CHAPTER 3: RESEARCH DESIGN AND SAMPLING

TECHNIQUE
Contents
 Research Design: A Blueprint for Your Study
 Key Components of a Research Design
 Sampling design
 Sample size determination
 Types of samplings
 Data collection approach in qualitative and quantitative research
 1. RESEARCH DESIGN: A BLUEPRINT FOR YOUR STUDY
A research design is essentially the blueprint for your research project.
It outlines the strategy you'll use to answer your research questions,
from inception to data analysis. A well-crafted research design ensures
your study is:
 Valid: Measures what it intends to measure.
 Reliable: Produces consistent results.
 Efficient: Uses resources effectively.
 Ethical: Adheres to ethical guidelines.
2. KEY COMPONENTS OF A RESEARCH DESIGN
A.Research Question(s)
Core of the Study: These are the fundamental inquiries that drive
the research.
Focus and Direction: They provide a clear focus and direction for
the entire research process.
Example: "What is the impact of social media on adolescent mental
health?"
B. Research Objectives
Specific Goals: These are the specific goals that the researcher aims to
achieve.
Breakdown of the Research Question: They break down the research
question into smaller, more manageable components.
Example: For the above research question, objectives might include:
o To identify the most popular social media platforms among adolescents.
o To examine the relationship between social media use and symptoms of
depression and anxiety.
o To explore the impact of social media on sleep patterns and academic
performance.
C. Research Methodology
Overall Approach: This refers to the broad approach to the research, such
as quantitative, qualitative, or mixed-methods.
o Quantitative: Relies on numerical data and statistical analysis.
o Qualitative: Relies on non-numerical data, such as text, images, or
videos.
o Mixed-Methods: Combines both quantitative and qualitative
approaches.
D. Research Design Type
Specific Design within Methodology: This specifies the particular design
within the chosen methodology.
Examples:
o Quantitative: Experimental, quasi-experimental, correlational, survey.
o Qualitative: Case study, ethnography, grounded theory,
phenomenology.
o Mixed-Methods: Sequential explanatory, sequential exploratory,
concurrent triangulation.
E. Data Collection Methods
Techniques for Gathering Data: These are the methods used to collect
data.
Examples:
o Surveys: Questionnaires administered to a sample of participants.
o Interviews: One-on-one or group discussions with participants.
o Observations: Observing behaviors or phenomena in a natural setting.
o Document Analysis: Analyzing existing documents, such as articles,
reports, or social media posts.
F. Data Analysis Techniques
Methods for Analyzing Data: These are the techniques used to analyze
the collected data.
Examples:
o Quantitative: Statistical analysis, such as t-tests, ANOVA, regression
analysis.
o Qualitative: Thematic analysis, content analysis, discourse analysis.
o Mixed-Methods: Data integration techniques, such as data
transformation and data merging.
G. Sampling Strategy
Selecting Participants or Data Sources: This involves determining how
participants or data sources will be selected.
Examples:
o Probability Sampling: Random sampling, stratified sampling, cluster
sampling.
o Non-Probability Sampling: Convenience sampling, purposive
sampling, snowball sampling.
H. Timeline
Research Schedule: This is a detailed schedule that outlines the various
stages of the research process, including data collection, analysis, and report
writing.
I. Ethical Considerations
 Ethical Guidelines: These are guidelines that ensure the research is conducted
ethically.
 Key Considerations:
o Informed consent
o Privacy and confidentiality
o Data security
o Beneficence
o Justice
 3. ADVANCED SAMPLING CONCEPTS
Component Description Advanced Considerations
Target Population The group whose characteristics the  Precise boundaries (temporal, spatial, demographic).
research aims to study.  Address hidden populations using adaptive methods
(e.g., snowball sampling).
Sampling Frame List of elements from which the sample is  Coverage error: Mismatch between frame and
drawn. population.
 Improved coverage via merged datasets or random-
digit dialing.
Sampling Units Levels of elements selected for the study  Multistage sampling combines levels for precision and
(e.g., districts, households, individuals). cost balance.
Sampling Methods of selecting samples (probability  Stratified sampling for precision.
Techniques and non-probability). - Cluster sampling for logistical ease.
 Respondent-driven sampling for network-based
populations.
Determining Calculation based on confidence level,  Use power analysis for effect size detection.
Sample Size margin of error, and population variability.  Oversample to account for non-responses.

