Probability Sampling
1. Definition
Probability sampling is a sampling method where each member of the
target population has a known, non-zero, and usually equal chance
of being selected into the sample. It is a cornerstone of quantitative
research, especially when generalization is the goal.
This technique relies on randomization, which helps eliminate selection
bias and ensures that the sample reflects the population as accurately as
possible.
Population: the entire group of interest (e.g., all college students in
the Philippines).
Sample: a subset drawn from the population (e.g., 1,000 randomly
chosen students).
Sampling Frame: the list from which the sample is selected (e.g.,
official student list from CHED).
2. Key Features of Probability Sampling
1. Random Selection – Ensures fairness, avoids human bias in
participant selection.
2. Known Probability – Each unit’s chance of being chosen can be
calculated.
3. Representativeness – Aims to mirror the population’s characteristics
(age, gender, income, etc.).
4. Generalizability – Findings from the sample can be extended to the
whole population.
5. Sampling Error Measurement – Allows computation of standard
error and confidence intervals.
3. Advantages
Minimizes selection bias.
Provides a statistically reliable foundation for hypothesis testing.
Ensures diverse subgroups are represented (through stratified
designs).
Enables valid comparisons across population segments.
Essential in large-scale surveys (e.g., census, public health
studies).
4. Disadvantages
Costly & Time-consuming: Requires more resources compared to
convenience sampling.
Requires a sampling frame: Difficult when population lists are
unavailable.
Complex design: Some methods (multistage, stratified) require
advanced statistical expertise.
Non-response issues: Even with randomization, people may refuse
to participate, affecting representativeness.
5. Types of Probability Sampling (with Examples)
a. Simple Random Sampling (SRS)
How it works: Every individual has an equal chance.
Methods: Lottery draw, random number generator, random tables.
Example: Picking 200 employees randomly from a company list of
2,000.
Strength: Easy to understand, unbiased.
Limitation: Impractical for large populations without a complete list.
b. Systematic Sampling
How it works: Selects samples at regular intervals (k).
Formula: k = N/n where N = population, n = desired sample size.
Example: From 10,000 households, select every 25th household to get
400 samples.
Strength: Easier than SRS when lists are large.
Limitation: May introduce bias if population has hidden patterns (e.g.,
every 25th person happens to be from the same block).
c. Stratified Random Sampling
How it works: Divide the population into strata (subgroups) based on
a variable (age, gender, income, education), then randomly select from
each.
Example: A school has 60% female and 40% male students. To sample
100 students, take 60 females and 40 males randomly.
Strength: Ensures important subgroups are represented.
Limitation: Requires prior knowledge of subgroup proportions.
d. Cluster Sampling
How it works: Divide the population into clusters (usually geographic
or organizational groups), then randomly select whole clusters to study.
Example: To survey farmers in a region, randomly select 10
municipalities (clusters) and survey all farmers in them.
Strength: Cost-effective for large and dispersed populations.
Limitation: Higher sampling error compared to SRS, because clusters
may be internally similar.
e. Multistage Sampling
How it works: Combination of methods applied in stages.
Example: Select provinces → select municipalities → select barangays
→ select households → select individuals.
Strength: Efficient for national surveys (e.g., NSO Census,
Demographic and Health Survey).
Limitation: Complex, requires multiple randomization steps and
statistical adjustments.
6. Applications of Probability Sampling
Government & Policy: Census, labor surveys, election polls.
Health & Medicine: Epidemiological studies (e.g., COVID-19
prevalence).
Market Research: Consumer preference studies.
Education: Assessment of student achievement across regions.
Social Science Research: Studying behavior, attitudes, or social
trends.
7. Comparison with Non-Probability Sampling
Non-Probability
Aspect Probability Sampling
Sampling
Selection Random, every unit has Non-random, chance
Method known chance unknown
Representative
High Low, prone to bias
ness
Generalizability Results can be generalized Limited generalization
More expensive, time-
Cost & Time Less expensive, quicker
consuming
Statistical Limited to descriptive
Allows inferential statistics
Validity statistics
Convenience or purposive
Example Stratified random sampling
sampling
8. Conclusion
Probability sampling is the gold standard in quantitative research when the
goal is accuracy, fairness, and generalizability. Despite its complexity and
cost, it provides a scientifically valid way to represent a population and
minimize bias. Researchers, policymakers, and organizations often rely on it
when making decisions that affect large groups of people.
