Theory of sampling 2019/Jyotika Minhas/Theory...Basis of sampling 1. The law of statistical...
Transcript of Theory of sampling 2019/Jyotika Minhas/Theory...Basis of sampling 1. The law of statistical...
Theory of SamplingPrepared by:
Jyotika Minhas
Assistant Professor
Department of Economics
Hans Raj Mahila Maha Vidyalaya
Jalandhar
Punjab
INTRODUCTION
For statistical enquiry we need reliable and
accurate data.
There are two ways to collect statistical
data.
Methods of enquiry
Census
method
Sampling
method
Universe or Population
Population or Universe in statistics is the
entire body of terms or items, about which we
want to obtain information.
TYPES OF UNIVERSE
Finite Infinite Existing Experimental
CENSUS METHOD
Under this technique each and every unit of
population is studied.
For example:- suppose we want to calculate the average income
of a colony.There are 10 houses in the colony and their total income are as follows:-
200,300,200,300,500,200,300,500,200,300
Their average income will be 300rs.
Merits of census method
Intensive study of Population
High Degree of accuracy and reliability
No chance of bias
Suitability of the method
Study of each and every unit of population
Demerits of census method
Time consuming
Costly method
Possibility of errors
Destructive population
Not suitable for infinite universe
Difficult to arrange large no. of trained
enumerators
This method is suitable when;
Universe is comparatively small.
Trained enumerators are available.
Intensive study is to be undertaken.
Accurate and exact results are required.
Sampling method
Sample:- A sample is a part of the universe or
population which contains the qualities of
universe.
In the sampling method only a part of
population is studied and on its basis valid
conclusions are made.
Examples:-
Basis of sampling
1. The law of statistical regularity:- if a large sample is
taken randomly from a population then the obtained
results are fairly closed to the census method.
2.The law of intersia of large numbers:-according to this
law ,larger is the size of sample,more chances of its
results are likely to be near to that of population.
Properties of good sample
Representativeness
Adequacy
Independence
Homogeneity
Sample should be in accordance with the
objective of investigation
Advantages of sampling method
Saves time and labour
Economical method
More dependable results
More detailed information can be collected
Destructive population
Sample investigation is easy to organise and
supervise
More scientific than census
Disadvantages of sampling method
Not suggested in all circumstances
It is not possible to achieve hundred percent
accuracy
Requires services of experts
May not be always economical
Difficult to obtain representative sample in
practice
Sampling errors
Difference between census method
and sampling method
1. Degree of accuracy
2. Scope
3. Cost
4. Convenience
5. Time
6.Reliability
7.Nature of items
8.Principles
9.Verification
10.Area
METHODS OF
SAMPLING
RANDOM SAMPLING
METHOD
In this method each and every
Unit has an equal or some definite
probability of being selected in a
sample.
Probability Sampling Methods
a) Simple Random Sampling
b) Restricted Random Sampling
I. Stratified sampling
II. Systematic sampling
III. Cluster sampling
Selection of Simple Random Sample
1. Lottery Method
2. Random Number Tables
Merits
1. There is no chance of
personal biasness.
2. It saves lot of time and
money.
3. It is a scientific method.
Demerits
1. Random sampling is not suitable for
heterogeneous data.
2. In lottery method if slips are not made
identically, so there is chances of
biasness.
3. The investigator has no control over the
selection of units.
Stratified Random sampling
Under this method the
universe is divided
into sub groups and
sample is taken from
each sub group. The
Sub group is known
as strata.
Methods of Stratified
sampling
1. Proportional method
2. Disproportional method
Merits
1. It helps in achieving high degree of accuracy.
2. There is a greater control of investigator in this method.
3. They are expected to be localised in a small area . So it is convenient method.
Merits
Demerits
1. There is possibility of biasness.
2. It is more expensive.
3. It is very difficult task to divide the
universe into homogeneous groups.
Systematic Random
Sampling
Under this method a sample is
taken from a list prepared
on a systematic arrangement
such as alphabetically ,
chronologically,
geographically order .
Merits
1. It is easy to understand.
2. It saves lot of time , energy and
finance.
3. If size of population is large then this
method is suitable.
Demerits
1. To apply this method a complete and
up to date list of units is required, which
is usually not available.
2. Any hidden periodicity in the list will
adversely affect the representative
character of the sample.
•This type of sampling is done in stages.
•Under this sampling the population is divided and sub – divided into groups called
clusters and sampling process is accomplished in
different stages.
Merits
1. This method is simple to understand.
2. Multistage sampling is more flexible
than other methods.
3. This method is systematic method
covering the divisions and sub-divisions
of entire population.
Demerits
1. If number of cluster large then
representative sampling is affected
2. This method requires higher cost and
time.
3. The results obtained through this
sampling is less accurate than simple
random sampling.
Non Probability Sampling is a technique
in which the selection of the sample
depends upon the personal judgement
of the investigator .
1. Judgement sampling
2. Convenience sampling
3. Purposive sampling
4. Quota sampling
TECHNIQUES OF NON PROBABILITY SAMPLING
In this process , sampling is
done on the judgement of
the investigator . This
method requires a deep
and thorough knowledge
of the universe on the part
of the investigator.
1. This method is applied when size of
sample is small.
2. If we have to study some of the known
characteristics of the universe then
judgement sampling helps us.
3. Judgement sampling is helpful in solving
day to day problems .
1. This method suffers from the personal judgement
of the investigator.
2. It is difficult to calculate sampling error because
the items are not based on random sampling
techniques.
3. It is a non-scientific method.
Convenience SamplingIn this method, the sample units are
selected as per the
Convenience of the investigator . He
selects only those
Units which are convenient for him.
Merits
1. This method is applied when
population is not clearly defined.
2. This method can be used when there is
non- availability of resource list.
Demerits
1. This is a non-scientific method.
2. Personal bias is there.
3. The results obtained are not true
representative of the population.
Quota Sampling
Under quota sampling the population is divided into some groups according to some characteristics like age , sex, occupation, religion, education level etc. Then investigator is then told to select from each of the sub-group according to some pre determined quota.
Merits 1. Quota sampling is often used in public
opinion surveys including all sections
of society.
2. In quota sampling , the cost of
sampling is less then other methods.
3. This method can give accurate results
if proper checks or control are
imposed.
Purposive Sampling
In such sampling the investigator has
some pre-determined objective In
his mind which he wants to impress
upon .Purposive sampling is also
known as Deliberate sampling.
Merits
1. The investigator may include necessary
items and remove the non-essential.
2. Such kind of sampling solves the purpose of
the investigator.
3. In this sampling, the investigator has full
control and can represent the samples in the
best possible way.
Demerits
1. It suffers from personal bias.
2. It requires complete knowledge of the
population , but investigator lack such
knowledge.
3. It is highly non-scientific in nature.
4. It does not represent the entire
population.