Sampling Bias Errors arising from chance (random) or  Minimized via randomization, weighting, and post-
and Errors design flaws (systematic). stratification.
Inclusion/Exclusi Rules defining eligibility to ensure  Use propensity score matching or screening surveys
on Criteria appropriateness. for focused participant selection.
Specialized Sampling in complex studies (e.g.,  Adaptive cluster sampling for underrepresented
Research longitudinal, big data, or hard-to-reach groups.
populations).  Dynamic designs adjust based on interim analysis.
Advanced Tools Techniques enhancing sampling precision.  Monte Carlo simulations test robustness.
 Machine learning predicts parameters.

Ethical and Balancing rigor, resource constraints, and  Ensure informed consent and cultural sensitivity.
Practical participant rights.  Conduct cost-benefit analyses for sampling strategy
 4. DETERMINING AN ADEQUATE SAMPLE SIZE
Determining an adequate sample size is critical for ensuring research
validity and reliability. Here are some basic approaches commonly
used to determine the appropriate sample size:
Basic points
 n:The required sample size
 Z: The Z-score (or Z-value) corresponding to the desired
confidence level.
 Example: For a 95% confidence level, Z≈1.96Z \approx
1.96Z≈1.96.
 P: The estimated proportion of the population with the
characteristic of interest.
 If no prior estimate is available, P=0.5P = 0.5P=0.5 is often used
as it maximizes the sample size (worst-case scenario).
 1−P:The complement of the proportion P, representing the
proportion without the characteristic of interest.
 E:The margin of error or the maximum allowable difference
between the sample proportion and the true population proportion.
Example: If a margin of error of ±5% is acceptable, E=0.05
 I. For Finite Populations
A. Cochran's Formula (Adjusted for Finite Population)
• This is one of the most widely used methods for calculating sample
size for a finite population. The finite population correction is
applied to the sample size calculated for an infinite population.
• The Initial Sample Size Formula (for an Infinite Population)
&Adjusted Sample Size (for a Finite Population) looks like the
followings
 Aspect Initial Formula (n0​) Adjusted Formula (n)
Formula

Purpose Calculates the sample size for an infinite Adjusts n0 to account for the finite
population. population size (N).
Confidence Level Uses the Z-value corresponding to the Relies on n0, which already includes
(Z) desired confidence level (e.g., the Z-value for confidence.
1.961.961.96 for 95%).
Proportion (p) Assumes an estimated proportion of the Proportion is already incorporated in
population with the characteristic of n0 so no additional input is required.
interest (p), typically 0.50.50.5 if
unknown.
Margin of Error Defines the maximum allowable error Margin of error is already considered
(e) between the sample estimate and the true in n0​.
population value (e.g., 0.050.050.05 for
±5%).
Population Size Assumes an infinite population; does Adjusts the sample size based on the
(N) not include the population size (N). total finite population size (N).
Key Adjustment No adjustment; assumes the population is Reduces the sample size
infinitely large, requiring more samples. proportionally when N is finite,
ensuring statistical validity without
over-sampling.
Output Produces n0n , the required sample size Produces n, the final adjusted sample
B. Yamane's Formula
This is a simplified formula specifically designed for determining
sample size for a finite population.

Where:
•n: Required sample size.
•N: Total population size.
•e: Margin of error (e.g., 0.05 for ±5%).
 C. Krejcie and Morgan Table

• This is a ready-made table that provides sample size estimates for

various population sizes and confidence levels (commonly 95%). It

simplifies the process by eliminating the need for calculations.

 Strength: Easy to use for most practical purposes.

 Limitation: Assumes standard conditions (e.g., 95% confidence,

5% margin of error).
 D. Sample Size for Proportions (Finite Population Correction)
If estimating a proportion with finite population, use:

Where:
•Z: Z-value for desired confidence level.
•P: Estimated proportion.
•EEE: Margin of error.
•N: Population size.
E. Sample Size for Means (Finite Population)

If estimating a mean with finite population:

Where:
• n: Required sample size.
• N: Population size.
• Z: Z-value for desired confidence level.
• σ\sigmaσ: Population standard deviation (if unknown, estimate from
a pilot study).
• E: Desired margin of error.
 II. For Infinite Populations

An infinite population assumes no defined size, or the population is

too large to feasibly consider as finite.
 A. Basic Sample Size Formula
This formula is used when the population size is considered infinite or
very large:

Where:
• N0 ​ = Required sample size for an infinite population
• Z = Z-value corresponding to the desired confidence level (e.g.,
1.96 for 95% confidence)
• p = Estimated proportion of the population with the characteristic of
interest (e.g., 0.5 for maximum sample size)
• e = Margin of error (e.g., 0.05 for a 5% margin)
B. Sample Size with Known Population Proportion
When the proportion (p) is known, the sample size can be calculated
as:

