How to Design Research Methodology
Daniel Ha · Seoul National University PhD StudentSelect a quantitative, qualitative, or mixed-methods approach based on your
research question, then design the methodology section in 6 steps: research
design, participants, data collection, analysis plan, validity and
reliability, and ethical considerations. The guiding principle is that the
question determines the method, not the other way around.
Why Does Methodology Design Matter?
Methodology determines how you will answer your research question, and it is the key basis on which reviewers assess the feasibility of your study. The same topic can become entirely different research depending on the methodology. If you are studying "the impact of remote work on productivity," you could survey 500 respondents and analyze the data statistically, or you could conduct in-depth interviews with 10 participants to explore their experiences. Which method is right depends on whether you are asking "How much does it affect?" or "How does it affect?"
Reviewers look for three things in the methodology section — alignment between the research question and the method, feasibility, and a logical rationale for why this method was chosen.
Which Research Approach Should You Choose?
Quantitative Research
Measures numerically and analyzes statistically. Suited for questions like "How much?", "What relationship?", and "Is there a difference?"
| Type | Description | Examples |
|---|---|---|
| Experimental | Manipulates variables and measures effects | A/B testing, pre-post tests |
| Survey | Collects large-scale data via questionnaires | Likert scale surveys, online surveys |
| Correlational | Statistically analyzes relationships between variables | Regression analysis, SEM |
| Meta-analysis | Synthesizes existing research results | Effect size aggregation, systematic reviews |
Strengths: Generalizable, objective, replicable. Limitations: Difficult to capture context and meaning.
Qualitative Research
Explores meaning and context through words, behaviors, and texts. Suited for questions like "Why?", "How?", and "What experiences?"
| Type | Description | Examples |
|---|---|---|
| Phenomenology | Explores the essence of experiences | In-depth interviews, experience descriptions |
| Grounded theory | Derives theory from data | Iterative coding, theoretical sampling |
| Case study | Analyzes specific cases in depth | Single/multiple case analysis |
| Ethnography | Understands behavior in cultural context | Participant observation, field notes |
Strengths: Deep understanding, contextual insight. Limitations: Difficult to generalize, potential researcher bias.
Mixed Methods
Uses both quantitative and qualitative methods together. Suited for complex questions that cannot be answered by a single method alone.
| Design | Sequence | When to Use |
|---|---|---|
| Convergent | Quantitative + qualitative simultaneously | Examining the same phenomenon from different angles |
| Sequential explanatory | Quantitative → qualitative | Explaining survey results through interviews |
| Sequential exploratory | Qualitative → quantitative | Validating exploratory findings at scale |
How Do You Match Methodology to Your Research Question?
Questions asking "how much?" call for quantitative methods; questions asking "why/how?" call for qualitative methods; and if you need both, mixed methods are the answer. The starting point for choosing a methodology is always the research question.
| Research Question Type | Suitable Approach | Data Collection | Analysis Method |
|---|---|---|---|
| "What is the effect of X on Y?" | Quantitative (experimental) | Experimental data, pre-post tests | t-test, ANOVA, regression |
| "What is the relationship between X and Y?" | Quantitative (correlational) | Structured questionnaires | Correlation analysis, SEM |
| "How do participants experience X?" | Qualitative | Semi-structured interviews, observation | Thematic analysis, phenomenological analysis |
| "What is the process of phenomenon X?" | Qualitative (grounded) | Interviews, document analysis | Grounded theory coding |
| "How does the effect of X work?" | Mixed | Surveys + interviews | Statistical + thematic analysis |
If you are unsure which methodology fits your research question, try entering it into Nubint AI's Methodology Advisor agent. It analyzes methods used in similar studies and compares the strengths and weaknesses of each approach.
Designing Data Collection
Research Participants (Sampling)
| Decision | Considerations |
|---|---|
| Define the population | Who is the target population for generalization? |
| Sampling method | Probability sampling (random) vs. non-probability (convenience, snowball, purposive) |
| Sample size | Quantitative: calculate via power analysis. Qualitative: until saturation |
| Inclusion/exclusion criteria | Who is included and excluded? Provide clear criteria |
Data Collection Instruments
For quantitative research, prioritize instruments with established validity and reliability. If you must develop a new instrument, a pilot test is essential.
For qualitative research, make sure interview guide questions directly connect to your research questions. Avoid leading questions and use open-ended questions.
Data Analysis Plan
Plan how you will analyze the data before you collect it. The approach of "let's collect the data first and see" leads to losing direction.
Quantitative Analysis
| Research Purpose | Analysis Method | Prerequisites |
|---|---|---|
| Group differences | t-test, ANOVA | Normality, homogeneity of variance |
| Variable relationships | Correlation, regression | Linearity, normality |
| Structural relationships | Structural Equation Modeling (SEM) | Sufficient sample size |
| Categorical data | Chi-square test | Expected frequency conditions |
| Mediation/moderation | Mediation analysis, moderated regression | Theoretical basis |
Qualitative Analysis
| Analysis Method | Suitable Design | Key Procedures |
|---|---|---|
| Thematic analysis | Most qualitative studies | Coding → categorization → theme identification |
| Phenomenological analysis | Phenomenology | Extracting meaning units → describing essential structure |
| Grounded theory coding | Grounded theory | Open coding → axial coding → selective coding |
| Narrative analysis | Narrative research | Story structure analysis, chronological reconstruction |
Validity and Reliability
These are the key elements that ensure the quality of your methodology.
Quantitative Research
- Internal validity: Did only the independent variable affect the dependent variable? (control variables)
- External validity: Can the results be generalized to other situations? (sample representativeness)
- Reliability: Would the same conditions produce the same results? (Cronbach's α ≥ .70)
Qualitative Research
- Credibility: Member checking, triangulation
- Transferability: Thick description
- Dependability: Audit trail
- Confirmability: Researcher reflexive journal
Ethical Considerations
Research involving human participants requires IRB (Institutional Review Board) approval.
| Ethical Principle | Implementation |
|---|---|
| Informed consent | Explain the purpose, procedures, and risks; obtain written consent |
| Confidentiality | Anonymize/pseudonymize data; store with encryption |
| Minimize risk | Design the study to avoid psychological or physical harm |
| Data management | Specify retention period and disposal method after collection |
Writing the Methodology Section
- Research design overview — What approach (quantitative/qualitative/mixed) you chose and why
- Participants — Population, sampling method, sample size and its rationale
- Data collection — Instruments, procedures, timeframe
- Data analysis — Analysis methods and their connection to research questions
- Validity/reliability — How you ensured research quality
- Ethical considerations — IRB approval, consent forms, data protection
Be sure to include "Why did you choose this method?" for each item. Referencing methodologies used in similar studies provides justification for your choices. Using the Literature Review agent to analyze prior research on your topic helps you identify commonly used methods and recent trends in your field.
Common Mistakes
| Mistake | Solution |
|---|---|
| Methodology does not match the research question | Trace back from the question type to the method |
| No rationale for "Why this method?" | Reference similar studies + compare pros and cons |
| No justification for sample size | Provide power analysis (quantitative) or saturation rationale (qualitative) |
| Deciding the analysis method after data collection | Establish an analysis plan before collection |
| No mention of validity/reliability | Report reliability coefficients (quantitative) or triangulation (qualitative) |
Summary
Research methodology is designed in the sequence: research question → approach selection → data collection design → analysis plan → validity measures. The most important principle is that "the question determines the method." Choose the method that best answers your research question, not the one that is trendy or familiar.
Once your methodology is finalized, use the How to Write a Research Proposal guide to document your full research plan. If your data collection requires a survey, try Nubint AI's Survey Generator agent.