S. Mathematics and Statistics for Economics. New Delhi: Vikas Publishing
House Pvt. Ltd., 2000.
M.Selvam 2014 Maritime Logistics and Documentation , Lakshmi Publications , New Delhi
M.Selvam 2013 Quality Management , Lakshmi Publications , New Delhi
Besterfield .H (2018) Total Quality Management

5.7 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress–1

1. The term trend means the general long-term movement in the time series
value of the variable (Y) over a fairly long period of time. Here, ‘Y’ stands for
such factors like sales, population and crime rate that we are interested in
evaluating for the future.
2. Regular swings or patterns that repeat over a long period of time are known
as cyclical fluctuations. These are usually unpredictable in relation to the
time of occurrence, the duration as well as the amplitude.
3. Factors like changes in climate and weather, and customs and traditions cause
seasonal variations.
Statistics for Management 165
Unit-5 Time Series Analysis

Check Your Progress–2

1. Those variations which are accidental, random or occur due to chance factors,
are known as irregular random variations.
2. The two methods adopted in smoothing techniques are:
(a) Moving Averages
(b) Exponential Smoothing
3. The methods used in measuring seasonal variation are:
(a) Simple Average Method
(b) Moving Average Method

5.8 POSSIBLE QUESTIONS

Short-Answer Questions
1. Differentiate between secular trend and cyclic fluctuation.
2. How is irregular variation caused?
3. Define seasonal variation.
4. What do you mean by trend analysis?
5. How will you measure cyclical effect?
6. Describe the simple average method of isolating seasonal fluctuations in
time series.

Long-Answer Questions
1. The following data shows the number of Lincoln Continental cars sold by a
dealer in Queens during the 12 months of 1994.
Month Number Sold
Jan 52
Feb 48
Mar 57
Apr 60
May 55
June 62
July 54
Aug 65
Sept 70
Oct 80
Nov 90
Dec 75

(a) Calculate the three-month moving average for this data.

(b) Calculate the five-month moving average for this data.
(c) Which one of these two moving averages is a better smoothing technique
and why?

166 Statistics for Management

Time Series Analysis Unit-5

2. The owner of six gasoline stations in New Jersey would like to have some
reasonable indication of future sales. He would like to use the moving average
method to forecast future sales. He has recorded the quarterly gasoline sales
(in thousands of gallons) for all his gas stations for the past three years. These
are shown in the following table.
Year Quarter Sales
1 1 38
2 58
3 80
4 30
2 1 40
2 60
3 50
4 55
3 1 50
2 45
3 80
4 70

(a) Calculate the three-quarter moving average.

(b) Calculate the five-quarter moving average.
(c) Plot the quarterly sales and also both the moving averages on the same
graph. Which of these two moving average seems to be a better
smoothing technique?
3. The following data represents the sales (in millions of dollars) in a large
department store for the twelve months of the last three years.
Month 1993 1994 1995
January 7.2 8.2 6.4
February 8.5 9.1 7.3
March 9.6 10.5 8.5
April 10.2 11.4 10.2
May 11.7 12.0 10.8
June 13.0 14.5 15.0
July 14.2 14.0 13.0
August 15.7 16.2 15.0
September 11.4 12.0 11.8
October 9.3 9.0 8.6
November 12.8 9.8 11.0
December 9.3 7.3 6.3
(a) Plot the data and check for any trends in the data.
(b) Compute a twelve-month moving average. Which components of the
time series do these values reflect?
4. The General Manager of NCO (National Computer Outlet) which is a chain
store in the North East wants to determine the sales trend for one of the stores
located in Flushing, Queens. He took the sales data for the last 20 years in
millions of dollars.

Statistics for Management 167

Unit-5 Time Series Analysis

The yearly data is given as follows:

Year Sales ($ million)
1975 3.5
1976 4.0
1977 5.7
1978 6.2
1979 5.0
1980 4.2
1981 3.4
1982 4.6
1983 5.6
1984 6.4
1985 7.2
1986 8.0
1987 6.9
1488 5.0
1989 4.2
1990 5.0
1991 6.2
1992 7.2
1993 8.2
1994 9.2

