Page 1 of 7

SAMPLING METHODS

POPULATION VS. SAMPLE:

• Population can be very broad or quite narrow; maybe you want to make inferences about the
whole adult population of a country or focus on customers of the APPLE company.

• This is an important term to define your target population.

• Should be according to the purpose and practicalities of a researcher’s project.

• If population is very large and diverse, it leads to difficulty to given access to a representative
sample.

• A lack of representative sample affects validity of the research and can lead to research biases,
particularly sampling basis.

• The Sampling Frame is the actual list of individuals that the sample will be drawn from.

• Ideally it should include the entire target population and nobody who is not part of that
population.

• The number of individuals you should include in your sample depends on various factors.

1. Size (I.e. number)

Of the population
2. Variability (includes population variance)

(A type of variance that is used to determine the variability of the population data with respect to the
mean.)

Sampling is the selection of a subset or a statistical sample of individuals from within a statistical
population to estimate characteristics of the whole population. The population can be defined in
terms of geographical location, age, income, etc.

For example: In "A study on Asian children," the statistical population refers to all individuals who fit
the criteria of Asian children. From this population, a smaller, representative group, referred to as the
sample, is chosen.

The population is the entire group that you want to draw conclusions about. To achieve this, you will
take a sample.

The sample is the group of individuals who will actually participate in the research. To draw valid
conclusions from your results, you have to carefully decide how you will select a sample that is
representative of the group as a whole. This process is called a sampling method.

There are two primary types of sampling methods:

1. Probability Sampling

2. Non-Probability Sampling
 Page 2 of 7

Aspect Probability Sampling Non probability sampling

Selection Process Participants are selected randomly, Selection is non-random, based on
giving each individual an equal chance convenience, judgment, or specific
of selection. criteria.

Generalisability Findings can be generalised to the entire Findings cannot be reliably generalised
population as the sample is due to potential lack of representation.
representative.

Purpose Commonly used in large-scale Used in exploratory and qualitative

quantitative research, such as public research, targeting specific populations.
opinion polls.

Bias and Accuracy Randomisation minimises selection bias, Prone to selection bias as it depends on
ensuring higher accuracy and reliability. the researcher's discretion or
convenience.

Complexity and Cost Requires more time, resources, and Simpler, quicker, and cost-effective for
careful planning. small-scale or initial studies.

WHY PROBABILITY FOR QUANTITATIVE AND NON-PROBABILITY FOR QUALITITATIVE

RESEARCH IS PREFERRED?

Probability sampling is favoured in quantitative research because it allows for random selection,
reducing bias and enabling statistical analysis, hypothesis testing, and generalization of results to a
larger population. This method supports the reliability and validity of findings, especially when large,
diverse samples are needed.

On the other hand, non-probability sampling is common in qualitative research because it prioritizes
depth and understanding over generalization. Researchers select participants based on specific
characteristics or experiences, aiming for rich, detailed insights rather than statistical
representativeness. This method is flexible and often more practical for small, focused groups,
allowing for a deeper exploration of complex phenomena. Additionally, qualitative research often
deals with subjective experiences and meanings that cannot be captured through random sampling
methods.
 Page 3 of 7

Aspect Simple Random Convenience Systematic Cluster

Sampling Sampling Sampling Sampling
Selection Participants are Participants are Every nth individual Population is
chosen randomly, chosen based on is selected from a divided into
Process ensuring equal ease of access and population list. clusters, and entire
chances for all. availability. clusters are
selected randomly.
Representation Ensures the May not represent Can provide Provides
sample is highly the population moderate representation if
representative of accurately. representation clusters are
the population. depending on the diverse.
interval.
Bias Minimal selection High selection bias Potential bias if the Cluster selection
bias due to due to lack of list order has a may not fully
randomisation. randomisation. pattern. capture population
diversity.
Cost and effort More time- Quick, easy, and Less time- Can be cost-
consuming and inexpensive. consuming than effective if clusters
resource-intensive. random sampling are geographically
but still structured. close.
Example A researcher A psychologist A researcher A psychologist
studying anxiety surveys students in selects every 5th studying mental
randomly selects a cafeteria about name on a list of health selects 4
participants from a social media use. therapy patients for schools randomly
university's student a study. and surveys all
registry. students in those
schools.

PROBABILITY SAMPLING

It means that every member of the population has a chance of being selected. It is mainly used in
quantitative research and when results that are representative of the whole population are desired.

It enables statistical validity and generalizability, and minimizes researcher bias. Quantitative research
focuses on measurable data that paves the way for standardization.

• Simple Random Sampling is when every member of the population has an equal chance of
being selected. Here the sampling frame should include the whole population. An example in
an experiment where it was used: Stroop Effect.

In a simple random sample is a subset of individuals chosen from a larger set in which a subset
of individuals is chosen randomly, all with the same probability.

It is unbiased: Every member has an equal chance, minimizing selection bias.

It provides a basis for Statistical Inference: Allows for accurate estimation of population
parameters.
 Page 4 of 7

• Systematic Sampling – Is similar to simple random sampling but slightly easier to conduct.
Systematic sampling is a probability method of selecting a random sample from a population
with a fixed interval.

Process:

1. Every member of the population is listed with a number like 1, 2, 3...

2. Instead of randomly choosing numbers, individuals are chosen at regular intervals.

Example: For instance, if from the first 10 numbers, No. 5 is selected, every 10th person on the list is
picked (e.g., 5, 15, 25...).

