Probability Sampling vs. Non-Probability Sampling

In statistics, sampling is when researchers determine a representative segment of a larger population that is then used to conduct a study. 

Sampling comes in two forms -- probability sampling and non-probability sampling.

Probability sampling uses random sampling techniques to create a sample.

Non-probability sampling methods use non-random processes such as researcher judgement or convenience sampling.

Probability Sampling

For a sampling method to be considered probability sampling, it must utilize some form of random selection. 

In other words, researchers must set up some process or procedure that ensures, with confidence, that the different units in their sample population have equal probabilities of being chosen. 

For example, if a researcher is dealing with a population of 100 people, each person in the population would have the odds of 1 out of 100 for being chosen. This differs from non-probability sampling, in which each member of the population would not have the same odds of being selected.

If you’re not an experienced researcher, chances are that you’re more familiar with probability sampling than you think. 

Remember the times in grade school that your class pulled a name out of a hat in order to determine who would give their presentation first? That’s probability sampling at work. 

However, researchers today tend to use computers as the primary mechanism for generating random numbers used in random selection.

Probability sampling allows researchers to create a sample that is accurately representative of the real-life population of interest.

The Methods of Probability Sampling

There are several types of probability sampling. Below we explain the basics of each, and address their advantages and disadvantages.

Simple Random Sampling

Simple random sampling is considered the easiest method of probability sampling. 

To perform simple random sampling, all a researcher must do is ensure that all members of the population are included in a master list, and that subjects are then selected randomly from this master list.

While simple random sampling creates samples that are highly representative of the population, it can be time consuming and tedious when creating large samples. 

Stratified Random Sampling

Stratified random sampling is also referred to as proportional random sampling. 

In stratified random sampling, the subjects are initially grouped into different classifications such as gender, level of education, or socioeconomic status. It’s important to note that these classifications should not have any overlapping subjects. 

From here, researchers randomly select the final list of subjects from the different defined categories to ensure a well rounded sample. 

This method of probability sampling is best used when the goal of the research is to study a particular subgroup within a greater population. It also results in more precise statistical outcomes than simple random sampling.

Stratified random sampling creates layers within a sample that are extremely accurate when it comes to representing the layers with the population, but it too can be time consuming and tedious while creating larger samples.

Systematic Random Sampling

Systematic random sampling is often compared to an arithmetic progression in which the difference between any two consecutive numbers is of the same value. 

For example, if you are a researcher examining a clinic that has 100 patients, the first step in systematic random sampling is to pick an integer that is less than the total number of the population. This will be the first subject. 

For the sake of this example let’s pick subject number 4.

The next step is to choose another integer, which will be the number of individuals between subjects. 

Let’s say we choose 6 in this example.

By carrying out the processes above, the subjects for our study would be patients 4, 10, 16, 22, 28, etc. 

Systematic random sampling allows researchers to create samples without using a random number generator, but the outcomes are not quite as random as they would be if a software program was used instead. 

Cluster (Area) Random Sampling

Cluster random sampling is conducted when the size of a population is too large to perform simple random sampling. 

Think of instances such as investigating the dietary trends amongst the entire population of Africa -- the population is just too large to manage effectively.

In cluster random sampling, the initial research identifies boundaries. Sticking with the example above, our boundaries would be the various countries in Africa. 

From here, the researcher randomly selects a number of identified boundaries. It’s important to note that each of the areas, in our case African countries, should have equal chances of being selected.

Finally, the researcher conducting the study can then include all of the individuals within the selected areas, or he or she can use simple random selection to select subjects from the identified countries.

The primary advantages of cluster random sampling are its convenience and ease of use, while its most notable disadvantage is that if members of clusters are not homogenous in nature, researchers will end up with less accurate data. 

Multi-Stage Sampling

Multi-stage sampling involves a combination of two or more of the probability sampling methods outlined above.

With more advanced research, using just one form of probability sampling does not ensure the randomization necessary to ensure confidence in results. 

By combining various probability sampling techniques at various stages of research initiatives, researchers are able to maintain confidence that they are mitigating biases as much as possible.


Now that you have an understanding of the various types of probability sampling, you can determine the best strategy for creating samples during your next research initiative.

Are you familiar with the strategies addressed in this article? How have they influenced your research? Feel free to share your story in the comments below!