The cluster sampling It is a type of sampling method that is used when homogeneous groups are evident in a statistical population, but they are internally heterogeneous. It is frequently used in market research.
With this sampling method, instead of immediately selecting all subjects from the entire population, the researcher takes several steps to gather his population sample. First, the researcher divides the total population into separate groups, called clusters. Then select a simple random sample of the population groups. Finally, it performs its analysis taking the sample data of these groups.
For a fixed random sample size, the expected error is lower when the greatest amount of variation in the population is present internally within the groups, and not between the groups..
A common reason for using cluster sampling is to lower costs by increasing sampling efficiency. This differs from stratified sampling, where the motive is to increase the accuracy..
Article index
- The population is divided into N groups, called clusters.
- The researcher randomly selects n groups to be included in the sample, where n is less than N.
- Each element of the population can be assigned to one, and only one cluster.
- Ideally, the population within a cluster should be as heterogeneous as possible, but there should be homogeneity between clusters. Each cluster has to be a representation of the total population on a small scale.
To choose which clusters to include in the study, a random sampling technique is used in any relevant cluster..
In one-stage cluster sampling, all elements within each of the chosen groups are included in the sample..
In two-stage cluster sampling, a subset of elements within the selected groups is randomly selected to be included in the sample..
It should be used only when economically justified, when the cost reduction outweighs the precision loss. This is more likely to occur in the following situations.
For example, it may not be possible to list all customers of a hardware store chain.
However, it would be possible to randomly select a subset of stores (stage 1) and then interview a random sample of customers who visit those stores (stage 2)..
For example, to conduct personal interviews with OR nurses, it might make sense to randomly select a hospital from a sample of hospitals (stage 1) and then interview all OR nurses in that hospital..
Using cluster sampling, the interviewer could conduct many interviews in a single day and in a single hospital.
In contrast, simple random sampling may require the interviewer to spend the entire day traveling to conduct a single interview in a single hospital..
May be cheaper than other sampling plans, for example less travel and administration costs.
This sampling method takes large populations into account. Since these groups are so large, implementing any other sampling method would be very expensive.
In this method, a great concern in spending, such as travel, is considerably reduced..
For example, compiling the information from an investigation in each household in a city would be very expensive, while it will be cheaper to compile information in several blocks of the city. In this case, trips will be greatly reduced.
When estimations are considered by any other method, a reduced variability in the results is observed. This may not be an ideal situation at all times.
When an all-item sampling frame is not available, only cluster sampling can be used.
If the group in the sample population has a biased opinion, it follows that the entire population has the same opinion. This may not be the real case.
There is a higher sampling error, which can be expressed in the so-called "design effect".
The other probabilistic methods give fewer errors than this method. For this reason it is not recommended for beginners.
Cluster sampling is used to estimate high mortalities in cases such as wars, famines, and natural disasters..
An NGO wants to establish a sample of children in five nearby towns to provide them with education.
Through one-stage cluster sampling, the NGO will be able to randomly select populations (groups) to create a sample to help uneducated children in those cities..
A business owner is looking to find out the statistical performance of his plants, which are distributed in various parts of the United States..
Taking into account the number of plants, the work done in each plant and the number of employees per plant, sampling in one stage would consume a lot of money and time..
Therefore, it is decided to carry out a sampling in two stages. The owner creates samples of workers from different plants to form the clusters. Then divide them into the size of a plant in operational state.
A two-stage cluster sampling was formed using other clustering techniques, such as simple random sampling, to begin the calculations..
Geographic cluster sampling is one of the most widely implemented techniques.
Each cluster is a geographic area. Since it can be expensive to conduct a survey in a geographically dispersed population, a greater economy can be achieved than with simple random sampling by grouping the different respondents into a cluster within a local area..
In general, achieving equivalent precision in estimates requires increasing the total sample size, but cost savings may make such an increase in sample size feasible..
For example, an organization intends to conduct a survey to analyze the performance of smartphones across Germany..
You can divide the population of the entire country into cities (clusters) and also select the cities with the highest population. Also filter those that use mobile devices.
Yet No Comments