The General Manager asked the company statistician to calculate the five-
year moving average.
(a) If you were the company statistician, how would you present the results
to the General Manager?
(b) Calculate the seven-year moving average and explain the difference
between the five-year moving average and the seven-year moving
average to the General Manager as to which of the two smoothes the
data better.
(c) Plot the actual sales and the seven-year moving average sales on the
same graph and interpret it.
5. An economist has calculated the variable rate of return on money market
funds for the last twelve months as follows:
Month Rate of Return (%)
January 6.2
February 5.8
March 6.5
April 6.4
May 5.9
June 5.9
July 6.0
August 6.8
September 6.5
October 6.1
November 6.0
December 6.0

168 Statistics for Management

Time Series Analysis Unit-5

(a) Using a three-month moving average, forecast the rate of return for
next January.
(b) Using exponential smoothing method and setting, = 0.8, forecast the
rate of return for next January.
6. The Indian Motorcycle Company is concerned about declining sales in the
western region. The following data shows monthly sales (in millions of dollars)
of the motorcycles for the past twelve months.
Month Sales
January 6.5
February 6.0
March 6.3
April 5.1
May 5.6
June 4.8
July 4.0
August 3.6
September 3.5
October 3.1
November 3.0
December 3.0

(a) Plot the trend line and describe the relationship between sales and time.
(b) What is the average monthly change in sales?
(c) If the monthly sales fall below $2.4 million, then the West Coast office
must be closed. Is it likely that the office will be closed during the next
six months?
7. The chief economist for New York State Department of Commerce reported
the following seasonally adjusted values for the consumption of durable goods
for the last twelve-month period.
Month Index
January 119
February 114
March 115
April 116
May 120
June 113
July 115
August 112
September 117
October 116
November 121
December 124

(a) Develop the linear trend model.

(b) Calculate the r2 value.
(c) Based on the value of r2, does it appear to be a good fit?

Statistics for Management 169

Unit-5 Time Series Analysis

8. An institution dealing with pension funds is interested in buying a large block

of stock of Azumi Business Enterprises (ABE). The president of the institution
has noted down the dividends paid out on common stock shares for the last
ten years. This data is presented as follows:
Year Dividend ($)
1985 3.20
1986 3.00
1987 2.80
1988 3.00
1989 2.50
1990 2.10
1991 1.60
1992 2.00
1993 1.10
1994 1.00

(a) Plot the data.

(b) Determine the value of regression coefficients.
(c) Estimate the dividend expected in 1995.
(d) Calculate the points on the trend line for the years 1987 and 1991 and
plot the trend line.
9. Rinkoo Camera Corporation has ten camera stores scattered in five areas of
New York city. The president of the company wants to find out if there is any
connection between the sales price and the sales volume of Nikon F-1 camera
in the various retail stores. He assigns different prices of the same camera for
the different stores and collects data for a thirty-day period. The data is
presented as follows. The sales volume is in number of units and the price is
in dollars.
Store Price Volume
1 550 420
2 600 400
3 625 300
4 575 400
5 600 340
6 500 440
7 450 500
8 480 460
9 550 400
10 650 310

(a) Plot the data.

(b) Estimate the linear regression of sales on price.
(c) What effect would you expect on sales if the price of the camera in store
number 7 is increased to $530?
(d) Calculate the points on the trend line for stores 4 and 7 and plot the
trend line.

170 Statistics for Management

Appendix

APPENDIX: STATISTICAL TABLES

(I) Some binomial distributions: P(r r0 /n, p)
(II) Values of e– (for calculating Poisson probabilities)
(III) Exponential functions
(IV) Area under standard normal distribution
(V) Cumulative normal distribution
(VI) Some critical values of ‘t’
2
(VII) Some critical values of
(VIII) Significance points of the Variance-Ratio (1% points of F)
(IX) Significance points of the Variance-Ratio (5% points of F)
(X) (a) Some two digit random numbers
(b) Some four digit random numbers

5.9 LEARNING OUTCOMES

• The time series

• The components of time series
• The influence of time series analysis
• The different methods of measuring trend
• The different methods of measuring seasonal variations

Statistics for Management 171

 Appendix
TABLE I
Some Binomial Distributions: P(r r0/n, p)

p p p p
n r0 0.10 0.25 0.40 0.50

1 0 0.9000 0.7500 0.6000 0.5000

1 1.0000 1.0000 1.0000 1.0000

5 0 0.5905 0.2373 0.0778 0.0313

1 0.9185 0.6328 0.3370 0.1875
2 0.9914 0.8965 0.6826 0.5000
3 0.9995 0.9844 0.9130 0.8125
4 0.9999 0.9990 0.9898 0.9687
5 1.0000 1.0000 1.0000 1.0000

