REPRESENTATION OF DATA

Structure
Objectives Introduction Stages of Statistical Inquiry Arrangement of Data
3.3.1 Simple Array 3.3.2 Frequency Array or Discrete Frequency Distribution 3.3.3 Continuous or Grouped Frequency Distribution 3.3.4 Various forms of Frequency Distributions

Tabulation of Data
3.4.1 Meaning and Types of Tables 3.4.2 Parts of a Table 3.4.3 Importance of Tables

Graphical Presentation of Data

3.5.1 Line Graphs 3.5.2 Histogram, Frequency Polygon and Frequency Curves 3.5.3 Cumulative Frequency Curves - Ogives

Diagrammatic Presentation of Data

3.6.1 One Dimensional Diagrams 3.6.2 Two Dimensional Diagrams or Area Diagrams 3.6.3 Pie Diagram or Pie Chart 3.6.4 Three Dimensional Diagrams 3.6.5 Pictograms and Statistical Maps

Let Us Sum Up Key Words Some Useful Books Answers or Hints to Check Your Progress Exercises

3.0 OBJECTIVES
On going through this Unit, you will be able to explain: stages of statistical inquiry after data have been collected; methods of organizing (classification and arrangement) and condensing statistical data;

concepts of frequency distribution and its various types; and different methods of presentation of statistical data such as tables, graphs, diagrams, pictograms, etc.

Data and Its Presentation

3.1 INTRODUCTION
In the preceding Unit, we discussed the methods of collection of data either by a statistical survey (or inquiry) or from some secondary source. Data collected either fiom census or sample inquiry, that is fiom primary source, are always hotchpotch and in rudimentary form. To start with, they are contained in hundreds and thousands of questionnaires. To make a head and tail out of them, they must be organised, (i.e., classified and arranged) and condensed or surnmarised. For this purpose we can use various methods like preparing master sheets in which various information are recorded directly h m the questionnaires. From these sheets small summary tables can be prepared manually. Now-a-days computers can be used for organisation and condensation of data more swiftly, efficiently and in much less time. Some computer softwares are available which help us to construct various types of graphs and diagrams. Data can be summarized numerically also. Here we use summary measures like measures of central tendency (such as Arithmetic, Geometric and Harmonic Means, Mode and Median); measures of dispersion (such as Range, Quartile Deviation, Mean Deviation, and Standard Deviation); measures of association in bivariate analysis (such as Correlation and Regression), Index Numbers, etc. In this Unit we plan to discuss how data can be summarized using tables and graphs. Numerical summarization will be discussed in subsequent Blocks (2,3 and 4). It must be kept in mind that a good summarization and presentation of data is not undertaken for its own sake. It is not an end in itself In fact it sets the stage for usehl analysis and interpretation of data. Again, a good presentation helps us to highlight significant facts and their comparisons. Figures can be made to speak out thereby making possible their intelligent use.
'

3.2 STAGES OF STATISTICAL INQUIRY

As studied in Section 2.3 of Unit 2, a statistical survey or inquiry is undertaken Unit in two stages, namely, the planning stage and the executing stage. In th~s we plan to concentrate on some aspects of the executing stage. This involves organising and condensing data in the form of simple array (ascending and descendmg order), fieqbency array and continuous frequency distributions, etc.; and presentation of statistical data in the form of tables and graphs.

3.3 ARRANGEMENT OF DATA

The mass of collected data is o h voluminous, unintelligible and boring. It seems totally uninteresting and is not easily inteqmtable. For example, if you are provided with monthly income figures of 1000 families in a village it is difficult for you to infer anything. But if you are told that the average monthly income of the village is Rs. 2540, it is quite interesting and you are in a position to compare it with other figures.
The first step in the analysis and interpretation of data is its classification and tabulation. The process of arranging data into groups according to their common characteristics is known as its classification. On the other hand tabulation implies s systematic presentation of data in raws and columns zccording to same salient

In Unit 2, a q~estionnaire was prepared on Family Planning. Suppose this questionnaire was used to collect information from 50 families of C-III Block of XYZ Colony, New Delhi. Let us assume that it producedthat following types of information as given in Tables 3.1 and 3.2. Can we make any head or tail out o it? f
Table 3.1 Number of Children per family in C-111 block, XYZ Colony, New Delhi

Tabulation end Graphical

Representation of Data

Table 3.2 Monthly Income of 50 families of C-111 block, XYZ Colony, New Delhi
547 622 691 684 567 586 680 578 583 578

As pointed out earlier, to make any head or tail out of the mass of raw data, such as presented above, we have to classi@ and arrange it. This can'be done either by forming a simple array or a frequency array (discrete frequency &stribution) or a continuous frequency distribution Sub-sections 3.3.1,3.3.2 and 3.3.3 attempt to explain this aspect.

3.3.1 Simple Array

It is an arrangement of given raw data in ascending or descending order. In the ascending order, the observations are arranged in increasing order of magmtude. For example, numbers 3,5,7,8,9,10 are arranged in ascenlng order. In descending order, it is the reverse. For example, the numbers 10,9,8,7,6,5,3 are in descending order. We can prepare both types of simple arrays fiom Table 3.1. In the following table, the figures have been arranged in ascending order. From the arrangement, it is clear that the lowest value is 0 and the highest one is 5.
Table 3.3 Number of Children per Family in c-111 Block of XYZ Colony, New Delhi Simple Array - Ascending Order

D a t a and

I ~ S Presentation

After the arrangement of data in ascending order as in Table 3.3 the raw data make some sense. The possible conclusions that can be drawn from this arrangement of data (see Table 3.3) are that five families are issueless, twelve fm'lies have one child each, fourteen have two children each, ten families posses three children each, six families have four children each and three families have five children each.

3.3.2 Frequency Array or Discrete Frequency Distribution

Here different observations are not repeatedly written as in simple array like 0, 0,0,0,0, 1, 1, 1, 1, 1, 1, 1, etc. We count the number of times (i.e., frequency) an observation repeats itself. For example, in Table 3.3 the observation 4 is repeated 6 times. Thus the frequency of 4 above is 6. The frequency array, for the simple array given in Table 3.3, will look like as given below in Table 3.3. A frequency array is a statistical table in which various observations are arranged in order of their magnitude along with their respective frequencies.

