Sampling methods.
Sample size.
Almaty, 2025
Part I. Theoretical Preparation
Definition of sampling
Sampling is the process of selecting a smaller group (sample)
from a larger population to represent it in a research study.
Importance: In educational research, it is often impossible to
study the entire population of students or teachers. Sampling
saves time, resources, and still provides reliable data if done
correctly.
1. Effect of sample size
A small sample size may lead to unreliable
results, low validity, and difficulty in generalizing
findings.
A large sample size increases reliability and
validity but requires more time and resources.
Aim of sampling
To make research manageable.
To ensure results can be generalized to the
larger population.
To improve efficiency without losing accuracy.
Diagram of Sampling Methods
Part II. In-Class Seminar Activity
1..Random sampling: unbiased, 2..Cluster sampling: practical
representative / but time- for large populations / but
consuming. less precise.
3..Stratified sampling: ensures
4..Convenience sampling:
subgroup representation / but
complex. easy, quick / but biased,
less reliable.
Sampling methods
Systematic Sampling
Simple Random Sampling
Strengths:
Strengths:
Eliminates selection bias (everyone has equal chance).
Easy to implement (select every kᵗʰ unit).
Results are highly representative if sample size is
large.
Good for large populations.
Easy to analyze with statistical methods. More practical than random sampling.
Weaknesses: Weaknesses:
Requires a complete population list (not always Risk of bias if hidden patterns in the
available).
population match the interval.
Can be costly and time-consuming for large
Still needs an ordered list of the
populations.
Randomization might produce unbalanced subgroups. population.
Sampling methods
Stratified Sampling
Cluster Sampling
Strengths: Strengths:
Ensures representation of all subgroups (e.g., Cost-effective for geographically dispersed
age, gender, region). populations.
Increases precision of estimates. Doesn’t require a full population list (just a list
of clusters).
Allows subgroup comparisons.
Useful for large-scale surveys.
Weaknesses:
Weaknesses:
Requires prior knowledge of population Less accurate than simple random or stratified
characteristics. (higher sampling error).
More complex and time-consuming to design. Clusters may not represent entire population.
Mistakes in stratification can lead to bias Requires larger sample size to achieve
precision.
Non-Probability Sampling Methods
Purposive (Judgmental)
Convenience Sampling Sampling
Strengths:
Focuses on participants with specific
Strengths:
characteristics.
Very easy, quick, and inexpensive. Useful for case studies, qualitative
Good for exploratory research. research.
Weaknesses: Efficient when studying small populations.
Weaknesses:
High risk of bias (not representative).
Strong researcher bias in selection.
Results cannot be generalized.
Not generalizable to wider population.
Limited reliability. Reliability depends on researcher’s
judgment.
Non-Probability Sampling Methods
Quota Sampling
Snowball Sampling
Strengths:
Strengths:
Ensures representation of key
Useful for hard-to-reach or hidden
characteristics (e.g., 50% male/female).
populations (e.g., minorities, special groups). Faster and cheaper than probability
Low cost, relies on social networks. methods.
Weaknesses: Practical in large surveys.
Sample is biased (friends of participants often Weaknesses:
Not truly random (interviewer bias
share similar traits).
possible).
Cannot calculate sampling error.
Over- or under-representation may occur.
Limited generalizability. Results less reliable than stratified sampling.
Part III. IWS – Independent Work of Students
150–200 words
Concluson
For this project, we decided that the most appropriate
sampling method is stratified random sampling. Our study
focuses on the effect of Communicative Language Teaching
(CLT) on the speaking fluency of secondary school students.
Since students in one school are often divided into different
grades and proficiency levels, stratified sampling allows us to
ensure that each subgroup is proportionally represented. This
approach increases the reliability of our results, as it prevents
the sample from being biased toward one particular group of
students.
Thank you