10 0 0.3487 0.0563 0.0060 0.0010

1 0.7361 0.2440 0.0463 0.0108
2 0.9298 0.5256 0.1672 0.0547
3 0.9872 0.7759 0.3822 0.1719
4 0.9984 0.9219 0.6330 0.3770
5 0.9999 0.9803 0.8337 0.6230
6 1.0000 0.9965 0.9452 0.8281
7 1.0000 0.9996 0.9877 0.9453
8 1.0000 1.0000 0.9983 0.9892
9 1.0000 1.0000 0.9999 0.9990
10 1.0000 1.0000 1.0000 1.0000

20 0 0.1216 0.0032 0.0000 0.0000

1 0.3917 0.0243 0.0005 0.0000
2 0.6768 0.0912 0.0036 0.0002
3 0.8669 0.2251 0.0159 0.0013
4 0.9567 0.4148 0.0509 0.0059
5 0.9886 0.6171 0.1255 0.0207
6 0.9975 0.7857 0.2499 0.0577
7 0.9995 0.8981 0.4158 0.1316
8 0.9999 0.9590 0.5955 0.2517
9 1.0000 0.9861 0.7552 0.4119
10 1.0000 0.9960 0.8723 0.5881
11 1.0000 0.9990 0.9433 0.7483
12 1.0000 0.9998 0.9788 0.8684
13 1.0000 1.0000 0.9934 0.9423
14 1.0000 1.0000 0.9983 0.9793
15 1.0000 1.0000 0.9996 0.9941
16 1.0000 1.0000 1.0000 0.9987
17 1.0000 1.0000 1.0000 0.9998
18 1.0000 1.0000 1.0000 1.0000
19 1.0000 1.0000 1.0000 1.0000
20 1.0000 1.0000 1.0000 1.0000

172 Statistics for Management

Appendix
TABLE II
–
Values of e (For computing Poission Probabilities)

e– e– e– e–
(Mean) (Mean) (Mean) (Mean)

0.1 0.90484 2.6 0.07427 5.1 0.00610 7.6 0.00050

0.2 0.81873 2.7 0.06721 5.2 0.00552 7.7 0.00045
0.3 0.74042 2.8 0.06081 5.3 0.00499 7.8 0.00041
0.4 0.67032 2.9 0.05502 5.4 0.00452 7.9 0.00037
0.5 0.60653 3.0 0.04979 5.5 0.00409 8.0 0.00034

0.6 0.54881 3.1 0.04505 5.6 0.00370 8.1 0.00030

0.7 0.49659 3.2 0.04076 5.7 0.00335 8.2 0.00027
0.8 0.44933 3.3 0.03688 5.8 0.00303 8.3 0.00025
0.9 0.40657 3.4 0.03337 5.9 0.00274 8.4 0.00022
1.0 0.36788 3.5 0.03020 6.0 0.00248 8.5 0.00020

1.1 0.33287 3.6 0.02732 6.1 0.00224 8.6 0.03018

1.2 0.30119 3.7 0.02472 6.2 0.00203 8.7 0.00017
1.3 0.27253 3.8 0.02237 6.3 0.00184 8.8 0.00015
1.4 0.24660 3.9 0.02024 6.4 0.00166 8.9 0.00014
1.5 0.22313 4.0 0.01832 6.5 0.00150 9.0 0.00012

1.6 0.20190 4.1 0.01657 6.6 0.00136 9.1 0.00011

1.7 0.18268 4.2 0.01500 6.7 0.00123 9.2 0.00010
1.8 0.16530 4.3 0.01357 6.8 0.00111 9.3 0.00009
1.9 0.14957 4.4 0.01228 6.9 0.00101 9.4 0.00008
2.0 0.13534 4.5 0.01111 7.0 0.00091 9.5 0.00007

2.1 0.12246 4.6 0.01005 7.1 0.00083 9.6 0.00007

2.2 0.1 1080 4.7 0.00910 7.2 0.00075 9.7 0.00006
2.3 0.10026 4.8 0.00823 7.3 0.00068 9.8 0.00006
2.4 0.09072 4.9 0.00745 7.4 0.00061 9.9 0.00005
2.5 0.08208 5.0 0.00674 7.5 0.00055 10.0 0.00005

