UNIT – 1

BUSINESS STATISTICS

What is Statistics?

The word “Statistics” has been derive from the Latin word “Status” or Italian word “Statista” or
German word “Statistika”. Each of these words means Political State. Initially, Statistics was
used to collect the information of the people of the state about their income, health, illiteracy and
wealth etc.

But now a day, Statistics has become an important subject having useful application in various
fields in day to day life.

Statistics in Plural Sense:--

In the plural sense, Statistics refers to information in terms of numbers or numerical data such as
Population Statistics, Employment Statistics etc. However any numerical information is not
statistics.

Example: Ram gets 繁 100 per month as pocket allowance is not Statistics. It is neither an
aggregate nor an average. Whereas average pocket allowance of the students of Class X is 繁 100
per month and there are 80 students in class XI & 8 students in Class XII are Statistics.

The following table shows a set of data that which is Statistics and which is not Statistics.

Data which are not Statistics Data which are Statistics

A cow has 4 legs. Average height of the 26 plus male people in
India is 6 feet compare to 5 feet in Nepal.
Ram has 200 rupees in his pocket. Birth rate in India is 18 per thousand compare
to 8 per thousand in USA.
A young lady was run over by a speeding truck Over the past 10 years, India has won 60 test
at 100 km per hour. matches in cricket and lost 50.

From above information we can say that “All Statistics are data, but all data are not
Statistics”

Definition:-

According to Bowley - “Statistics are numerical statements of facts in any department of

enquiry placed in relation to each other.”
According to Yule and Kendall ----- “By Statistics we mean quantitative data affected to
marked extent by multiplicity of causes.”

Characteristics of Statistics in Plural Sense

Main characteristics of Statistics in terms of numerical data are as follows:

(1) Aggregate of Facts – A single number does not constitute Statistics. We can not draw
any conclusion from single number. We can draw any conclusion by the aggregate
number of facts.
For example, if it is stated that there are 1,000 students in our college then it has no
significance. But if it is stated that there are 300 students in arts, 400 students in
commerce and 300 in science in our college. It makes statistical sense as this data convey
statistical information. Similarly if it is stated that population of India is 130 crore or the
value of total exports from India is 繁 11, 66,439 crore then these aggregate of facts will
be termed as Statistics.
(2) Numerically Expressed - Statistics are expressed in terms of numbers. Qualitative
aspects like small or big, rich or poor etc. are not statistics. For instance if we say that
Irfan Pathan is tall Sachin is short then this statement has no statistical sense. However if
it is stated that height of Irfan Pathan is 6 ft and 2 inch and the height of Sachin is 5 ft and
4 inch then these numerical will be called Statistics.
(3) Affected by Multiplicity of Causes – Statistics are not affected by any single factor but
it is affected by many factors. For instance 30% rise in prices may have been due to
several causes like reduction in supply, increase in demand, shortage of power, rise in
wages, rise in taxes, etc.
(4) Reasonable Accuracy - A reasonable degree of accuracy must be kept in view while
collecting statistical data. This accuracy depends on the purpose of investigation, its
nature, size and available resources.
(5) Pre-determined Purpose - Statistics are collected with some pre-determined
objective. Any information collected without any definite purpose will only be a
numerical value and not Statistics. If data pertaining to the farmers of a village is
collected, there must be some pre-determined objective. Whether the statistics are
collected for the purpose of knowing their economic position or distribution of land
among them or their total population. All these objectives must be pre – determined.
(6) Collected in a Systematic Manner – Statistics should be collected in a systematic
manner. Before collecting the data, a plan must be prepared. No conclusion can be drawn
from data collected in haphazard manner. For instance, data regarding the marks secured
by the students of a college without any reference to the class, subject, examination, or
maximum marks, etc will lead no conclusion.
 Statistics in Singular Sense

In a singular sense, statistics means science of statistics or statistical methods. It refers

to techniques or methods relating to collection, classification, presentation, analysis
and interpretation of quantitative data.
Definition
------Statistics may be defined as the collection, presentation, analysis and interpretation
of numerical data.----------------------------------CROXTON AND COWDEN
------Statistics is the science which deals with the collection, classification and tabulation
of numerical facts as a basis for the explanation, description and comparison of
phenomena.------------------------------------------LOVITT

Subject Matter of Statistics

Subject matter of statistics includes two components:

1. Descriptive Statistics 2. Inferential Statistics
1. Descriptive Statistics: Descriptive Statistics refers to those methods which are used
for the collection, presentation as well as analysis of data. These methods relate
measurement of central tendencies, measurement of dispersion, measurement of
correlation etc. For Example: Descriptive statistics is used when you estimate
average height of the secondary students in your school. Descriptive statistics is also
used when you find the marks in science and mathematics of the students in all
classes are intimately related to each other.
2. Inferential Statistics: Inferential Statistics refers to all such methods by which
conclusion are drawn related to the universe or population on the basis of a given
sample. For example: If your class teacher estimate average weight of the entire
class on the basis of average weight of only a sample of sample of students of the
class then we use the inferential statistics.

Important terminology in statistics

1. Population: By population we mean a well defined set or group of all the objects for a
particular study. The objects may be persons, plants, books, fishes in ponds, shops etc.
the population will consist of certain elements like the plants of a certain kind in a
specified field, the fishes in a pond, the unemployed person in India, books in library and
so on. For instance, if we want to study the properties of students in a school then the
population consists of all the students of school. For instance if we want to study about
the books in a library then the population includes all the books of the library etc. if the
number of elements are limited then the population is finite. On the other hand if the
 number of elements is not limited then the population is infinite. Mostly we deal with
finite population.

2. Sample: It is a part of the population selected by some sampling procedure. The process
of selection of sample is known as sampling. The number of objects in the sample is
called the size of the sample. It is believed that a sample is best representative of the
population.

For instance, suppose a research worker is required to study the weight of fishes in a
pond after a particular period of growth. For this purpose suppose that there are 3,000
fishes in the pond, he may either measure the weight of all the fishes in the pond or he
may decide to select a small group of fishes and measure their weights. The first
approach of measuring the weight of all fishes is called complete enumeration or
census. Another approach in which only a small group of fishes is considered is called
sample survey. In brief we can say that in complete enumeration, information is
collected on all the units of the universe and in sample survey, only a part of the universe
is considered.
3. Variable: A property of objects is known as variable which differ from object to object
and is expressible numerically, in terms of numbers.
For instance: the marks in Mathematics of students in a class can be expressed in the
term of marks obtained by the students. So it is a variable property which is expressible
quantitatively.
4. Attribute: A property and characteristic of objects is known as attribute which are not
expressible quantatively in number. We can express the data qualitatively. For example,
smoking, color, honesty etc.

CHARACTERISTICS OF STATISTICS

1. Statistics are aggregate of facts.

2. Statistics are numerically expressed.
3. Statistics are affected to a marked extent by multiplicity of causes.
4. Statistics are either enumerated or estimated with reasonable standard of accuracy
5. Statistics are collected in a systematic manner.
6. Statistics are collected for a pre-determined purpose.
7. Statistics should be placed in relation to each other.

In the absence of the above characteristics numerical data can’t be called Statistics and hence “all
statistics are numerical statements of facts but all numerical statements of facts are not statistics.”

According to above Definitions, Statistics is both a science and an art. It is related to the
study and application of the principles and methods applicable in the collection,
 presentation, analysis, interpretation and forecasting of data. Or statistical facts
influenced by several factors and related to any area of knowledge or research so that
concrete and intelligent decisions may be taken in the phase of uncertainty

NATURE OF STATISTICS

Statistics as a science: science refers to a systematized body of knowledge. It studies cause and
effect relationship and attempts to make generalizations in the form of scientific principles or laws.
“Science, in short, is like a light house that gives light to the ships to find out their own way but
doesn’t indicate the direction in which they should go.” Like other sciences, Statistical Methods are
also used to answer the questions like, how an investigation should be conducted. In what way the
valid and reliable conclusions can be drawn? Statistics is called the science of scientific methods.

In words of Croxton and Cowden, “Statistics is not a science, it is scientific

methods.”According to Tippet, “as science, the statistical method is a part of the general scientific
method and is based on the same fundamental ideas and processes.”

Statistics as an art: we know that science is a body of systematized knowledge. How this
knowledge is to be used for solving a problem is work of an art. An art is an applied knowledge. It
refers the skill of handling facts so as to achieve a given objective. It is concerned with ways and
means of presenting and handling data, making inferences logically and drawing relevant conclusion.
Art aspects of statistics tell, ‘how to use statistical rules and principles to study the problems and
finding their solutions. ‘Collections of statistics (data) its use and utility are itself an art.

Statistics is both science and art: After studying science and art aspects of statistics, it is used
not only to gain knowledge but also to understand the facts and draw important conclusions from it. If
science is knowledge, then art is action. Looking from this angle statistics may also be regarded as an
art. It involves the application of given methods to obtain facts, derive results and finally to use them
for devising action.

STAGES IN A STATISTICAL INVESTIGATION 5 stages -

Collection Organisation Presentation

Analysis Interpretation

1. Collection: This is the primary step in a statistical study and data should be collected
with care by the investigator. If data are faulty, the conclusions drawn can never be
reliable. The data may be available from existing published or unpublished sources or
 else may be collected by the investigator himself. The first hand collection of data is one
of the most difficult and important tasks faced by a statistician.
2. Organization: Data collected from published sources are generally in organized form.
However, a large mass of figures that are collected from a survey frequently needs
organization. In organizing, there are 3 steps as
(A) Editing (B) Classify (C) Tabulation.
(A) Editing: The collected data must be editing very carefully so that the omissions,
inconsistencies irrelevant answers and wrong computation in the returns from a survey
may be corrected or adjusted.
(B) Classify: Classification is the process of arranging the data according to some common
characteristics possessed by the items constituting the data.
(C) Tabulation: To arrange the data in columns and rows.

Hence collected data is organized properly so that the desire information may be highlighted and
undesirable information avoided.

3. Presentation: Arranged data is not capable to influence a layman. Thus, it is necessary

that data may be presented with the help of tables, diagrams and graphs. By these devices
facts can be understood easily.
4. Analysis: A major part of it is developed to the methods used in analyzing the presented
data, mostly in a tabular form. For this analysis, a number of statistical tools are
available, such as averages, correlation, regression etc.
5. Interpretation: the interpretation of a data is a difficult task and necessitates a high
degree of skills and experience in the statistical investigation because certain decisions
made on the basis of conclusions drawn.

