NOTES AND ASSIGNMENT

COLLECTION OF DATA

Key words
Investigator: He is the person who plans and conducts the statistical investigation. He may
conduct the investigation himself or with the help of others.
Enumerator: Enumerator is the person who goes out to the field and collects desired statistical
information He is the person who actually collects data.
Respondent: Respondent is a person who provides the actual data. He responds by answering
the questions included in the questionnaire.

To study any economic problem the relevant data is required, which can be collected through
various ways. The source of information may either be primary or secondary. Thus statistical
data may be either being (i) primary data or (ii) secondary data.
Difference between primary and secondary data.

Primary data Secondary data

1 Primary data are those data which are Secondary data refers to those data which
collected originally. are collected through other sources.
2 Primary data in the form of raw material Secondary data are in the form of
to which to which statistical methods are finished product as they have already
applied. been statistically applied.
3 Primary data are collected for a definite Secondary data are collected from
purpose. published and unpublished sources.
4 Primary data are strictly in accordance Secondary data are made according to the
with the investigation. object of the investigation.

Methods of collecting primary data

Primary data may be collected by using any of the following methods
1 Direct oral investigation.
2 Indirect oral investigations.
3 Information through correspondents.
3 Information through mailed questionnaire.
4 Questionnaire through enumerators.
Direct personal investigation: Under this method data are collected by asking questions relating
to the enquiry. Suppose the investigator wants to collect the data about the wages of a particular
factory he would go to that factory and obtained required information by contacting the workers
personally.
Indirect oral investigation: Under this method the investigator collects the facts by interviewing
persons not those from whom information is to be collected instead he makes contacts with some

Page 1 of 4
other persons who are directly or indirectly in touch with them eg in case of a road accident the
investigator can ask the person present at or near the place of accident.
Information through correspondents: The investigator does not get the information personally
rather he appoints local agents or correspondents in different parts of the field of enquiry . These
agents or correspondents would collect and send the required information to the central office
where the data is processed.
Information through mailed questionnaire: Under this method a questionnaire is ( a list
consisting of questions) is sent to the informants by post along with a covering letter to fill up the
questionnaire and send it back within a specified time.
Essentials of a good questionnaire
A questionnaire is a list of questions relating to the field of enquiry and provides space for
answers. It may be defined as an instrument of collecting primary data from a large number of
persons. A good questionnaire must possess the following:
1 The questions asked should be short and simple.
2 The number of questions should be small as far as possible.
3 Questions should be properly arranged.
4 The questions should be directly related to the investigation.
5 The questions which require calculations should be avoided.
6 Personal questions should be avoided.
Main sources of secondary data:
Most of the investigators make use of secondary data in their respective studies The main
sources of the secondary data can broadly be classified into two parts:
Published sources: The various sources of published data are given as follows,
(i) Government publications are the major sources of secondary data. Various
government departments regularly collect and publish data which are useful and
reliable eg economic surveys, annual budgets, national sample survey etc.
(ii) Semi official publications of various local bodies such as municipal committees,
corporations etc.
(iii) Many international institutions such as IMF world bank etc provide useful statistical
information about various countries of the world .
(iv) Private publications such as stock exchange, labour unions, financial and economic
journals etc.
(v) Publications of universities, research institutes and individuals fall in this category.
Methods of collecting data
There are two methods of collecting the required information , these are (a) census method (b)
sampling method.
Census method or population: Under t5his method the data is collected from each and every unit
of the investigation, this method is also called the method of complete enumeration.

Page 2 of 4
Sampling method: In the sampling method only a part of the universe instead of every unit of
the universe is studied. On the basis of this part of study conclusions are drawn for the whole
universe.
Meaning of sample and universe:
In the field of any statistical inquiry, there may be persons, households, items, shops or any other
unit. The aggregate of all such similar units under investigation is called census or population or
universe.
Sample is a miniature picture of population from which it is taken eg inspection of a handful of
grain to judge the quality of wheat, rice or pulses etc.

Difference between census and sampling method

Census sampling Sampling method
1 Under this method each and every unit of Under this method only a few units of the
the universe is studied. universe are selected and studied.
2 High degree of accuracy is obtained. High degree of accuracy is not obtained.
3 It is more suitable where the scope of It is suitable where the area of enquiry is
investigation is limited. wide.
4 Census method require a large number of
enumerators. Sampling can be taken for out by the quite the
5 The result of the investigation cannot be smaller number of enumerators.
verified. Result can be tested by taking out another
6 It is suitable where all items of the universe sample.
are not homogenous. It is not suitable where units of the universe
are heterogeneous.

Methods of Sampling
Random sampling
Under this method selection of item for a sample depends upon chance. Every item of the
universe has equal chance of being selected e g Lottery method
Lottery method is a very common and popular technique of taking a random sample. Under this
method all items of the universe are given separate numbers on separate slips. Now these slips
are mixed up in a container or drum. Then the selection is made blindly of the number of slips
that are required for the sample. The withdrawn slips constitute the random sample.
Deliberate sampling: Under this method the sample is taken out from the universe according to
the need of the investigator, he selects the items which he thinks are full representative of the
universe.
Stratified sampling: Under this method the population is divided into several homogenous
groups, e g a sample of the 20 students is to be taken from the total of 2000 students of a school .
The school will be divided into various streams such as science commerce and arts . Then
sampling strength of 20students should be distributed from all three faculties.
Systematic sampling: Under this method all units of the universe are arranged in a systematic
manner on some basis. The basis can be numerical, alphabetical or geographical or some other

Page 3 of 4
basis. This method is generally used in those areas where a complete list of the whole universe is
available.
Multi stage sampling: Under this method sample is taken out in several stages Suppose we have
to select a sample of 1000 households from the state Uttar Pradesh. At the first stage the state is
to be divided into districts and a few districts are selected at random. Then in the second stage
each district is subdivided into villages and few villages are selected are random and at the third
stage number of households is selected from each village, in this way sample goes on decreasing.
Quota sampling: Under this method the whole universe is divided into various groups on the
basis of different characteristics such as sex, occupation, age, income etc. Quota of a person is
fixed for each group of each investigator.
Sampling and non sampling errors
In a sample survey only the small part of population is studied, therefore while drawing
conclusions about the population there would be some amount of inaccuracies and errors. In a
census study such errors disappear altogether. Thus the difference between samples
Result and a census result under the same circumstances is called sampling errors.

ASSIGNMENT QUESTIONS
1 How are samples drawn through table of random numbers?
2 What is non-random sampling?
3 Write a short note on the following
(a) Judgment sampling
(b) Quota sampling
(c) Convenience sampling
4 Do you agree that sampling is a better than census method of collecting information. Discuss
any six valid points to prove your point.
5 What are the limitations of sampling method?
6 Prepare a questionnaire with 10 questions about the consumer’s preference of ABC Television
set.

Page 4 of 4