Statistics for Management 173

 Appendix
TABLE III
Exponential Functions

x ex e–x x ex e–x
0.0 1.000 1.000
0.1 1.105 0.905 5.1 164.0 0.0061
0.2 1.221 0.819 5.2 181.3 0.0055
0.3 1.350 0.741 5.3 200.3 0.0050
0.4 1.492 0.670 5.4 221.4 0.0045
0.5 1.649 0.607 5.5 244.7 0.0041
0.6 1.822 0.549 5.6 270.4 0.0037
0.7 2.014 0.497 5.7 298.9 0.0033
0.8 2.226 0.449 5.8 330.3 0.0030
0.9 2.460 0.407 5.9 365.0 0.0027
1.0 2.718 0.368 6.0 403.4 0.0025
1.1 3.004 0.333 6.1 445.9 0.0022
1.2 3.320 0.301 6.2 492.8 0.0020
1.3 3.669 0.273 6.3 544.6 0.0018
1.4 4.055 0.247 6.4 601.8 0.0017
1.5 4.482 0.223 6.5 665.1 0.0015
1.6 4.953 0.202 6.6 735.1 0.0014
1.7 5.474 0.183 6.7 812.4 0.0012
1.8 6.050 0.165 6.8 897.8 0.001 1
1.9 6.686 0.150 6.9 992.3 0.0010
2.0 7.389 0.135 7.0 1096.6 0.0009
2.1 8.166 0.122 7.1 1212.0 0.0008
2.2 9.025 0.111 7.2 1339.4 0.0007
2.3 9.974 0.100 7.3 1480.3 0.0007
2.4 11.023 0.091 7.4 1636.0 0.0006
2.5 12.18 0.082 7.5 1808.0 0.00055
2.6 13.46 0.074 7.6 1998.2 0.00010
2.7 14.88 0.067 7.7 2208.3 0.00045
2.8 16.44 0.061 7.8 2440.6 0.00041
2.9 18.17 0.055 7.9 2697.3 0.00037
3.0 20.09 0.050 8.0 2981.0 0.00034
3.1 22.20 0.045 8.1 3294.5 0.00030
3.2 24.53 0.041 8.2 3641.0 0.00027
3.3 27.11 0.037 8.3 4023.9 0.00025
3.4 29.96 0.033 8.4 4447.1 0.00022
3.5 33.12 0.030 8.5 4914.8 0.00020
3.6 36.60 0.027 8.6 5431.7 0.00018
3.7 40.45 0.025 8.7 6002.9 0.00017
3.8 44.70 0.022 8.8 6634.2 0.00015
3.9 49.40 0.020 8.9 7332.0 0.00014
4.0 54.60 0.018 9.0 8103.1 0.00012
4.1 60.34 0.017 9.1 8955.3 0.00011
4.2 66.69 0.015 9.2 9897.1 0.00010
4.3 73.70 0.014 9.3 10938 0.00009
4.4 81.45 0.012 9.4 12088 0.00008
4.5 90.02 0.011 9.5 13360 0.00007
4.6 99.48 0.010 9.6 14765 0.00007
4.7 109.95 0.009 9.7 16318 0.00006
4.8 121.51 0.008 9.8 18034 0.00006
4.9 134.29 0.007 9.9 19930 0.00005
5.0 148.4 0.0067

174 Statistics for Management

Appendix
TABLE IV
Area Under Standard Normal Distribution
Between the Mean and Successive Value of Z

Z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0.0 0.0000 0.0040 0.0080 0.0120 0.0160 0.0199 0.0239 0.0279 0.0319 0.0359
0.1 0.0398 0.0438 0.0478 0.0517 0.0557 0.0596 0.0636 0.0675 0.0714 0.0753
0.2 0.0793 0.0832 0.0871 0.0910 0.0948 0.0987 0.1026 0.1064 0.1103 0.1141
0.3 0.1179 0.1217 0.1255 0.1293 0.1331 0.1368 0.1406 0.1443 0.1480 0.1517
0.4 0.1554 0.1591 0.1628 0.1664 0.1700 0.1736 0.1772 0.1808 0.1844 0.1879
0.5 0.1915 0.1950 0.1985 0.2019 0.2054 0.2088 0.2123 0.2157 0.2190 0.2224

0.6 0.2257 0.2291 0.2324 0.2357 0.2389 0.2422 0.2454 0.2486 0.2517 0.2549
0.7 0.2580 0.2611 0.2642 0.2673 0.2704 0.2734 0.2764 0.2794 0.2823 0.2852
0.8 0.2881 0.2910 0.2939 0.2967 0.2995 0.3023 0.3051 0.3078 0.3106 0.3133
0.9 0.3159 0.3186 0.3212 0.3238 0.3264 0.3289 0.3315 0.3340 0.3365 0.3389
1.0 0.3413 0.3438 0.3461 0.3485 0.3508 0.3531 0.3554 0.3577 0.3599 0.3621