SCOPE OF STATISTICS

In early stages, the scope of statistics was very limited. It was confined mainly to the
administration of government and was, therefore, called the ‘Science of Kings’. But in modern time,
the scope of statistics has widened usually all those facts come in the purview of statistics, which are
expressed in quantitative terms directly or indirectly. That is why Croxton & Cowden observed,
“Today there is hardly a phase of endeavor which does not find statistical devices at least
occasionally useful.” It is not unfair to say, science without statistics bears no fruit and statistics
without science have no root.” The applications of statistics are so numerous that it is often
remarked, “Statistics is what statisticians do.” Now let us examine a few fields or areas in which
statistics is applied.
1. Statistics and the State: in recent years the functions of the state have increased
tremendously. The concept of the state has changed from that of simply maintaining law
and order to that of a welfare state. Statistical data and statistical methods are of great
help in promoting human welfare. The government in most countries is the biggest
collector and user of statistical data. These statistics help in framing suitable policies.
2. Statistics in Business and Management: with growing size and increasing competition,
the problems of business enterprises have become complex. Statistics is now considered
as an indispensable tool in the analysis of activities in the field of business, commerce
and industry. The object can be achieved by properly conducted market survey and
research which greatly depends on statistical methods. The trends in sales and production
can be determined by statistical methods like time-series analysis which are essential for
future planning of the phenomena. Statistical concepts and methods are also used in
controlling the quality of products to satisfaction of consumer and the producer. The
bankers use the objective analysis furnished by statistics and then temper their decisions
on the basis of qualitative information.
3. Statistics and Economics: R.A.Fisher complained of “the painful misapprehension that
statistics is a branch of economics.” Statistical Data and methods are of immense help in
the proper understanding of the economic problems and in the information of economic
policies. In the field of exchange, we study markets, law of prices based on supply and
demand, cost of production, banking and credit instruments etc. The development of
various economic theories own greatly to statistical methods, e.g., ‘Engel’s law of family
expenditure’, ‘Malthusian theory of population’. The impact of mathematics and statistics
has led to the development of new disciplines like ‘Econometrics’’ and ‘Economic
Statistics’. In fact, the concept of planning so vital for growth of nations would not have
been possible in the absence of data and proper statistical analysis.
4. Statistics and Psychology and Education: Statistics has found wide application in
psychology and education. Statistical methods are used to measure human ability such as;
intelligence, aptitude, personality, interest etc. by tests. Theory of learning is also based
on Statistical Principles. Applications of statistics in psychology and education have led
to the development of new discipline called ‘Psychometric’.
5. Statistics and Natural science; Statistical techniques have proved to be extremely useful
in the study of all natural sciences like biology, medicine, meteorology, botany etc. for
example- in diagnosing the correct disease the doctor has to rely heavily on factual data
like temperature of the body, pulse rate, B.P. etc. In botany- the study of plant life, one
has to rely heavily on statistics in conducting experiments about the plants, effect of
temperature, type of soil etc. In agriculture- statistical techniques like ‘analysis of
variance’ and ‘design of experiments’ are useful for isolating the role of manure, rainfall,
watering process, seed quality etc. In fact it is difficult to find any scientific activity
where statistical data and methods are not used.
 6. Statistics and Physical Science: The physical sciences in which statistical methods were
first developed and applied. It seems to be making increasing use of statistics, especially
in astronomy, chemistry, engineering, geology, meteorology and certain branches of
physics.
7. Statistics and Research; statistics is indispensable in research work. Most of the
advancement in knowledge has taken place because of experiments conducted with the
help of statistical methods. Statistical methods also affect research in medicine and public
health. In fact, there is hardly any research work today that one can find complete without
statistical methods.
8. Statistics and Computer: The development of statistics has been closely related to the
evolution of electronic computing machinery. Statistics is a form of data processing a
way of converting data into information useful for decision-making. The computers can
process large amounts of data quickly and accurately. This is a great benefit to business
and other organizations that must maintain records of their operations. Processing of row
data is extensively required in the application of many statistical techniques.

CLASSIFICATION OF STATISTICS

Statistics can be divided into 3 parts;

DESCRIPTIVE STATISTICS
INFERENTIAL STATISTICS
APPLIED STATISTICS
1. Descriptive Statistics: Descriptive statistics is related to numerical data or facts. Such
data are collected either by counting or by some other process of measurement. It is also
related to those methods, includes editing of data, classification, tabulation, diagrammatic
or graphical presentation, measures of central tendency, measures of dispersion,
correlation etc., help to make the description of numerical facts simple, systematic,
synoptic understandable and meaningful.
2. Inferential Statistics: Inferential statistics help in making generalizations about the
population or universe on the basis of study of samples. It includes the process of
drawing proper and rational conclusion about the universe. Among these methods,
probability theory and different techniques of sampling test are important.
3. Applied Statistics; It involves application of statistical methods and techniques to the
problems and actual facts. For example-statistics related to national income, industrial
and agricultural production, population, price etc. are called applied statistics. It can be
divided into 2 parts-(1) Descriptive Applied Statistics- it deals with the study of the dat
 which are known and which naturally relate. Its main object is to provide descriptive
information either to the past or to the present for any area. For example- price index
number and vital statistics comes under the category of descriptive applied statistics. (2)
Scientific Applied Statistics- under this branch of statistical science, statistical methods
are used to formulate and verify scientific laws. For example-an effort is made by an
economist to establish the law of demand, quantitative theory of money, trade circle etc.
These are established and verify by the help of scientific applied statistics.

IMPORTANCE OF STATISTICS

In recent days, we hear talking about statistics from a common person to highly qualified person.
It only show that how statistics has been intimately connected with wide range of activities in daily
life. They realize that work in their fields require some understanding of statistics. It indicates the
importance of the statistics. A.L.Bowley says, “Knowledge of statistics is like knowledge of foreign
language or of algebra. It may prove of use at any time under any circumstances”.

1. Importance to the State or Government; In modern era, the role of state has increased
and various governments of the world also take care of the welfare of its people.
Therefore, these governments require much greater information in the form of numerical
figures. Statistics are extensively used as a basis for government plans and policies. For
example-5-years plans are framed by using reliable statistical data of different segments
of life.
2. Importance in Human Behavior; Statistical methods viz., average, correlation etc. are
closely related with human activities and behavior. For example-when a layman wishes
to purchase some article, he first enquiries about its price at different shops in the market.
In other words, he collects data about the price of a particular article and aims at getting
idea about the average of the prices and the range within which the price vary. Thus, it
can be concluded that statistics play an important role in every aspect of human activities
and behavior.
3. Importance in Economics; Statistics is gaining an ever increasing importance in the
field of economics. That is why Tugwell said, “The science of economics is becoming
statistical in its method.” Statistics and economics are so interrelated to each other that
the new disciplines like econometrics and economic statistics have been developed.
Inductive method of generalization used in economics, is also based on statistical
principle. There are different segments of economics where statistics are used-

(A) Consumption- By the statistics of consumption we can find the way in which people in
different group spend their income. The law of demand and elasticity of demand in the field
of consumption are based on inductive or inferential statistics.

(B) Production- By the statistics of production supply is adjusted according to demand. We

can find out the capital invested in different productive units and its output. The decision
 about what to produce, how much to produce, when to produce is based on facts analyzed
statistically.

(C) Distribution- Statistics play a vital role in the field of distribution. We calculate the
national income of a country by statistical methods and compare it with other countries. At
every step we require the help of figures without them. It is difficult to move and draw
inferences.

4. Importance in Planning; for the proper utilization of natural and manual resources,
statistics play a vital role. Planning is indispensable for achieving faster rate of growth
through the best use of a nation’s resources. Sometimes said that, “Planning without
statistics is a ship without rudder and compass.” For example- In India, a number of
organizations like national sample survey organization(N.S.S.O.), central statistical
organization (C.S.O.) are established to provide all types of information.
5. Importance in Business: The use of statistical methods in the solution of business
problems dates almost exclusively to the 20 th century. Or now days no business, large or
small, public or private, can prosper without the help of statistics. Statistics provides
necessary techniques to a businessman for the formulation of various policies and
planning with regard to his business. Such as-
(A) Marketing- In the field of marketing, it is necessary first to find out what can be sold
and them to evolve a suitable strategy so that goods reach the ultimate consumer. A
skillful analysis of data on population, purchasing power, habits of people, competition,
transportation cost etc. should precede any attempt to establish a new market.
(B) Quality Control- To earn the better price in a competitive market, it is necessary to
watch the quality of the product. Statistical techniques can also be used to control the
quality of the product manufactured by a firm. Such as - Showing the control chart.
(C) Banking and Insurance Companies- banks use statistical techniques to take
decisions regarding the average amount of cash needed each day to meet the
requirements of day to day transactions. Various policies of investment and sanction of
loans are also based on the analysis provided by statistics.
(D) Accounts writing and Auditing- Every business firm keeps accounts of its revenue
and expenditure. Statistical methods are also employed in accounting. In particular, the
auditing function makes frequent application of statistical sampling and estimation
procedures and the cost account uses regression analysis.
(E) Research and Development- Many business organizations have their own research
and development department which are responsible for collection of such data. These
departments also prepare charts groups and other statistical analysis for the purpose.

FUNCTIONS OF STATISTICS

Statistics performs the functions of making the numerical aspects of facts simple, precise,
comparable and reliable. In fact, the various functions performed by statistics are the basis of its
utility. R.W. Burgess says, “The fundamental gospel of statistics is to push back the domain of
ignorance, prejudice, rule of thumb, arbitrary and premature decisions, tradition & dogmatism and to
increase the domain in which decisions are made. Principles are formulated on the basis of analyzed
quantitative facts.”

1. Numerical and definite expression of facts: The first function of the statistics is the
collection and presentation of facts in numerical form. We know that the numerical
presentation helps in having a better understanding of the nature of a problem. One of the
most important functions of statistics is to present general statements in a precise and
definite form. Statements and facts conveyed in exact quantitative terms are always more
convincing than vague utterances.
2. Simplifies the data (condensation): Not only does statistics present facts in a definite
form but it also helps in condensing mass of data into a few significant figures.
According to A.E.Waugh, “the purpose of a statistical method is to simplify great bodies
of numerical data.”In fact, human mind cannot follow the huge, complex and scattered
numerical facts. So these facts are made simple and precise with the help of various
statistical methods like averages, dispersion, graphic or diagrammatic, presentation,
classification, tabulation etc. so that a common man also understand them easily.
3. Comparison of facts: Baddington states, “The essence of the statistics is not only
counting but also comparison.” The function of comparison does help in showing the
relative importance of data. For example- the pass % of examination result of a college
may be appreciated better when it is compared with the result of other college or the
results of previous years of the same college.
4. Establishment of relationship b/w two or more phenomena; to investigate the
relationship b/w two or more facts is the main function of statistics. For example-demand
and supply of a certain commodity, prices and wages, temperature and germination time
of seeds are interrelated.
5. Enlarges individual experiences: In word of Bowley, “the proper function of statistics
indeed is to enlarge individual experience.” Statistics is like a master key that is used to
solve problems of mankind in every field. It would not be exaggeration to say that many
fields of knowledge would have remained closed to the mankind forever but for the
efficient and useful techniques and methodology of the science of statistics.
6. Helps in the formulation of policies: statistics helps in formulating policies in different
fields, especially in economic, social and political fields. The government policies like
industrial policy, export-import policies, taxation policy and monetary policy are
determined on the basis of statistical data and their movements, plan targets are also fixed
with the help of data.
7. Helps in forecasting: statistical methods provide helpful means in estimating the
available facts and forecasting for future. Here Bowley’s statementis relevant that, “a
statistical estimate may be good or bad, accurate or the reverse; but in almost all cases it
is likely to be more accurate than a casual observer’s impression.”
 8. Testing of hypothesis: statistical methods are also employed to test the hypothesis in
theory and discover newer theory. For example-the statement that average height of
students of college is 66 inches is a hypothesis. Here students of college constitute the
population. It is possible to test the validity of this statement by the use of statistical
techniques.

