STATISTICS DEFINED

There have been many definitions of the term

articles have carefully collected together 'Statistics'-indeed
hundreds of
scholarly
have defined Statistics as
statistical definitions. Some
data (plural sense) whereas
statistical methods (singular sense). Afew others as
examined below. definitions are analytically
Statistical Data
Quantitative or numerical information may be found almost
business, economics and many other areas. It is everywhere in
to refer to data in
quantitative form as probably more common
numerical data is statistical and hence it isstatistical data. But not all
necessary to examine a few
definitions of statistics to understand the characteristics
According to Webster "Statistics are the classified factsof statistical data.
conditions of the people in a State. especially those factsrepresenting
which can be
the
stated in numbers or in any tabular or classified
The above definition is too narrow as it arrangement".
confines
to only such facts and figures which relate to the the scope of statistics 1
people in aState. conditions of the
DS STATISTICS-WHAT AND WHy

yule and Kendall defined statistics thus: "By Statistics we meai

S. duantitative data affected to a marked extent by multipllcity of causes.
ns This definition is less comprchensive than the one given by PrOl.
he Horace Secrist who defined statistics as follows:
"By Statistics we mean aggregate of facts affected to a marked extent by
or estirmated
iultiplicity of causes, numerically expressed, enumerated
are according to reasonable standards of accuracy, collected in a systematiç
he manner tor a predetenined purpose and placed in relation to each other."
C.. This definition clearly points out certain characteristics which numerical
to data must possess in order that they may be called statistics. These are
as follows:
of Wacheist
ble OStatistics are Aggregate ¡f Facts. Single and isolated fîgures are not
to statistics or the simple reason that such figures are unrelated and cannot
be compared. To illustrate, if it is stated that the income of Mr. X is
it.
has
Rs. 5 lakh per annum, this would not constitute statistics although it is a
yse
numerical statement of fact. Similarly, a single figure relating to production,
sale, birth, employment, purchase, accident, etc. cannot be regarded statistics
although aggregates of such figures would be statistics because of their
that comparability and relationship as parts of a common phenomenon.
es is
962). (it) Statistics are Affected to a Marked Extent by Multiplicity of Causes.
stics, Generally speaking, facts and figures are affected to a considerable extent
ortant by a number of forces operating together. For example, statistics of
ution production of rice are affected by the rainfall, quality of soil, seeds and
and manure, method of cultivation, etc. It is very difficult to study separately
the effect of each of these forces on the production of rice. The same is
v of true of statistics of prices, imports, exports, sales, profits, etc. In the
st of experinental sciences like Physics and Chemistry it is possible to isolate
more the effect of various forces on a particular event. Ways and meansforces are
also being devised in 'Statistics' for segregating the effect of various
in statistical
on an event. However, it has proved to be a difficult task
studies of phenomena which are influenced by a complex variety of
factors, many of whích are not measurable.
larly statistics are numerical
Some (iy Statistics are Nunerically Expressed. All
s as statements of facts, Le., expressed in numbers. Qualitative statements
cally such as 'the population of India is rapidly increasing': or the production
not
of wheat is not sufficient'; or 'India is a developing country' do and
constitute statistics. The reason is that Such statements are vague the
hand,
one cannot nake out anything from them. On the other crore
statement, "The population of India is expected to increase from 121
re in in 201l to 140 crore in 2021 is a statistical statement.
mon
t all are Enumerated or Estimated According to Reasonable Standards
o StatisticsFacts
few of Accuracy. viZ.. byand figures about any phenomenon can be derived in
actual counting and measurement or by estimate.
data. two Ways,
g the Estimates cannot be as precise and accurate as actual counts or
measurements. For example, an estimate that 5 lakh people witnessed the
n be
Republic Day parade does not mean exactly 5 lakh; it may be a few
istics
hundreds or thousands more or less. On the other hand, if we count the
number of students in a class and say lat there are 60 students, this
f the
figure would be l00 per cent accurate. In many cases, 100 per cent
 CAL METHODS