1.1 0.3643 0.3665 0.3686 0.3708 0.3729 0.3749 0.3770 0.3790 0.3810 0.3830
1.2 0.3849 0.3869 0.3888 0.3907 0.3925 0.3944 0.3962 0.3980 0.3997 0.4015
1.3 0.4032 0.4049 0.4066 0.4082 0.4099 0.4115 0.4131 0.4147 0.4162 0.4177
1.4 0.4192 0.4207 0.4222 0.4236 0.4251 0.4265 0.4279 0.4292 0.4306 0.4319
1.5 0.4332 0.4345 0.4357 0.4370 0.4382 0.4394 0.4406 0.4418 0.4429 0.4441

1.6 0.4452 0.4463 0.4474 0.4484 0.4495 0.4505 0.4515 0.4525 0.4535 0.4545
1.7 0.4554 0.4564 0.4573 0.4582 0.4591 0.4599 0.4608 0.4616 0.4625 0.4633
1.8 0.4641 0.4649 0.4656 0.4664 0.4671 0.4678 0.4686 0.4693 0.4699 0.4706
1.9 0.4713 0.4719 0.4726 0.4732 0.4738 0.4744 0.4750 0.4756 0.4761 0.4767
2.0 0.4772 0.4778 0.4783 0.4788 0.4793 0.4798 0.4803 0.4808 0.4812 0.4817

2.1 0.4821 0.4826 0.4830 0.4834 0.4838 0.4842 0.4846 0.4850 0.4854 0.4857
2.2 0.4861 0.4864 0.4868 0.4871 0.4875 0.4878 0.4881 0.4884 0.4887 0.4890
2.3 0.4893 0.4896 0.4898 0.4901 0.4804 0.4906 0.4909 0.4911 0.4913 0.4916
2.4 0.4918 0.4920 0.4922 0.4925 0.4927 0.4929 0.4931 0.4932 0.4934 0.4936
2.5 0.4938 0.4940 0.4941 0.4943 0.4945 0.4946 0.4948 0.4949 0.4951 0.4952
2.6 0.4953 0.4955 0.4956 0.4957 0.4959 0.4960 0.4961 0.4962 0.4963 0.4964
2.7 0.4965 0.4966 0.4967 0.4968 0.4969 0.4970 0.4971 0.4972 0.4973 0.4974
2.8 0.4974 0.4975 0.4976 0.4977 0.4977 0.4978 0.4979 0.4979 0.4980 0.4981
2.9 0.4981 0.4982 0.4982 0.4983 0.4984 0.4984 0.4985 0.4985 0.4986 0.4986
3.0 0.4987 0.4987 0.4987 0.4988 0.4988 0.4989 0.4989 0.4989 0.4990 0.4990

Statistics for Management 175

176

TABLE V
Cumulative Normal Distribution
[Extract from the Table Concerning the Area under the Normal Curve]

Z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0.0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7703 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133
0.9 0.8159 0.8186 0.8212 0.8238 0.8264 0.8289 0.8315 0.8340 0.8365 0.8389
1.0 0.8413 0.8438 0.8461 0.8485 0.8508 0.8531 0.8454 0.8577 0.8599 0.8621
1.1 0.8643 0.8665 0.8686 0.8708 0.8729 0.8749 0.8770 0.8790 0.8810 0.8830
1.2 0.8849 0.8869 0.8888 0.8907 0.8925 0.8944 0.8962 0.8980 0.8927 0.90147
1.3 0.9032 0.9049 0.90658 0.90824 0.90988 0.91149 0.91309 0.91466 0.91621 0.91774
1.4 0.91924 0.92073 0.92220 0.92364 0.92507 0.92647 0.92785 0.92922 0.93056 0.93189
1.5 0.93319 0.93448 0.93574 0.93699 0.93822 0.93943 0.94062 0.94179 0.94295 0.99406
1.6 0.94520 0.94630 0.94738 0.94845 0.94950 0.95053 0.95154 0.95254 0.95352 0.95449
1.7 0.95543 0.95637 0.95728 0.95818 0.95907 0.95994 0.96080 0.96164 0.96246 0.96327
1.8 0.96407 0.96485 0.96562 0.96638 0.96712 0.96784 0.96851 0.96926 0.96995 0.97062
1.9 0.97128 0.97193 0.97257 0.97320 0.97381 0.97441 0.97500 0.97558 0.97615 0.97670
2.0 0.97725 0.97778 0.97831 0.97882 0.97932 0.97982 0.98030 0.98077 0.98125 0.98169
2.1 0.98214 0.98257 0.98300 0.98341 0.98382 0.98422 0.98461 0.98500 0.98537 0.98574
2.2 0.98610 0.98645 0.98679 0.98713 0.98745 0.98773 0.98809 0.98840 0.98870 0.98899
2.3 0.98928 0.98956 0.98983 0.990097 0.990358 0.090613 0.990863 0.991106 0.991344 0.991576
2.4 0.991802 0.992024 0.992240 0.992451 0.992656 0.992857 0.993053 0.993244 0.993431 0.993613