LIMITATIONS OF STATISTICS

Newsholme states, “Statistics must be regarded as an instrument of research of great value but
having several limitations which are not possible to overcome and as such they need out careful
attention.”

1. Statistics does not study qualitative facts: Statistics means aggregate of numerical
facts. It means that in statistics only those phenomena are studied which can be expressed
in numerical terms directly or indirectly. Such as- (1) directly in numerical terms like age,
weight and income of individual (2) no directly but indirectly like intelligent of students
and achievements of students (3) neither directly nor directly like morality, affection etc.
such type of facts don’t come under the scope of statistics.
2. Statistics doesn’t study individual: According to W.I.King, “Statistics from their very
nature of subject cannot and will never be able to take into account individual causes.
When these are important, other means must be used for their study.” These studied are
done to compare the general behavior of the group at different points of time or the
behavior of different groups at a particular point of time.
3. Statistical results are true only on the average: The statistical laws are not completely
true and accurate like the law of physics. For example – law of gravitational forces is
perfectly true & universal but statistical conclusions are not perfectly true. Such as the
average age of a person in India is 62 years. It does not mean that every person will attain
this age. On the basis of statistical methods we can say only in terms of probability and
not certainty.
4. Statistics as lack of complete accuracy: According to Conner, “Statistical data must
always be treated as approximations or estimates and not as precise measurements.”
Statistical result are based on sample or census data, are bound to be true only
approximately. For example – according to population census 2001, country’s population
is 1,02,70,15,247 but can real population may not be more or less by hundred, two
hundred and so on.
5. Statistics is liable to be misused: Statistical deals with figures and it can be easily
manipulated, distorted by the inexpert and unskilled persons it is very much likely to be
misused in most of the cases. In other words, the data should be handled by experts. Thus
it must be used by technically sound persons.
 6. Statistics is only one of the methods of studying a phenomenon; According to Croxton
& Cowden, “It must not be assumed that the statistical method is the only method to be
used in research; neither this method be considered the best attack for every problem.”
The conclusions arrived at with the help of statistics must be supplemented with other
evidences.
7. Statistical results may be misleading; Without any reference, statistical results may
provide doubtful conclusions. For example – on the basis of increasing no. of prisoners in
the prison, it may be conclude that crime is increasing. But it may be possible that due to
rude behavior of police administration the number of prisoners is increasing but crime is
decreasing.

Therefore, it is worth-mentioning that every science based on certain assumption and limitations.
This does not reduce the importance of the subject but lays emphasis on the fact that precautions
should be taken while dealing with statistical analysis and interpretations.

DISTRUST OF STATISTICS

For practical view point statistics is very useful and important science. We know that utility of
statistics lies not merely in data but in correct analysis and proper interpretation of data. Several times
due to ignorance and bias, people misuse this delicate tool of knowledge and it creates distrust about
data.

Opinion about distrust of statistics;

1stOpinion–Statisticians fully trust on the statistical conclusions because data is collected,

edited, analyzed and interpreted on the basis of statistical methods. Thus there is no reason to doubt
on it and said that “Figures don’t lie” or “Figures can prove nothing.”

2nd Opinion –The statistics is looked upon with a suspicious eye and is quite often condemned
as “Figures are tissue of flesh hood. Discardi remarks that there are three kinds of lies- lies, damned
lies and statistics or “There are black lies, white lies, multi-chromatic lies and statistics is rainbow of
lies.”

Many persons feel that data are false, confusing and incorrect and with their help truth can be
proved wrong and lies can be put as truth. Hence it is said that “Statistics can prove anything” or
“Statistics are like clay of which you can make God or Devil, as you please.”

In this context, the observation is worth quoting that “Statistician is the person who is deeply
involved in statistical data. He can freely play with them, misuse them and can cheat common people.
So he is just magician who shows the games of tricks of hand through statistical data. His result can
be surprising but not trustworthy.
 TYPES OF DATA

Data are the foundation stones and basic raw material in relation to any statistical investigation
that can be counted, classified, measured or quantified.

Types of Data are following;

ON THE BASIS OF CHARACTERISTICS OF FACTS

Data may be divided into two types;

1. Quantitative Data or Numerical Data: These types of data can be measured

directly such as age, income, production, marks etc. those facts are called variables
and variables may be discrete or continuous.
Discrete variable– Those variables whose values are individually distinct and discontinuous.
There is a definite difference between two variables. According to Boddington, “Discrete
variables is one where the variables (Individual values) differ from each other by definite
amounts.” For example – number of students of a class, number of children in a family, number
of cattle’s etc. It takes integral values such as 0, 1, 2, 3, 4 …etc.
Continuous variable – A continuous variable is one which assumes all values with in an
interval. That is no definite breaks are visible in this type of series. For example – age, weight,
height……
Questions; State which of the following represents Discrete data or Continuous data?
I. No. of accidents on each day in a month
II. Lengths of 1,000 bolts produced in a factory
III. Speed of an automobile in kilometer per hour
IV. No. of books on a library shelf

2. Qualitative Data or Categorical Data: They include data relating to such facts
which can‘t be measured directly but are counted or categorized to the basis of
attributes such as literates, illiterates, unemployed, honest etc. are called attributes.
For example- population can be classified on the basis of males and females or males
may be classified on the basis of marital status, i.e. married or unmarried. Qualitative
Data may further be classified into two categories

ON THE BASIS OF VARIABLES

On the basis of variables, also data may be of two types;

(1) Univariate Data: When the frequencies are determined on the basis of one variable. For
example – no. of workers on the basis of wages, no. of persons on the basis of age etc.

(2) Bivariate Data: When the data are edited or presented on the basis of two variables
simultaneously. For this two-way frequency table is constructed, one variable is placed
horizontally and the second one vertically. For example – to present the number of
students in one table on the basis of marks obtained in two subjects, to tabulate the no. of
persons in one table on the basis of two variables i.e. height and weight.

ON THE BASIS OF ARRANGEMENTS

Data may be categorized into two types;

(1) Raw Data: When the data is arranged and analyzed. It is called ‘Raw’ because it is
unprocessed by statistical methods.

(2) Arrange Data: When the data is processed and is arranged, summarized, classified and
tabulated in proper way.
 Terms like ‘Data Point’ and ‘Data Set’ are also used in order to distinguish between the
numbers relating to individual or single facts and the aggregate of facts. For example– the
data of production of sugar for ten years will be termed as ‘Data Set’ and the figures for
production of one year will be as ‘Data Point’.

CLASSIFICATION

After collection and editing of data the first step towards further processing the same is classification.
Classification is a process in which the collected data are arranged in separate classes, groups or
subgroups according to their characteristics. According to Secrist, “Classification is the process of
arranging data into sequences and groups according to their common characteristics or separating
them into different classes.”

It concludes that classification means the arrangements and systematization of data into different
classes and these classes are determined on the basis of nature, objectives and scope of the enquiry.

OBJECTIVES OF CLASSIFICATION
Classification is a method or technique for extracting the essential information supplied by the raw
data.
(1) To condense the data: the main objective of classification is to condense and simplify the
statistical material, so that the same may be easily understandable.
(2) To bring out points of similarities and dissimilarities of data: classification brings out
clearly the points of similarity and dissimilarities of statistical facts because data of similar
characteristics are placed in one class i.e., males and females, literates and illiterates, married
and unmarried etc.
(3) To make facts comparable: by arranging the data according to the points of similarity and
dissimilarities, it helps in comparison.
(4) To bring out relationship: classification helps in finding cause and effect relationship in the
data. For example- based on literacy and criminal tendency of a group peoples, it can be
established whether literacy has any impact on criminal tendency or not.
(5) To prepare ground for tabulation: tabulation is the basis of statistical analysis and
classification is the basis for tabulation.

It concludes that classification occupies an important place in the process of statistical investigation.
The fact is that the process of tabulation, presentation and analysis can’t even be shorted without
classification.

CHARACTERISTICS OR RULES FOR A GOOD CLASSIFICATION

(1) Unambiguity: the various classes should be so defined that there is no roomfor doubt and
confusion. For example–population is classified as literates or illiterates.
(2) Exhaustive and mutually exclusive: classification should be so exhaustive (clear all aspects)
and one item may not be find place in more than one class. For example – students of a
 college are classified into three groups – urban, rural and hostlers. This classification is not
mutually exclusive because among hostlers some may be urban and some other rural.
(3) Stability: the classification of data into various classes must be stable over be a period of time
of investigation.
(4) Suitability: the classification should confirm to the objectives of enquiry. For example–to study
the relationship between sex and university education, there is no need to classify on the basis
of age and religion.
(5) Flexibility: a good classification should be flexible so that adjustments may be easily be made
in classes according to changed situations. An ideal classification is one that can adjust itself
to these changes and yet retains its stability.

METHODS OF CLASSIFICATION
There are 4 methods of classification;

Geographical Classification

Chronological Classification
Qualitative Classification

Quantitative Classification

TABULATION
Tabulation is the next step of classification of the data and is designed to summaries lots of
information in a simple manner. In common language tabulation is the process of arranging data
in a systematic manner in the form of rows and columns. According to Blair, “Tabulation in its
broadest sense is any orderly arrangement of data in columns and rows.”
 OBJECTIVES OF TABULATION
1. To simplify complex data
2. To facilitate comparison
3. To economies Space
4. To facilitate presentation
5. Help in analysis of data
6. To help in reference

DIFFERENCE BETWEEN CLASSIFICATION & TABULATION

Basis of Difference Classification Tabulation

Presentation Classifies into different Classifies into row and columns.
classes.
Sequence First step Second step
Methods Method of statistical Method of data Presentation
Analysis.
Use of data Original data are used. Derivatives like percentages, coefficients
Proportion, etc. may also be used.

