top of page
Search

Sampling methods in Statistics

As part of various research projects, researchers want to study a huge group of people/members having some common characteristic/behavior. It is impossible to study every person in the group , So researcher picks a subset of overall population that represents the population as close as possible. Researcher can generalize the research results from sample to whole population. These methods of selecting a subset is calling Sampling and we will learn few such methods in this blog.

Lets learn some basic concept about first

Population: It is the entire group that you want to draw conclusions about.

Sample: It is the specific group of individuals that you will collect data from. It is subset of whole population

Methods are divided into two broad categories depending on whether the sample selection is based on randomization or not.

1) Probability Sampling : In this method, every member of the population has a fair chance of being selected. It is very useful in quantitative research. It could be time consuming

2) Non-Probability Sampling: This technique does not work on randomization that mean not every person has fair chance to be picked. Its on researcher ability/understanding to choose sample. It is quite cheap method, but has great chances to introduce bias in the research. It is easy and less time consuming

Below are the sampling methods which will be covered in this blog:

Probability Sampling methods:

1) Simple Random sampling : Every member of population has equal chance of getting selected. You can use tools like random number generator etc.

E.g. If research is doing a study (Sample size = 100) on all high school students (Total population is 1000). He can assign a number to every student and generate 100 random numbers to make his sample for analysis.

2) Systematic sampling : It is quite similar to simple random sampling, but it is usually slightly easier to conduct. Every member of the population is listed with a number, but instead of randomly generating numbers, individuals are chosen at regular intervals.

E.g If we need to conduct research on 1000 students, each would be numbered from 1-1000, Instead of picking randomly. We will pick every 10th member. Suppose we start from 7, so below would be our sample (7,17,27,37,47,57â€¦â€¦..)

3) Stratified Sampling: It involves dividing population in different sub-groups (called strata) based on relevant characteristics (e.g Gender, Age, Income range etc) and then pick up members from each group by Random or systematic sampling. Based on overall proportion for those sub-groups in the population, you decide how many samples to be taken from each subgroup

E,g Company X has 1500 employees (1000 Female, 500 Male), you can pick 100 Females and 50 Males from those two groups to accurately represent your population genderwise

4) Cluster sampling: In this method, researchers divide the entire population into sections or clusters that represent a population. Each cluster shall have similar characteristics as per overall population. Instead of sampling from individual clusters, you choose entire clusters for analysis.

E.g. A company has offices in 10 states with similar distribution of roles of employees etc. Instead of studying offices in all states, researcher can pick up offices at 2-3 States (clusters) for analysis

Non-Probability Sampling methods:

1) Convenience Sampling : This method includes selecting the individuals who happen to be most accessible to the researcher.

E.g. Professor chooses his class members to participate in survey for elective subjects etc (rather than all classes in the university)

2) Purposive Sampling: In this method, the researcher using their expertise to choose a sample that is most useful to the purposes of the research.

E.g. Researcher wants to study some heart related conditions, so he will choose only heart patients rather than selecting all random patients at clinic.

3) Snowball sampling: In this method, Existing people in group are asked to nominate other known people , so that sample size increase like a rolling snowball

e.g. You are researching experiences of homelessness in your city. Since there is no list of all homeless people in the city, probability sampling isnâ€™t possible. You meet one person who agrees to participate in the research, and she puts you in contact with other homeless people that she knows in the area.

4) Quota Sampling: In this type of sampling, researcher chooses members based on predetermined characteristics of the population

E.g selecting females over age of 60 to make a sample.

Hope you got some idea about different sampling techniques , thanks for reading.

Reference used:

https://www.simplypsychology.org/sampling.html