Professional Documents
Culture Documents
Chapter 3-2
Chapter 3-2
COLLECTION OF DATA
LEARNING OBJECTIVES
3.1 INTRODUCTION
3.1 INTRODUCTION
business world. Data is the base
Statistics has gained significant place in the modern complex
success and tailure of
on which the superstructure of
statistical investigation is made. The
data.
the quality, adequacy and accuracy of
investigation mainly depends upon
Statistical Enquiry
by any agencywherein relevant quantitative
Bystatistical enquiry, we mean aninvestigation ona topic
information is collected.
with the aid of statistical methods of
Thus, statistical enquiry implies search of truth
collection, compilation, analysis, interpretation,etc.
is to collect facts and figures relating to
For any statistical enquiry, the basic problem
*************
3.2
**usse*********************
Collection of data is the first step in any statistical investigation. So, it is a very important
1970-71 108
1978-79 132
1979-80 108
1990-91 176
1996-97 199
1997-98 194
2001-02 212
2018-19
283
As in Table 3.1, the food grains production varies from
seen
year to year. As these values vary
they are called variable. A variable refers to quantity or attribute
whose value varies from
one
investigation to another. The variables are generally
In Table 3.1, the represented by the letters X, Y or L.
years are
represented by variable X and the production of food
variable Y. These values of the variables X
and Y are the 'data', from grains Dy
information about the trend of the which we can obtan
in understanding production of food grains in India.
'Data' is a tool, which hep»
problems by providing information.
In thefollowing sections, we will discuss the sources of
data collection. Let us first discuss data, types of data, methods and instrumenis
the sources
of data.
SOURCES OF COLLECTION OF
DATA
Sources of Data
Broadly, there are two main sources of data: Internal Sources
External Sources
G) Internal Sources; (ii) External
Sources.
Secondary Data
Primary Data
Collection of Data
3.3
Internal Sources
In an organisation, when the data is collected from its reports and records, it is known as
internal sources of data.
.For example, a company publishes its annual report on profit and loss, total sales, loans,
wages, etc.
Data collected and compiled through internal sources is called internal data.
External Sources
External sources refer to the data collected from outside the organi
For exanple, if a Tour and Travels Company obtains information on 'UP Tourism' from
Uttar Pradesh Transport Corporation, it would be known as external sources of data.
Data collected and compiled through external sources is called external data.
.External Data can be collected from primary as well as secondary sources.
Collection of data is ofutmost importance in any statistical enquiry as no useful and valid conclusions
can be drawn in case of any inadequacies in data. Data may be collected either from a primany
rce, known as Primary Data' or from a secondary source, known as 'Secondary Data.
Pripnary Data
agency for the first
Primary data is the data which is originally collected by an investigator
or
Setgndary Data
The data which is not directly collected but rather obtained from the publishedor unpublished
sources, is known as Secondary Data.
It is also known as Second Hand Data.
the enumerators or investigators
themselves do not collect
These are not original data since
of the data collected by the others.
these data. They simply make use
For example, National Accounts Statistics published by
the CSO, EconomicSurvey published
by Government of India.
3.4 Statistics for conomics
Between Primary Data and Secondary Data
Dreronce
Basis Primary Data Secondary Data
Originality Primary data are original because they are Secondary data are not origine
collected by investigator himself. investigatormakes use of the datajina)s ince
by other agencies. collecte
Source Primarydata are collected by some agency Secondary data are already colle.
or
person
by using the data processed by some
method of person or aand
collection. ready for use.
is and
Time Factor Itrequires longer time for data collection. Secondary data require less time
Cost Factor The collection of primary data requires The collection of
secondary data is o
a considerable amount of
cheape
money and as it is taken from published or unpuhi
personals as whole plan of investigation
is initiated.
material. lishet
Reliability and Primary data is more reliable and suitable Secondary data is less reliable and
Suitability to theenquiry as the investigator himself suitable as less
someone else has
collects it. the data which may
collere
ected
Precautions There is no great need not serve the purpose
for precautions while Secondary data should be used with area
using primary data. care and caution. at
Organisation Collection of primary data
requires
Factor
organisation set up. elaborate There is
no need for
organisational set uo
in case of
The basic difference
between primary data and
secondary data.