accuracy of numbers may be dMcult to attain. The degree of accuracy

desired largely depends upon the nature and object of the enquiry. For
example, in measuring heights of persons even 1/10th of of a cm. is
material whereas in measuring distance between two places, say Delhi and
Mumbai, even fraction of a kilometre can be ignored. Hernce, in many
statistical studies mathematical accuracy cannot be attained. However, it
is important that reasonable standards of accuracy should be attained,
otherwise numbers may be altogether misleading.
( ) Statistics ae Collected in a Systematic Manner. Before
collectirng
statistics a suitable plan of data collection should be prepared and the
work carried out in a systematic manner. Data collected in a haphazard
manner would very likely lead to fallacious conclusions.
(wi Statistics are Collected for a Predetermined Purpose. The purpose
of collecting data must be decided in advance. The purpose should be
specific and well-defined. A general statement of purpose is not enough. For
example. if the objective is to collect data on prices, it would not serve any
useful purpose unless one knows whether he wants to collect data on
wholesale or retail prices and what are the relevant commodities in view.
lvio Statistics Should be placed in Relation to Each Other. If numerical
facts are to be called statistics, they should be comparable. Statistical
data are often compared period-wise or region-wise. For instance, the
population of India at a particular point of time may be compared with
that of earlier years or with the population of other countries, say the
U.S.A., the U.K., China, etc. Valid comparisons can be made only if the
data are homogeneous, Le., relate to the same phenomenon or subject
and only likes are compared with likes. It would be meaningless to
compare the height of elephants with the height of human beings.
In the absence of the above characteristics, numerical data cannot be
called statistics and hence "all statistics are numerical statements of facts
but all numerical statements of facts are not statistics."
It may be noted that the statistician's job is to prepare
meaningful
numbers, often involving summaries of numbers or other data. The
statistician must present such data in a form that is useful to the
recipient of those data,.
 Diagrammatic and
Graphical Representation
Collection of data is the basis for any statistical inquiry and the collected
data should be first classified. Then the classified statistical data may be
presented in the formof tables as well as diagrams and graphs. Diagrams
andgraphs are more attractive and have a visual appeal. Hence they are
more impressive even to laymen and help one in making quick and accurate
comparison of data.
TYPES OF DIAGRAMS
There are a large number of diagrams by which a statistical data can be
presented in practice. They are grouped under the following types:
() One-dimensional or bar diagrams
(ii) Two-dimensional or area diagrams
(ii) Three-dimensional or volume diagrams
(iv) Pictograms and cartograms
One-dimensional or Bar Diagrams
One-dimensional or bar diagrams are most popular form of diagramatic
representation. This type of diagrams consist of a series of rectangular
bars standing on a common base. The length of the bars are proportional
to the magnitude of the variable. The comparison of bars is based only on
lengths. The bars may be either vertical or horizontal. But the vertical bars
are preferredas they give a better look and facilitate comparison. The bar
diagrams are classified into the following five types:
Simple bar diagrams It represents the magnitude of a single variable
like population, sale, production, profits, etc.
 DGP-2 Business Statistics and Operations Rese arch

Example

250

200
200

(SBuuo u) 150
150

100 75

50

Copper Iron Alumnium

Asimple bar diagram showing India 's export as on 2003

Multiple bar diagrams It is used to compare two or more sets of
statistical data. So in amultiple bar diagram two or more bars are constructed
side by side to represent set of values for comparison. Since more than
one phenomenon is represented, different shades or colors are used to
distinguish between the bars.
Example

500
450

+ Rice
400 390 390
+Wheat
300
300 ’Barley
220
200

100 100

2002 2003

Multiple bar diagram showing the production of food crops

in India

Subdivided bar diagrams These diagrams are used to represent the

various parts of awhole. In this type, each bar representing the total value
is subdivided into its different component parts. This enables comparison
 DGP:3
Diagrammatic and Graphical Representation

between different components and also between a component and the

whole.
Example

100
100
90 + Science
thousands)
(In 80
+Commerce
70
60
60 Arts

E 40

20

2000 2001 2002 2003

Subdivided bar diagram showing the number of students enrolled

in auniversity over a periodof 4years
Percentage bar diagrams This is another form of component bar
equal to 100 and the
diagram. In this type, the length of the bars are kept
components are subdivided in these equal bars.
Example

100
Transport

80 + Health
(Percentage)
’ Education

Agriculture
40

20

2002 2003 Year

2000

Percentage bar diagram showing the percentage of government

different services
financial resources spent on
Deviationbar diagrams These are popularly used for representing
or deficit net exports or
loss, excess
net quantities such as net profit, net
 Operations Rese arch
DGP4 Business Statistics and
have both positive and negative values. Positive
impots, etc.These bars can
above the base line and negative values below it.
values areshown
Example
Profit

15
15
Loss

10

10

2001 2003

2000 -2002

-5

-8
-10
-14

Deviation bar diagram showing the profit and loss of a textile

industry
Two-dimensional Diagrams
In this type, the length as well as the width of the bars are considered.
Thus the area of the bars represent the given data. These are also known
as surface diagrams or area diagrams. Rectangles, squares, and circles
come under this category. The area diagram is more difficult to read than to
construct because it is difficult tojudge the areas and the relative
magnitudes with precision.
Pie diagram A pie diagram is pictorial representation of a statistical
data with severalsubdivisions in a circular form, The various component
values are converted into degrees by taking 360° as awhole. Acircle is
drawn withconvenient radius and different sectors are drawn with angles
at the center to represent the different components. The areas of these
sectors are proportional to the magnitude of the components.
 Diagrammatic and Graphical Representation DGP:5

