Stratified Random Sampling

Stratified random sampling is a method that divides people into minor groups, which are called strata. Groups are usually combined on the grounds of attributes or characteristics that members of the future group have in common. For example, it can be education or income. Stratified random sampling is applied in many areas, for instance, it may help with the studying life expectancy or the demographics of the population. The method’s second name is proportional random sampling or quota random sampling.

Main principles of Stratified Random Sampling

Research usually requires a large group of people. The requirement is driven by the need to make the research comprehensive, representative and accurate at the same time. However, the large number of people may also be disadvantageous, since it’s rather difficult to conduct such research. It takes more time and requires more money. For this reason, researchers are prone to lessen the amount of people studied to the smallest group. Despite the smaller quantity, the group must remain representative. Such small representative groups are described as sample size. In order to divide people in such groups, stratified random sampling is created along with other methods.

The method is based on dividing the whole population into similar groups. The groups are called strata. Then, the researchers choose random samples out of each stratum. For instance, a researcher has an intention to study the favorite color of the kids of 6 years old. There are numerous 6 years old kids. Thus, the researcher takes 10,000 kids in order to facilitate the study, then the division into groups takes place. The division is based on such characteristics as gender, race, country, etc. To make the study accurate, the researcher must be accurate himself or herself. In order to accomplish this goal, the proportions must be kept. Thus, the random sample is proportional to the stratum’s size. The next step of the study is to join subsets with each other to create a random sample and analyze it. The last step is a conclusion, which is based on the conducted analysis.

Difference between simple random sampling and Stratified Random Sampling

Among statistical measurement methods, two methods are distinguished. One is the topic of the current article, the other is called simple random samples. The last is utilized as a representation of the all data population. On the other hand, stratified random sampling uses stratification as a tool and divides the population into smaller representative groups. As it was stated before, the stratified random sampling requires much time and more money to fulfill its goals. For these reasons, this method is considered rather more complicated than the simple random samples.

However, there are several cases in which using the simple random sampling is a way more advantageous than the stratified random sampling. Firstly, the simple random sampling is suitable in cases with a small amount of available data. Secondly, there may be too many criteria to form groups on their base. Thirdly, the reverse situation, when the division characteristic is the only one.

In case the division isn’t the main focus of researchers, and they are satisfied if the study consists of randomly chosen samples, these researchers may use a simple random sample method. For instance, a company needs an assessment of the new product to estimate whether it will be generally successful or not. Thus, it conducts research based on the simple random sample method, it randomly takes 50 potential customers and asks for their opinion.

Advantages of Stratified Random Sampling

The stratified random sampling method has numerous advantages compared to other methods, some of beneficial aspects are presented below:

• It seizes the key characteristics of each sample.
• It gives proportional characteristics to the overall population.
• It’s well applied to a population with many different characteristics.
• It has almost no estimation errors and is distinguished by the higher accuracy.

Disadvantages of Stratified Random Sampling

Despite the numerous advantages the method is too far from to be called universal, it can’t be applied to every study. In order to apply the method properly, researchers must fulfill a list of conditions. From these conditions the disadvantages appear, some of them are presented below:

• It isn’t always possible to identify and classify every part of a studied material. Thus, the main condition, which is division into subgroups, can't be met.
• In case it’s possible to divide the material, another challenge appears. The division itself may take too much time.
• Overlapping may also become a problem, since there are often parts or members that may be assigned to different groups.