FREQUENCY DISTRIBUTION

The tabular arrangement of data showing the frequency of each item is called a frequency
distribution. According to Croxton and Cowden, “Frequency distribution is a statistical table in
which different values of variable are shown in the sequence of magnitude along with
corresponding frequencies.”

TYPES OF FREQUENCY DISTRIBUTION

(1) Discrete frequency distribution: It is a discontinuous frequency distribution, where

observations are independent to each other. Each observation is different and separates
from others. Example – the no. of two children in 20 families;

1,1,2,3,4,3,2,1,1,4,5,2,4,2,2,1,3,3,2,5

Construction of discrete frequency distribution

1. A table with three columns is prepared.
2. In 1st column of this table all possible values of the variable are placed.
3. In the 2nd column, tally marks or tally bars are put, keeping in view the repetition of each
value.
4. After putting tally marks, they are counted and this counting is shown in the 3rd column,
entitled frequency.
 No. of Tally Bars Frequency
children
1 |||| 5
2 |||| | 6
3 |||| 4
4 ||| 3
5 || 2

(2) Continuous frequency distribution: A continuous frequency distribution is such a

distribution in which data are arranged in classes or groups which are not exactly
measureable. Groups or class-intervals are always in a continuous form from the
beginning of the frequency distribution, till the end, within a given range of the data. For
example–
39,25,5,33,19,21,12,48,13,21,9,1,10,
9,8,12,17,40,12,46,37,17,27,30,6,2,23

Construction of continuous frequency distribution

(1) Find out the range = maximum marks – minimum marks

(2) Class size = decide the class size
(3) Width of class interval = 慶軍契傾祁
卦珪軍慧慧 慧兄渓祁
(4) Mid value =
掲径径祁慶 卦珪軍慧慧 珪兄型兄憩袋珪形敬祁慶 卦珪軍慧慧 珪兄型兄憩
匝
(5) Class frequency = the no. of observations corresponding to a particular class is known
as frequency of the class.

Marks Tally Bars No. of

Students
0 – 10 |||||| 7
10 – 20 |||| ||| 8
20 – 30 |||| 5
30 – 40 |||| 4
40 – 50 ||| 3

There are two types of series according to class interval;

(1) Inclusive form; A frequency distribution in which each upper limit of each class is also
included. Such as; 0-9, 10-19, 20-29……………..
(2) Exclusive form; In which the upper limit of the next class-interval. Such as; 0-10, 10-20,
20-30…………
 MEASURES OF CENTRAL TENDENCY

OBJECTIVES
After going through this unit, you will learn:
• the concept and significance of measures of central tendency
• to compute various measures of central tendency, such as arithmetic mean, median, mode and quartiles
• the relationship among various averages.

INTRODUCTION

The objective here is to find one representative value which can-be used to locate and summarise the
entire set of varying values. This one value can be used to make many decisions concerning the entire
set. We can define measures of central tendency (or location) to find some central value around which the
data tend to cluster.

SIGNIFICANCE OF MEASURES OF CENTRAI TENDENCY

Measures of central tendency i.e. condensing the mass of data in one single value , enable us to get an
idea of the entire data. For example, it is impossible to remember the individual incomes of millions of
earning people of India. But if the average income is obtained, we get one single value that represents the
entire population.
Measures of central tendency also enable us to compare two or more sets of data to facilitate comparison.
For example, the average sales figures of April may be compared with the sales figures of previous
months.
PROPERTIES OF A GOOD MEASURE OF CENTRAL TENDENCY

[1] It should be easy to understand and calculate.

[2] It should be rigidly defined.
[3] It should be based on all observations.
[4] It should be least affected by sampling fluctuation.
[5] It should be capable of further algebraic treatment.
[6] It should be least affected by extreme values.
[7] It should be calculated in case of open end interval.
Following are some of the important measures of central tendency which are commonly used in business
and industry.
 Arithmetic Mean
 Weighted Arithmetic Mean
 Median
 Quantiles(quartiles, deciles and percentiles)
 Mode
 Geometric Mean
 Harmonic Mean
ARITHMETIC MEAN

The arithmetic mean (or mean or average) is the most commonly used and readily understood measure of
central tendency. In statistics, the term average refers to any of the measures of central tendency.
Ungrouped data/Raw data
The arithmetic mean is defined as being equal to the sum of the numerical values of each and every
observation divided by the total number of observations. Symbolically, it can be represented as:

𝑁
デ 散𝑖
μ
=

where,
∑X indicates the sum of the values of all the observations, and N is the total number of
observations.
For example, let us consider the monthly salary (Rs.) of 10 employees of a firm x
2500, 2700, 2400, 2300, 2550, 2650, 2750, 2450, 2600, 2400
If we compute the arithmetic mean, then 2500+2700+2400+2300+2550+2650+2750+2450+2600+2400 =
25300
Mean=25300 /10= Rs. 2530.
Therefore, the average monthly salary is Rs. 2530.

Discrete data

When the observations are classified into a frequency distribution, Therefore, for discrete data; the
arithmetic mean is defined as

𝑁
デ 𝑓散𝑖
μ
=
Where, f is the frequency for corresponding variable x and N is the total frequency, i.e. N = ぇ f.

X f fx
10 12 120
20 23 460
30 35 1050
40 47 1880
50 38 1900
60 29 1740
70 16 1120
Sum 200 8270

Mean=8270/200= 41.35
Continuous Data
When the observations are classified into a frequency distribution, Therefore, for grouped data; the
arithmetic mean is defined as

デ𝑓 𝑖
μ 𝑁
=
Where X is midpoint of various classes, f is the frequency for corresponding class and N is the total
frequency, i.e. N = ぇ f.
This method is illustrated for the following data which relate to the monthly sales of 200 firms.
the midpoint of the class interval would be treated as the representative average value of that class.

Monthly Sales Mid point No. of firms fX

(Rs. Thousand) X f
300-350 325 5 1625
350-400 375 14 5250
400-450 425 23 9775
450-500 475 50 23750
500-550 525 52 27300
550-600 575 25 14375
600-650 625 22 13750
650-700 675 7 4725
700-750 725 2 1450
Sum 200 10200

Mean=102000/200=510

MERITS OF MEAN

[1] It is easy to understand and calculate.

[2] It is rigidly defined.
[3] It is based on all observations.
[4] It is least affected by sampling fluctuation.
[5] It is capable of further algebraic treatment.

DEMERITS OF MEAN

1) It is highly affected by extreme values.

2) It is not calculated in case of open end interval.
MEDIAN

A second measure of central tendency is the median. Median is that value which divides the distribution
into two equal parts. Fifty per cent of the observations in the distribution are above the value of median and
other fifty per cent of the observations are below this value of median. The median is the value of the
middle observation when the series is arranged in order of size or magnitude (Ascending order).

UNGROUPED DATA

If the number of observations is odd, then the median is equal to one of the original observations (Middle).

𝑁袋 1
Median =岾 峇 th value
2

For example, if the income of seven persons in rupees is 1100, 1200, 1350, 1500, 1550, 1600, 1800, then

Median =岫岫ば髪な岻�に岻 =4th value

Median =1500

If the number of observations is even, then the median is the arithmetic mean of the two middle
observations.

𝑁 𝑁
峽岾 峇 袋岾 1 袋峇峺
鉄
Median 鉄
th
2
= value

For example, if the income of eight persons in rupees is 1100, 1200, 1350, 1500, 1550, 1600, 1800,1850,
then the median income of eight persons would be 1500+1550/2= 1525

DISCRETE SERIES

First we find cumulative frequency.then locate (N+1/2) the value in cumulative frequency.corresponding
that value of x is median.

X f cf
10 12 12
20 23 35
30 35 70
40 47 117
50 38 155
60 29 184
70 16 200
Sum 200
N=200
N+1/2=100.5
Median= 40
 CONTINUOUS DATA

For continuous data, First we find cumulative frequency. then locate (N+1/2) the value in cumulative
frequency. corresponding class interval is median class.the following formula may be used to locate the
value of median.

Median=l1+ {(N/2 に cf)*h} /f

where l1 is the lower limit of the median class, cf is the preceding cumulative frequency to the median class,
f is the frequency of the median class and h is the width of the median class.

Consider the following data which relate to the age distribution of 1000 workers in an industrial
establishment.
The location of median value is facilitated by the use of a cumulative frequency distribution as shown
below in the table.

Age (Years) No. of workers Cumulative frequency

f c.f
Below 25 120 120
25-30 125 245
30-35 180 425(cf)
35-40 160(f) 585
40-45 150 735
45-50 140 875
50-55 100 975
55 and Above 25 1000

N=1000
Median Class=(1000+1)/2=500.5th =(35-40)
l1 =35,cf =425,f=160,N/2=500,h=5

Median =35+{(500-425)*5}/160 =35+375/160=35+2.34=37.34

MERITS OF MEDIAN

[1] It is easy to understand and calculate.

[2] It is rigidly defined.
[3] It is not affected by extreme values.
[4] It is calculated in case of open end interval.
[5] It is located by graphically also.

DEMERITS OF MEDIAN

[1] It is not based on all observations.

[2] It is affected by sampling fluctuation.
[3] It is not capable of further algebraic treatment.
 MODE
The mode is the typical or commonly observed value in a set of data. It is defined as the value
which occurs most often or with the greatest frequency. The dictionary meaning of the term mode is most
usual'.

For example, in the series of numbers 3, 4, 5, 5, 6, 7, 8, 8, 8, 9, the mode is 8 because it occurs the
maximum number of times. That means in ungrouped data mode can find by inspection only.

DISCRETE DATA

Mode is the value of X which has highest frequency.

For example,

X f
10 12
20 23
30 35
40 47
50 38
60 29
70 16
Sum 200

Mode=40

CONTINUOUS DATA

First, we find Modal class = corresponding to highest frequency.

𝑓1 𝑓宋
MODE = l1+ { 𝑓1 𝑓𝑓匝 }*h

where l1 is lower limit of the modal class, f1 is the frequency of the modal class,f0 the frequency of the
preceding class, f2 is the frequency of the succeeding class, h is the size of the modal class.

To illustrate the computation of mode, let us consider the following data.

Age (Years) No. of workers

f
Below 25 120
25-30 125(f0)
30-35 180(f1)
35-40 160(f2)
40-45 150
45-50 140
50-55 100
55 and Above 25

Modal class=(30-35)

l1 =30, f1 =180, f0 =125, f2 =160,h=5

Mode =30+{(180-125)/(2*180-125-160)}*5 =30+(55/75)*5 =30+3.667=33.67
MERITS OF MODE

[1] It is easy to understand and calculate.

[2] It is not affected by extreme values.
[3] It is calculated in case of open end interval.
[4] It is located by graphically also.

