There are different methods used to draw a sample
from a given population. These methods are broadly classified into two
categories namely –
2. Non-Probability
Sampling
Non-probability sampling is based on the personal
judgement of the researcher. In this method a desired number of sample units
are selected based on the object of inquiry so that the important items
representing the true characteristics of the population are included in the
sample. Purposive sampling, Quota sample and convenience sampling are the
methods of non probability sampling.
Probability Sampling
I. Simple
Random Sampling:
In this method, each and every unit of the population has
an equal chance of being selected in the sample. Selection of a unit is
independent of other units in the sample.
A number of techniques are used to
select a random sample. Lottery method is the simplest method where all the
units are numbered on identical slips of paper. All the sips are folded and
mixed in a drum thoroughly. Required numbers of slips are drawn from the drum
blindfold.
Random sample can also be selected
by constructing lists of all the units by assigning numbers 1,2,3 … and
selecting the units from the random number tables constructed for the purpose.
Merits and Demerits of Simple
Random Sampling:
The advantages of this method are –
1. It
is a more scientific method and eliminates personal bias of the researcher
2. No
knowledge of the characteristics of the population is required
3. Sample
drawn is more representative of the universe.
4. Accurate
results can be expected
5.
Very simple to follow and it saves time,
money and labour.
The random sampling method has
certain difficulties and limitations. These are –
1. This
method requires complete lists of the population. For some types of enquiries,
lists may not be available. In such cases this method cannot be used.
2. If
the data is to be collected from a wider geographical area, then this method is
difficult to use as it consumes more time and requires lot of money.
3.
If the sample size is too small, then it
may not be representative of the population.
2. Stratified Random Sampling
When the population is heterogeneous
then stratified sampling can be used to obtain more accurate results.
Stratification means division of the universe into groups based on
geographical, sociological or economic characteristics.
Process
of Stratification Stratified Random
Sampling involves the following steps.
o
The universe is first divided into
sub-groups or strata and the required units are selected at random from each
sub-group
o
The stratification should be conducted
in such a way that the items in each group or strata should be similar to each
other but they vary significantly from items in other groups.
o
The size of each group or stratum must
be large enough so that it is possible to select the sample units using random
sampling.
o
Size of the sample from each stratum can
either be proportional or disproportional to the size of each stratum.
Merits
and Demerits of Stratified Random Sampling:
This method has several advantages.
These are –
o
If the stratification is done correctly,
then even a small number of units will form a representative sample
o
In this method all the groups are
equally represented and no one group is left unrepresented
o
Bias can be controlled to a great extent
in this method. Greater accuracy can be
achieved.
o
It saves time and cost of data
collection as the sample size can be less.
The disadvantages of this method
are –
o
Dividing the population into homogeneous
groups is a difficult task
o
If the strata are overlapping or
disproportionate, then the sample may not be representative of the population
o
If the stratification is faulty then it
is not possible to correct the sampling errors.
3. Systematic Sampling:
In this method, the sample is
selected from a list prepared in a systematic arrangement either on the basis
of alphabetic order or on house number or any other method. In this method,
first the entire population is arranged in serial numbers 1 to N. Then the size
of the sample should be decided. By dividing the population by the size of the
sample, the first sample unit is obtained. Using the same interval, any number
of sampling items can be selected. For example, if the population is 1000 and
the sample size is 250 then the first sample unit is obtained by dividing
1000/250=4. Then starting from 4 and leaving the same interval, items numbered
4-8-12-16-20-24…. are selected in the sample.
Merits and Demerits of
Systematic Sampling:
Merits
·
Easy to follow and checking can be done quickly.
·
Sample will be representative due to
randomness in selection.
Demerits
Demerits
·
This method can be used only when
complete and uptodate sample frame is available and if the units are randomly
arranged.
·
If the lists are faulty then it will
affect the sample.
In this method the
total population is divided into some sub-divisions which are termed as
clusters using simple random method. All the units in each cluster are
surveyed. The cluster should be as small as possible. When the cluster is small
will be easy to study and it saves time and money. The number of units in each
cluster should be the same.
If the
clusters are many, then the probability and representativeness of the sample is
affected. The results will be less accurate if the number of sampling units in
each cluster are not same.
This method is
generally used in selecting a sample from a very large area. Multi-stage
sampling refers to the technique which is carried out in various stages. The
entire population is regarded to consist of number of primary units. The
primary units are again composed of secondary units which further consist of
third stage units and so on until the researcher ultimately obtains the desired
sample. At each stage random sampling techniques are used. The sample size may
be proportional or disproportional.
This
method is particularly useful in surveys of under developed areas where up to
date sample frame is not available. However, errors are likely to be large in
this method compared to other methods.