COLLECTION OF DATA

Collection of data constitute the first step in

statistical investigation. The must be
collected with utmost care because they
form the foundation of statistical analysis. If
the data are faulty, then the whole analysis
will me misleading and lead to a wrong
conclusion.
 Types of statistical data
Statistical Data may be of two types
2.Primary Data
3.Secondary Data
 Primary Data

Primary Data are those which are collected for

specific purpose directly from the field of enquiry
and original in nature. Such data are published
by authorities who themselves are responsible for
their collection.
 For Example :

• “Reserve Bank of India Bulletin” issued monthly

by the Reserve Bank of India, Bombay.
• “Census Report” published by Registrar General
and Census Commissioner, India.
 Secondary Data

When the investigator uses the data which has

already been collected by others for some different
purpose, such data are called Secondary Data.
Infact the data collected by someone when used by
another or collected for one purpose when used for
another, will be called Secondary Data. The same data
are primary in the hands of collecting authority, but
are secondary in the hands of another.
Secondary Data can be obtained from
Journals,reports,government publications,
publications of research organization etc.
 For Example:-
The Census figures published by Registrar
General and Census Commissioner, India
will be primary data, while the same data
contained in any other publication will be
called secondary data.
“ Monthly Abstract of Statistics", issued by
C.S.O.
Some aspects of Secondary Data:
a) Whether the available data are suitable for
the purpose of investigation:-

The suitability is being judge by the objective of

the investigation.
For Example :
If the objective of investigation is to study the
wage level including allowances of workers and
the data which we have collected relate to basic
wage alone, then such data will not serve our
purpose of investigation.
If the objective is to study the annual sales
volume of LG T.V in India, in 2007 and the
available data covers only W.B or the data
available up to 2006 , then it would not serve the
purpose.
 b) Whether the data are reliable:-

The reliability of the data is being judge by

considering the following aspects,

Whether the enumerators are properly

trained?
- Accuracy or the data mostly depend on this
factor.
 Whether the collecting data are biased or
unbiased?
-
For example:- Since in WB the enrolment rate
at class I is taken as an indicator for assessing
the overall performance of the school authority,
there is a natural tendency of the school
authority to inflate the enrolment rate beyond
its actual figure and hence the enrolment data
in WB are highly inaccurate and biased.
• If the enumeration was based on sample, was
the sample representative?

Was the checking/editing on the accuracy of

the data carefully and properly done?
Proper scrutiny of data can rectify the errors
and hence lead to more accurate and reliable
data.
Primary Data vs. Secondary Data
It is generally preferable to make use of primary data
because, they are more reliable than the secondary data,
since the enquiring authority themselves conduct the
survey. So they will put more effort in order to make the
data more accurate since if the collected data are
erroneous then the whole analysis based on the data will be
misleading.
But inspite of all these advantages due to limitations of
time and money the secondary data are used.
Methods of collecting primary data:
The following methods are generally used
for collection of primary data:-

3.Direct personal observation.

4.Questionnaire.
 1.Direct personal observation

The investigator collects the data personally

through observation or by measurement. No
question is being asked here.
For Example:- In order to study the conditions of
students in Calcutta hostels, the investigator visits
the hostels and collect necessary data after a
personal study.
This method is time-consuming and costly, but
yields very accurate results. It is therefore suitable
for such studies when the field of enquiry is small.
 2.Questionnaire Method
Definition:- The questionnaire is a
proforma containing a sequence of
questions relevant to the statistical enquiry.

Here the data are collected by asking

questions from people who are thought to
have the desired information. A formal list
of such questions is called a Questionnaire.
 For Example:-
Questionnaire of a study entitled ‘Consumer survey
on T.V set’ was conducted by a BBA students of NBS.
Questionnaire

1.Do you have T.V set? Yes No

If yes,
What brand ?................................
What is the size of the screen ?.............................
When did you buy it ? (Month and Year)……………
Approximate, price paid ?...............
2.Rank the following factors according to your
preference at the time purchasing the TV set (rank
from 1 to 8):

Picture quality Sound

Material of cabinet shape of cabinet

Color of cabinet Size of screen

Price Warranty period

3. Did you make a special survey of the market
before buying your TV set?

Yes No

If yes, please state why (tick whichever boxes are

appropriate):

To see performance of various brands of TV set

To check which dealer offered lower price
To check which dealer offered more incentives
To check which dealer offered better after-sales services
The questionnaire is used in two ways:

6.Personal Interview.
7.Mailed Questionnaire.
 1.Personal Interview

Here the enumerators (i.e. people employed for

collection of data) personally meet the informants
and fill up the questionnaire by interviewing the
informants. The information obtained by this
method is likely to be more accurate because the
interviewer can clear up doubts, can cross
examine the informants and thereby obtain the
accurate information.
Example :- This method is applied during the
Census of population.
 2.Mailed Questionnaire.

Very often, instead of personal interview,

questionnaires are sent by post to individual
informants who are suppose to fill it and return
the completely filled in questionnaire again by
post. This method is very popular and usually
referred to as the mailed “questionnaire method”
of investigation.
The following points should be observed
while drafting a questionnaire:-
 1) The questionnaire should be as short as
possible:-

• If many questions are included, then the

questionnaire will be unduly lengthy and as
a consequence the respondent will feel
bored and reluctant to answer all these
questions.
 2) The individual questions should
be short and simple:-

The questions should be short,simple,easy to

understand and convey one and only one
sense. Lengthy questions should cause
irritation, resulting in careless and inaccurate
replies.
If a particular question needs clarification, the
explanation or definition of these terms must
be provided with the questionnaire.
 For Example:-
In a socio-economic survey of a village, in order to know the economic
condition of the households the following questions are being asked.
Q. Quality of dwelling:-

a. kancha b.semi pucca c. pucca

• kaccha bari: it should be made of materials like

bamboo, mud, grass etc.
• semi-pucca: the walls are made with materials
like cement and brick but the ceiling / roof is made up
with “tin”, “tali” and “asbestos”.
• pucca house : all walls and roof are made up with
solid materials like brick and cement, stones, metals
etc.
Q. Type of loans taken in last one year.

a. Consumption loan b. Production loan

Consumption loan: loan for meeting household expenses on

food, clothing etc., medical expenses, and expenses incurred
for performing ceremonies, like marriage, sraadhi or meeting
festival expenses during puja, loans for purchasing consumer
durables, housing loans. It refers to any loan that does not
lead to generation of income.
Production loans: loans taken for the purpose of self-
employment or buying any capital goods (machinery) and for
the expansion of existing business or for meeting working
capital expenses. It refers to loans, which lead to generation of
output and income.
3.Questionnaire should be logically arranged.

• The questions must be arranged in a logical order

so that a spontaneous flow of answers come out.
For example:- It is undesirable to ask a man about
his monthly salary before asking him whether he
is employed or not.
They should not skip back and forth from one
topic to another topic.
• Questions supplying identification and description
of the respondent should come first ( consist of
information about religion,sex,age,marital status,
educational qualification) since the background
questions are easier to answer and can ease the
respondent into the questionnaire.
 4.Questionnaire should be capable of
objective answers
Various types of questions that may form
part of a questionnaire can be grouped
under three categories:-
b)Dichotomous questions.
c)Multiple choice questions.
d)Open-ended or free answer question.
 a) Dichotomous questions.
Dichotomous questions are fixed
alternative questions in which only two
alternative are listed. The respondent has to
tick one of these alternatives. Such
questions can usually be answered in ‘yes’
or ‘no’.
Q. Do you have T.V ?
Yes No
Q. Do you have vehicle?
Yes No

If Yes, then mention the type(tick all applicable)

Cycle
Motorcycle
Scooter
Bike
Car
others (please specify)…………………..
 b)Multiple choice questions.

In Multiple choice questions all possible

answers to a questions are listed and
respondent have to choose the appropriate
one. Most of survey questions are Multiple
choice types since they are generally the
easiest for a respondent to answer and the
easiest to analyze.
Q. Do you have T.V ?
Yes No
If yes, then mention the brand name (tick category)
Sony
LG
Samsung
Phillips
Videocon
Onida
Others (please specify)…………………..
Here the respondent have to tick one out of these
alternatives.
Q.Why did you purchase Sony TV? (tick all
applicable)

Price Lower than other brands.

It presents best quality.
Picture is better.
Warranty period is longer.
After-sales service is better.
Any other (please specify)…………………..
But here the respondent may like to tick more than
one alternative. For example:- One might brought
the Sony TV not only because of the best quality,
but longer period of warranty and better after-
sales service.
 c) Open-ended or free answer
question.
In the free answer form or open-ended questions,
the respondents are asked to answer a question in
his or her own words in essay form.
For example:- The MBA students after completion
of the course may be asked questions like
 What is your opinion of the quality of teaching?
 What do you feel about the facilities offered by the
faculty?
 What do you think of the practical usefulness of
the course you have undergone?
 5.Question of sensitive nature should be
avoided.

Questions affecting the pride or sentiments

of informants must not be asked i.e.
questions on personal matters like income,
amount of profit, property should be
avoided as much as possible.
6.Answer of the questions should not
require calculations
Questions should not require calculations to
be made.
For example – Informant should not be
asked information about the yearly income,
since in most cases they are paid monthly
and so they can no readily provide the
annual income figure, in order to provide
the annual income figure they have to make
calculation, which may lead to errors.
7.The units of measurement should be
properly mentioned.

The units in which the information is to be

collected should be clearly mentioned in the
questionnaire
For example:- If there is a question on
weight ,its units of measurement should be
also specified. whether the weight is to be
expressed in pounds or kilogram.
 8.Cross-checks
If possible, one or more cross-checks should be
incorporated into the questionnaire, in order to
judge the reliability of the answers.
For example:- After collecting the information on
the income of individual member of the family, in
order to cross-check we may collect the amount of
family income.
9.Pre-testing of the questionnaire
After the questionnaire has been prepared, it is
desirable to test it on few individuals. The
procedure, which is known as pilot survey, is
useful in detecting the shortcomings of the
questionnaire, so that the necessary modification
can be done before it is used in the actual enquiry.
 10.Incentives to the respondents
Some incentives in filling up the questionnaire
should be provided to the respondent. It may be in
the form of gift coupons or a sample of a product
which the company wanted to introduce etc.
 Merits of mailed questionnaire:-
• Mailed questionnaire method of
investigation is the least expensive, since
the enumerators do not have to travel to
the informants’ place. The required
information can be collected only at a
fraction of cost needed in the other
method.
• For the same reason, a vast area can be
covered for the investigation.
• The informations can be gathered at a
comparatively short period of time, because
the questionnaire can be sent out
simultaneously to many informants.
• The method is useful for collecting
information on such matters which the
informants may not like to disclose in the
presence of investigator.
Demerits of Mailed questionnaire
• Mailed questionnaire method is not very
satisfactory due to the low proportion of
response. In many cases the respondents are
reluctant to fill it.
• The quality of the collected information is also
very poor because in many cases the respondents
do not carefully fill in the questionnaire and give
haphazard answers and keep it incomplete.
• Respondents may not fully understand the
meaning of the questions and provide an
inaccurate answer.