Table 3.4
Number of Children :
Number of Families :
L

0
5

1 12

4 6

5
3

Total
X)

14

10

When the number of observations is large enough, the counting process is often undertaken by the use of tally bars. In this method, all possible values of the variable are written in a column. For every observation, a tally bar denoted by ( ( ) is noted against its corresponding values. Every fifth repetition is marked by crossing the previous four bars as ( m ) . In this way, we get blocks of five which simpli& counting at the end. Thus a nurnber or an observation repeated fourteen times will be marked as (Mm I I II).Note that after representing each observation by a bar on the tally sheet, the same will be ticked (3) or crossed (5) so that it is not duplicated. The data of Table 3.1 is rewritten in the form of frequency distribution as shown in Table 3.5 below:

Table 3.5 Frequency Distribution of Number of Children per Family

No. of Children

Tally Sheet

Frequency

3.3.3 Continuous or Grouped Frequency Distribution

Numbers like 1, 2, 3,4, 5,20,40, etc. are discrete numbers and are used where no value between the two consecutive numbers is possible. As in the case of the number of children, it will be impossible as well as h y to say that a particular family has 2.083 or 2.1 or 2.75 number of children. The family can have either 2 or 3 children and not a fraction in between. Out of the two examples of raw data mentioned in Section 3.3, the number of children (Table 3.1) is an example of discrete data while monthly income (Table 3.2) is an example of continuous variable giving continuous data. In this Sub-section we propose to illustrate the construction of continuous or grouped fi-equencydistribution fiom the raw data of Table 3.2 on monthly income of the 50 families. To construct a grouped frequency distribution, the range of the given.data, i.e., the difference of the highest and the lowest observations, is divided into various mutually exclusive and exhaustive sub-intervals, also h o w n as class-intervals. The frequency of each class interval is then counted and written against it.
Table 3.6 Frequency Distribution of Monthly Income of Families

Tnbulatlon and Graphical Representation of Data

Monthly Income

Tally Sheet
'THL ?t+lI

(Rs.1
500 - 550 550 - 600 600 - 650 650 - 700 700 - 750 750 - 800 800 - 850
Total

No. of Families (Frequency) 5 6 1 0 1 2 9

7HITHI 7HI?t+lII

'THLIIII
'THL

5
3 50

Ill

In Table 3.6 we have completed an exercise where the variable "incomeo the f family" has been grouped in order to reduce it to a manageable form called grouped data or Continuous Frequency Distribution. However, prior to the construction of any grouped fi-equencydistribution, it is very important to find answers to the following questions: 1) What should be the number of class intervals? 2) What should be the width s f each class interval? 3) How will the class limits be designated? 1) What should be the number of class intervals? Though there is no hard and fast rule regarding the number of classes to be formed, yet their number should be neither too small nor too large. If the number of classes is too small, i.e., width of each class is large, there is likelihood of greater loss of information due to grouping. On the other hand, if the number classes is very large, the distribution may appear to be too fiagrnented and may not reveal any pattern of behaviour of the variable. Based on experience, it has been observed

Data and Its Presentation

that the minimum number of classes should not be less than 5 or 6 and in any case, there should not be more than 20 classes. Usually the formula to determine the number of classes is given by Number of classes = 1+ 3322 x log,, N , where N is the total number of observations.

In our example of raw data on incomes of 50 families, the number of classes can be calculated as under:
Number of classes
= 1+ 3322 x log,, 50
=1
=

1 + 3.322

x 1.6990

+ 5.644 = 6.644= 7.

2) What should be the width of each class interval? As far as possible, all the class intervals should be of equal width. However, when a frequency distribution, based on equal class intervals, does not reveal a regular pattern of behaviour of observations, it might become necessary to re-grobp the observations into class intervals of unequal width. By a regular pattern of behaviour we mean that there are no classes, with possible exclusion of extreme classes, where there are nil or very few observationswhile there is concentration of observations in their adjoining classes. The approximate width of a class can be determined by the following formula:
Width of a Class = Largest Observation - Smallest Observation Number of Class Intervals

However, the final decision, regarding width of class intervals, should also take into account the following points.

0 As far as possible, the width should be a multiple of 5, because it is easy to grasp numbers like 5, 10, 15, ..... etc.
i) It should be convenient to find the mid-value of a class.
iii) The observations in a class should be uniformly distributed.

3) How will the class limits be designated?

The srpallest and the largest observations of a class interval are known as class limits. Thlsebre also termed as the lower and upper limits of a class, respectively. Since the mid-value of a class, which is used to compute mean, standard deviation, etc., is obtained fiom the class limits, it is necessary to define these limits in an unambiguous manner. The following points should be kept in mind while defining iis class lmt: a) It is not necessary that the lower limit of the first class be exactly equal to the smallest observation of the data. In fact it can be less than or equal to the smallest observation. Similarly, the upper limit of the last class.may be geater than or equal to the largest observation of the data. b) It is convenient to have the lower limit of a class either equal to zero or some multiple of 5 or 10. c) The chosen class limits should be such that the observations in a class are uniformly distributed.

The class limits can be defined in either of the following methods:

Tabulation and Graphical Representation of Data

i) Exclusive Method, and ii) Inclusive Method.

i) Exclusive Method : In this method, the upper limit of a class is taken to be equal to the lower limit of the following class. In order to keep various class intervals as mutually exclusive, it is decided that the observations with magnitude greater than or equal to lower limit but less than the upper limit of a class are included in it. For example, the class 500 - 550 shall include all observationswith magnitude greater than or equal to 500 but less than 550. An observation with magnitude equal to 550 will be included in the next class, i.e., the class 550 - 600. The major benefit of exclusive class intervals is that it ensures continuity of data because the upper limit of one class is the lower limit of the next class. In our example on monthly income (Table 3.6), there are 5 families whose income lies between Rs. 500 to Rs. 550, i.e., Rs. 500 to 549 and 6 families whose income lies between Rs. 550 to Rs. 600, i.e., Rs. 550 to 599, and so on. Based on this presumption we can rewrite this fi-equencydistribution in the form of Table 3.7 also.
Table 3.7 Exclusive Class Intervals

Monthly Income (Rs.)