Note: When Z = 4.09, the table value is 0.99997843 (i.e. = 1 for all practical purposes).
(This table contains the area of left half of the curve (i.e., 0.5) plus the relevant area of the right half for different values of ‘Z’)
Statistics for Management

Appendix
Appendix
TABLE VI
Some Critical Values of ‘t ’

Level of Significance

Degrees of Freedom 1% 5% 10%

1 63.657 12.706 6.314

2 9.925 4.303 2.920
3 5.841 3.182 2.353
4 4.604 2.776 2.132
5 4.032 2.571 2.015

6 3.707 2.447 1.943

7 3.499 2.365 1.895
8 3.355 2.306 1.860
9 3.250 2.262 1.833
10 3.169 2.228 1.812

11 3.106 2.201 1.796

12 3.055 2.179 1.782
13 3.012 2.160 1.771
14 2.977 2.145 1.761
15 2.947 2.131 1.753

16 2.921 2.120 1.746

17 2.898 2.110 1.740
18 2.878 2.101 1.734
19 2.861 2.093 1.729
20 2.845 2.086 1.725

21 2.831 2.080 1.721

22 2.819 2.074 1.717
23 2.807 2.069 1.714
24 2.797 2.064 1.711
25 2.787 2.060 1.708

26 2.779 2.056 1.706

27 2.771 2.052 1.703
28 2.763 2.048 1.701
29 2.756 2.045 1.699
2.576 1.960 1.645

Note: These table values of ‘t’ are in respect of two-tailed tests. If we use the t-distribution for one-
tailed test then we are interested in determining the area located in one tail. So to find the
appropriate t-value for a one-tailed test say at a 5% level with 12 degrees of freedom, then we
should look in the above table under the 10% column opposite the 12 degrees of freedom row.
(This value will be 1.782). This is true because the 10% column represents 10% of the area under
the curve contained in both tails combined, and so it also represents 5% of the area under the
curve contained in each of the tails separately.

Statistics for Management 177

 Appendix
TABLE VII
2
Some Critical Values of for Specified Degrees of Freedom

Level of Significance
Degrees of
Freedom 10% 5% 1%

1 2.706 3.841 6.635

2 4.605 5.991 9.210
3 6.251 7.815 11.345
4 7.779 9.488 13.277
5 9.236 11.071 15.086

6 10.645 12.592 16.812

7 12.017 14.067 18.475
8 13.362 15.507 20.090
9 14.684 16.919 21.666
10 15.987 18.307 23.209

11 17.275 19.675 24.725

12 18.549 21.026 26.217
13 19.812 22.362 27.688
14 21.064 23.685 29.141
15 22.307 24.996 30.578

16 23.542 26.296 32.000

17 24.769 27.587 33.409
18 25.989 28.869 34.805
19 27.204 30.144 36.191
20 28.412 31.410 37.566

21 29.615 32.671 38.932

22 30.813 33.924 40.289
23 33.007 35.172 41.638
24 33.196 36.415 42.980
25 34.382 37.652 44.314

26 35.363 38.885 45.642

27 36.741 40.113 46.963
28 37.916 41.337 48.278
29 38.087 42.567 49.588
30 40.256 43.773 50.892

Note: For degrees of freedom greater than 30, the quantity 2x2 2v 1 may be used as a
normal variate with unit variance.

