Basics of Research Process

Stratified Random Sampling: Definition, Methods & Examples

Stratified sampling
Worried about writing a unique paper?

Use our free
Readability checker

Stratified sampling requires dividing a population into smaller sub groups or strata based on certain characteristics. From each stratum, a sample is then randomly selected. Stratified random sampling is particularly useful when the population exhibits significant diversity or variability in certain characteristics. It allows researchers to obtain a representative sample from each subgroup.

This comrehensive guide provides a detailed information on stratified random sampling and its methods. Once you finish reading this article, you will be a master of stratified sampling. So let’s get started! Or, if the topic sounds like alien technology to you, simply go for our research paper writing help

What Is Stratified Sampling: Definition

Stratified sampling is a method, where researchers use strata (plural of stratum) to divide a population into homogeneous sub populations depending on distinct features. Every person in the population involved in your survey is assigned to one of such strata. Researchers test each stratum using a different probability sampling approach, such as cluster or simple randomization to estimate statistical measures for each subpopulation.

This method is also known as quota random sampling as researchers take quotas based on properties such as race. Only then, they conduct a secondary testing.

Stratified sampling is widely used if the population's features are diverse and researchers want to guarantee that every attribute is accurate. Different properties yield different results. For that reason, this method is utilized only in selective use cases.  

Stratified vs Cluster Sampling

The difference between stratified and cluster sampling is fundamental. In stratified sampling the sizable number of populations is split into distinct homogenous strata, from which members are picked randomly. In the cluster method, the target demographic is split into several groups. To construct the target data, some of these clusters are randomly picked, or a two-step or multistage sampling is carried out. The elements of stratified sampling will be unique ensuring that your whole population is used. In cluster one, that’s not the case.  

Systematic vs Stratified Sampling

Systematic random sampling and stratified random sampling are again fundamentally different as well. Systematic method requires that you use a k value as an interval to select data from the population. Stratified method helps you obtain samples from generated proportions of data through groups of elements by properties. Once you identify such properties, you should conduct randomization to get results. In stratified sampling, members are randomly picked from strata that are non-overlapping and homogenous.

When to Use Stratified Sampling?

To determine when to use stratified random sampling we must conceptualize how researchers use stratum in their studies and surveys. This approach is preferable when you are working with diverse populations. The method allows to reduce selection bias and guarantees that the whole demographic group is presented.

However, conducting statistical surveys is quite challenging since you will need a large population. To save time and money, a more practical method would be to choose a smaller group or sample size to test the total population. Over the course of data collection, this method is also a good way to improve accuracy and population portrayal.  

Stratified Sampling: Formula

The stratified random sampling formula can be represented as follows:

Stratified Sampling Formula

In this formula, nh is the sample size for the hth stratum and Nh is a population size. N is the entire population along with n as the entire sample size.

Researchers will need great precision in conducting this hence strata should be considered when determining samples. In this method, accuracy determines how efficient this approach will be. Overrepresented or underrepresented stratum may cause invalid results. For this reason, you should be as precise as possible.  

Stratified Sampling: Example

Here’s an example of stratified sampling. For instance, let’s assume some team does research on the demographics of college students in the United States. It discovers that 13% of students are majoring in English, 28% are majoring in Science, 24% are majoring in Computer Science, 21% are majoring in Engineering, and 15% are majoring in Mathematics.

In this case, five strata are formed. The team will then do possible statistical analysis. Researchers check whether the population stratum is proportional to a sample stratum. However, they discover that these proportions are not equal. This team must then randomly pick 480 English, 1,120 Science, 960 Computer Science, 840 Engineering, and 600 Mathematics students from the population of 4,000 students.  

How to Do Stratified Sampling

To determine how to do stratified random sampling one must reflect a demographic group under investigation. A researcher picks a small sample size with similar characteristics. A survey population may be too vast to evaluate individually. Therefore, it is divided into groups with similar features to save money and possible effort.

This method has a wide range of applications, including calculating income for various populations, polling elections, and predicting life expectancy. When looking for a suitable stratum, look for one that reduces variance in the variables under examination while maximising variation between strata.

Then, you should choose a sample size. It should be large enough to make statistical calculations. Finally, you should apply another probability method to gather data from each stratum.  

Pros and Cons of Stratified Sampling

The advantages and disadvantages of stratified sampling range for various reasons. Some of these are small while some differentiate in method and the variance caused in your final data set. Below you can find both pros and cons of this method.

Advantages of Stratified Sampling

Advantages of stratified random sampling allow for a standard statistical base to be adopted in a large study like a case study. Since average researchers stratify the whole population before using random methods, stratified random sampling properly reflects the group being researched. In other words, it guarantees that each subgroup is accurately represented. As a consequence, this method provides greater population coverage since researchers have more control over subgroups and can ensure that all of them are included.  

Disadvantages of Stratified Sampling

The disadvantage of stratified sampling unfortunately lies in the approach which cannot be applied in all studies. It must meet a number of requirements in order to be effective. Every member of a population being researched must be identified and classified into one, and only one, subpopulation. As a result, stratified this method is inefficient when researchers are unsure whether or not every population member belongs to a subgroup. This can affect the research outcomes and the entire study may be invalid. It might also be difficult to compile a complete and conclusive list of a whole population.  

Final Thoughts on Stratified Sampling

With a diverse population that can be separated using supplementary data, stratified random sampling works best. Feel free to use this method to ensure similar variance or reduce a general variance in your population. By the way, our dissertation writing service includes stratified sampling.

Need expert help?

If you are struggling with a project that requires stratified sampling to be done, look no further! We offer top-notch academic writing service and guarantee quick turnaround.  

Article posted on:May 10, 2022
Article updated on:May 11, 2023


Leave your comment here:
6/25/2022 9:48 AM
What is the difference between multistage sampling and stratified sampling? I know this question is basic, but I just can't see any difference between these two sampling methods. Could you provide some examples, if possible?
6/25/2022 11:47 AM
Hi Karen,

In both multistage and stratified sampling you need to divide a large population into clusters. However, unlike multistage sampling, in stratified sampling all clusters should be sampled.

For example, if you want to study long-term memory in high-school students in the USA, you will divide all 50 states into 10 regions and then randomly select students from each region. In multistage sampling, you will select only 10 states out of all 50 states, and then you will choose only 30 counties out of all counties in 10 states and so on.
7/7/2022 8:27 AM
How to ensure that stratified sampling is proportional? I want the results to be valid and represent the entire population, but I'm confused how to make it proportional.
7/7/2022 9:00 AM
Hi John,

To ensure proportional stratified sampling you should check whether the size of your sample strata is proportional to the population size of the stratum. In other words, each selected stratum should have exactly the same sampling portion.