500 but 550 but 600 but 650 but 700 but 750 but 800 but less than less than less than less than less than less than less than 550 600 650 700 750 800 850

Number of Families (Frequency)

1 1

10 12

I I

9 5

3
50

Total

ii) Inclusive Method :In this method, all the observationswith magnitude greater than or equal to the lower limit but less than or equal to the upper limit of a class is included in it. Now observe Table 3.8. Income of Rs. 549 is included in the class 500 to 549 so that an income of Rs. 550 automatically goes to the next class of 550 to 599. Since the upper limit of one class is not equal to the lower limit of the following class, this saves us fiom the conhsion whether Rs. 550 goes to (500 to 549) or (550 to 599) class.
Table 3.8 Inclusive Class Intervals Monthly Income (Rs.)
500 - 549

Number of Families (Frequency)

Data and Its Presentation

I he cmolce between excZusive and incluszve rriclhads depends upon whether we are dealing with continuous variable like income, heights, weights, etc. or a discrete variable like number of children in a family. For a continuous variable it is desirable to construct fiequency distribution by the exclusive method because, as we have seen earlier, it ensures continuity. For a discrete variable like number of children in a family or number of students getting h t division, the frequency distributions should be constructed by using inclusive type of class intervals.

Table 3.9 Class Boundaries of Inclusive Class Intervals

Monthly Income (Rs.)

Number of Families (Frequency)

499.5- 549.5 549.5 - 599.5 599.5 - 6493 649.5 - 699.5 699.5- 749.5 749.5 - 799.5 799.5- 849.5
Total

5
6

1 0

1 2

9 5
3
50

Mid-Value of a Class
In exclusive type of class intervals,the mid-value or class mark of a class is defined as the arithmetic mean of its lower and upper limits. However, in case of inclusive class intervals, there is a gap between the upper limit of a class and the lower limit of the following class. This gap is eliminated by adding half of the gap to the upper limit and subtracting halfofthe gap from the lower limit. The new class limits, thus obtained, are known as class boundaries. The class boundaries of the inclusive class intervals in Table 3.8 are given in Table 3.9.

3.3.4 Various Forms of Frequency Distributions

Here we propose to intraduce the meaning of the following iiquency distributions: a) Open End Frequency Distribution b) Frequency Distribution with Unequal Class Width c) Cumulative Frequency Distribution d) Relative Frequency Distriiution
a) Open End Frequency Distribution Open-en4 fiequency distribution is one which has at least one of its ends open. Either the lower limit of the first class or upper limit of the last class or both are not specified. The words "below" or "less than" and "above7' or "more than" are used. In the former the value extends to - m and in the latter to + m. Example of such a fiequency distribution is given in Table 3.10.

a n

Table 3.10 Open-end Class Frequency Class Frequency

Table3.11 u;equal Class Frequency Class Frequency

Tabulation and Graphical Representation of Data

Below 25

20 - 25 25 - 30

25 - 30 30 - 40 40 - 50 50 and above

3
5 2 1
12

3
5

. 30-40
40 - 55 5 - 60 5

2 1
12

Total

Total

b) A Frequency Distribution with Unequal Class Width The classes of a fiequency distribution may or may not be of equal width. A fiequency distributionwith unequal class width is reproduced in Table 3.11. Here, the width of lst, 2nd and 5th classes is 5, while that of 3rd is 10 and that of 4th is 15. As we will see in Unit 4, mode is not a representative value in such types of series and hence not defined. c) Cumulative Frequency Distribution Suppose that, with reference to data given in Table 3.6, we ask the following questions:

How many families have their monthly income less than or equal to Rs.700? n How many families have their monthly income greater than or equal to ) Rs. 600? The answers to the above questions can be ehsily obtained by forming an appropriate cumulative kquency distribution. To answer t h question, we need k t to form a "less than type" cumulative fkquency distribution while a "great& than type" cumulative frequency distribution is required for answering the second question. These distributions are given in Tables 3.12 and 3.13 respectively.
Table 3.12 "Less-than type" Cumulative Frequency Distribution Monthly Income (Rs.) Frequencies

Simple Less than 550 Less than 600 Less than 650 Less than 700 Less than 750 Less than 800 Less than 850

Cumulative

5
6
5+6

5 11 2 1
I

1 0
12
9

5410 5+6+10+12 5+6+10+12i9 561+2!5 ++01+% 5+6+10+12+9+5+3

3 3
42

.5 3

47
5 0

Data and Its Presentation

Table 3.13 "More-than type" Cumulative Frequency Distribution Monthly Income (Rs.)
Slmple More than 500 More than 550 More than 600 More than 650 More than 700 More than 750

Frequencies
Cumulative

5 6

10
1 2 9 5
3 3

More than 800

d) Relative Frequency Distribution So far we have expressed the fiequency of a value or that of a class as the number o! times an observation is repeated. We can also express these frequencies as a f fraction or apercentage of the total number of observations. Such frequencies are known as the relativefrequencies. Table 3.14 demonstrates the construction of relative fiequency distribution.
Table 3.14 Relative Frequency Distribution of Monthly Income of 50 Families
Class Frequency Relative Frequency As a fiaction As a percentage

500 - 549 550 - 599 600 - 649 650 - 699 700 - 749 750 - 799 800 - 849
Total

5 6 1 0

5 t 50 = 0.10 6 + 50 = 0.12 10 + 50 = 0.20 12 + 50 = 0.24 9 + 50-0.18 5 + 50 = 0.10

0.10 + 100 = 10 0.12 + 100 = 12 0.20 + 100 = 20 0.24 + 100 = 24 0.18 + loo= 18 0.10 + 100 = 10 0.06 + 100 = 6
100

12
9

5
3
50

3 + 50 = 0.06
1

firom the above table it is clear that sum of the relative ikquencies should be either 1 (in case of fraction) or 100 (in case of percentage).

Check Your Progress 1

1) Distifiguish between the following, giving at least two points of distinction.

a) Discrete and continuous frequency distributions b) Simple and cumulative kequency distributions c) Exclusive and inclusive class intervals d) Simple a d frequency array

.............................................................................................................

Tabulation and Graphical Representation of Data

.............................................................................................................
2) Explain the following terms giving examples: a) Ungrouped data b) Class mark c) Open end classes d) Class limits e) Class boundaries f) Class fiequencies g) Tally bar h) Relative fiequencies

.............................................................................................................
3) Build a hypothetical fkequency distribution on monthly pocket money of 20 students belonging to the lower middle class of a college. Prepare a relative frequency distribution fi-om it.

