Blog
Use Stratified Random Sampling When the Population Is Not Entirely Homogeneous
Sampling techniques are very important, especially when we’re observing a specific population. By taking samples, we can save on costs, time, and effort—yet still obtain results that represent the population being studied.
Generally, if the population members are homogeneous, we can simply take a random sample from the population. But what if the population is not entirely homogeneous? For instance, the population might be divided into several strata, and each stratum is internally homogeneous.
In this article, I’ll walk you through the stratified random sampling technique. Let’s dive in together.
What Is Stratified Random Sampling?
Stratified random sampling is a probability sampling method where the population is divided into strata or subgroups that are homogeneous based on certain characteristics. Then, from each of these strata, samples are selected randomly.
This sampling technique is suitable when we know there are important variations within the population being observed, and we want to avoid bias.
So, why choose stratified random sampling over simple random sampling? The reason is because the population contains important subgroups or strata that should be represented. By using this method, we ensure that each key subgroup is proportionally represented in the research sample.
Types of Stratified Random Sampling
Next, we need to understand that stratified random sampling is divided into two types:
- Proportionate stratified sampling, and
- Disproportionate stratified sampling.
The difference lies in how the samples are taken:
- In proportionate stratified sampling, the number of samples drawn from each stratum is proportional to the size of that stratum within the population.
- In disproportionate stratified sampling, the number of samples from each stratum does not have to be proportional. This is usually applied when there are small but important strata that need to be included—even if their population size is relatively small.
Case Study Example
To make it easier to understand this sampling technique, let me give you a simple case study.
A researcher is observing high school students’ satisfaction with a digital learning program in City ABC. Based on the data, City ABC has 10 high schools. The research population consists of Grade 11 students from all 10 schools, totaling 2,000 students.
However, there is a clear division: some of the schools are public, while others are private. The researcher suspects there are differences in the implementation of digital learning between these two types of schools. So, the population is divided into two strata: public and private schools.
Initial data shows that there are 1,400 students from public schools and 600 students from private schools. Using Slovin’s formula, the minimum sample size is calculated to be 200 students.
Then, using the proportional method, we calculate the number of students to sample from each group:
- 140 students from public schools
- 60 students from private schools
This number of samples is proportional to the size of each subgroup in the population. The next step is for the researcher to randomly select students within each stratum. This can be done using a random number generator, Excel, or any other randomization tool.
Final Thoughts
Alright, that was our discussion on the stratified random sampling technique. I hope this helps you understand when to use simple random sampling and when stratified random sampling might be the better choice.
Hopefully, this article adds value and provides new insights for all loyal readers of Kanda Data. Thanks for reading, and see you in the next article from Kanda Data!