178 Statistics for Management

Appendix
TABLE VIII
Significance Points of the Variance-Ratio ‘F’
1 per cent Points of F
v 1 2 3 4 5 6 8 12 24
1
v
2

1 4052 5000 5403 5625 5764 5859 5982 6106 6235 6366
2 98.50 99.00 99.17 99.25 99.30 99.33 99.37 99.42 99.46 99.50
3 34.12 30.82 29.46 28.71 28.24 27.91 27.49 27.05 26.60 26.13
4 21.20 18.20 16.69 15.88 15.52 15.21 14.80 14.37 13.93 13.45

5 16.26 13.27 12.06 11.39 10.97 10.67 10.29 9.89 9.47 9.02
6 13.75 10.92 9.78 9.15 8.75 8.47 8.10 7.72 7.31 6.88
7 12.25 9.55 8.45 7.85 7.46 7.19 6.84 6.47 6.07 5.65
8 11.26 8.65 7.59 7.01 6.63 6.37 6.03 5.67 5.28 4.86
9 10.56 8.02 6.99 6.42 6.06 5.80 5.47 5.12 4.73 4.31

10 10.04 7.56 6.55 5.99 5.64 5.39 5.06 4.71 4.33 3.91
11 9.65 7.21 6.22 5.87 5.32 5.07 4.74 4.40 4.02 3.60
12 9.33 6.93 5.95 5.41 5.06 4.82 4.50 4.16 3.78 3.36
13 9.07 6.70 5.74 5.21 4.86 4.62 4.30 3.96 3.59 3.17
14 8.86 6.51 5.56 5.04 4.69 4.46 4.14 3.80 3.43 3.00

15 8.68 6.36 4.42 4.89 4.56 4.32 4.00 3.67 3.29 2.87
16 8.53 6.23 5.29 4.77 4.44 4.20 3.89 3.55 3.18 2.75
17 8.40 6.11 5.18 4.67 4.34 4.10 3.79 3.46 3.08 2.65
18 8.29 6.01 5.09 4.58 4.25 4.01 3.71 3.37 3.00 2.59
19 8.18 5.93 5.01 4.50 4.17 3.94 3.63 3.30 3.92 2.49

20 8.10 5.85 4.94 4.43 4.10 3.87 3.56 3.23 2.86 2.42
21 8.02 5.78 4.87 4.37 4.04 3.81 3.51 3.17 2.80 2.36
22 7.95 5.72 4.82 4.31 3.99 3.76 3.45 3.12 2.75 2.31
23 7.88 5.66 4.76 4.26 3.94 3.71 3.41 3.07 2.70 2.26
24 7.82 5.61 4.72 4.22 3.90 3.67 3.36 3.03 2.66 2.21

25 7.77 5.57 4.68 4.18 3.85 3.63 3.32 2.99 2.62 2.17
26 7.72 5.53 4.64 4.14 3.82 3.59 3.20 2.96 2.58 2.10
27 7.68 5.49 4.60 4.11 3.78 3.56 3.26 2.93 2.45 2.13
28 7.64 5.45 4.57 4.07 3.75 3.53 3.23 2.90 2.52 2.06
29 7.60 5.42 4.54 4.04 3.73 3.50 3.20 2.87 2.49 2.03

30 7.56 5.39 4.51 4.02 3.70 3.47 3.17 2.84 2.47 2.01
40 7.31 5.18 4.31 3.83 3.51 3.29 2.99 2.66 2.29 1.80
60 7.08 4.98 4.13 3.65 3.34 3.12 2.82 2.50 2.12 1.60
120 6.85 4.79 3.95 3.48 3.17 2.96 2.66 2.34 1.95 1.38
6.64 4.60 3.78 3.32 3.02 2.80 2.51 2.18 1.79 1.00

v1 = Degrees of freedom for greater variance.

v2 = Degrees of freedom for smaller variance.

Statistics for Management 179

 Appendix
TABLE IX
Significance Points of the Variance-Ratio ‘F’
5 per cent Points of F

v1 1 2 3 4 5 6 8 12 24
v2

1 161.4 199.5 215.7 224.6 230.2 234.0 238.9 243.9 249.0 254.3
2 18.51 19.00 19.16 19.25 19.30 19.33 19.37 19.41 19.45 19.50
3 10.13 9.55 9.28 9.12 9.01 8.94 8.84 8.74 8.64 8.53
4 7.71 6.94 6.59 6.39 6.26 6.16 6.04 5.91 5.77 5.63
5 6.61 5.79 5.41 5.19 5.05 4.95 4.82 4.68 4.53 4.36