by one wil be primary, but the same secondary data
data used by another will make
is of 'degree'. The data
collected
collected by Ministry of Human it secondary. For
Resource with example, data
used by us, it becomes regard
if the same data is to birth rate in the
country is primary data. However,
secondary
3.3 METHODS OF
cOLLECTING PRIMARY DATA
There are various methods
of
collection of primary data are collecting primary data. The methods
presented
in the
following chart
commonly used for the
METHODS OF COLLECTING
PRIMARY DATA
Direct Indirect
Personal Information from
Oral
Correspondents Telephonic Mailed
Investigation Investigation Interviewss Questionnaires
Questionnaire Filled by
Direct Personal Method Enumerators
Investigation (Fersonal
Under this method, the
investigator Interview)
collects
contact with the data by
informant and conducts having direct
.The investigator
goes
to the on-the-spot enquiry.
field
respondents and convince them in personally, contact the
under study. giving the information
.The investigator or the
of interviewer has the
explaining the study and opportunity
respondents. answering any query of the
Collection of Data
*********
****************** ************
3.5
.The interviewer can
request the respondent to expand on answers that
important. The visual reactions of
the respondents can
are
particularly
Example:Hindustan Unilever Limited (HUL) wishes to conductprovide supplementary intormation.
a statistical
soap used by people of Delhi. For this, HUL will enquiry regarding bathing
select some investigators, who will
people of Delhi and collect the approach tne
necessary data. It is necessary that in such cases, the investigator
has a keen sense of observation and he is
very polite and courteous. He must be skilful, tactful,
accurate, amiable and neutral.
investigator.
Demerits of Direct Personal
Investigation is
Areas: This method is not suitable when thearea ofcoverage
1. Not Suitable for Wide are to be contacted.
respondents
or when many
considerably wide and time consuming, particularly
This method is expensive
consuming:
2
Expensive andtime is large
when the field of investigation
3.6
Statistics for Economics
3. Fersonal P'rejudice: The bias or preiudice of investigators can affect the accuracy.fect the
accuracy of dat
tO a large extent. There is a possibility of influencing respondents.
4. Trained l'ersonal: The entire success of this method depends upon
nterviewers must be thoroughly trained and unbiased, otherwise they may
the
investigator.sSo, the
ay not
not be able to
collect the information.
DIRECT PERSONAL INVESTIGATION
Merits Demerits
3.8
to a s c e r t a i n the accuracy;
to make croSs check system
OrEs should be made
their approach and the wit.
interviewers are not subjective in
5. It must be ensured that view.
point of
c O Prejudicial
in favour of or against a particular
Investigation
Direct Personal Investigation Vs Indirect Oral
Indirect Oral lInvestigation
Basis Direct Personal Investigation
This method can be used to covera
Coverage This method is suitable for limited area.
area of investigation.
wide
The data collected is original in character. This method lacks originality as data is
Originality collected from the witnesses.
3. Suitable tor Special Purpose: This method is suitable for some special purpose or
investigations.
4. Continuity: It is very useful for collecting regular information.
Telephonic Interviews
interview over the telephone with the
Under this method, data is collected through an
interviewer.
Good Morning Sir, I
in the cases where the
Telephone interview is better would like to ask certain
nomica
Merits of Telephonic Interviews
for
TELEPHONIC INTERVIEWS
1. Wide Coverage: This method is useful Data is collected through an interview
telephone with the interviewerthe
conducting enquiry over a large area.
2. Economical: This method is cheaper as it
5. Maintains secrecy: This method is suitable for sensitive questions and maintains the
anonymity of respondent.
5. Time Consuming: The process is time consuming, particularly when the information is to
be obtained by post.
Precautions for Mailed Questionnaire Methobd
1. Questionnaire should be simple, short and attractive.
2. The questions should not hurt the informants and should not be very personal.
3. There should be a system, a sequence, and proper planning in the questions.
1. Wide Coverage:
Filled
This technique
by
of data
collection
from vast
can
areas.
be used
direct
informati
coming has a conta.
of people
from large
number
information:
Since the investigator
information.
with he
and reliable reliable
2. Accurate
accurate
and
to get the respondents in in vingthe
it is possible induce
respondents,
enumerator may
Better Response: The presence ofthe in case of Mailed Questionn
ire Methoa
(like
chances of non-response
intormation. So,
is less. limited possibility for partiality o
tiality or biase
Partiality: There is very
4. Limited Chance of
information. Questionnaire Metho
Unlike the 'Mailed
of illiterate respondents:
5. Useful in case
illiterate.
are
method can be used even if the respondents
Pilot Surveys
From practical point of view, it is found
useful to conduct a pre-test known
on a small scale before as Pilot Survey
starting the main survey.