Example

Two wheelers

35% Lorries

’Cars

Others

Pie diagram showing the percentage of vehicles in chennai

Three-dimensional Diagrams
In these type of diagrams, three things, length, width, and height are taken
into account. These are also known as volume diagrams and consists of
cubes, cylinders, spheres, etc. These diagrams are not recommended for
statistical presentation as they are difficult to read and compare. Volume
between smallest
diagrams are used in those cases where the difference
and largest values to be represented is very large.
Pictograms and Cartograms Pictograms and cartograms are very
informative and
elementary from of visual presentation. But they are more
are drawn to
more effective than other forms. In pictograms, pictures give only an
can
represent the data. Although they are attractive, they
overall picture and difficult to construct.
Example
Banks

Offices

Houses 100 Computers

i00 Air condtioners

afirm
Number of computers and air-conditioners distributed by
in a town
DGP6 Business Statistics and Operations Research

Cartogranms are statistical maps, used to give quantitative information

on geographical basis. The quantitatives on the map can be shown in
many ways such as shades, colors, dots, etc.
GRAPHICAL REPRESENTATION OF FREQUENCY
DISTRIBUTIONS
Afrequency distribution can be presented graphically in any one of the
following ways:
(0) Histogram
(i) Frequency polygon
(ii) Frequency curve
(iv) Ogives or cumulative frequency curve
(v) Lorenz Curve
Histogram
It is one of the most important and useful methods of presenting frequency
distribution of acontinuous series. In a continuous series, the class intervals
may be equal as well as unequal.
When the class intervals are equal, we construct adjacent rectangles
or bars with the heights indicating the corresponding frequencies (along
Y-axis) and the width indicating the size of the variable (along X-axis).
Example
14
14

12 12

10 10

8
No. of firms 6
6
5

10 20 30 40 50 60 70 80 90

Sales (in lakhs)

Histogram showing the distribution of firms according to their sales
 Diagrammatic and Graphical Representation DGP7

Whenthe class intervals are unegual, acorrection must be made. For

' making the adjustment, we take the class which has lowest class inierval
Ifone class interval istwice as wide as the one having lowest class interval
we divide the coresponding frequency by two, if it is thrice more, we
divide thc coresponding frequency by three and so on.
Example
Let us consider the following distribution of workers as per their weekly
wages:

Wcckly wages (in Rs.) 10- 15 I5- 20 20 - 25 25-35 35 -45

No. of workers 7 13 4

Since the class intervals are unequal, a correction must be made. The
lowest class difference is 5. So the frequency of the class 25 - 35 and
35 -45 is divided by two. The histogram is drawn as follows:

14 13

12

10

7
No. of workers 6 5

4
2
2

20 25 30 35 40 45
10 15

Daily wages (in Rs.)

workers according to their
Histogranshowingthe distribution of
weekly wages
Frequency Polygon
polygon is a graph generally obtained by connecting the
The frequency histogram by a straight line.
midpoints of the top of each rectangle ofa polygon is to take the mid
frequency
Another method of constructing and then plot the frequency
various class intervals
points of the these by straight lines. Frequency
corresponding to each point and join all
open end classes.
polygon cannot be used for distributions having
 Research
DGP-R Business Statistics and Operations

Example

Students
ofNo. 50
41
40 37
30
30

20
YO
10

10 20 30 40 50 60 70

Marks

Histogram and frequencypolygon

Frequency Curve
When we smoother a frequency polygon we get frequency curve. The
curve is drawn freehand through the various points of the polygon such
that the total area under the curve is equal to that of polygon.
Example

36
35

30
Frequency 25
25
20
20

15
15

10

100 110 120 130 140 150 160 170

Variable

Histogram andfrequency curve

 Diàgrammatic and Graphical Representation 0GP9
Cumulative Frequency Curve or Oglves
When frequencies are added, we call it as cumulative frequenuy. When
such cumulative frequencies are plotted in a graph, we get a curve called
cumulative trequency curve or ogive. There are two methods of
constructing ogive.
Less than method In this method,the cumulative frequercy is related
tothe upper limits of theclass and is obtained by adding the frequencies
of the class. When these frequencies are plotted we get a rising curve
called less than ogive or cumulative frequency curve.
More than method In this method, the cumulative frequency is related
to the lower limit of the classes and is obtained by subtractipg each
frequency of the class from the cumulative total. When these frequencies
are plotted, we get a declining curve known as more than ogive.
Example
30-40 40-50 50-60 60-70
C.I 10-20 20-30
7 12 9 4
F 2 5

Less than 20 30 40 50 60 70
2 7 14 26 35 39
Cf
20 30 40 50 60
More than 10
39 37 32 25 13
Cf

Le
tshs
a
ogni v e
40

35 More than ogive

30
Frequency
25

20

15

10

50 60 70 80
30 40O
10 20
Variable