6 5.99 5.14 4.76 4.53 4.39 4.28 4.15 4.00 3.84 3.67
7 5.59 4.74 4.35 4.12 3.97 3.87 3.73 3.57 3.41 3.23
8 5.32 4.46 4.07 3.84 3.69 3.58 3.44 3.28 3.12 3.93
9 5.12 4.26 3.86 3.63 3.48 3.37 3.23 3.07 2.90 2.71
10 4.96 4.10 3.71 3.48 3.33 3.22 3.07 2.91 2.74 2.54

11 4.84 3.98 3.59 3.36 3.20 3.09 2.95 2.79 2.61 2.40
12 4.75 3.88 3.49 3.26 3.11 3.00 2.85 2.69 2.50 2.30
13 4.67 3.80 3.41 3.18 3.02 3.92 2.77 2.60 2.42 2.21
14 4.60 3.74 3.34 3.11 2.96 2.85 2.70 2.53 2.35 2.13
15 4.54 3.68 3.29 3.06 2.90 2.79 2.64 2.48 2.29 2.07

16 4.49 3.63 3.24 3.01 2.85 2.74 2.59 2.42 2.24 2.01
17 4.45 3.59 3.20 2.96 2.81 3.70 2.55 2.38 2.19 1.96
18 4.41 3.55 3.16 2.93 2.77 2.66 2.51 2.34 2.15 1.92
19 4.38 3.52 3.13 2.90 2.74 2.63 2.48 2.31 2.11 1.88
20 4.35 3.49 3.10 2.87 2.71 2.60 2.45 2.28 2.08 1.84

21 4.32 3.47 3.07 2.84 2.68 2.57 2.42 2.25 2.05 1.81
22 4.30 3.44 3.05 2.82 2.66 2.55 2.40 2.23 2.03 1.78
23 4.28 3.42 3.03 2.80 2.64 2.53 2.38 2.20 2.00 1.76
24 4.26 3.40 3.01 2.78 2.62 2.51 2.36 2.18 1.98 1.73
25 4.24 3.38 2.99 2.76 2.60 2.49 2.34 2.16 1.96 1.71

26 4.22 3.37 2.98 2.74 2.59 2.47 2.32 2.15 1.95 1.69
27 4.21 3.35 2.96 2.73 2.57 2.46 2.30 2.13 1.93 1.67
28 4.20 3.34 2.95 2.71 2.56 2.44 2.29 2.12 1.91 1.65
29 4.18 3.33 2.93 2.70 2.54 2.43 2.28 2.10 1.90 1.64
30 4.17 3.32 2.92 2.69 2.53 2.42 2 .27 2.09 1.89 1.62

40 4.08 3.23 2.84 2.61 2.45 2.34 2.18 2.00 1.79 1.51
60 4.00 3.15 2.76 2.52 2.37 2.25 2.10 1.92 1.70 1.89
120 2.92 3.07 2.68 2.45 2.29 2.17 2.02 1.83 1.61 1.25
3.84 2.99 2.60 2.37 2.21 2.09 1.94 1.75 1.52 1.00

v1 = Degrees of freedom for greater variance.

v2 = Degrees of freedom for smaller variance.

180 Statistics for Management

Appendix
TABLE X
(a) Some Two Digit Random Numbers

(Extract of 80 random numbers lying between 00 and 99 from R.A. Fisher and F. Yates’ table of
random numbers)

44 16 09 83 19
84 07 99 11 32
82 77 81 45 14
50 66 97 65 31
85 50 30 34 96
40 20 36 89 03
96 50 75 12 93
88 95 72 64 16
16 58 79 86 62
16 44 83 46 24
97 77 07 32 08
92 11 00 76 38
39 08 42 07 88
33 38 13 51 74
83 87 97 25 47
42 45 16 36 00

(b) Some Four Digit Random Numbers

2952 9525 2924 0957 8011

3170 2693 1458 6247 2075
7203 6107 5476 3790 1836
3408 9025 1952 3895 1625
0560 7979 8875 5627 4223
6641 1545 3207 5841 3772
5624 2370 0924 2188 4660
5356 6913 3453 4996 9784
2769 6008 8459 9631 6470
5246 5911 7724 8409 5630
3992 1396 9937 6706 6726
4167 7483 3621 4423 7538
1300 7691 9190 6977 1005
3563 8126 4989 5547 9819
1112 3581 8591 8513 4135
9792 2011 2379 2455 1728

Statistics for Management 181