CLUSTER SAMPLING

Cluster sampling is a sampling in which the entire population is divided into groups with the proper sampling of the groups or clusters and are selected on random basis. It can be used when a researcher does not get the exact information or the complete list but gets the information in groups or clusters. This sampling is more economical in comparison to the other samplings.

EXAMPLE

Suppose that Department of Food wants to identify the number of people consuming wheat in South India. First of all a sample will be taken to identify the number of countries in South India, then a sample will be prepared to identify the consumption of rice in different stores and the number of people consuming rice, the amount will be matched with the consumption of wheat in the particular country. It can be seen that if the researcher gets the number of people, then it will be easier for him to take out the people who are consuming wheat in their food.

REQUIREMENTS FOR CLUSTER SAMPLING

1. It should be entirely heterogeneousi.e dissimilar in elements or parts.
2. Every cluster should have the tendency to represent the information on small scale so that it is small and less than the presentation of the entire population.

SUBTYPES OF CLUSTERCSAMPLING

There are two subtypes of cluster sampling

Single stage cluster sampling – In a single stage cluster sampling the sampling is done only once. For example one researcher has to search the A1 students in class tenth, so he or she will select the sections of class tenth randomly and then he or she will select the A1 students in that particular section.

Two stage cluster sampling – The two stage cluster sampling is similar to the first stage except that a random sample is taken from each cluster which is termed as sub -sampling. Refering to the above example it can be said that the researcher will select the subset of students in that particular section who can score A1.

ADVANTAGES OF CLUSTER SAMPLING

To consume less time , money and labour geographically designed cluster sampling is a boon.

It is easier to implement in comparison to other samplings.

In some samplings proper frame is required for maintaining a design but in cluster sampling no frame is required.

It is much easier for implementation.

DISADVANTAGES OF CLUSTER SAMPLING

In representation, simple random sampling is more presentable.

In the field of analysing data cluster sampling is difficult to analyse data in comparison to simple random sampling.

Since there are more stages in cluster sampling so it invite more number of errors to enter into the statics data.

Novel information is restricted in cluster sampling because the same information is repeated many times in different clusters.