Stratified random sampling refers to making a layer or classes while classifying the population units into subgroups based on their similar characteristics. This process is called strata or stratification. Stratified random sampling is used, if the population is divided in heterogeneous nature but still after dividing the population into several layers, then it needs to be converted into homogeneous nature by using simple random sampling for getting the accurate results from the filtered samples.
What describes the word “Random” in Stratified random sampling?
The word “Random” describes the instant action taken at the same time without showing and performing any partiality or without bias. At the time of stratification, random shows distributed the individuals into classes, subgroups without thinking so much or searching enough, that’s why it is used in a heterogeneous nature which is complex in nature that involves difficulty.
On the other hand, simple random sampling refers to taking the action based on characteristics or attributes, categories. At the time of stratification, random shows distributed the individuals into samples after examining or surveying them does not distribute into subgroups or classes, that’s why it is used in a homogeneous nature which is apparent in nature that involves feasibility.
When stratified random sampling is used?
Stratified random sampling is used in the nature of a heterogeneous population but after converting the heterogeneous population into multiple strata, it takes the nature of homogeneous population through simple random sampling because it does not exist without homogeneous and simple stratified sampling is the end of constructing result from the distributed individuals in multiple varieties into multiple samples
Understand with a theoretical example
For example, when a company decide to choose the best employee of the year, they have to follow some procedure like:
- They have to collect data of employees
- Then, they will make different sections for distributing the employees based on their characteristics, experience, and work opportunities.
- After classifying, they need to take out an employee’s name who gives the best in the whole year and choose a deserving name from samples, filtered samples include the filtered results in which it is easy to find the name.
We hope now you will understand the process of stratification with the above example. In simple words, stratification refers to classifying with distributing the population units into multiple groups after collecting the population data based on their characteristics.
When do researchers need to avoid overlapping?
Researchers are those who continue the process of sampling through the collection of data so that they can survey which one is best and accurate. It is done by dividing the population units into several layers, classes, which is known as strata, and samples are taken out from stratum so that they can reach the result but at the time of happening this procedure, they need to avoid overlapping so that data cannot be mixed up or we can say that divided population units cannot be lied on each other or being cannot be confused.
Researchers can be advantageous for stratifying processes because they have full control on population division to get out better coverage of population and take out a better representation.
Stratification is an economic part of statistics that are divided into groups by researchers to stratify the study by using strategies. Overlapping creates confusion or involvement of high risk for individuals.
Advantages of SRS (Stratified Random Sampling)
- Representation contains accuracy if researchers apply stratified random sampling rules.
- Follow study material by researchers before acquiring any method of sampling.
- Researchers have full authority to study with all available details.
Disadvantages of SRS (Stratified Random Sampling)
- Overlapping is the biggest disadvantage of SRS.
- This method can’t be used in every economic study.
- It doesnot divide the population into multi subgroups.
- The result could be uninformative and unsatisfied with lack of accuracy.