1. Simple Random Sampling
A university has 5,000 students. The researcher wants to select 200 students
by randomly assigning numbers and using a computer to pick them. Which
type of probability sampling is this?
A. Stratified Sampling
B. Cluster Sampling
C. Simple Random Sampling
D. Systematic Sampling
✅ Answer: C. Simple Random Sampling
📘 Bloom’s Level: Knowledge/Remembering
2. Systematic Sampling
A company with 1,000 employees wants to survey 100 workers. The HR
department selects every 10th employee from the company roster. Which
method is being used?
A. Simple Random Sampling
B. Systematic Sampling
C. Multistage Sampling
D. Cluster Sampling
✅ Answer: B. Systematic Sampling
📘 Bloom’s Level: Understanding
3. Stratified Sampling
A researcher is studying high school students’ study habits. Since the
population has 60% females and 40% males, the researcher selects a sample
that preserves this ratio. Which method ensures representation of both
genders?
A. Simple Random Sampling
B. Cluster Sampling
C. Stratified Random Sampling
D. Multistage Sampling
✅ Answer: C. Stratified Random Sampling
📘 Bloom’s Level: Understanding
4. Cluster Sampling
A nationwide study on farmers is too costly to cover every region. Instead,
the researcher randomly selects 5 provinces and surveys all farmers in
those provinces. What method is applied?
A. Stratified Sampling
B. Cluster Sampling
C. Simple Random Sampling
D. Systematic Sampling
✅ Answer: B. Cluster Sampling
📘 Bloom’s Level: Applying
5. Multistage Sampling
The National Statistics Office conducts a household income survey by first
selecting provinces, then municipalities, then barangays, then households,
and finally individuals. This is an example of:
A. Cluster Sampling
B. Multistage Sampling
C. Stratified Sampling
D. Simple Random Sampling
✅ Answer: B. Multistage Sampling
📘 Bloom’s Level: Applying
6. Identifying Bias
A researcher wants to evaluate college satisfaction but selects only students
from the engineering department. Why is this not an example of probability
sampling?
A. It is too expensive
B. Not all students had equal chance of being selected
C. It requires stratification
D. It is cluster sampling instead
✅ Answer: B. Not all students had equal chance of being selected
📘 Bloom’s Level: Analyzing
7. Practical Decision
A school principal wants to evaluate teaching effectiveness. She divides
teachers into groups based on subject taught (Math, English, Science) and
then randomly selects from each group. Which method best ensures
representation across subjects?
A. Cluster Sampling
B. Stratified Sampling
C. Systematic Sampling
D. Multistage Sampling
✅ Answer: B. Stratified Sampling
📘 Bloom’s Level: Applying
8. Evaluating Appropriateness
A researcher plans to study consumer buying behavior in a city of 2 million
people. Which sampling method is most practical if he has limited time and
budget?
A. Simple Random Sampling
B. Stratified Sampling
C. Cluster Sampling
D. Multistage Sampling
✅ Answer: C. Cluster Sampling
📘 Bloom’s Level: Evaluating
9. Error in Sampling
If a researcher uses systematic sampling with k = 20, but the list of names is
arranged by income level (lowest to highest), what problem might occur?
A. The sampling will be unbiased
B. The sample may overrepresent certain income levels
C. It will automatically stratify the sample
D. It will become a cluster sample
✅ Answer: B. The sample may overrepresent certain income levels
📘 Bloom’s Level: Analyzing
10. Choosing the Best Method
A health researcher wants to estimate diabetes prevalence in the Philippines.
Which probability sampling method should be used to ensure
representativeness across the whole country?
A. Simple Random Sampling
B. Stratified Sampling
C. Multistage Sampling
D. Convenience Sampling
✅ Answer: C. Multistage Sampling
📘 Bloom’s Level: Evaluating/Creating