Data and Its Presentation

4) What points are to be kept in mind while taking decisions for preparing a frequency distribution in respect of :

a) The number of classes, and b) Width of the class interval?

...................................................................................................................
5) Construct less than and more than type cumulative frequency distributions

from the following data: Class: 10 - 20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 Frequency: 5 8 10 12 8 7

...................................................................................................................
6) Construct a relative frequency distribution for the data given in question 5.

3.4 TABULATION OF DATA

Good presentation of data is as important as their satisfactory collection and arrangement. In fact, satisfactory collection and arrangement of data must be followed by good presentation. However, good presentation is not an end in itself. It may be necessary for satisfactory analysis and interpretation. A satisfactory is& presentation helps us in more than one ways F r t it helps to highhght significant
fwt~ rnntaind in t

h data .CPrnndh, it farilitatec the cnmnnrisnn nf data Finalhr. ~

We will discuss presentation of statistical data under three heads.

1") Formal tables

Tabulation and Graphical Representation o f Data

ii) Graphic methods which will include line graphs, histograms, fiequency polygon and curves, and cumulative fiequency curves.

iiii Geometric forms, pictures and statistical maps, which will include pie diagrams, bar diagrams, area and volume diagrams, etc.

In thls Section we concentrate on tabular forms of presentation.

3.4.1 Meaning and Types of Tables

A table or a statistical table is a systematic arrangement of related statistical data in columns and rows, with a given predetermined and a well decided objective. A row of a table represents a horizontal while a column represents a vertical arrangement of data. To explain the natwe of information given in a table, its rows and columns are designated by appropriate stubs and captions (or headings or sub-headings) respectively. Presentation of data in a tabular form should be simple, planned, unambiguous and logical.
'

Table 3.15 is based on hypothetical figures of exports and imports of country X with country B' for three years 1995, 1996 and 1997.
Table 3.15 Imports and Exports o f i with Country B during 1995 1997

(In Crore of Rupees)

Country
I

Imports

1995

Exports

Imports

1996

Exports

Imports

1997@

Exports

Total
@

195

202

225

235

240

230

Note : Figures are quick estimates. Source : Trade Bulletin, 1998, Ministry of Foreign Trade of X.

In this table it is clear that the purpose is to show the imports and exports of country X vis-a-vis the rest of the world. Note that a particular entry of the table refers t o ~ o l u m n a row. For example, an entry at the intersection of second row and and fourth column indicates that in 1996 country X imported goods and services worth Rs.60 crore b m country B. This figure then can be compared with other import and export figures to seek important interpretations.

Types of Tables
Basically, we have two types of tables: 1) Reference tables or general purpose tables
2) Text tables or special purpose tables. 1) Reference tables are a general purpose tables and are a store of information with the aim of presenting detailed statistical information. From these tables, we can derive our information (i.e., secondary source). Tables presented by different government departments, ministries, Reserve Bank of India, Economic Surveys, etc. are reference tables and are a routine work of these departments.

Data and

I ~ S Presentation

Another important example is the Population Census tables prepared by the Registrar General of India giving detailed information on the dernographik features of India. Students are advised to consult the latest issue of "Economic Survey" which is issued every year along with the union budget of India. Prepare fiom it a table on exports and imports of India to USA, UK, Russia, Canada and Germany for three or four years. 2) Text tables are the special type of tables. They are smaller in size and are prepared fiom the reference tables. Their aim is to analyse only a particular aspect to bring out a specific point or to answer a particular question. For example fiom the Population Census tables we may pick out information on the number of people in Bombay and Delhi who speak different languages (mother tongue), profess different religions and come fiom different states of India. Similarly fiom various publications of Reserve Bank of India, we may be able to extract dormation, in tabular form, on money supply, rate of interest and bank rate for the last ten years or so. Tables can be simple and one way, like the tables'given in Section 3.3, where we deal with only one variable, say, income. Alternatively, it is called a univariate frequency distribution. In addition to this, we can have two-way or multi-way tables where we deal with two or more related characteristics (for example, Table 3.15).

3.4.2 Parts of a Table

Parts ar elements of a table vary fi-om table to table depending upon the nature of data and purpose of tabulation. Yet some points are common. These are:

1) Table number is required for the identification of a table particularly when there are more than one tables in a particular analysis. Table number is always mentioned in the centre at the top. 2) Title of the table gives the indication of the type of information contained in the body of the table. It is said that the title is to the table what heading is to an essay. Next to the table number, we mention the title of the table. Its purpose is to answer the questions like: a) b) c) d)
What is in the table? Where is it in the table? m e n did a particular information occur? How has a particular information been arranged?

In respect of a sample of a table on exports and imports, (Table 3.15), these qyestiqns will be answered as below: a) The table contains values of exports and imports of country X. b) Mormation contained in the body of the table shows exports (sales to) and imports (purchases fi-om) four countries A, B, C and D. c) These exports and imports occurred in 1995,1996 and 1997. d) Idormation on exports and imports has been arranged according to year and countries. Dos and Don'ts of the Title Don't opt for long sentences. Title should be brief and to the point. Present the title in bold letters and/or in capital letters. Expressions used should not convey more &an one meaning. Avoid the expressions like 'Table Presents ..........' or 'A

1i

Detailed Comparison of Data Relating to .........', etc. It should be like a telegraphic message. 3) Head note, also called prefactory note, is written just below the title. It shows conients and unit of measurement like (rupees crore) or (lakh tomes) or (thousand bales). It should be written in brackets and should appear on right side top just below the title. However, every table does not need a head note, like number of students in each class. 4) Stubs are used to designate rows. They appear on the left hand column of the table. Stubs consist of two parts: a) Stub head describes the nature of stub entry. b) Stub entry is the description of row entries.

Tabulation and Graphical Representation of Data

5) Captions, also called box heads, designate the data presented in the columns of the table. It may contain more than one column heads, and each column head may be sub-divided in more than one sub-head. For example, we can divide the students of a college into hostelers and non-hostlers and then again i l into males and females. This w l help us to know the number of male hostelers in, say, first year, second year and third year.

