Learning Objectives
By the end of this lesson, learners should be able to:
- Explain probability sampling.
- Apply simple random, systematic, stratified, and cluster sampling.
- Evaluate the strengths and limitations of each method.
Meaning Of Probability Sampling
In probability sampling, every population member has a known, non-zero chance of selection. This allows statistical inference from the sample to the population.
Probability sampling is preferred when researchers want reliable, generalizable results.
Simple Random Sampling
Every member has an equal chance of selection.
Procedure
- Number all population members.
- Use random numbers or software to select participants.
Example
Selecting 500 customers from a database of 50,000 customers using a random number generator.
Advantages
- Easy to understand,
- Minimizes selection bias,
- Supports statistical analysis.
Limitations
- Requires a complete sampling frame,
- May be costly for geographically dispersed populations.
Systematic Sampling
Select every kth member after a random starting point.
Example
Population = 10,000 customers
Desired sample = 1,000
Sampling interval = 10,000 ÷ 1,000 = 10
Choose a random start between 1 and 10, then select every 10th customer.
Advantages
- Simple and efficient,
- Good for ordered lists.
Limitation
Can be biased if the list has a hidden pattern.
Stratified Sampling
The population is divided into homogeneous groups (strata), and samples are drawn from each stratum.
Example
A global bank divides customers by region:
|
Region |
Population Share |
Sample Share |
|
North America |
35% |
35% |
|
Europe |
30% |
30% |
|
Asia-Pacific |
25% |
25% |
|
Middle East & Africa |
10% |
10% |
Advantages
- Ensures representation of important groups,
- Often increases precision.
Limitation
Requires accurate population information.
Cluster Sampling
The population is divided into clusters, and entire clusters are selected.
Example
A hotel chain selects 20 hotels worldwide and surveys all guests staying in those hotels during one week.
Advantages
- Lower travel and administrative costs,
- Useful for geographically dispersed populations.
Limitation
Less precise if clusters differ substantially.
Choosing A Probability Sampling Method
|
Situation |
Recommended Method |
|
Complete customer list available |
Simple random |
|
Ordered database available |
Systematic |
|
Important subgroups must be represented |
Stratified |
|
Population widely dispersed geographically |
Cluster |
International Case Study
A multinational insurance company wants to estimate customer satisfaction across 15 countries. Because customer expectations differ by country, the company uses stratified sampling by country to ensure each market is represented fairly.
Common Mistakes
- Using systematic sampling on periodic data,
- Ignoring small but important strata,
- Treating cluster samples as simple random samples,
- Failing to randomize within strata.
Practical Exercise
Design a probability sample for a survey of employees in an organization with offices in Toronto, London, Dubai, Singapore, and Johannesburg.
Learning Materials / Reference Materials
- Cochran, W. G. Sampling Techniques.
- Lohr, S. Sampling: Design and Analysis.
- ESOMAR Sampling Guidelines.
Lesson Summary
Probability sampling methods provide a scientific basis for selecting respondents and allow researchers to make valid conclusions about the population.