DEMERITS OF MODE

[1] It is not based on all observations.

[2] It is highly affected by sampling fluctuation.
[3] It is not capable of further algebraic treatment
[4] It is not rigidly defined.

RELATIONSHIP AMONG MEAN, MEDIAN AND MODE

MODE = 3MEDIAN-2 MEAN
QUARTILES
Quartiles are those values which divide the total data into four equal parts. Since three points divide the
distribution into four equal parts, we shall have three quartiles. Let us call them Q1, Q2, and Q3. The first
quartile, Q1, is the value such that 25% of the observations are smaller and 75% of the observations are
larger. The second quartile, Q2, is the median, i.e., 50% of the observations are smaller and 50% are
larger. The third quartile, Q3, is the value such that 75% of the observations are smaller and 25% of the
observations are larger.

UNGROUPED DATA

If the number of observations is odd, then the median is equal to one of the original observations (Middle).

𝑁袋 1
Q1=岾 替 峇 th value

𝑁袋 1
Q3=ぬ 岾 替
峇 th value

For example, if the income of seven persons in rupees is 1100, 1200, 1350, 1500, 1550, 1600, 1800,1850
then

Q1 =岫岫ぱ髪な岻�ね岻 =2.25nd value

Q1=1200+0.25(1350-1200) =1200+37.5 =1237.5

Q3=岫ぬ岫ぱ髪な岻�ね岻 =6.75th value

Q3= 1600+.75(1800-1600) =1600+150 = 1750

DISCRETE SERIES
 First we find cumulative frequency. then locate (N+1/4 ) and 3(N+1/4)the value in cumulative frequency
.corresponding that value of x is Q1 and Q2 respectively.

First X f cf
quartil 10 12 12
e 20 23 35
30 35 70
40 47 117
Third 50 38 155
quartil 60 29 184
e 70 16 200
Sum 200
N=200
Q1=N+1/4=50.25th value
Q1 = 30
Q3 =3*(N+1/4)= 150.75th value
Q3=50

CONTINUOUS DATA

For continuous data, First we find cumulative frequency. then locate (N+1/4) and 3(N+1/4)the value in
cumulative frequency. corresponding class interval is first quartile class and third quartile class
respectively.the following formula may be used to locate the value of quartiles.

Q1=l1+ {(N/4 に cf)*h} /f

where l1 is the lower limit of the first quartile class, cf is the preceding cumulative frequency to the first
quartile class, f is the frequency of first quartile class and h is the width of the first quartile class .

Q3=l1+ {(3N/4 に cf)*h} /f

where l1 is the lower limit of the third quartile class, cf is the preceding cumulative frequency to the third
quartile class, f is the frequency of third quartile class and h is the width of the third quartile class .

Consider the following data which relate to the age distribution of 1000 workers in an industrial
establishment.
The location of quartile value is facilitated by the use of a cumulative frequency distribution as shown
below in the table.
First
Age (Years) No. of workers Cumulative frequency quartile
f c.f class
Below 25 120 120
25-30 125 245
30-35 180 425(cf)
35-40 160(f) 585
40-45 150 735
45-50 140 875 Third
50-55 100 975 quartil
55 and Above 25 1000 e
class
 N=1000
Q1=(1000+1)/4=250.25th =(25-30)
l1 =25,cf =120,f=125,N/4=250,h=5
Q1=25+{(250-120)*5}/125 =35+390/125=35+3.12=38.12

Q3=3(1000+1)/4=750.75th =(45-50)
l1 =45,cf =735,f=140,3N/4=750,h=5
Q3=45+{(750-735)*5}/140=45+75/140=45+0.53=45.53

1] Following is the cumulative frequency distribution of preferred length of study-table obtained from
the preference study of 50 students.

A manufacturer has to take decision on the length of study-table to manufacture. What length would
you recommend and why?

2] An incomplete distribution of daily sales (Rs. thousand) is given below. The data relate to 229 days.
Daily sales No. of days Daily sales No. of days
(Rs. thousand) (Rs. thousand)
10-20 12 50-60 ?
20-30 30 60-70 25
30-40 ? 70-80 18
You are told that the median value is 46. Using the median formula, fill up the missing frequencies and
calculate the arithmetic mean of the completed data.
CHAPTER

7
Correlation

As the summer heat rises, hill

Studying this chapter should stations, are crowded with more
enable you to:
• understand the meaning of and mor e visitors. Ice-cream sales
the term correlation; become mor e brisk. Thus, the
• understand the nature of temperature is related to number of
relationship between two visitors and sale of ice-creams.
variables;
Similarly, as the supply of tomatoes
• calculate the differ ent
measures of correlation; increases in your local mandi, its
• analyse the degree and direction price dr ops. When the local harvest
of the relationships. starts reaching the market, the
price of tomatoes drops from a
1. INTRODUCTION princely Rs 40 per kg to Rs 4 per kg
or even less. Thus supply is related
In pr evious chapters you have to price. Correlation analysis is a
learnt how to construct summary
means for examining such
measures out of a mass of data and
relationships systematically. It
changes among similar variables.
Now you will lear n how to examine deals with questions such as:
the relationship between two • Is there any relationship
variables. between two variables?

2015-16
92 STATISTICS FOR ECONOMICS

integral part of the theory of

demand, which you will study in
Class XII. Low agricultural
productivity is related to low
rainfall. Such examples of
relationship may be given a cause
and effect interpretation. Others may
be just coincidence. The relation
• If the value of one variable between the arrival of migratory
changes, does the value of the birds in a sanctuary and the birth
other also change? rates in the locality cannot be given
any cause and effect interpretation.
The relationships are simple
coincidence. The relationship
between size of the shoes and
money in your pocket is another
such example. Even if
relationship exist, they are difficult
to explain it.
In another instance a third
variable’s impact on two variables
• Do both the variables move in may give rise to a relation between
the same direction? the two variables. Brisk sale of ice-
creams may be related to higher
number of deaths due to drowning.
The victims are not drowned due to
eating of ice- creams. Rising
temperature leads to brisk sale of
ice-creams. Moreover, large number
of people start going to
swimming pools to beat the heat.
This might have raised the number of
deaths by drowning. Thus
• How strong is the relationship? temperature is behind the high
correlation between the sale of ice-
2. TYPES OF RELATIONSHIP creams and deaths due to
drowning.
Let us look at various types of
relationship. The relation between What Does Correlation Measure?
movements in quantity demanded
and the price of a commodity is Correlation studies and measures
an the direction and intensity of
relationship among variables.
Correlation measures covariation,
not causation. Correlation should
never be
2015-16
CORRELATION 93

interpreted as implying cause and 3. T ECHNIQUES FOR M EASURING

effect relation. The presence of CORRELATION
correlation between two variables X
and Y simply means that when the Three important statistical tools
value of one variable is found to used to measure correlation are
change in one direction, the value of scatter diagrams, Karl Pearson’s
the other variable is found to change coefficient of correlation and
either in the same direction (i.e. Spearman’s rank correlation.
positive change) or in the opposite A scatter diagram visually
direction (i.e. negative change), but presents the nature of association
without giving any specific
in a definite way. For simplicity we
numerical value. A numerical
assume here that the correlation, if
measure of linear relationship
it exists, is linear, i.e. the relative
between two variables is given by
movement of the two variables can
Karl Pearson’s coefficient of
be represented by drawing a correlation. A relationship is said to
straight line on graph paper. be linear if it can be represented
by a straight line. Spearman’s
Types of Correlation coefficient of correlation measures
Correlation is commonly classified the linear association between
into negative and positive ranks assigned to indiviual items
correlation. The correlation is said to according to their attributes.
be positive when the variables Attributes are those variables
move together in the same which cannot be numerically
measured such as intelligence
direction. When the income rises,
of people, physical appearance,
consumption also rises. When
honesty, etc.
income falls, consumption also
falls. Sale of ice- cream and
Scatter Diagram
temperature move in the same
direction. The correlation is A scatter diagram is a useful
negative when they move in technique for visually examining the
opposite directions. When the price form of relationship, without
of apples falls its demand increases. calculating any numerical value. In
When the prices rise its demand this technique, the values of the two
decreases. When you spend more variables are plotted as points on a
time in studying, chances of your graph paper. From a scatter
failing decline. When you spend less diagram, one can get a fairly good
hours in your studies, chances of idea of the nature of relationship. In
scoring low marks/grades increase. a scatter diagram the degree of
These are instances of negative closeness of the scatter points and
correlation. The variables move in their overall direction enable us to
opposite direction. examine the relation-
2015-16
 94 STATISTICS FOR ECONOMICS

ship. If all the points lie on a line, A careful observation of the

the correlation is perfect and is said scatter diagram gives an idea of
to be unity. If the scatter points are the nature and intensity of the
widely dispersed around the line, relationship.
the correlation is low. The
correlation is said to be linear if the Karl Pearson’s Coefficient of
scatter points lie near a line or on a Correlation
line. This is also known as product
Scatter diagrams spanning over moment correlation and simple
Fig. 7.1 to Fig. 7.5 give us an idea correlation coefficient. It gives a pr
of the relationship between two ecise numerical value of the
variables. Fig. 7.1 shows a scatter degree of linear relationship
around an upward rising line between two variables X and Y. The
indicating the movement of the linear relationship may be given by
variables in the same direction. Y = a + bX
When X rises Y will also rise. This is This type of relation may be
positive correlation. In Fig. 7.2 the described by a straight line. The
points are found to be scattered intercept that the line makes on the
around a downward sloping line. Y-axis is given by a and the slope of
This time the variables move in the line is given by b. It gives the
change in the value of Y for very
opposite directions. When X rises Y
small change in the value of X. On
falls and vice versa. This is negative
the other hand, if the relation
correlation. In Fig.7.3 there is no
cannot be represented by a
upward rising or downward sloping straight line as in
line around which the points are Y = X2
scattered. This is an example of no the value of the coef ficient will be
correlation. In Fig. 7.4 and Fig. 7.5 zero. It clearly shows that zero
the points are no longer scattered correlation need not mean absence
around an upward rising or of any type of relation between the
downward falling line. The points two variables. Let X , X , ..., X be N
1 2 N
themselves are on the lines. This is values of X and Y1, Y2 ,..., YN be the
referred to as per fect positive corresponding
correlation and perfect negative values of Y. In the subsequent
correlation respectively. presentations the subscripts
indicating the unit are dropped for
the sake of simplicity. The
arithmetic means of X and Y are
Activity
defined as
• Collect data on height, weight
ΣX ΣY
and marks scored by students X= ; Y=
in your class in any two N N
subjects in class X. Draw the and their variances are as follows
scatter diagram of these
variables taking two at a time. Σ(X −X)2 ΣX 2
What type of relationship do σ 2
x = = −X 2

you find? N N
2015-16
CORRELATION 95

2015-16
 96 STATISTICS FOR ECONOMICS

ΣY2 Properties of Correlation Coefficient

and σ2 Σ(Y − = −
= Let us now discuss the properties of
y Y)2 Y2
N the correlation coefficient
N
The standard deviations of X • r has no unit. It is a pure
and number. It means units of
Y respectively are the positive measurement are not part of r. r
square roots of their variances. between height in feet and
Covariance of X and Y is defined as
weight in kilograms, for

Cov(X,Y) Σ( X − X)(Y − Y) instance, could be say 0.7.