6) Main body of the table, also calledjeld of the table, is its most important and bulky part. It contains the relevant numerical information about which a hint is already contained in the title of the table. In our example of Table 3.15 the title amply suggests that the body of the table contains numerical information on exports and imports of co.untry X for a period of three years.
7) Foot Note, is a qualifying statement put just below the table (at the bottom). Its purpose is to caution about the limitations of the data or certain omissions~' For example, in Table 3.15, the foot note reads that "@ figures are quick . estimates". This implies that the figures for the year 1997 where a superscript '@' is given are not final.

8) Source of data may be the last part of a table, yet it is important. It speaks about the authenticity of the data quoted. It also offers opportunity to the reader to check the data if (s)he so desires and get more of it.
Taking all these points into consideration, the format of a hypothetical table is presented below:

Table 3.16
(

TITLE

1
(In Crore of Rupees)

Stub Head

Caption

Stub Entnes

Column Head I Sub-head Sub-head

Column Head I1 Sub-head Sub-head

MAIN

BODY OF

TIE

TABLE

Total Footnote(s) : Source :

Data and Its Presentation

3.4.3 Importance of Tables

Numerical information arranged in tabular form has distinct advantage over other forms of presentation. First, tabulated data are easy to understand and interpret. Secondly, one can make quick comparison between different characteristics, for example, 'Are imports greater than exports over all the three years?' or 'Are exports increasing?' Thirdly, it opens doors for further investigations. Fourthly, they have a more lasting impression on human mind than the textual statements. Needless to say, that the statistical tables are used extensively in almost all fields of human inquiry.

Check Your Progress 2 1) Distinguish between

a) Caption, stub-head and stub-entries b) One-way and two-way tables

c) Reference tables and text tables d) Column entry and row entry e) Head note and foot note

...................................................................................................................
2) Comment on the statement: "Title is to the table what heading is to an essay".

...................................................................................................................
3) Enumerate the various parts of a Statistical table.
.

...................................................................................................................
4) Make a sketch of a two-way table to show the following information: For a college divide the students according to
a) 1st Year, 2nd Year and 3rd Year students b) Hosteler and non-hostelers

c) Male and female students

Take hkothetical data.

Tabulation and Graphical Representation of Data

...................................................................................................................
f
L

3.5 GRAPHICAL PRESENTATION OF DATA

Besides formal tables, statistical data can also be presented in the form of various types of graphs. Graphs are a useful way of conveying information very quickly and briefly. With the same ease and efficiency, they help in comparing data over time and space. They are visual aids and have a p o w e f i impact on the people. It is ofien said, "a picture is worth a thousand words". They attract a reader's attention to what they are supposed to convey about the data. Further, they may help us to estimate some values at a glance, and serve as a pictorial check on the accuracy of our solutions. However, graphical presentation of data, although usem in different ways mentioned above, is only one method of describing data. T ' s cannot and is not a substitute for other forms of presentation as well as firher statistical analysis. In the following, we discuss some of the graphical methods of presentation.
3.5.1 Line Graphs

i
I
I

Although there are four quadrants on a plane, in economics we usually draw our diagrams only in the first quadrant where both the quantities measured on X-axis and Y-axis are positive. Economic quantities like price, quantity demanded and supplied, national income, consumption, production and host of other such variable are non-negative ( 2 0 ). Let us take a demand schedule and plot it on the graph. The resultant curve on i l joining different points,assumingcontinuity,wl give us line graph expmsing relation between price and quantity demanded. Such a line graph in Economics is called a demand curve. Note that price is measured on Y-axis and quantity demanded on X-axis. The demand curve for data given in Table 3.17 is given in Fig. 3.1.
Table 3.17 Demand Schedule Table 3.18 Time Series Data Year
1990 1991
1992

R. Price of X ( s )
5 10 1 5
20 25

Quantity of X demanded
16 12 8
4 2

Production of Steel (tons)

10
'

25

20
40

1993 1994 1995 19%

50

30

45

60

Data and Its Presentation

Demand Curve

Quantity Demanded

Fig. 3.1

A line graph may be used to show changes in some economic variable, say, steel praduction over time. In other words, if out of the two variables, one happens to be time (months, years, etc.), we get a line graph over time or simply time series graph or historigram. A time series expresses behaviour of an economic variable over time. An example of time series data is given in Table 3.18. Measuring years on X-axis and steel production on Y-axis, we can plot time series data on a graph, as shown in Fig.3.2.

Historigram of Production of Steel

Years

Fig. 3.2

3.5.2 Histogram, Frequency Polygon and Frequency Cuwe

Histogram (do not confuse with historigran discussed earlier) is a very common type of graph for displayingclassified data. It is a set of rectangles erected vertically It has the following features: a) It is a rectangular diagram. b) Since the rectangles are drawn with specified width and height, histogram is a two dimensional diagmn. The width of a rectangle equals tihe class interval & ~height d - Class frequency x Width of the shortest class interval in the data Width of the class interval c) The area of each rectangle is proportional to the fiequency of the respective class.

Construction of Histogram
To plot a histogram of the frequency distribution given in Table 3% on a graph paper, we mark off class intervals like 500 - 550,550 - 600, etc. on the horizontal axis. Similarly, we mark off fiquencies on the vertical axis. Since all the classes

are of equal width, the height of each rectangle is taken to be equal to the Ikquency of the respective class. The histogram is shown in Fig. 3.3.
Histogram

and G r q h i c a l Reprerentation of Data

Monthly Income (Rs.)

Fig. 3 3

Advantages of histogram are:

1) The width of various rectangles show the nature of classes in the distribution, i.e., whether of equal width or not.
2) Area of a rectangle shows the proportion of the class frequency in the total.
I

Frequency Polygon
Frequency Polygon has been derived finm the word "polygon" which means many sides. In statistics, it means a graph of fiquency distribution. A fquency polygon is obtained from a histogram by joining the mid-points of the top of various rectangles with the help of straight lines, as shown in Fig. 3.4. In order that total area under the polygon remains equal to the area under histogram, two arbitrary classes, each with zero frequency, are added on both ends, as shown below.

Frequency Polygon

450-500

550-600

650-700

750-800

8%-900

Monthly Income

Fig. 3.4

Frequency Curve
Ethe points, obtained in the case of frequency polygon are joined with the help of a smooth curve, we get a frequency curve 8 shown in Fig. 3.5.

