STRATIFIED SAMPLING: WHAT IS IT, TYPES, ADVANTAGES AND DISADVANTAGES - ECONOMY - 2025

The stratified sampling, or stratification, is a sampling method that involves dividing a population into smaller subgroups, known as strata. In turn, these strata are formed based on the shared attributes or characteristics of the members, such as income or educational level.

It is used to highlight the differences between groups in a population, as opposed to simple sampling, which treats all members of a population as equal, with the same probability of being sampled.

Source: needpix.com

The goal is to improve the precision of the sample by reducing the sampling error. It can produce a weighted mean with less variability than the arithmetic mean of a simple sample of the population.

Stratification is the process of fragmenting members of a population into homogeneous subsets prior to sampling. Through the strata, a population distribution is defined.

That is, it must be collectively exhaustive and mutually exclusive, so that a single stratum must be assigned to each element of the population. Then a systematic or simple sampling is applied within each stratum.

Important considerations

It is important to note that the layers should not be juxtaposed. Having overlapping subgroups will give some people a greater chance of being selected as subjects. This totally dulls the notion of stratified sampling as a sampling prototype.

It is equally important that the researcher must use simple sampling within the different strata.

The most common strata used in stratified sampling are age, gender, socioeconomic status, religion, nationality, and educational level.

What is stratified sampling?

When completing the analysis on a group of entities with similar characteristics, an investigator may find that the population size is too large to complete the investigation.

To save time and money, a more feasible perspective can be taken by selecting a small group from the population. This small group is called the sample size, which is a subset of the population used to represent the entire population.

A sample from a population can be selected in several ways, one of which is with stratified sampling. This involves dividing the total population into homogeneous groups called strata. Then random samples are selected from each stratum.

Process for performing stratified sampling

- Divide the population into subgroups or smaller strata, according to the attributes and characteristics shared by the members.

- Take a random sample from each stratum in a number that is proportional to the size of the stratum.

- Group the subsets of strata to form a random sample.

- Carry out the analysis.

For example, consider a researcher who wants to know the number of business students who received a job offer within three months of graduating in 2018. They will soon discover that there were almost 200,000 business graduates that year.

You could decide to simply take a random sample of 5,000 graduates and conduct the survey. Better yet, you could divide the population into strata and take a random sample from those strata.

To do this, you would create population groups based on age, race, nationality, or professional background.

A random sample would be taken from each stratum, in proportion to the size of the stratum with respect to the total population. These subsets would be grouped together to form a sample.

Types

Proportional stratified sampling

In this type, the sample size for each stratum is proportional to the population size of the stratum when compared to the total population. This means that each stratum has the same sampling rate.

When a characteristic of individuals is selected to define the strata, the resulting subgroups are often of different sizes.

For example, you want to study the percentage of the Mexican population that smokes, and you decide that age would be a good criterion to stratify because it is believed that smoking habits can vary significantly according to age. Three strata are defined:

- Under 20 years old.

- Between 20 and 44.

- Over 44.

When the population of Mexico is divided into these three strata, the three groups are not expected to be the same size. In fact, the actual data confirms this:

- Stratum 1: 42.4 million (41.0%).

- Stratum 2: 37.6 million (36.3%).

- Stratum 3: 23.5 million (22.7%).

If proportional stratified sampling is used, the sample should consist of strata that maintain the same proportions as the population. If you want to create a sample of 1,000 individuals, the samples must have the following sizes:

It is very similar to gathering a smaller population, determined by the relative proportions of strata within the population.

Uniform stratified sampling

In this type, the same sample size is assigned to all defined strata, regardless of the weight of these strata within the population.

A uniform stratified sampling taking the previous example would produce the following sample for each stratum:

This method favors the strata that have less weight in the population, by granting them the same level of importance as the more relevant strata.

This reduces the overall effectiveness of the sample, but allows the individual characteristics of each stratum to be studied with greater precision.

