Sampling design refers to the method or procedure used to select a subset of elements (sample) from a larger group (population) for research. Sampling designs are broadly classified into two categories:

? 1. Probability Sampling (Random Sampling)

In probability sampling, every element in the population has a known, non-zero chance of being selected. This method is scientific, unbiased, and allows for generalization of results to the entire population.

a. Simple Random Sampling

Every unit has an equal chance of being selected. Selection is often done using a random number table or computer software.

Example: Picking 100 names at random from a list of 10,000 registered students.
✅ Best when population is homogeneous and complete list is available.

b. Systematic Sampling

Select every kᵗʰ unit from a list after a random starting point. k = N/n where N = population size, n = sample size.

Example: From a list of 1,000 people, choosing every 10th person after starting from 3rd.
✅ Simple and quicker than simple random sampling.

c. Stratified Sampling

Divide the population into homogeneous subgroups (strata) and perform random sampling within each stratum.

Example: Selecting 50 students from each class (stratum) in a school.
✅ Improves accuracy when population is diverse.

d. Cluster Sampling

Divide the population into groups or clusters (usually by geography or institution) and randomly select a few clusters. Study all units within them.

Example: Selecting 5 schools randomly and surveying all students in those schools.
✅ Useful when the population is spread over a large area.

e. Multistage Sampling

A combination of different sampling techniques applied in stages. Often used in large-scale surveys.

Example: First select states → then districts → then schools → then students.
✅ Reduces cost and complexity in nationwide studies.

? 2. Non-Probability Sampling

In non-probability sampling, not every element has a known or equal chance of being included. It is often used for exploratory research, where generalization is not the main goal.

a. Convenience Sampling

Sample is selected based on ease of access or availability.

Example: Interviewing people at a mall.
❌ High risk of bias but useful for quick, preliminary studies.

b. Judgmental or Purposive Sampling

Researcher uses their judgment to select units believed to be most representative.

Example: Choosing experienced employees for feedback on a new system.
❌ Subjective and may not reflect full diversity of population.

c. Quota Sampling

Divide the population into subgroups and select a fixed number from each group (like stratified), but selection within each group is non-random.

Example: Surveying 30 men and 30 women regardless of how they are chosen.
❌ Fast but may suffer from selection bias.

d. Snowball Sampling

Existing participants refer other participants, creating a chain.

Example: Used in studying rare groups like drug users or victims of abuse.
✅ Effective for hard-to-reach or hidden populations.

✅ Comparison Table: Probability vs Non-Probability Sampling

? Conclusion

Choosing the right sample design depends on research objectives, population characteristics, available resources, and the need for generalization. While probability sampling ensures scientific accuracy, non-probability sampling is more practical for qualitative or exploratory research.