Once the draft questionnaire is
ready, it is advisable to conduct a try-out with a siau
group which is known as Pilot Survey
.It helps the
or
Pre-Testing of the
questionnaire.
investigator to have a rough idea of the information and
technique. the suitability O
.The investigator is also able to
determine
shortcomings and drawbacks of the ques ons.
.Pilot survey also helps
in assessing the suitability of
Derformance of enumerators and the cost and questions, clarity of instrucu S
time involved in the
actual survey.
Collection of Data
3.13
3.4 CONSTRUCTION OF QUESTIONNAIRE OR SCHEDULE
A questionnaire ora schedule is list
of questions relating to the problem under
a
investigation.
.The construction of a Schedule
Questionnaire
or a is
specific The art.
questionnaire have strong bearing upon the quality framing
a
and overall drafting of a of questions
the
of questionnaire.
Theretore, it is to beprepared on such a
informants.
pattern that it evokes the desired response from
Preparing a
questionnaire is a technical job and requires great deal of skill,
and practice.
a
expertise
The follouving general principles should be followed while designing drafting the questionnaire:
or
1. Covering Letter: A
very polite covering letter should be sent to the respondents along with
the questionnaire.
It should
clearly explain in brief, the objectives and scope of the survey.
It should ensure
respondents about their secrecy.
In case of mailed
questionnaire method, a self-addressed stamped envelope should be
enclosed.
To ensurequick and better response, the respondents may also be offered certain
incentives (free gifts, coupons, etc.).
2. Decision regarding Questions: To frame
questionnaire or schedule, first of all, we have to
take decision regarding various
questions to be incorporated. The selection of questions
depends upon the purpose of enquiry. So, precaution should be adopted to avoid irrelevant
or unnecessary questions.
OWn words. The questions should be such that it is possible to answer them in.few
3. Number of
Questions: The number of questions should be
as small as
possible. Long
questionnaires discourage people from completing them. Only those questions which
have a direct relevance to the
problem be included.
4. Simple and short
Questions: The questions should be clear, brief, and
question should be framed in such a manner that their answers are unambiguous. The
specific and precise.
5. Avoid Question requiring calculations: The
or questions which force the respondent to questions relating to mathematical computations
For example, informants should not be
recollect from his memory should not be asked.
asked yearly income, since in
are paid monthly. most of the cases they
pertaining
beasked.Sensitive questions, which may irritate the
to their
private affairs should never
ould nev
should not be asked. respondents or may hurt
urt their
their feelings
tee
Coiection of Data
3.15
s. Avoid leading
Questions: The question should not be
clue about how the leading question, which gives
a
respondent should answer. a
3. Address:
4. Sex: Male Female
5. Phone: Landline Mobile
6. Monthy Family Income:
Less than 30,000 O30,000 to 40,000
D
40.000 toR 50,000 More than 50,000
7. What kind of food do you normally eat at home?
Family
Friends Colleagues
Others
10. Any specific days when you go out:
Demographic data on birth and death rates, literacy, workforce, life expectancy, size and
composition of population, etc. are collected and published by the Registrar General of India.
The famous example of this method is the Census of India conducted after every 10 years.
The last Census of India was held in 2011.
Census method is also known as Complete Enumeration' or "100% Enumeration' or Complete
Survey.
Census ofIndia 2011
According to Census 2011, the population of India stood at 1,21,01,93,422 comprising of
62,37,24,248 males and 58,64,69,174 females. India is the second country in the world after
China to cross the one billion mark.
Sampling Method
these
are selected and data collected from
When only some representative items of a population
method is knouwn as Sampling Method.
items are used for the analysis, the
Population or Population'
is used. Population
is defined.
edas
term 'Universe'
in the field of statistics, the
composed of peopie
or tnings, which are
items, whether
he aggregate of individual
subjected to a statistical investigation. the 5,000 studente
by all
or example, if we want to study the monthly
expenditure incurred
ents
the universe or population.