= Σxy = • A negative value of r indicates an
N N inverse relation. A change in one
Where x = X −X and y = Y − variable is associated with
change in the other variable in
Y are the deviations of the i th value
the opposite direction. When
of X and Y from their mean
price of a commodity rises, its
values respectively.
demand falls. When the rate of
The sign of covariance between
interest rises the demand for
X and Y determines the sign of the
funds also falls. It is because
correlation coefficient. The standard
now funds have become costlier.
deviations are always positive. If the
covariance is zero, the correlation
coefficient is always zero. The pr
oduct moment correlation or the
Karl Pearson’s measure of
correlation is given by
Σxy
r Nσx σy ...(1)
=
or
Σ(X − X) ( Y − Y)
r = ...(2)
Σ(X − Σ(Y −
or X)2 Y)2
(ΣX )(ΣY) • If r is positive the two variables
ΣXY − N
r= move in the same direction.
When
(ΣY)2 ... the price of coffee, a substitute
ΣX)2 2−
(ΣX ΣY2 − (3) N of tea, rises the demand for tea
N
also rises. Improvement in
or irrigation facilities is associated
with higher
NXY – ( X)( Y)
r ...(4)
NX 2 – (X)2  NY2 – (Y)2
yield. When temperature
rises the sale of ice-
creams becomes brisk.

2015-16
CORRELATION 97

• The value of the correlation before correlation is calculated and

coefficient lies between minus interpreted. An epidemic spreads in
one and plus one, –1  r  1. If, in some villages and the government
any exercise, the value of r is sends a team of doctors to the
outside this range it indicates affected villages. The correlation
error in calculation. between the number of deaths and
• If r = 0 the two variables are the number of doctors sent to the
uncorrelated. There is no linear villages is found to be positive.
relation between them. However Normally the health care facilities
other types of relation may be provided by the doctors are
there. expected to reduce the number of
• If r = 1 or r = –1 the correlation deaths showing a negative
is perfect. The relation between correlation. This happened due to
them is exact.
other reasons. The data relate to a
• A high value of r indicates
specific time period. Many of the
strong linear relationship. Its
reported deaths could be terminal
value is said to be high when
it is close to cases where the doctors could do
+1 or –1. little. Moreover, the benefit of the
• A low value of r indicates a weak presence of doctors becomes visible
linear relation. Its value is said only after some time. It is also
to be low when it is close to possible that the reported deaths
zero. are not due to the epidemic. A
• The magnitude of r is unaffected tsunami suddenly hits the state and
by death toll rises.
the change of origin and change Let us illustrate the calculation
of scale. Given two variables X of r by examining the relationship
and Y let us define two new between years of schooling of
variables. farmers
X–A Y–
CU= ;V=
B D and the annual yield per acre.
where A and C are assumed means
Example 1
of
X and Y respectively. B and D are
common factors and of same sign. the need for understanding the
Then data properly
rxy = uv
r
This. property is used to
calculate correlation coefficient
in a highly
simplified manner, as in the step
deviation method.
As you have read in chapter 1,
the
statistical methods are no substitute
for common sense. Here, is
another example, which highlights
No. of years Annual yield per 6 10
of schooling acre in ’000 8 10
(Rs) of farmers 10 8
12 7
0 4
2 4 Formula 1 needs the value of
4 6
Σxy, σx , σ y

2015-16
 98 STATISTICS FOR ECONOMICS

From Table 7.1 we education, higher will be the yield

get, per acre. It underlines the
importance of farmers’ education.
Σxy = 42,
To use formula (3)
σx Σ(X − = 112 ,
X)2 N ΣXY (ΣX )(ΣY)
= 7 −
(ΣX )2 N (ΣY)2
r = ΣX2 − ΣY 2 −
Σ(Y − Y)2 38 N N ...(3)
σy = =
N 7
Substituting these values in the value of the following
formula (1) expressions have to be
42
calculated i.e.
ΣXY, ΣX 2 , ΣY 2 .
r= =
Now apply formula (3) to get the
0.644
7
112 38 value of r.
7 7
The same value can be Let us know the interpretation of
obtained differ ent values of r. The
from for mula (2) also. correlation coefficient between
marks secured in English and
Σ (X − X)(Y − Y)
r= Statistics is, say, 0.1. It means that
...(2)
Σ (X − Σ (Y − though the marks secured in the
X)2
Y)2 two subjects are positively
42 0.644 correlated, the strength of the
r= relationship is weak. Students with
= 112 38
high
Thus years of education of marks in English may be getting
farmers and annual yield per relatively low marks in statistics.
acre are positively correlated. The Had the value of r been, say, 0.9,
value of r is also large. It implies students with high marks in English
that more the number of years will invariably get high marks in
farmers invest in Statistics.

TABLE 7.1
Calculation of r between years of schooling of farmers and
annual yield
Years of (X– X ) (X– X )2 Annual yield (Y– Y ) (Y– Y )2 (X– X )(Y– Y )
Education per acre in ’000 Rs
(X) (Y)
0 –6 36 4 –3 9 18
2 –4 16 4 –3 9 12
4 –2 4 6 –1 1 2
6 0 0 10 3 9 0
8 2 4 10 3 9 6
10 4 16 8 1 1 4
12 6 36 7 0 0 0

Σ X=42 Σ (X– X )2=112 Σ Y=49 Σ (Y– Y )2=38 Σ (X– X )(Y– Y

)=42

2015-16
CORRELATION 99

An example of negative
correlation is the relation between TABLE 7.2
arrival of vegetables in the local Year Annual growth Gross Domestic
mandi and price of vegetables. If r is of National Saving as
–0.9, vegetable supply in the local Income percentage of GDP

mandi will be accompanied by 1992–93 14 24

lower price of vegetables. Had it 1993–94 17 23
1994–95 18 26
been –0.1 large vegetable supply
will be accompanied by lower price, 1995–96 17 27
1996–97 16 25
not as low as the price, when r is –
1997–98 12 25
0.9. The extent of price fall depends
1998–99 16 23
on the absolute value of r. Had it
1999–00 11 25
been zero there would have been
2000–01 8 24
no fall in price, even after large
2001–02 10 23
supplies in the market. This is also
a possibility if the increase in supply Source: Economic Survey, (2004–05) Pg. 8,9
is taken care of by a good transport transformation of the variables X
network transferring it to other and Y as follows:
markets. X–A Y–B
U= ;V=
h k
Activity where A and B are assumed means,
• Look at the following table. h and k are common factors and
Calculate r between annual have same signs.
growth of national income at Then rUV = XY
current price and the Gross r
This can be illustrated with the
Domestic Saving as exercise of analysing the correlation
percentage between price index and money
of GDP.
supply.
Step deviation method to calculate Example 2
correlation coefficient. Price 120 150 190 220 230
index (X)
When the values of the variables
Money 1800 2000 2500 2700 3000
are large, the burden of calculation supply
can be considerably reduced by in Rs crores (Y)
using a property of r. It is that r is The simplification, using step
independent of change in origin and deviation method is illustrated
scale. It is also known as step below. Let A = 100; h = 10; B =
deviation method. It involves 1700 and k = 100
the

2015-16
 100 STATISTICS FOR ECONOMICS

The table of transformed Activity

variables is as follows: • Using data related to India’s
Calculation of r between price population and national
index and money supply using step income, calculate the
deviation method correlation between them
using step deviation method.
TABLE 7.3
U V Spearman’s rank correlation
 X − 100  Y −1700 Spearman’s rank correlation was
    U2 V2 UV
10   100 developed by the British
2 1 4 1 2 psychologist
5 3 25 9 15 C.E. Spearman. It is used when the
9 8 81 64 72 variables cannot be measured
meaningfully as in the case of price,
12 10 144 100 120
income, weight etc. Ranking may
13 13 169 169 169
be more meaningful when the
U = 41; U = 35; U2 = 423; measurements of variables are
V2 = 343; UV = 378 suspect. Consider the situation
where we are required to calculate
Substituting these values in formula
the correlation between height and
(3)
weight of students in a remote
village. Neither measuring

UV  UU rods nor weighing scales are

r= available. The students can be
N
easily ranked in terms of height and
2 U  2
V 2 (3)
weight without using measuring
U  V  2
N  rods and weighing scales.
There are also situations when
you are requir ed to quantify
41 × 35
qualities such
as fair ness, honesty etc. Ranking
may

= 378 − be a better alternative to quantifica-

5
2 2 tion of qualities. Moreover,
sometimes the correlation
(41)
423 − (35)
343 −
= 5 5 coefficient between two variables
0.98 with extreme values may be quite
different from the coefficient
without the extreme values. Under
these circumstances rank
correlation
The strong positive correlation rises.
between price index and money
supply is an important premise of
monetary policy. When the money
supply grows the price index also
provides a better alternative to
simple correlation.
Rank correlation coefficient and
simple correlation coefficient have
the same interpretation. Its formula
has

2015-16
CORRELATION 101

been derived from simple

correlation coefficient where not utilised. The first differences of
individual values have been replaced the values of items in the series,
by ranks. These ranks are used for arranged in order of magnitude, are
the calculation of correlation. This almost never constant. Usually the
coefficient provides a measure of data cluster around the central
linear association between ranks values with smaller differences in
assigned to these units, not their the middle of the array. If the
values. It is the Product Moment first differences were constant,
Correlation between the ranks. then r and r would give identical
Its results. The
k
formula is first difference is the difference of
... consecutive values. Rank
(4) correlation is preferred to
Pearsonian
where n is the number of coefficient when extreme values are
observations and D the deviation of present. In general
ranks assigned to a variable from rk is less than or equal to r.
those assigned to the other The calculation of rank correlation
variable. When the ranks are will be illustrated under three
repeated the formula is situations.

6 ΣD (
3− m ) (m3 − m )
1
2 1
 1. The ranks are given.
2 +m + 2 ... 