In the example, if you want to make a specific statement about the population of stratum 3 (over 44), you could reduce the sampling errors using a sample of 333 units, instead of a sample of 227 units, as obtained from proportional stratified sampling.

Advantages and disadvantages

Stratified sampling works well for populations that have a variety of attributes, but will otherwise not be effective if subgroups cannot be formed.

- Advantage

Collect key features

The main advantage of stratified sampling is that it collects the key characteristics of the population in the sample.

Similar to a weighted average, this sampling method produces characteristics in the sample that are proportional to the total population.

Higher statistical precision

Stratification gives less error in the estimation than the simple sampling method. The greater the difference between the strata, the greater the gain in precision.

There is a higher statistical precision when compared with simple sampling. This is due to the fact that within the subgroups the variability is lower, when compared with the variations that occur with the total population.

Smaller sample size

As this technique has high statistical accuracy, it also means that it requires a smaller sample size, which can save researchers a lot of effort, money and time.

- Disadvantages

Unfortunately, this research method cannot be used in all studies. The disadvantage of the method is that several conditions must be met for it to be used correctly.

Difficulty finding strata

The main disadvantage is that it can be difficult to identify appropriate strata for a study. In addition, finding a complete and definitive list of an entire population can be challenging.

Complexity to organize

A second disadvantage is that it is more complex to organize and analyze the results compared to simple sampling.

Researchers must identify each member of a study population and classify it into only one subpopulation. As a result, stratified sampling is disadvantageous when researchers cannot confidently classify each member of the population into a subgroup.

Juxtaposition can be a problem if there are subjects that fall into multiple subgroups. When simple sampling is performed, those in multiple subgroups are more likely to be chosen. The result could be a misrepresentation or an inaccurate reflection of the population.

Examples like college students, graduates, men and women, make it easy, as they are clearly defined groups.

However, in other situations it could be much more difficult. You can imagine incorporating characteristics such as race, ethnicity, or religion. The classification process would become more difficult, making stratified sampling an ineffective method.

Example

Suppose a research team wants to determine the grade point average of college students in the United States.

The research team has obvious difficulties in collecting this data from the 21 million college students. Therefore, you decide to take a sample from the population, using only 4,000 students.

The team looks at the different attributes of the sample participants and wonders if there is a difference between the grade point average and the students' specialization.

It is found in the sample that 560 students are English students, 1,135 of science, 800 of computer science, 1,090 of engineering and 415 of mathematics.

The team wants to use proportional stratified sampling, where the sample strata are proportional to the population sample.

Creation of strata

To do this, the team investigates the statistics of university students in the US and finds the official percentage of students who specialize: 12% in English, 28% in science, 24% in computer science, 21% in engineering and 15% in mathematics.

Therefore, five strata are created from the stratified sampling process. The team must confirm that the population stratum is proportional to the sample stratum. However, he finds that the proportions are not equal.

Consequently, the team needs to resample the population of 4,000 students, but this time randomly selecting 480 (12%) English learners, 1,120 (28%) science, 960 (24%) computer science, 840 (21%) in engineering and 600 (15%) in mathematics.

With this, we have a proportional stratified sample of university students, which provides a better representation of university students in the US.

Researchers will be able to highlight a specific stratum, observe the various studies of US college students, and observe the different grade point averages.

References

1. Adam Hayes (2019). Stratified Random Sampling. Taken from: investopedia.com.
2. Wikipedia, the free encyclopedia (2019). Stratified sampling. Taken from: en.wikipedia.org.
3. Explorable (2019). Stratified Sampling Method. Taken from: explorable.com.
4. Survey Gizmo (2019). What is Stratified Sampling & When is it Used? Taken from: surveygizmo.com.
5. Ashley Crossman (2019). Understanding Stratified Samples and How to Make Them. Thought Co. Taken from: thoughtco.com.
6. Carlos Ochoa (2017). Random sampling: stratified sampling. Taken from: netquest.com.