O a school, then 5,000 students constitute
Method
Merits of Sampling collection
is economical than the technique as the task of
census
1. Reduced Cost: It
more
unsatisfactory.
5. Not possible in case ofheterogeneous population: A population is said tobe heterogeneous
if its items or sampling units are not similar with regard to the characteristics under
investigation. In case of heterogeneous population, any sample may not correctly reflect
the characteristics of the population. In this case, sampling is not advisable.
METHODS OF CONDUCTINGSURVEY
Census Method Sampling Method
(Data is collected from (Only some representative items
each and every element of a population are selected and
of the population) data collected from them are
used for the analysis)
Exceptional Applicability
Collection of Data
**************s***ines*************isiesse* **.****nea t****d************** * * *
3.21
Census
Sample
-
arise in the collection of data is error of bias. sampling apart from error of bias.
Organisation and It is very difficult to organize and supervise Sampling method is comparatively easy to
organize and supervise.
census method.
Supervision
SAMPLING)
3.6 TYPES OF SAMPLING (METHODS OF methods
a sample froma population. The
various types or
There are various ways of selecting
understood from the following chart:
of sampling can be
TYPES OF SAMPLING
Non-Random Sampling9
Random Sampling
Convenience
Quota
Restricted Random Judgement Sampling
Simple Random Sampling
Sampling Sampling
Sampling
Cluster Multistage
Stratified Random Systematic Sampling
Sampling Sampling
Sampling
a known
3.7 RANDOM SAMPLING the universe has
which every item in
method refers to a method in
Random sampling
the sample
chance of being chosen for of the person making
the study.
items is independent
T h e selection of sample
Sampling8.
It is also known as 'Probability
Statistics for Economics
3.22
of the population.
(i) Simple
Random Sampling:
There are two methods under random sampling:
being selected
This method is known as Unrestricted Random Sampling as method of selection shall not
favour one item or another.
Under this method, the sample is selected by the following two methods:
(i) Lottery Method: This is the simplest method of selecting a random sample. Steps for
conducting Lottery Method:
These 12 students, whose serial numbers are mentioned above, will be selected for
investigation.
than 100, each 4-digit random number is broken
However, if the units in the universe are less
into a two-digit number and then, from
these two-digit figures, samples are selected. For
out of 60 students, then we may choose serials of
example, if 6 students are to be selected
29, 52, 39, 31, 41 and 15 respectively.
Other Popular Tables of Random Numberss
However, in case of heterogeneous population, simple random sample may not be a true
representative of the population.
In such cases, it becomes necessary to select a random sample under certain
Such restrictions,
samples are known as 'Restricted Random
Samples'.
Important Methods under this category are:
(a) Stratified Random Sampling
(b) Systematic Sampling
(c) Cluster Sampling
(d) Multistage Sampling
Let us now discuss the
important methods of restricted random sampling in detail.
(a) Stratified Random Sampling
In this method, the universe or the entire
population S divded into
then certain numbers of items are taken from each group at random.a( number of groups or 'Strata' and
Its basic purpose is to ensure that all the
characteristics of a
adequately represented in the sample. heterogeneous population are
It also helps in increasing the accuracy of the results of
investigation
Three Types of Stratified Random Sampling
1. Proportionate manner: Underthismanner, sample
size of the different strata
at certain percentage or ratio, or in respective proportion of the
size of the population.
size fixed
of the stratum int
the
2. Disproportionate manner: Under this manner, sample sizes of t
of the sizes different strata are
fixed arbitrarily without any regard for the sizes of the differents
3.
strata.
Optimum manner: Under this manner, size of the samples ofthe se
of the
on the principle of obtaining maximum benefits at the minimum
m cost. ditferent strata is fixed
Randomn Sampling
Merits of Stratified
YMore Representative:
Since thepopulation is first divided into
into
represen
divided various strata
every group gets
a representation in
the sample, Ore
a more
under this
representative
method
method as
as sample and then is
2Greater Precision: There is greater accuracy
extent under this
method. variability in eachobtained.
is reduced to a
considerable
stratum
Collection of Data 3.25