Sampling Techniques in Simple Words

math notebook

Before getting acquainted with sampling techniques, it is necessary to name the basic terms. So, let’s define them.

3 Essential Definitions:

  1. “Population” (a total amount of specific elements that possess the pre-defined set of features);
  2. “Sample” (a subgroup of the elements inside the population)
  3. “Sampling” (the process of assorting samples)

The three linked terms must be differentiated further:

Population= global array of data

Sample=  specific subset

Let us bring an example. Imagine that there is a field. Many flowers are growing there.

The percentage of dandelions is 40%, 40% = sample, while “all the flowers in the field” =population.

You can do sampling in different ways, like take 10% of each kind, or choose only tulips.


This process is created to subdivide the whole group (population) into several valuable parts. These parts are usually differentiated either by the targeted principle or arbitrarily. The technique of sampling defines the method of segmentation.

If you wish to know more about advanced AI and software development technologies, check this blog, I got many insights from their posts.

Two approaches to selection:

The Probability selection (Sampling) (‘Simple (Random) selecting’; ‘Stratified selecting’; ‘Systematic selecting’; ‘Cluster selecting’);

  • Simple (Random) selecting

The probability principle of this method is similar to the bottle game when you need to spin the bottle to choose the person to kiss. The principal of ‘playing a lot’. When there is no specified information that can point to the target population, we use Simple (Random) selection. Each element has the same chance to be selected.


-Lack of bias;

-It is easy to exploit and can work as the representation of the general population for further analysis.

  • Stratified selecting.

Indifference from the previous kind, stratified selection presupposes subdivision based on the pre-defined target info. The subdivision is based on definite characteristics (age, nationality, education, etc.) that can categorize elements according to their similarity.

These subgroups are called strata, and each of these subgroups is further sampled by a random selection of subjects.

As before, we can bring the same example here. Although, there will be one modification. This time we are at a party. All the players here are subdivided into small groups (strata) with common interests. People still play together, but only in small groups with their friends. Then some group members are chosen randomly by the bottle over time.


-Greater precision

-Money-saving (cost-effective)

-Proper representation

  • Systematic selecting;

Systematic selection uses arbitrary selection at the first stage and selects the following elements at fixed intervals. The process functions as a chain with the elements of the population that have an equal chance of selection.


-Lack of bias (there’s the order of selecting elements);

-Simple in use

  • Cluster selecting;

There are three kinds of Cluster selection:

-Single Stage Cluster selecting (where a cluster is selected arbitrarily= no specific features are taken into consideration before sampling);

-Two-Stage Cluster selecting (1-arbitrary clusters- 2arbitrary selection of the elements);

-Systematic Clustering (The first element within a segmentation process is arbitrary; The following elements are sampled according to the principal o regular gaps between the elements in the subgroups).

-Multi-staged selecting (uses a complex way of selecting: 1 or more methods are combined);


-Not variable

-An expanded area of sampling

-It is easy to exploit

-Very reliable

2) Non- Probability selecting (‘Convenience Sampling’; ‘Purposive Sampling’; ‘Quota Sampling’; ‘Referral/ snowball Sampling’);

  • Convenience selecting

It means that convenient/ available sources are taken into consideration. This is the method of easy-fallen elements (segments). Specified segmentation is achieved by the accessibility of the groups within the population. The choice is made on the accessibility of the source.

In everyday life, it is something like the example when we buy products from definite departments in a store and choose the products from shelves at the level of our eyes, which is very convenient, but you can miss bread or salt, as they are placed differently. The negative side is that such selection results do not correctly represent the general population.


-Data is collected quickly;


-Easy to exploit;

  • Purposive selecting

This approach is oriented toward the goal and intention of research. The Purposive Sampling corresponds to the qualitative analysis. It is called the ‘judgment process’ because the investigator, who participates in the study, excludes the samples from the general population. The choice I made on the purpose of the study.



-Expedite the purpose of the study

-Simple in carrying

  • Quota selecting

=The approach that is conducted using specific acquired characteristics. The researchers usually create the attributes on their own.

=There is no randomization in this case.

=These self-created characteristics are made to enlist the elements/ samples by this or that principle and categorize them.

=Special features demanded


-Cost-effective (can be used in cases when the budget is limited)

-This selection process works faster and saves time

-It is easy to carry out

  • Referral/ Snowball selecting.

This method fits the population that is not specified.

As in the recruitment process, a segment (like a person) of research of referral sampling could address the elements around (people from close surroundings if it comes to the recruitment process) and ask them. It helps to reach more specified members. Thus, the unknown population (with no detailed info) is segmented into more accurate samples.

(From abstract – to concrete= from a small snowball into a giant Snowman)

The referral sampling method is similar to building a Snowman. If we want to make a giant snowman, we need to start with a small snowball and roll it (adding the surrounding snow, using the space around). When the snowball is big enough, we can shape the snowman.


-Quick in finding samples

-Simple exploit

-Precise data

Wish to go deeper into modern science trends, visit the Academy page and watch video lectures.

The difference between the Probabilistic and Non-Probabilistic Sampling

Firstly, we should note that in the probabilistic sampling methods, there is an equal opportunity for each unit of the population to be selected. Furthermore, probabilistic sampling aims at statistical analysis. Thus, it is perfect for quantitative research.

Indifference from the probabilistic approach, the non-probabilistic approach possesses different goals and advantages. This approach has nothing in common with an arbitrary principle. Moreover, not all the members included in the population are usually selected.

It depends on the topic of interest/purpose of study (research). Self-selection is also a characteristic feature of non-probable sampling. This technique doesn’t fit quantitative research, but it’s ideal for qualitative research.