Sample Size for a Finite Population (Adjusted Formula)



If you need to adjust for a finite population, you use:

Where:
•NNN = Total population size
•nnn = Adjusted sample size for a finite population
 C. Sample Size for Estimating a Mean
If you are estimating the mean with known standard deviation (σ\sigmaσ):

Where:
•σ\sigmaσ = Standard deviation of the population
D. Sample Size for a Proportion with Adjusted Confidence Interval
If the desired confidence interval and margin are known:

These formulas are essential tools for determining the appropriate sample size
when conducting surveys, experiments, or studies with infinite or very large
populations.
 Population Approach Description
Finite Cochran's Formula (Adjusted for Finite Adjusts sample size for finite populations using finite
Population) population correction.
Yamane's Formula Simplified formula for determining sample size in
finite populations.
Krejcie and Morgan’s Table Provides pre-calculated sample sizes for various
population sizes and confidence levels.
Sample Size for Means (Finite For estimating the mean with finite populations,
Population) considering population standard deviation.
Sample Size for Proportions (Finite Uses Z-scores and margin of error, adjusted for finite
Population Correction) populations.
Sample Size for a Finite Population Adjusts sample size based on Cochran's formula for
(Adjusted Formula) finite populations.
Infinite Basic Sample Size Formula Calculates sample size assuming an infinite or very
large population.
Sample Size with Known Population Used when the proportion (p) is known for sample
Proportion size calculation.
Sample Size for Estimating a Mean Calculates sample size for mean estimation with
known standard deviation.
Sample Size for a Proportion with Adjusts sample size for specific confidence intervals
Adjusted Confidence Interval and margin of error.
Power Analysis Determines sample size needed to achieve statistical
power for detecting effects.
Effect Size Uses expected effect size to estimate the required
sample size.
Bayesian Sampling Adapts sample size dynamically based on prior
knowledge and Bayesian analysis.
 5. TYPES OF SAMPLING METHODS
Sampling methods are broadly categorized into probability sampling
and non-probability sampling, based on whether each member of the
population has a known chance of being selected. Below are the types of
sampling within each category.
I. Probability Sampling
In probability sampling, every member of the population has a known
and non-zero chance of selection. This ensures that the sample is
representative of the population.
A. Simple Random Sampling
o Description: Every individual in the population has an equal
chance of being selected.
o Example: Using a random number generator to pick participants
from a list.
o Advantages: Eliminates selection bias, highly representative for
large samples.
o Limitations: Requires a complete list of the population, can be
time-consuming.
B. Systematic Sampling
o Description: Selects every kthk^{th}kth member of the
population after a random starting point.
o Example: Selecting every 10th student from a school roster.
o Advantages: Easier and quicker than simple random sampling.
o Limitations: Can introduce bias if there is a hidden pattern in the
population.
C. Stratified Sampling
o Description: Divides the population into subgroups (strata) based
on specific characteristics, then randomly samples from each
stratum.
o Example: Sampling equal numbers of men and women in a
survey.
o Advantages: Ensures representation of all subgroups.
o Limitations: Requires detailed population information.
D. Cluster Sampling
o Description: Divides the population into clusters (e.g., geographic
regions), then randomly selects entire clusters for sampling.
o Example: Surveying all households in randomly selected villages.
o Advantages: Cost-effective and practical for large populations.
o Limitations: Higher chance of sampling error compared to
stratified sampling.
E. Multi-Stage Sampling
o Description: Combines multiple sampling methods, such as
selecting clusters first and then using random sampling within
clusters.
o Example: Selecting cities randomly, then households within those
cities.
o Advantages: Flexible and manageable for large populations.
o Limitations: Can be complex and involve multiple sources of
error.
 Sampling Description Example
Method
Simple Every individual has an equal chance Using a random number
Random of selection. generator to pick
Sampling participants from a list.
Systematic Selects every kthk^{th}kth member Selecting every 10th student
Sampling of the population after a random from a school roster.
starting point.
Stratified Divides the population into Sampling equal numbers of
Sampling subgroups (strata) based on specific men and women in a
characteristics, then samples survey.
randomly from each stratum.
Cluster Divides the population into clusters Surveying all households in
Sampling (e.g., geographic regions) and randomly selected villages.
randomly selects entire clusters for
sampling.
Multi-Stage Combines multiple sampling Selecting cities randomly,
Sampling methods, such as selecting clusters then households within
first and then sampling within those cities.
clusters.
 II. Non-Probability Sampling
In non-probability sampling, individuals are selected based on non-
random criteria, and not every member of the population has a chance
of being included. This is often used in exploratory or qualitative
research.
A. Convenience Sampling
o Description: Selects individuals who are easiest to access.
o Example: Surveying people at a mall.
o Advantages: Quick and inexpensive.
o Limitations: High potential for bias, not representative of the
population.
B. Purposive (Judgmental) Sampling
o Description: Participants are selected based on the researcher’s
judgment about who is most relevant to the study.
o Example: Interviewing experts in a specific field.
o Advantages: Useful for targeted studies.
o Limitations: Subjective and prone to researcher bias.
C. Quota Sampling
o Description: Ensures that specific subgroups are represented in the sample
by setting quotas for each subgroup.
o Example: Ensuring that 50% of respondents are women in a study about
workplace diversity.
o Advantages: Ensures subgroup representation without requiring random
selection.
o Limitations: Can still introduce selection bias.
D. Snowball Sampling
o Description: Existing participants recruit others, often used for hard-to-
reach populations.
o Example: Interviewing members of an underground community by
referrals.
o Advantages: Effective for studying rare or hidden populations.
o Limitations: Can lead to biased samples due to reliance on participant
networks.
E. Voluntary Response Sampling
o Description: Participants self-select into the study.
o Example: Online polls where people choose to respond.
o Advantages: Easy to conduct, especially for large-scale studies.
o Limitations: Likely to over represent individuals with strong opinions.
In summary
Sampling Method Description Example

Convenience Selects individuals who are easiest to Surveying people at a

Sampling access. mall.
Purposive Participants are selected based on the Interviewing experts in
Sampling researcher's judgment of their a specific field.
relevance to the study.

Quota Sampling Ensures specific subgroups are Ensuring 50% of

represented in the sample by setting respondents are
quotas for each subgroup. women.

Snowball Sampling Existing participants recruit others, Interviewing members

often used for hard-to-reach of an underground
populations. community by referrals.
Voluntary Participants self-select into the study. Online polls where
Response Sampling people choose to
respond.
6. COMPARISON OF PROBABILITY AND NON-PROBABILITY
SAMPLING

Criteria Probability Sampling Non-Probability Reference

Sampling
Representatio More representative of Less representative, as Cochran, W. G. (1977).
n the population. selection is not Sampling Techniques.
random.
Bias Low risk of bias due to High risk of bias due to Etikan, I., Musa, S. A., &
random selection. non-random selection. Alkassim, R. S. (2016).
Complexity Requires complete Easier and quicker to Kumar, R. (2019).
population data, which implement. Research Methodology.
can be time-consuming
and costly.
Cost and More costly and time- Less costly and time- Fowler, F. J. (2009).
Time intensive. consuming. Survey Research
Methods.
Suitability Best for quantitative and Suitable for qualitative Patton, M. Q. (2002).
large-scale studies. and exploratory Qualitative Research &
research. Evaluation Methods.
 7. TYPES OF DATA COLLECTION IN QUALITATIVE AND
QUANTITATIVE RESEARCH
Data Collection Qualitative Research Quantitative Research
Method
Interviews In-depth, semi-structured, or unstructured interviews to Structured interviews using standardized
explore participants’ experiences, beliefs, and perceptions.
questions, often with closed-ended responses,
to gather numerical data.
Focus Groups Facilitated group discussions to gather diverse perspectives Rarely used in quantitative research unless
on a topic. followed by structured questionnaires.
Observations Non-participant or participant observation to capture Systematic observation with predefined
natural behaviors, contexts, and interactions. checklists or coding systems to record
quantifiable data.
Surveys/Questi Open-ended questionnaires to collect detailed, narrative Close-ended surveys with predefined
onnaires responses. response options to collect large-scale
numerical data.
Document Examination of textual or visual documents (e.g., reports, Analysis of existing records or datasets to
Analysis emails, diaries, videos) to extract qualitative themes. quantify specific variables.

Case Studies Detailed, holistic examination of a single case or a small Comparative analysis of multiple cases with
number of cases. measurable variables.
Experiments Not typically used but can involve exploration of real- Controlled experiments with manipulated
world interventions with a focus on understanding variables to establish cause-effect
processes. relationships.
Ethnography Immersive observation and interaction within a cultural or Rarely used in quantitative research due to its
social group to understand their practices and meanings. subjective and context-dependent nature.
Content Thematic analysis of text, audio, or visual data to identify Quantitative content analysis involves coding
Analysis patterns and insights. and counting specific elements within the
data (e.g., frequency of keywords).
Secondary Review and interpretation of qualitative secondary data Statistical analysis of quantitative datasets
Data Analysis (e.g., transcripts, archival records). collected by others (e.g., census data, health
surveys).