 12 12 + 
2. The ranks are not given. They
r s = 1– n(n2 −1) have to be worked out from the
data.
3. Ranks are repeated.
where m , m , ..., are the number of
1 2
m3 m Case 1: When the ranks are given
1
repetitions of ranks 1
..., Example 3
and 12
their corresponding corr ection
Five persons are assessed by three
factors. This correction is needed
judges in a beauty contest. We have
for every repeated value of both
to find out which pair of judges has
variables. If thr ee values are
the nearest approach to common
repeated, there will be a correction
perception of beauty.
for each value. Every time m1
indicates the number of times a Competitors
value is repeated. Judge 1 2 3 4 5
All the properties of the simple A 1 2 3 4 5
correlation coefficient are applicable B 2 4 1 5 3
here. Like the Pearsonian C 1 3 5 2 4
Coefficient of correlation it lies There are 3 pairs of judges
between 1 and –1. However, necessitating calculation of rank
generally it is not as accurate as the correlation thrice. Formula (4) will
ordinary method. This is due the be used.
fact that all the information
concerning the data is
2015-16
102 STATISTICS FOR ECONOMICS

r =1− Case 2: When the ranks are not

6ΣD2 ... given
(4)
s
n3−
Example 4
n
The rank correlation between A
and We are given the percentage of
B is calculated as follows: marks, secured by 5 students in
Economics and Statistics. Then the
A B D D2 ranking has to be worked out and
the rank correlation is to be
1 2 –1 1
calculated.
2 4 –2 4
3 1 2 4
4 5 –1 1 Student Marks in Marks in
5 3 2 4 Statistics Economics
(X) (Y)
Total 14
A 85 60
B 60 48
Substituting these values in C 55 49
formula (4) D 65 50
E 75 55
rs = 1 6ΣD2
n3− ...(4)
− Student Ranking in Ranking in
n
=1 6× 84 Statistics
− =1− = 1 − 0.7
14 Economics (R )
= 0.3 (RY )
53− 120 x

A 1 1
5
The rank correlation between A B 4 5
and C is calculated as follows: C 5 4
D 3 3
E 2 2
A C D D2
1 1 0 0 Once the ranking is complete
2 3 –1 1
formula (4) is used to calculate rank
3 5 –2 4
4 2 2 4
correlation.
5 4 1 1
Case 3: When the ranks are repeated
Total 10
Example 5
Substituting these values in
The values of X and Y are given as
formula (4) the rank correlation is 0.5. X 25 45 35 40 15 19 35 42
Similarly, the rank correlation Y 55 60 30 35 40 42 36 48

between the rankings of judges B tastes.

and C is 0.9. Thus, the perceptions
of judges A and C are the closest.
Judges B and C have very different
In order to work out the rank
correlation, the ranks of the values
are worked out. Common ranks are
given to the repeated items.
The

2015-16
CORRELATION 103

common rank is the mean of the

ranks which those items would have The necessary correction thus is
assumed m3 − m
23 − 2 1
if they were slightly different from each = =
12 12 2
other. The next item will be
assigned Using this equation
the rank next to the rank already  (m 3 − m)
assumed. The formula of 6 ΣD2 + 

Spearman’s
rank correlation coefficient when the  12  ...(5)
ranks are repeated is as follows rs = 1 − n3−n
rs = 1 Substituting the values of these
− expressions
 (m31 − m1) (m32 − m2 ) 
6 ΣD2 + + +...
 12 12  6(65.5 + 0.5) 396
rs = 1 − =1−
n(n 2 − 1) 8 −8
3 504
where m , m , ..., ar e the =1 − 0.786 =
number 0.214
1 2
of Thus there is positive rank correlation
repetitions of ranks and
between X and Y. Both X and Y
m 1 − m1
3
move in the same direction.
..., their corresponding However, the relationship cannot be
12
correction factors. described as strong.
X has the value 35 both at the
4th and 5th rank. Hence both are
Activity
given the average rank i.e., •
Collect data on marks scored by
4+5 10 of your classmates in class IX
th
th = 4.5 rank and X examinations. Calculate
2 the rank correlation
coefficient between them. If
2 your data do not have any
X Y Rank of Rank of Deviation in D
XR' YR' Ranking
repetition, repeat the exercise
' D=R'–R'' by taking a data set having
repeated ranks. What are the
25 55 6 2 4 16
circumstances in which rank
45 80 1 1 0 0 correlation coefficient is
35 30 4.5 8 3.5 12.25 preferred to simple correlation
40 35 3 7 –4 16 coefficient? If data are precisely
measured will you still prefer
15 40 8 5 3 9
rank correlation coefficient to
19 42 7 4 3 9 simple correlation? When can
35 36 4.5 6 –1.5 2.25 you be indifferent to the choice?
42 48 2 3 –1 1 4. CONCLUSION
Total ΣD = 65.5 We have discussed some
techniques for studying the
relationship between
2015-16
104 STATISTICS FOR ECONOMICS

two variables, particularly the linear be measured precisely, rank

relationship. The scatter diagram correlation can be used. These
gives a visual presentation of the measures however do not imply
relationship and is not confined to causation. The knowledge of
linear relations. Karl Pearson’s correlation gives us an idea of the
coefficient of correlation and direction and intensity of change in
Spearman’s rank correlation a variable when the correlated
measure linear relationship among variable changes.
variables. When the variables
cannot

Recap
• Correlation analysis studies the relation between two variables.
• Scatter diagrams give a visual presentation of the nature
of relationship between two variables.
• Karl Pearson’s coefficient of correlation r measures numerically
only
linear relationship between two variables. r lies between –1 and 1.
• When the variables cannot be measured pr ecisely Spearman’s
rank correlation can be used to measure the linear
relationship numerically.
• Repeated ranks need correction factors.
• Correlation does not mean causation. It only
means covariation.

EXERCISES

1. The unit of correlation coefficient between height in feet and weight

in kgs is
(i) kg/feet
(ii) percentage
(iii) non-existent
2. The range of simple correlation coefficient is
(i) 0 to infinity
(ii) minus one to plus one
(iii) minus infinity to infinity
3. If rxy is positive the relation between X and Y is of the type
(i) When Y increases X increases
(ii) When Y decreases X increases
(iii) When Y increases X does not change
2015-16
CORRELATION 105

4. If rxy = 0 the variable X and Y are

(i) linearly related
(ii) not linearly related
(iii) independent
5. Of the following three measures which can measur e any type of relationship
(i) Karl Pearson’s coefficient of correlation
(ii) Spearman’s rank correlation
(iii) Scatter diagram
6. If precisely measured data are available the simple correlation
coefficient is
(i) more accurate than rank correlation coefficient
(ii) less accurate than rank correlation coefficient
(iii) as accurate as the rank correlation coefficient
7. Why is r preferred to covariance as a measure of association?
8. Can r lie outside the –1 and 1 range depending on the type of data?
9. Does correlation imply causation?
10. When is rank correlation more precise than simple correlation coefficient?
11. Does zero correlation mean independence?
12. Can simple correlation coefficient measure any type of relationship?
13. Collect the price of five vegetables from your local market every day
for a week. Calculate their correlation coefficients. Interpret the result.
14. Measure the height of your classmates. Ask them the height of their
benchmate. Calculate the correlation coefficient of these two variables.
Interpret the result.
15. List some variables where accurate measurement is difficult.
16. Interpret the values of r as 1, –1 and 0.
17. Why does rank correlation coefficient differ from Pearsonian correlation
coefficient?
18. Calculate the correlation coef ficient between the heights of fathers in inches
(X) and their sons (Y)
X 65 66 57 67 68 69 70 72
Y 67 56 65 68 72 72 69 71
(Ans. r = 0.603)
19. Calculate the correlation coefficient between X and Y and comment on
their relationship:
X –3 –2 –1 1 2 3
Y 9 4 1 1 4 9
(Ans. r = 0)

2015-16
106 STATISTICS FOR ECONOMICS

20. Calculate the correlation coefficient between X and Y and comment on

their relationship
X 1 3 4 5 7 8
Y 2 6 8 10 14 16
(Ans. r = 1)

Activity
• Use all the formulae discussed here to calculate r between
India’s national income and exports taking at least ten
observations.

2015-16
Regression
Analysis
By: Dr.
 Regression - Introduction

The statistical technique of estimating the unknown value

of one variable (called dependent variable) from the
known value of other variable (called independent
variable) is called Regression Analysis.
 Utility of Regression
•
It provides a functionaAl fonrm

aoflryelastioinsship between two related

variables with the help of which the unknown value
of one variable can be estimated from the other
known variable.

•
It provides a measure of coefficient of correlation
(denoted by ’r') between the two variables, which is
the square root of the product of two regression
coefficients, i.e.,
 Utility of Regression Analysis
•
It establishes the rate of change in one variable in terms of
the change in another variable.

•
In regression analysis, one variable is assumed to be cause
and the other is the effect.

•
The coefficient of determination “r2” can be obtained by
multiplying the two regression coefficients. i.e

r2 = bxy . byx
CORRELATION
Correlation &
Regression – Diff.
REGRESSION
Correlation explains the degree &
Regression explains the cause and effect
direction of relationship between two
relationship between two variables.
variables.

In correlation, there is no dependent In regression, one variable is dependent

and independent variable. and other variable is independent.

Correlation coefficient value lies between It shows if there is change of 1 unit in the
±1. value of one variable, how much
average change will be there in another
variable.
Its use is limited upto linear relationship It studies the linear or non-linear
of two variabales. relationship between two or more
variables.
 Types of
•
Simple &

Mult RipleeRgegrreesssiosn:ion

In simple regression, two variables are

involved (one dependent and other independent).


In multiple regression, more than two variables
are involved (one dependent and
more than two independent).
 Types of
•
Linear and Non-linear Regression
Regression:


After plotting the data on graph paper, scatter diagram is
obtained. These dots follow some path, which may be
described by a line or a curve.


If dots gives no message, then variables are not related.
 Regression
•
In bivariate data,there are
lines of
l i ntewos
alwa y s
regression: Regression line of Y on X and
Regression line of X on Y.

•
In regression line of X on Y, Y is independent variable
and X is dependent variable.

•
In regression line of Y on X, X is independent variable
and
 Properties of
•
Regression
Both the regression lines intersect Lines
each other at mean point
(mean of X and Y).

•
If both the regression lines are perpendicular to each other,
then there is no correlation between X and Y. (r = 0)

•
If both the regression lines coincide each other, then there
is a perfect correlation between the variables X and Y. (r=1, -1)

•
There are always two regression lines.
 Methods of
•
Graphic studying
Method:
Regression
- Under this method, the pair of points are plotted on a graph
paper and a scatter diagram is obtained.

- Thereafter, a regression line is drawn in between these points.

 Methods of
Mathematical studying
Regression
•

Method:
There are two methods used for obtaining regression lines:

(a) Regression equations through Normal Equations.

(b) Regression equations through Regression Coefficients.