1)nln and I r s Presentation

Frequency Curve
l4 T

450-500

550-600

650-700

750-800

850-900

Monthly Income Fig. 3.5

3.5.3 Cumulative Frequency Curve - Ogives

The graph of a cumulative eequency distribution is known as cumulative fkquency curve or ogive. Since a cumulative fkequency distribution can be of 'less than' or 'greater than' type, accordingly, we can have 'less than' or 'greater than' type of ogives. Ogivcs can be used to locate, graphically, certain partition values. We can also determine the percentage of observations lying between given limits. The ogives for the cumulative frequency distributions given in Tables 3.12 and 3.13 are drawn in Fig. 3.6. . Note that to draw a less than type ogive, we add a class interval of 'less than 500'lwith fkquency equal to zero. Similarly, we add a class interval of 'more than 900' with fkequency zero for the construction of a greater than type ogive.
'Less than' and 'More than' type Ogives

500

550

600

650

700

750

800

850

Monthly Income Fig 3.6

3.6 DIAGRAMMATIC PRESENTATION OF DATA

A diagram is a visual form for the presentation of statistical data. Diagram refers to bars, squares, circles, maps, pictorials, cartograms, etc. Diagrams are different f o graphs as the former are used only for presentation while the later can be rm

3.6.1 One Dimensional Diagrams

These are also known as bar diagrams. A bar is defined as a thick line, often made thicker to attract the attention of a reader. The height of the bar highlights the value of the variabre with width presenting nothing. Therefore, it has nothing to do with the area of the bar. It is different fiom the histogram where both the width as well as the height of the bar are important. Further, the bars of the bar diagram are separated fiom one another so that the gap between the successive bars is same, whereas in histogram they are placed adjacent to one another with out gap. Finally, in histogram the bars are always vertically placed whereas in bar diagram they can be placed both vertically as well as horizontally. Let us take a simple example to demonstrate the construction of a bar diagram.
Table 3.19 Number of students in four zones of a country

~abulation and Graphical Representation of Data

Zone North South East West No. of students (lakhs) 6 10 2 4

The bar diagram of the above data is drawn in Fig. 3.7. To make the bar diagram beautiful we can either colour the bars or shade them in different ways. This is left to the aesthetic taste of the investigator.

Bar Diagram

North

South

East

West

Zone
Fig. 3.7

Sub-divided or Component Bar Diagram

A sub-divided bar diagram is used when it is desired to represent the comparative values of different components of a phenomenon. In this diagram, the bars, comspondmgt each phenomenon, isdivided into various unnponhts. The portion o of the bar occupied by each component denotes its share in the total. The subut divisions of different bars m s always be done in the &me order and these should be distinguished fiom each other by using different colours or shades. A subdivided bar diagram for the hypothetical data on sales of T.V. sets, given in Table 3.20 is drawn in Fig. 3.8.

Data and Its Presentation

Table 3.20 Lone-wise sale of T.V. sets (1995-1997)' Zone Number of T.V. Sets sold (lakhs) 1995 1996 1997

NO&

12
8
5

20

28

South East West

.

9
7

15 10 11
64

6
31

8
44

Total

SuMvided or component Bar Diagram

60

m "0 50 m 40

f!
9

b,

i 30 :
20
10

W South

z"

Years
Fig. 3.8

Multiple Bar Diagram

This diagram is used when comparisons are to be shown between two or more s e l of data. A set of bars for a period, place or a related phenomenon are drawn side by side without gap. Different bars are distinguished by different shades or colours. Amultiple bar diagram for the hypothetical data given in Table 3.21 is drawn in Fig. 3.9.
Table 3.21 Total revenue, total cost and profit of MIS XYZ (1990-92) (Rupees thousand)

Year
1990

Total Revenue 30

Total cost
25

Profit
5

Multiple Bar Diagram

Tabulation and Graphicai Representation of Data

40 --

6 c: 2

35 -30
--

$
$

--

.Total
Profit

Cost

25 20 -1510 -5 -

t
1990
a

1991

1992

Years

Fig. 3.9

3.6.2 Two Dimensional Diagrams or Area Diagrams

In the case of one dimensional diagrams only the height of the bar is important, and the width can be chosen according to convenience or aesthetic taste of the investigator.But in the case of two dimensional diagrams, area is more important. That is why they are also known as Area diagrams. There are three types of area diagrams.
a) Rectangles, where area equals width (or base) multiplied by the length (or height) of the rectangle. b) Squares where area equals square of side (or base). c) Circles where area equals x f ,with x =22/7 and r
= radius.

Let us consider data on, say, average salaries of three categories of university teachers, and prepare all the three types of area diagrams.
Table 3.22 Average Salaries of University Teachers as on 1/1/1998 Class of teachers
Professors Readers Lecturers

Average Salaries (Rs.)

w,ooO 16,000 9,000

a) For drawing rectangles, a common base of, say, 100 is taken. Accordingly, the heights can be determined as: 1) Salary of Rs.25,000 2) Salary of Rs. 16,000 3) Salary of Rs. 9,000
= =
=

100 (base) x 250 (height) 100 (base) x 160 (height) 100 (base) x 90 (height)

Now take a scale of 2 cm = 100, so that the first rectangle has dimensions of 2 cm. x 5 cm, the second one has the dimensions of 2 cm x 3.2 cm and the third one has the dimensions of 2 cm x 1.8 cm. After this, we are in a position t draw the rectangles as area diagrams (Fig. 3.10). o

Data and Its Presentation

Average Salaries of University Teachers ( s ) R.

Professors

13
Readers Rs. 25,000
Rs. 16,000

Lecturers

Rs. 9,000

Scale : 2 crns = Rs. 100

Fig. 3.10: Area Diagram (rectangles)

b) For drawing squares, we find the square root of various incomes. We have,

Chose a scale 1 cm = 50 so that the first square has each side approximately . equal to 3.2 cm. (since 158.114150= 3.2), second has the side of 2.53 cm. and the third has the side of 1.9 cm. The relevant squares are drawn in Fig. 3.11.
Average Salary of University Teachers (Rs.)

Professors

El
Rs. 25,000

Readers Lecturers

Rs. 16,000

Rs. 9,000

Scale : 2 crns = Rs.50

Fig. 3.11: Area Diagram (squares)

c) For drawing Circles we take the squares of their radii in the ratio of areas, i.e., 25000: 16000: 9000 or 25: 16: 9. This is based on the property of the circles that area of a circle is proportional to the square of its radius. Let r,, r2 and r, denote the radii of the three circles, then we can write' r12: r22: r,2 = 25 : 16 : 9 or rl : r2 : r3 : = 5 : 4 : 3. Taking 2.5 units = 1 cm the radii of the three circles will be 2.0, f.6 and 1.2 cms respectively. Let us draw the required circles..

