Lecture 4 - Sampling Methods
Sampling is the process of selecting a subset of individuals, units, or
observations from a larger population to represent that population in a
research study. Since it is often impractical or impossible to collect data from
every member of a population, researchers use sampling techniques to
obtain representative data that can be used to make valid conclusions about
the entire population. The choice of sampling method depends on the
research objectives, study design, available resources, accessibility of the
target population, and the level of accuracy required.
Sampling methods are broadly classified into probability sampling and
non-probability sampling.
4.1 Probability Sampling Methods
Probability sampling is a sampling technique in which every member of the
target population has a known and non-zero chance of being selected.
Because selection is based on randomization, probability sampling minimizes
selection bias and allows researchers to generalize findings from the sample
to the entire population. These methods are commonly used in quantitative
research.
4.1.1 Simple Random Sampling
Simple random sampling is the most basic form of probability sampling.
Every individual in the population has an equal chance of being selected, and
the selection of one individual does not influence the selection of another.
A sampling frame containing all members of the population is required.
Researchers then use random methods such as lottery techniques, random
number tables, or computer-generated random numbers to select
participants.
Advantages
Easy to understand and implement.
Minimizes selection bias.
Produces representative samples when the population is
homogeneous.
Suitable for statistical analysis.
Disadvantages
Requires a complete and accurate sampling frame.
May be costly and time-consuming for large populations.
May not adequately represent important subgroups.
Example
A university has 2,000 registered students. A researcher wishes to select 200
students for a survey. Each student is assigned a unique number, and
computer-generated random numbers are used to select the participants.
4.1.2 Systematic Sampling
Systematic sampling involves selecting participants at regular intervals from
an ordered list after randomly choosing a starting point.
The sampling interval (k) is calculated as:
k=Nnk=\frac{N}{n}k=nN
Where:
N = Population size
n = Desired sample size
After randomly selecting the first participant, every kth individual is included
in the sample.
Advantages
Simple and quick to implement.
Ensures even coverage of the population.
Less time-consuming than simple random sampling.
Disadvantages
Can introduce bias if the list follows a repeating pattern.
Requires an ordered sampling frame.
Example
A clinic has 1,000 patient records and requires a sample of 100 participants.
k=1000100=10k=\frac{1000}{100}=10k=1001000=10
If the first participant selected is number 7, the sample will include
participants numbered 7, 17, 27, 37, and so forth.
4.1.3 Stratified Random Sampling
Stratified random sampling divides the population into distinct subgroups
(strata) that share similar characteristics, such as age, gender, profession, or
geographic location. A random sample is then selected from each stratum.
Sampling may be:
Proportionate stratified sampling, where sample sizes reflect the
size of each stratum.
Disproportionate stratified sampling, where unequal numbers are
selected from each stratum to ensure adequate representation.
Advantages
Improves representativeness.
Ensures inclusion of important subgroups.
Increases statistical precision.
Reduces sampling error.
Disadvantages
Requires detailed information about the population.
More complex and time-consuming than simple random sampling.
Example
A hospital employs 600 healthcare workers consisting of:
300 nurses
180 clinicians
120 laboratory staff
If a sample of 120 workers is required, proportionate sampling would select:
Nurses: 60
Clinicians: 36
Laboratory staff: 24
Random sampling is then conducted within each professional group.
4.1.4 Cluster Sampling
Cluster sampling involves dividing the population into naturally occurring
groups or clusters, such as schools, villages, hospitals, or districts. Instead of
selecting individuals directly, researchers randomly select clusters and then
study all members or a sample of members within those clusters.
Cluster sampling is particularly useful when populations are geographically
dispersed.
Advantages
Reduces travel costs and data collection time.
Suitable for large populations.
Practical when no complete sampling frame exists.
Disadvantages
Less statistically efficient than simple random sampling.
Higher sampling error due to similarities within clusters.
Example
A national health survey randomly selects 20 districts from all districts in the
country. Households within the selected districts are then surveyed.
4.1.5 Multistage Sampling
Multistage sampling combines two or more sampling techniques in
successive stages. Researchers first select larger units and progressively
sample smaller units.
Advantages
Efficient for large and geographically dispersed populations.
Reduces operational costs.
Flexible for complex study designs.
Disadvantages
More complicated to design and analyze.
Sampling errors may accumulate across stages.
Example
A researcher studying childhood nutrition may:
1. Randomly select counties.
2. Select sub-counties within those counties.
3. Select villages within sub-counties.
4. Select households within villages.
5. Select one eligible child per household.
4.2 Non-Probability Sampling Methods
Non-probability sampling involves selecting participants without
randomization. The probability of selection is unknown, making it difficult to
generalize findings to the broader population. These methods are commonly
used in qualitative, exploratory, and pilot studies.
4.2.1 Convenience Sampling
Convenience sampling involves selecting participants who are readily
available and willing to participate.
Advantages
Fast and inexpensive.
Easy to conduct.
Suitable for pilot studies.
Disadvantages
High risk of selection bias.
Poor representativeness.
Limited generalizability.
Example
A researcher interviews patients attending a clinic during one week because
they are easily accessible.
4.2.2 Purposive (Judgmental) Sampling
Purposive sampling involves deliberately selecting participants who possess
specific knowledge, experience, or characteristics relevant to the research
question.
It is commonly used in qualitative research.
Advantages
Provides rich, detailed information.
Focuses on knowledgeable participants.
Useful for specialized populations.
Disadvantages
Subjective selection.
Potential researcher bias.
Limited generalizability.
Example
A study exploring leadership experiences selects hospital managers with at
least five years of management experience.
4.2.3 Quota Sampling
Quota sampling divides the population into categories and recruits
participants until predetermined quotas are achieved. Unlike stratified
sampling, participants are selected non-randomly.
Advantages
Ensures representation of important groups.
Faster than probability sampling.
Useful when sampling frames are unavailable.
Disadvantages
Susceptible to selection bias.
Does not allow estimation of sampling error.
Example
A researcher studying consumer preferences recruits:
100 males
100 females
Participants are enrolled until each quota is reached.
4.2.4 Snowball Sampling
Snowball sampling is used when studying hidden or difficult-to-reach
populations. Initial participants identify additional eligible participants,
causing the sample to grow like a snowball.
Advantages
Effective for hard-to-reach populations.
Builds trust among participants.
Useful for sensitive topics.
Disadvantages
High risk of sampling bias.
Participants may recruit individuals with similar characteristics.
Limited representativeness.
Example
Researchers studying injection drug users recruit initial participants, who
then refer other eligible participants.
4.2.5 Consecutive Sampling
Consecutive sampling involves recruiting every eligible participant who
meets the inclusion criteria over a specified period until the desired sample
size is achieved.
It is frequently used in clinical and hospital-based studies.
Advantages
More representative than convenience sampling.
Simple to implement.
Reduces investigator selection bias.
Disadvantages
Time-consuming.
May still not represent the broader population.
Example
A researcher enrolls every eligible patient diagnosed with diabetes attending
a hospital clinic between January and June until 250 participants have been
recruited.
Comparison of Probability and Non-Probability Sampling
Probability
Characteristic Non-Probability Sampling
Sampling
Selection method Random Non-random
Chance of selection Known Unknown
Sampling bias Low Higher
Generalizability High Limited
Statistical inference Possible Limited
Cost Usually higher Usually lower
Quantitative Qualitative and exploratory
Common use
research research