Sampling Methods

Introduction to Sampling
• Sampling is the process of selecting a subset from a larger population to
analyze and draw conclusions about the whole.
• It is essential in research because studying an entire population is often
impractical due to cost, time, and effort constraints.
• The key goal of sampling is to obtain a representative group that accurately
reflects the characteristics of the population
Types of Sampling Methods
Sampling methods are broadly classified into two categories:
[Link] Sampling – Each unit has a known and non-zero chance of
selection, ensuring a representative sample.
[Link]-Probability Sampling – Selection is based on convenience, judgment,
or other non-random criteria, which may introduce bias.
Probability Sampling Overview
Probability sampling ensures that each element in the population has a
measurable chance of being selected. It is ideal for statistical research,
hypothesis testing, and when generalizability is crucial.

Methods of Probability Sampling:

• Simple Random Sampling
• Stratified Sampling
• Systematic Sampling
• Cluster Sampling
• Multistage Sampling
Simple Random Sampling
Definition: Every individual in the population has an equal chance of being
chosen.
When to Use: When the population is homogeneous, and no subgroup
classification is necessary.
• Example: If a company wants to survey employee satisfaction, it can
randomly select 100 employees using a lottery method or a computer-
generated random list.
Stratified Sampling
Definition: The population is divided into homogeneous subgroups (strata) based
on characteristics like age, gender, or income level, and a proportional sample is
taken from each.
When to Use: When subgroups within a population need to be represented
proportionally.
• Example: In a university with 60% undergraduate and 40% graduate students, a
researcher ensuring fair representation might randomly select 60 undergraduates
and 40 graduate students.
Systematic Sampling
Definition: Every nth element from an ordered list is selected after choosing a
random starting point.
When to Use: When the population is organized in a list and a systematic
approach is preferred.
• Example: A company conducting quality control in a factory can inspect every
10th product coming off the production line.
Cluster Sampling
Definition: The population is divided into naturally occurring groups (clusters),
and entire clusters are randomly selected for study.
When to Use: When population elements naturally form groups, and it is difficult
to sample individuals separately.
• Example: A nationwide educational study might randomly select certain schools
instead of selecting individual students from across the country.
Multistage Sampling
Definition: A combination of different probability sampling techniques used in
multiple stages.
When to Use: When sampling must be conducted in large populations or across
different hierarchical levels.
• Example: In a national survey, researchers first select cities (cluster sampling),
then neighborhoods within cities (random sampling), and finally individuals
within neighborhoods (systematic sampling).
Non-Probability Sampling Overview
In non-probability sampling, elements are selected using non-random
criteria, often leading to bias. These methods are useful in exploratory
research, when quick data collection is required, or when targeting specific
individuals.

Methods of Non-Probability Sampling:

• Convenience Sampling
• Purposive (Judgmental) Sampling
• Snowball Sampling
• Quota Sampling
Convenience Sampling
Definition: Participants are selected based on their accessibility and
availability.
When to Use: When time and resources are limited, and representativeness is not
the primary concern.
• Example: A researcher conducting a quick survey might only interview people
who walk past a university campus.
Purposive Sampling (Judgmental Sampling)
Definition: The researcher selects participants based on their knowledge,
expertise, or specific characteristics.
When to Use: When studying specialized groups or expert opinions are required.
• Example: A study on AI in healthcare may involve interviewing only medical
professionals experienced in AI applications.
Snowball Sampling
Definition: Existing participants help recruit additional participants, forming a
growing chain of respondents.
When to Use: When researching hidden, rare, or hard-to-reach populations.
• Example: A study on drug addiction might begin with a few known addicts, who
then refer other addicts to participate.
Quota Sampling
Definition: Researchers ensure that certain characteristics of the population
are represented in the sample, but selection is not entirely random.
When to Use: When ensuring representation of specific groups while using a non-
random approach.
• Example: A market research study might require the sample to include 50% men
and 50% women but allow recruiters to select participants at their discretion.
Choosing the Right Sampling Method
• Use probability sampling when representativeness and generalizability are
critical (e.g., political polling, large-scale research).
• Use non-probability sampling when conducting exploratory studies, when
targeting specific groups, or when resources are limited.
• Always consider bias, cost, and feasibility when selecting a sampling
method.
Case 1: Employee Satisfaction Survey
• A multinational company with 50,000 employees across 10 countries
wants to measure job satisfaction.
• The company aims to get a representative sample but must also consider
cultural differences between different countries. They also have
constraints on budget and time, so conducting individual interviews with a
random sample of all employees might not be feasible.
• What sampling method should they use?
Case 1: Employee Satisfaction Survey
• Recommended Sampling Method: Stratified Sampling
• Explanation: Since employees are located in different countries with
varying work cultures, a simple random sample may not accurately
represent all regions. Stratified sampling is suitable as each country can
be considered a stratum, and employees can be randomly selected
within each stratum in proportion to the total workforce. This ensures a fair
representation of employees across different cultural and economic
backgrounds.
Case 2: Medical Study on a Rare Disease
• A research team wants to study the effects of a newly developed drug on
patients with a rare genetic disorder. Only about 1 in 100,000 people have
this disorder, and they are spread across different regions. Many patients are
reluctant to participate, and the researchers have limited access to them.
• What would be the most effective sampling method in this situation?
Case 2: Medical Study on a Rare Disease
• Recommended Sampling Method: Snowball Sampling
• Explanation: As the disease is rare, patients are difficult to reach. Snowball
sampling is the best choice because it starts with a few known patients, who
then refer other patients within their network. This method is particularly
useful in research involving hidden, small, or hard-to-reach populations,
such as people with rare genetic conditions or stigmatized illnesses.
Case 3: Customer Feedback in a Retail Chain
• A large retail company with hundreds of stores across different cities wants
to collect customer feedback about their shopping experience. They need
a cost-effective way to ensure the results reflect the opinions of their
customers in various locations, without having to survey every single
customer.
• How should they conduct their sampling?
Case 3: Customer Feedback in a Retail Chain
• Recommended Sampling Method: Cluster Sampling
• Explanation: The retail chain operates in multiple locations across different
cities, making it impractical to survey individual customers from all stores.
Cluster sampling is an effective solution:
• The entire set of stores can be divided into clusters (e.g., by city or region).A
random selection of stores is chosen, and all customers within those
selected stores are surveyed.
• This method reduces logistical and financial burdens while maintaining
representative results.
Case 4: Political Polling for an Upcoming Election

• A polling organization wants to predict the outcome of an election by

surveying potential voters. However, voter demographics such as age,
gender, and political preference are not evenly distributed across different
regions. The goal is to ensure the poll results are representative of the entire
voting population.
• Which sampling method should be used?
Case 4: Political Polling for an Upcoming Election

• Recommended Sampling Method: Quota Sampling or Stratified Sampling

• Explanation: Voters differ in age, gender, political views, and regional
distributions, so ensuring demographic representation is crucial. Two possible
approaches:
• Stratified Sampling: Voters are categorized into strata based on key
demographic factors (e.g., age groups, regions, or education levels), and
random samples are drawn from each group.
• Quota Sampling: Instead of random selection, researchers set quotas for
each demographic category (e.g., 50% female, 30% urban, 20% rural) and
then select participants non-randomly within those quotas. Quota sampling
is faster and more practical but may introduce some bias.
Case 5: The Future of Space Tourism

• A research team is conducting a study on the future of space tourism and

wants to gather insights on its feasibility, risks, and potential market growth.
While many people have opinions about space travel, only a select group of
experts—such as astronauts, aerospace engineers, space industry
executives, and policymakers—have the knowledge to provide credible and
informed insights.
• Since interviewing random travelers or space enthusiasts wouldn’t yield
reliable or technically sound information, what is the most suitable
sampling method for this study?
Case 5: The Future of Space Tourism

• Recommended Sampling Method: Purposive Sampling

• Explanation: Since space tourism requires expert insights, purposive
sampling is the best approach. Instead of surveying the general public, the
study will focus on astronauts, aerospace engineers, space industry executives,
and policymakers who have direct experience in the field. This ensures reliable
and informed responses, though it may introduce some selection bias.