Average Salary of University Teachers (Rs.)

Tabulation and Graphical Representatlon of Data

Professors Readers

Rs.25,000
Scale : 1 crns = 2.5 unit.
Fig. 3.12: Area Diagram (circles)

3.6.3 Pie Diagram or Pie Chart

It is also known as angular diagram. It is used to represent percentage break downs onr of the given data. For example, the exports of a c u t y to different countries and continents of the world can be expressed into ratios or percentages. These ratios or percentages can then be converted into angles by the fon-da
Share of the sub - division x 360 Total

Table 3.23 Exports of country X to countries A, B, C and D in 1990 Country

A

Exports

Percentage Share
(300 X 100) s 800 = 37.50 (250 x 100)
i 800 = 31.25

Degree
(37.5 X 3600) + 100 = 135O (31.25 X 360)
t

300

B
C

20 5
150 100 800

10Q= 12.5O

(150 X 100) + 800 = 18.75 (100 X 100) + 800 = 12.50 100

(18.75 X 3600) + 100 = 67S0 (12.5 X 360") + 100 = 45O 360

D
Total

Pie Diagram Representing Exports of X

Fig. 3.13

Data and Its Presentation

Steps in the construction of Pie diagram

1) Find the total of all components. 2) Find ratio or percentage of the share of sub-division to the total and multiply by 360 to get the angle corresponding to each sub-division. 3) Draw a circle of a suitable size. 4) Use protractor to draw different angles at the centre. Preferably start with the largest one.

5) Shade the different segments with different colours or shades.

6) Write the components with percentage values in the marked, shaded or coloured areas.

3.6.4 Three Dimensional Diagrams

These diagrams are not very popular and are used rarely. Since these diagrpns are three dimensional (involving length, breadth and width), they denote volumes. They can take the fonn of boxes, cubes, blocks, spheres and cylinders. They are very usehl when the variations in magnitudes of the observations are very.marked. Here we will explain only the presentation of data by cubes for which we take the following steps: 1) Find cube-root of each figure. 2) Take a convenient scale, preferably in centimeters. 3) Ciaw cubes, dimensions of which are calculated below for an example consisting of two classes of families : Poor and Very Rich.
Table 3.24 Income class
1.

Income (Rs.)
216

Cube-root

Side of cube
1.5 cms.

Poor
Very Rich

2.

3375

m= 15
=4

==6

3.75 cms.

Scale : 1 cm

units.

4) Now draw two cubes with sides equal to 1.5 cms. and 3.75 cms. respectively.
Income Levels of Poor-and Very Rich People (Rs.)

Very Rich

Poor

Rs.216

Rs. 3375
Fig. 3.14

3 6 5 Pictograms and Statistical Maps ..

These are also known as catrograms. Pictures are more attractive to laymen than other forms of graphic presentations. But these are not suitable everywhere. It may suit cases involving population of people of a state or number of vehicles in a metropolitan city like Delhi or Murnbai. For showing population of human beings, we draw human figures. Here also we have a scale. We may represent 1 lakh people by one human figure so that a population of three and half lakhs is shown by drawing 3%human figures, as given in Fig. 3.15.

Tabulation and Graphical Representation of Data

Fig. 3.15

Pictograms suffer fiom a limitation that they present only approximate values. For more accurate presentations bar diagrams are preferable.

Check Your Progress 3

1) Disthguish between the following giving at least two points of distinction. a) Histogram and historigram. b) Histogram and bar diagram. c) Histogram and frequency polygon, d) "Less-than" and "More-than" Ogives. e) Pie diagram and circle.

.............................................................................................................
2) Prepare a sub-divided bar chart and a pie diagram from the following data.
Academic Year Expenditure on Books Economics 5200 8000 Commerce 10000 14000 Maths 5000 7000 Languages 4800 6000 Total 25000 35000

1996 - 97 1997 - 98

Data and Its Presentation

.................................................................................................................
3) Explain the following terms:
a) Line graph b) Bar diagram c) Sub-divided or component bar diagram d) Multiple bar diagram e) Area diagram
f ) Volume diagram

.............................................................................................................

.............................................................................................................
4) Fill in the blanks with a suitable word out of those given in brackets: a) A pie diagram is also called ........................... diagram. (bar, angular, multiple bar).

b) In the case of vertical bars, the variable is measured on the ........................... (X- axis, plane, Y- axis). c) Bar diagrams, rectangles, squares, circles and pie charts are ........................... forms of presenting data. (geometric, arithmetic, horizontal). d) By joining the mid-points of the top of each rectangle of a histogram, we get ........................... (an ogive, a frequency curve, a frequency polygon) e) Graph of "morethan" cumulative kquency distribution is also called "more - than" ........................... (ogive, frequency polygon, frequency curve) f ) The caption of a table labels data presented in the ...........................of a table. (rows, columns, foot-note)
5) Are the following statements true or false? If false, what should be the correct statement?

1) A picture is worth a thousand words: 2) Squares and circles are examples of area diagrams. 3) We can have only vertical bar to present some data having one variable. 4) The graph of an ordinary fiequency distribution is called ogive. 5) A time series graph is known as historigram. 6) Histogram is same as bar diagram,

Tabulation and Graphical Representation o f Data

3.7 LET US SUM UP

Collected data are unorganised and complex mass of figures. To draw some meaninghl conclusions, they must be arranged in an orderly manner. This can be done in many ways, such.as by forming simple and frequency array, discrete and continuous frequency distributions, etc. Sometimes, it serves a usehl purpose to form what is called "less-than" or "morethan" cumulative frequency distributions. The former is arrived at by successive totaling of fi-equenciesh m above and the latter by successive totaling ii-om below. After collection and condensation of data, good presentation of data is important. A good presentation helps to highlight important points of the data and makes possible usell comparisons and their intelligent use. This can be done through five statistical tools. These are: i) formal tables - one-way and two-way; ii) line graphshistograms, fiquency polygon and fi-equencycurves; iii) cumulative distributions"less-than" and "more-than" ogives; iv) one, two and three dimensional diagrams such as bar diagrams, rectangles, squares, circles, cubes and pie diagrams; and v) statistical maps. While using diagrams, their limitations must always be kept in mind. Diagrams give only a approximate idea of the problem and can portray only a h t e d number of characteristics.Unlike a graphicpresentation, the main limitation of a diagrammatic presentation is that it cannot be used as a tool of analysis. The level of accuracy of a graphic method is oftea lower than that of mathematical method.

Condensation of data: It is a process of classifjrlng and arranging complex and unorganised mass of data to make them fit for comparison and analysis. Array: An array is an arrangement of data in ascending or descending order. It is also called a simple array. Frequency array: It is an array or series formed by writing various possible values of the variable along with their respective Gequencies. Discrete frequency distribution: A discrete distribution or discrete series is formed where the variable can take only discrete values like 1,2,3,..... Number of children in a family, number of students in a university, etc. are examples of discrete variable. Continuous frequency distribution:A continuous frequency distribution is formed where the variable can take any value between two numbers. For example, height, weight, income and temperature. Inclusive type class interval: A class interval in which all observations lying between and including the class limits are included. Exclusive type class interval: A class interval which includes all observations that are greater than or equal to the lower limit but less than the upper limit.

Data and Its Presentation

Open-end class: A class in which one of the limits is not specified. kequency polygon: It is a broken line graph to represent a kequency distribution and can be obtained either from a histogram or directly from the frequency distribution. Frequency curve: It is a smoothened graph of a fiequency distribution obtained fkom fkequency polygon through h e hand tracing in such a way that the area under both of them is approximately the same. Class and class limits:It is a decided group of magnitudes having two ends called class limits or class boundaries. Class range: Also called class interval. It is the difference between two limits of a class. It is equal to upper limit minus lower limit. It is also called class width. Mid-point: Also called mid-value. It is the average value of two class limits. It falls just in the middle of a class. Relative frequency distribution:It is a fiquency distribution where the fiquency of each value is expressed as afiaction or apercentage of the total number of observations. Cumulative frequency distribution: It is obtained by successive totaling of the simple fkequencies of a discrete or continuous frequency distribution. This totaling can be done either h m above (we get "less-than" cumulative fiquency distribution) or from below (we get "more-than" cumulative frequency distribution). Ogivt?:It is the graph of cumulative fkequency. Graph of "less-than" cum~llative fkequencies gives "less-than" ogive and that of '=more-than"gives "more-than" ogive. Tabulation: It is a systematic presentation of data in rows and columns. Caption:It is a part of a table and labels data presented in the column of a table. It is also called box head. It may contain one or more than one column head. Stub: It is a part of a table. It consists of stub head and stub entries. Each stub entry labels a given data placed in the rows of the table. Both stub head and stub entries appear on the left-hand column of a table. They describe the row heads. Main body of the table: It is certainly the most important part of the table and contains numerical information about which a hint is already made clear by the title. It is also calledfield o the table. f
Line graph: It is the locus of different points obtained with the combinations of X arid Y coordinates measured on X-axis and Y-axis respectively.

Historigram: The line graph of a time series is c'alled historigram (For example, steel production since 1950). Histogram: It is a set of adjacent rectangles presented vertically with areas proportional to the frequencies. Bar diagram: It is often defined as a set of thick lines corresponding to various values of the variable. It is different from histogram where width of the rectangle is important.

one variable can be presented. A sub-divided bar diagram is used to show various . components of a phenomenon.

Representation o f Data

Pie diagram: It is a circle sub-divided'into components to present proportion of different constituent parts of a total. It is also called pie chart. Area diagrams: These are two dimensional diagrams. Here both the height and the base of the diagram are important. That is why they are known as area diagrams. They can be either rectangles, or squares or circles.
Volume diagrams: These are three dimensional diagrams. In their constsuction length, width and height are used. They consist of boxes, cubes, blocks, spheres and cylinders.

Pictographs: Here the data are presented in the form of pictures.

3.9 SOME USEBUL BOOKS

Elhance, D. N. and V. Elhance, 1988,Fundamentals of Statistics, Kitab Mahal, Allahabad. Nagar, A. L. and R. K. Dass, 1983,Basic Statistics, Oxford University Press,

Delhi
Mansfield, E., 1991, Statistics for Business and Economics: Methods and Applications, W. W. Norton and Co. Yule, G. U. and M. G. Kendall, 1991, An Introduction to the Theory of Statistics, Universal Books, Delhi.

3.10 ANSWER OR HINTS TO CHECK YOUR PROGRESS EXERCISES

Check Your Progress 1
1) a) See Sub-section 3.3.2 and 3.3.3
b) See Sub-section 3.3.4

c) See Sub-section 3.3.3

d) See Sub-section 3.3.1 and 3.3.2

2) You may give examples h m your surrounding. For exact meaning of the terms refer to Section 3.3. 3) In the text we have converted the monthly income data in Table 3.2 to a frequency distribution in Table 3.6. From this you can take a clue. 4) Refer to Sub-section 3.3.3 5) Refer to Sub-section 3.3.4(c) 6) Refer to Sub-section 3.3.4(d)

Check Your Progress 2

1) Refer to Table 3.16 and Sub-section 3.4.2 for different parts of a table.
-\

n-r-..

4-

C'.-l.

2 A 9/91 '

Data and Its Presentation

3) Refer to Table 3.16 4) It can be presented in more than one ways. We have given one below. Try another.
Division of Students of XY College
Year
Male
First Year

Hostelers Female

Non-Hostelers Male Female

Second Year
Third Year

Check Your Progress 3 1) a) See Sub-section 3.5.1 and 3.5.2 b) See Sub-section 3.5.2 and 3.6.1 c) See Sub-section 3.5.2 d) See Sub-section 3.5.3 e) See Sub-section 3.6.2 and 3.6.3 2) Refer to Sub-sections 3.6.1 and 3.6.3 3) a) See Sub-section 3.5.1 b) c) d) e) See Sub-section 3.6.1 See Sub-section 3.6.1 See Sub-section 3.6.1 See Sub-section 3.6.2 See Sub-section 3.6.4

4) a) an& b) y-axis c) geometric d) a hquency polygon e) ogive 9 ~01umns 5) Tnpe: 1,2,5 False: 3,4,6