 Regression
•
Regression Line of Y Equations
on X: through
Normal Equations
In this, X is the independent variable and Y is the
dependent variable. The form of equation is given by:

Y (dependent) = a + bX(independent)

Above equation shows that Y is a linear function of

X. ’a' is called the intercept

 Regression
Equations
The values of ’a' and ’b' are through
determined by the method of least
squares. The following are Normal
the two normal Equations
equations:

ΣY = Na + bΣX............................................................(1)

ΣXY = aΣX + bΣX2......................................................(2)

Solving these equations, the value of ‘a' and ‘b' are obtained and
regression line is obtained.
 Regression
•
Equations
Regression Line of X on Y:
through
Normal Equations
In this, X is the dependent variable and Y is the
independent variable. The form of equation is given by:

X (dependent) = a + bY (independent)

Above equation shows that X is a linear function of

Y. ’a' is called the intercept

 Regression
Equations
The values of ’a' and ’b' are through
determined by the method of least
squares. The following are Normal
the two normal Equations
equations:

ΣX = Na + bΣY............................................................(1)

ΣXY = aΣY + bΣY2......................................................(2)

Solving these equations, the value of ‘a' and ‘b' are obtained and
regression line is obtained.
 Regression
(a)
Equations
Regression Equation of Y
through
on X:
Regression
Coefficients
The above equation can also be rewritten

as (Y - Y ) = byx (X - X )

(•)
Here, ơx = Standard deviation of X , ơy = Standard deviation of Y
 Calculation of
Regression
Coefficients byx
 Regression
(a)
Equations
Regression Equation of X
through
on Y:
Regression
Coefficients
The above equation can also be rewritten as

(X - X ) = bxy (Y - Y )

(•)
Here, ơx = Standard deviation of X , ơy = Standard deviation of Y
 Calculation of
Regression
Coefficients bxy
Kurtosis

Dr. Yamini Negi

 Kurtosis - Meaning
• Kurtosis is a statistical measure that defines
how heavily the tails of a distribution
differ from the tails of a normal
distribution.

• Kurtosis is a measure of the shape of a

distribution.

• It is a measure of the relative peakedness of

its frequency curve in relation to the normal
curve.
 Kurtosis – Meaning (Contd.)
• It indicates the extent to which a distribution is
more peaked or flat tapped than the normal
curve.

• Frequency curves can be divided into

three categories depending upon the
shape of their peak.

• The three shapes are termed as

Leptokurtic, Mesokurtic and
Platykurtic.(Theory*)
• Skewness measures the symmetry of the
distribution, while kurtosis determines the
heaviness of the distribution tails.(difference
*)
• In finance research, kurtosis is used as a
measure of
financial risk.
• High value of kurtosis signifies high level of risk
of an investment. (i.e. high probability of
extremely large or extremely small returns).
• Small kurtosis value shows moderate level of risk
as the probability of extreme returns are relatively
low.
 Excess Kurtosis
• The kurtosis of a normal distribution equals to
3. Thus, the excess kurtosis is found using the
formula below:

Excess Kurtosis = Kurtosis value - 3

 Types of Kurtosis
• The types of kurtosis are determined by the
excess kurtosis of a particular distribution.

• The excess kurtosis can take positive (4-3) or

negative (2-3) values as well as values close
to zero (3-3)
 Mesokurtic
• Data that follows a mesokurtic distribution
shows an excess kurtosis or zero or close to
zero.

• It means that if the data follows a

normal distribution, it follows a mesokurtic
distribution.
 Leptokurtic
• Leptokurtic indicates a positive excess
kurtosis. Leptokurtic distribution shows heavy
tails on either side, indicating the large outliers.

• In finance, a leptokurtic
distribution shows that
investment returns may
be prone to extreme
values
(investments are considered to be riskty).
 Platykurtic
• A platykurtic distribution shows a negative
excess kurtosis.

• In finance, platykurtic distribution shows that

there is a small probability that the investment
would experience extreme returns.
• Mesokurtic = Normal Curve = 3, excess kurtosis
=0

• Leptokurtic = More peaked than Normal

curve (positive excess kurtosis)

• Platykurtic = Less peaked than Normal curve

(negative excess kurtosis)
 SAMPLING:
TYPES & ERRORS

Dr. Yamini Negi

• A population can be defined as including all people or items with
the characteristic one wishes to understand.

• As there is very rarely enough time or money to gather information

from everyone or everything in a population, the goal becomes
finding a representative sample of that population.
Approaches of
Investigation:
• Census Investigation:
 When information on each and every unit of the population is
collected, it is called census investigation.

 The results in this approach are expected to be

accurate, representative and reliable.

 Example: Knowing the average life time of a batch of 5000 bulbs

produced by a certain machine on a given day.
 Approaches of
Investigation
(contd):
• Sample Investigation:
 When each and every unit of the population is not investigated but
only a part of the population called sample is investigated, it is
called sample investigation.

 Example: Choosing 100 bulbs and determining their average life

and inferences can be applied to all the bulbs.
Assumptions underlying in
sample
investigation:
• Homogeneity Amidst Complexity: If two units were alike in any
respect, the sampling would have been impossible.

• Possibility of Representative Selection: Each unit of the

population should get equal chance of being selected.

• Absolute Accuracy not essential: 95% accuracy is sufficient for

large scale studies.

• Independency: All items in a sample should be independent of

each other.
 Sample & Census
Investigation:
Comparison

• Cheaper

• Time saving

• Errors

• Preference

• Obtaining data
Steps in Statistical
Investigation
•
Objective of the survey
•
Population to be sampled
•
Data to be collected
•
Methods of measurement
•
Selection of the sample
•
The Pretest
•
Summary & Analysis of the data
•
Information gained for future surveys
 Terminology:
•
Sampling Fraction: The ratio of sample size ‘n’ to the population
size ‘N’.
•
Sampling Unit: An identifiable part of the population on which
some information has to be collected.
•
Sampling list or Source list: Ordered representation of the
sampling units of a population from which the sample is
selected.
•
Population Parameter: Mean of population
•
Estimator: Mean of sample.
 Methods of Sampling
•
Random Sampling or Probability Sampling

 Simple Random Sampling

 Stratified Random Sampling

 Systematic Sampling

 Multi-stage sampling

 Cluster Sampling
•
Non-Random Sampling or non-Probability sampling

 Judgement Sampling

 Quota Sampling

 Convenience Sampling
• Simple Random Sampling:

 Each unit of the universe has an equal chance of its selection.

 No item is given preference.

 With replacement sampling/without replacement sampling.

 Methods: By lottery, by random numbers.

• Stratified Random Sampling:

 Population under consideration is divided into number of sub-

populations called strata.

 Samples are selected independently from each stratum.

 Types: Equal allocation, Proportional allocation.

• Systematic Sampling:

 involves selecting only first unit at random, the rest being selected
according to some predetermined pattern.

 This sampling is also called “Quasi-random Sampling”.

 Here, k = N/n

where k = sampling interval,

N = Population

n = Sample
• Cluster Sampling:

 It involves grouping the population units and selected some of these

groups by random sampling technique.

 In cluster sampling, within cluster variation is more and

between cluster variation is less.
• Multi-stage sampling:

 This sampling is used to reduce the cost of survey.

 It involves selection of sample at various levels: from large-

sized sample to selection of smaller number of units.
• Non-Random Sampling:

 Sample is selected with a definite purpose and choice of the

sampling units depends entirely on the personal judgement of
the investigator.

 The sample size selected may not be the true representativeness

of the population.
• Judgement Sampling:

 This sampling is also called purposive sampling.

 In this, the sample is selected by the personal judgement of the

investigator.

 sample selected by this method is not a random one.

• Convenience Sampling:

 The sample is taken from a group of people who are easy to contact or to
reach. For example, standing at a mall or a grocery store and asking
people to answer questions would be an example of a convenience sample.

 This sampling is also called accidental sampling or grab sampling.

• Snowball Sampling:

 The research starts with a key person and introduce the next one to
become a chain.

 Make contact with one or two cases in the population.

 Ask these cases to identify further cases.

• Self Selection Sampling:

 It occurs when each case is allowed to identify their desire to take part
in the research.

 Need for the cases is publicized, either by advertising or by asking them to

take part.

 Collect data from those who respond.

• Sampling Errors:

 The difference between a sample estimate (mean of the sample) and the
population parameter (mean of the population) obtained is called the
sampling error.

 Sampling error = Sample Estimate – Population parameter

 These errors can be reduced by increasing sample size.

 A sample with minimum fluctuations is taken as the best representative of

the population.
• Non - Sampling Errors:

 The errors mainly arising at the stages of ascertainment and processing

of data is termed as non-sampling error.

 This error is common both to complete enumeration (census

investigation and sample surveys.

 Non sampling errors can of ‘Non Response Error’ and ‘Response

Error’.
• Non - Response Errors:

 A non-response error occurs when units selected as part of the sampling

procedure do not respond in whole or in part.
• Response Errors:

 A response or data error is any systematic bias that occurs during data
collection, analysis or interpretation.

 It can be any types like Respondent error, Interviewer error, Recording

errors, Poorly designed questionnaires, Measurement error.
• Respondent Error:

 A respondent gives an incorrect answer, e.g. due to prestige, or due to

social undesirability of question.

 Respondent misunderstands the requirements.

 Question requires memory/recall.

 Proxy respondents are used, i.e. taking answers from someone other than
the respondent.
• Interviewer Bias:

 Different interviewers administer a survey in different ways.

 Differences occur in reactions of respondents to different interviewers.

 Inadequate attention to the selection of interviewers.

 Too much workload on the interviewer.

• Measurement Error:

 The question is unclear, ambiguous or difficult to answer.

 Requested information assumes a framework unfamiliar to the respondent.

Test of
SignifiCanCe:
Hypothesis Testing
 Hypothesis
•
A hypothesis is a tentative
generalisation, the validity of which
remains to be tested.

•
In its most elementary stage, the
hypothesis may be a any guess,
imaginative idea, which become the basis
 BasiC ConCepts:
Null Hypothesis:

•
A null hypothesis is a statement about
population parameter (such as µ), and the
test is used to decide whether or not
accept the hypothesis.
•
The null hypothesis is always expressed in
the form of mathematical statement which
includes:

H0 : µ(≤, , ≥) µ0
•
Alternate Hypothesis:
The alternate hypothesis states that specific
population parameter value is not equal to the value
stated in the null hypothesis and can be written as:

H1 : µ ³ µ0
Procedure in Hypothesis Testing:
(1) Formulate a Hypothesis

(2) Set up a suitable signifiCanCe level.

(3) Select Test Criterion (z test, T test, F test,

Chi- square test)

(4) Compute the

Types of Errors in Hypothesis Testing:

When a hypothesis is tested, there are four

possibilities:

(1) Hypothesis is true but our test leads to its

rejection.
Type I Error:

Hypothesis is true but our test leads to its

rejection.

Type II Error: