2 STATISTICAL

INVESTIGATION

INTRODUCTION
By the term investigation (or enquiry) we mean the search for information or knowledge.
Statistical investigation, thus, implies search for knowledge with the help of statistical devices

enquiries have always required considerable skill on the part of the statistician, rooted in a broad
knowledge of the subject matter area and combined with considerable ingenuity in overcoming

the numerical facts since statistical analysis is not possible without them.
For example, if an investigation is made into accounts of a college hostel, then the investigation
will mainly cover:
1. Income from residents as seat rent, meal charge, any grant from college.
2. Expenditure as hostel rent, overhead charges.
3. Expenditure on dry rations, broad meals, including special meals.
4. Expenditure for annual functions.

ORGANISATION OF STATISTICAL INVESTIGATION

Statistical investigation is a long and comprehensive process. It extends over various stages from

1. Planning of statistical investigation.

2. Collection of data.
3. Editing of data.
4. Presentation of data.

6. Interpretation of data.
7. Preparation of the report.
6 Biostatistics

PLANNING OF STATISTICAL INVESTIGATION

execution. Careful planning is essential to get the best results at the minimum cost and time. It
is essential to consider the following points while planning a statistical investigation.
1. Objective of the enquiry should be fully known.
2. Scope of the enquiry should be determined.
3. Nature of information to be collected should be decided.

5. Source of data collection or type of data to be used, that is, primary or secondary should
be decided.
6. Method of data collection, that is, census or sampling method, should be decided
beforehand.
7. Choice of frame should be made.

COLLECTION OF DATA

‘respondents’.

of two types, namely (1) Primary data and (2) Secondary data.

PRIMARY DATA
It is the data collected by a particular person or organisation for his own use from the primary
Wessel

Methods of Collecting Primary Data

1. Direct personal investigation: Under this method, the investigator collects the
data personally. He has to go to the spot for conducting enquiry and has to meet the
persons concerned. It is essential that the investigator should be polite, tactful and has
a sense of observation.
Merits
1. Original data is collected.
2. Reliable information is obtained.
3.
 Statistical Investigation 7
4.
5. It is most useful when the area of investigation is very small.
Demerits
1. It is a very costly method.
2. It is very time-consuming.

4. It can give wrong results.

5. It can be used only when the area of investigation is small.
2. Indirect oral investigation: Here the investigator gets the information indirectly.
In this method a third person is contacted who is expected to know the necessary details

commissions and enquiry committees appointed by the government.

Merits
1. It is a simple and convenient method of investigation.
2. Intensive and extensive investigation is possible.
3. It is economical.
4.
Demerits

doubt.

degree of accuracy can be affected.

3. Information through correspondents: In this method, the investigator appoints

not strictly comparable and the person who does not deal with this data may know the
exact nature of the variation in the reported data.
Merits
1. It is the most economical and extensive method of investigation.
2. It is useful where the information is needed from time to time.
Demerits

4. Questionnaire method: In this method, the necessary information is collected from

8 Biostatistics

(a) Questionnaire sent through post: In this method, the questionnaire is mailed
to the informant. It also carries a request for the quick response and lays down the
aims and objectives of collecting the information. In this method, the return postage

area under investigation is very large and when the informants are educated and
are likely to cooperate.
Merits
1. It can conveniently be used when the area of investigation is very large.
2. It is economical and involves less labour and time.
3. Collected information is more reliable.
4. Original data can be obtained directly from the informants.
Demerits
1. It involves uncertainty about the response.
2. Information can be collected from the literates only.
3. Supplementary questions cannot be asked.
(b) Questionnaires sent through enumerators: In this method, the enumerators
are appointed by the investigators and they contact the informants, get replies and

illiterate.
Merits
1.
informants.
2. Personal contact makes it possible to collect reliable data.
3. Personal bias is minimised.
4. Illiterate informants can also be approached.
Demerits
1. It is costly because the enumerator has to be paid.
2. It is time-consuming.
3. It requires trained enumerators.
4. It can be employed only by big organisations.
5. Technique of drafting of a good questionnaire:
be kept in mind while drafting a questionnaire:

3. Possible answers could be suggested.

5. Irrelevant questions should be avoided.

6. Questions should be free from personal bias and should not injure the feelings of
the respondents.
7. Questions involving the mathematical calculations should be avoided.
 Statistical Investigation 9

SECONDARY DATA

Sources of Secondary Data

Published sources:

1. International Publications:
10 Biostatistics

Individual research scholars and other allied agencies

also supply material on matters of importance.
Semi-government institutions like municipalities,
corporations, district boards, panchayat samities, publish reports and data regarding
birth and death, health and other related subjects.
Unpublished data:

in the universities and other research institutions.

Precautions in the Use of Secondary Data

Before making use of the secondary data, the investigator must satisfy himself about the
following.

2. Whether the data is suitable for the purpose?

3. Whether the data is adequate?

Primary data Secondary data

1. Needs more funds. 1. Needs less funds.
2. Investigating agency collects the data. 2. Some other investigating agency collects it for
its own use.
3. It requires longer time for collection. 3. It requires less time for collection.
4. 4.
more reliable and suitable to the enquiry. has done the collection which may or may not
serve the purpose.
5. It requires elaborate organisation. 5. No need of any organisational set-up.
6. No extra precautions are required. 6. Secondary data needs more care and attention.

METHODS OF DATA COLLECTION

population’ has a different meaning in Statistics than in ordinary usage. Ordinarily,
the word population is used to mean the number of the people living in an area, a region or a
country. In Statistics the word ‘population’ or ‘universe’ refers to the number of observations
connected with the enquiry.

representative of the whole population.

Sample Method.
When the entire population is taken into account for study, it is known as Census Method.
When a small group is taken into account as representative of the whole, it is known as Sample
Method.
 Statistical Investigation 11
Census Method
Census method means the complete enumeration of the population. In the world, every country
conducts the population census at a regular interval of time. In our country the population census
is conducted every 10 years.
Merits

Demerits

may not be used successfully.

Sample Method
In the sampling technique instead of every unit of the universe only a part of it is studied and the

Bowley in 1912 to study the extent of the poverty of labourers. In 1934, the Indian government
also adopted this method.
(a) Random sampling: Random sampling refers to the sampling technique in which each
and every item of the population is given an equal chance of being included in the sample. U.M.

In this method, a sample is selected by which each

item of the population has an equal and independent chance of being included in the
sample. Under this method certain number of items are chosen at random without any

population.
In this technique units are
drawn from the population one after the other. Before drawing any unit, the

in SRSWR there is the possibility for the units to be repeated into the sample.
Nn n’ can be drawn from a population
containing ‘N’ units and therefore the probability of selecting any one of them
1
is equal to n
N
In this technique, the
unit drawn at any given draw is not replaced into the population before
n’ that can be drawn
from a population containing ‘N’ units is equal to Ncn and the probability of
1
selecting any one of them is equal to .
Ncn
12 Biostatistics

when the population is heterogeneous. So in this method the population which is to

be sampled, is divided into homogeneous groups and these groups are named strata
possessing the similar characteristics. Samples are selected by taking equal proportion
of items from each group.

complete list of the population from which the sample is to be drawn is available. Under
this method, units to be drawn into the sample are selected at evenly spaced intervals.

unit is selected by following any of the simple random techniques. Subsequent units are
drawn at an equal sampling interval.
(b) Non-random sampling:
probability. Considerations such as expert judgements, convenience, etc., determine the choice
of sample. In other words the choice of sample is not free, but is determined by a number of
considerations. Some of the important methods of non-random sampling are as follows.
In this method, the selection of items to be included in a

be representative of a given population are selected for studying the characteristics of

a population. Here the probability that an individual item will be selected is not known
because the inclusion of items in a sample depends exclusively upon the preference
or nature of the investigator. In judgement sampling, the accuracy of results depends
upon the investigator. If he is not biased and has full knowledge of the population to be

Quota sampling is a type of judgement sampling. In this method

salary groups, etc. Each interviewer is asked to collect information accordingly. But the

and availability or resources, etc.

one has to make a choice between the two. So, when the population is small and precise

needed with minimum cost, sampling method is the most appropriate.