It was used in MMPI development in Psychology.

Simple and Easy: Relatively straightforward to implement.

Efficient: Can be quicker than simple random sampling for large populations.

Can be More Representative: If the sampling frame is ordered randomly, systematic sampling can
provide a more evenly distributed sample compared to simple random sampling.

• Stratified Sampling

It involves dividing the population into sub-populations that differ from each other. For example:
population of Asian Americans are divided into subgroups called strata based on the relevant
characteristics like: Gender identity, Age range, Job roles

It is based on the overall proportions of the population → how many people should be sampled
for each subgroup is decided.

This is used to maintain the male/female ratio in a research as well. Another example where it
is also used are studies like: The Sleep & Circadian Rhythms Study

• Cluster Sampling
It is similar to stratified sampling where there are subgroups, BUT each subgroup here has
similar characteristics to the whole group. Instead of sampling individuals from each
subgroup, you randomly select entire subgroups.

If clusters are large → individuals from each cluster are also sampled. This is called Multistage
Sampling.

An example of where it is used: National Comorbidity Survey - The National Comorbidity Survey (NCS)
is a survey mandated by Congress to study psychiatric disorders in the United States.
 Page 5 of 7

Aspect Snowball Quota Convenience Judgemental Voluntary

Sampling sampling sampling Sampling sampling
Selection Participants Participants Participants are Researchers Participants self-
recruit others are selected to chosen based select select to
Method from their meet on ease of participants participate.
network. predefined access. based on
quotas. expertise or
relevance.
Purpose Used for Ensures Used for quick Focuses on Used in surveys or
hard-to-reach representation and low-cost individuals research with open
or hidden of specific data collection. relevant to the participation.
populations. subgroups. study.

Bias High potential May not reflect High risk of Risk of Prone to self-
for network- the population selection bias researcher- selection bias from
based outside the due to lack of introduced bias motivated
selection quotas. randomisation. in selection. participants.
bias.

Key Results Limited Not Heavily reliant Results often lack

heavily generalisability representative of on researcher representativeness.
limitation depend on beyond quota the broader judgment.
initial criteria. population.
participants.

NON-PROBABILITY SAMPLING

Individuals are selected based on non-random criteria, and not every individual has a chance of being
included. This type of sample is easier and cheaper, but there remains a high risk of sampling bias. This
is because of inferences made about a population are weaker than with probability samples, and could
be assumed as good evidence. Here, conclusions may be more limited.

It is often used in qualitative research as the aim is to develop an initial understanding of a small or
under-researched population, and not to test a hypothesis.

• Quota sampling is a non-random sampling technique best used when researchers want to
ensure that specific subgroups are adequately represented in a study.

It relies on the non-random selection of a predetermined number or proportion of units. This

is called a Quota. For example, in psychology, it can be applied to study gender differences in
emotional recognition by dividing the population into subgroups (strata) like males and
females and selecting a predetermined number of participants from each group. This method
is particularly useful in situations where the goal is to control specific traits or characteristics
in the sample. However, because it does not involve random selection, it may introduce bias,
limiting the generalizability of the findings.

The aim here is to control the sample.

 Page 6 of 7

• Snowball sampling is most effective when researchers need to study hard-to-reach or hidden
populations. For instance, it is commonly used in psychology to study individuals with rare
psychological disorders.

It is also called referral sampling, and is a unique approach researchers use to recruit
participants in qualitative research. The technique involves identifying a few initial
participants who meet the eligibility criteria and asking them to refer others they know who
also fit the requirements. The sample size grows as referrals are added, creating a chain-like
structure.

In this method, the researcher recruits a few initial participants who then help identify and
recruit others from their social networks, creating a “snowball effect.” This approach works
well when the target population is small or difficult to access. It is used in psychology to study
rare psychological disorders.

However, snowball sampling can lead to sampling bias and lack of representativeness, as it
relies heavily on participants’ networks, which might not reflect the diversity of the
larger population.

• Convenience Sampling involves selecting individuals who are most accessible to the
researcher. While it is easy to implement, it has significant drawbacks:

However, there is no way to ensure that the sample is representative of the entire population.
The results may not be applicable to the broader population due to this lack of
representativeness. Additionally, the method is prone to both sampling bias and selection
bias, which can compromise the validity of the study.

For example, using college students for stress studies, as they are readily available.

• Voluntary response sampling is somewhat similar to convenience sampling but involves

participants self-selecting for the study. A voluntary response sample is a sample made up of
individuals who volunteer to be included in the sample.

People who choose to volunteer may differ significantly from those who do not, introducing
self-selection bias. Some individuals may be more likely to volunteer than others, leading to
unrepresentative results.

For example, online surveys where participants voluntarily provide their responses.

• Purposive sampling or judgmental sampling is a method researchers use to pick specific

people to participate in a study because they have particular traits, experiences, or knowledge
that are important for the research.

In psychology, this method helps researchers focus on participants who can provide
meaningful insights about a specific topic. For example, if a psychologist wants to study the
 Page 7 of 7

effects of trauma on firefighters, they would specifically choose firefighters who have
experienced traumatic events instead of randomly selecting participants. This makes the study
more targeted and relevant.

It's a practical way to gather detailed information, but it might not represent the entire
population, so researchers often use it for specialized or in-depth studies. Example: Studies
for expert testimony.