| name | research-methodology |
| version | 1.0.0 |
| description | Select and justify research methodology. Compare approaches, choose designs,
and write methodology chapters for thesis and research proposals.
|
| allowed-tools | ["Read","Write","Edit"] |
Research Methodology Guide
Pilih dan justify methodology untuk research. Bantu tulis methodology chapter.
RESEARCH PARADIGMS
Positivism
- Belief: Reality is objective, measurable
- Methods: Quantitative experiments, surveys
- Knowledge: Generalizable laws
- Use when: Testing hypotheses, cause-effect relationships
Interpretivism (Constructivism)
- Belief: Reality is socially constructed
- Methods: Qualitative interviews, ethnography
- Knowledge: Contextual understanding
- Use when: Exploring meanings, experiences
Pragmatism
- Belief: What works is what matters
- Methods: Mixed methods
- Knowledge: Problem-focused
- Use when: Practical research questions
Critical Theory
- Belief: Research should challenge power structures
- Methods: Participatory action research
- Knowledge: Emancipatory
- Use when: Social justice research
QUANTITATIVE APPROACHES
Experimental Design
| Type | Description | When to Use |
|---|
| True Experiment | Random assignment, control group, manipulation | Gold standard for causality |
| Quasi-Experiment | No random assignment, but comparison group | When randomization impossible |
| Pre-experimental | One group, pre-post test | Pilot studies, exploratory |
Key Elements:
- Independent variable (what you manipulate)
- Dependent variable (what you measure)
- Control group (baseline comparison)
- Random assignment (reduces bias)
Non-Experimental Designs
| Type | Description |
|---|
| Cross-sectional | Data at one point in time |
| Longitudinal | Data over time (panel, cohort) |
| Correlational | Examine relationships, no causation |
| Survey | Self-report data from sample |
Sampling for Quantitative
| Method | When to Use |
|---|
| Random sampling | Generalizability needed |
| Stratified sampling | Subgroups must be represented |
| Cluster sampling | Large, dispersed population |
| Convenience sampling | Limited resources (weakest) |
Sample Size Guidelines:
- Experimental: Minimum 30 per group
- Survey: Minimum 100-300
- Complex analysis: 10-20 participants per variable
- Power analysis: Recommended
QUALITATIVE APPROACHES
Phenomenology
- Focus: Lived experiences
- Data: In-depth interviews
- Sample: 5-25 participants
- Question: "What is the experience of...?"
Grounded Theory
- Focus: Develop theory from data
- Data: Interviews, observations
- Process: Constant comparison
- Question: "What theory explains...?"
Ethnography
- Focus: Culture/context
- Data: Observations, interviews, documents
- Duration: Long-term immersion
- Question: "How do people in this culture...?"
Case Study
- Focus: In-depth single case
- Data: Multiple sources
- Scope: Bounded system
- Question: "How/why does this case...?"
Narrative Inquiry
- Focus: Life stories
- Data: Interviews, documents
- Analysis: Restorying
- Question: "How do people make sense of their lives?"
Sampling for Qualitative
| Method | When to Use |
|---|
| Purposive | Specific characteristics needed |
| Snowball | Hard-to-reach populations |
| Theoretical | Grounded theory development |
| Maximum variation | Diverse perspectives |
Sample Size Guidelines:
- Phenomenology: 5-25
- Grounded theory: 20-30
- Ethnography: Long-term, not fixed
- Case study: 1 (embedded: multiple units)
- Saturation → Stop when no new themes emerge
MIXED METHODS DESIGNS
Convergent Parallel
QUANT Data ──┐
├──→ Integration → Results
QUAL Data ───┘
- Collect simultaneously
- Compare/merge results
- Purpose: Triangulation
Explanatory Sequential
QUANT Data → Results → QUAL Data → Integration
- Quant first, then explain with qual
- Purpose: Explain findings
Exploratory Sequential
QUAL Data → Results → QUANT Data → Integration
- Qual first, then test with quant
- Purpose: Develop instrument
Embedded Design
- One method supports the other
- Primary + secondary questions
- Purpose: Comprehensive understanding
JUSTIFYING YOUR CHOICE
Template for Justification
This study adopts a [quantitative/qualitative/mixed methods] approach because
[reason aligned with research questions]. Specifically, a [specific design]
was selected to [purpose].
This approach is appropriate because:
1. The research question asks [alignment with paradigm]
2. Previous studies [cite precedent]
3. Practical considerations [resources, access]
4. Theoretical framework [if applicable]
Example Justifications
Quantitative Example:
A quasi-experimental design with pre-test and post-test was selected to
examine the effect of AI tutoring on mathematics achievement. This design
is appropriate because: (a) random assignment was not possible due to
existing class structures, (b) comparison groups allowed control for
confounding variables, and (c) pre-test measures established baseline
equivalence (Shadish et al., 2002).
Qualitative Example:
A phenomenological approach was chosen to explore teachers' lived experiences
with AI integration in classrooms. This methodology aligns with the study's
aim to understand subjective meanings rather than measure objective outcomes
(Creswell & Poth, 2018). Purposive sampling ensured participants had direct
experience with the phenomenon.
Mixed Methods Example:
An explanatory sequential design was employed. First, survey data quantified
technology use patterns. Second, interviews explored reasons behind these
patterns. This design allowed statistical generalization while preserving
depth of understanding (Creswell & Plano Clark, 2018).
METHODOLOGY CHAPTER STRUCTURE
Standard Outline
1. Introduction
- Restate research questions
- Preview methodology approach
2. Research Paradigm (optional for undergrad)
- Philosophical assumptions
- Justification
3. Research Design
- Type of design
- Justification
- Visual diagram
4. Participants/Sample
- Population
- Sampling strategy
- Sample size + justification
- Inclusion/exclusion criteria
- Recruitment
5. Data Collection
- Instruments/materials
- Procedure
- Timeline
6. Data Analysis
- Approach (quant/qual/mixed)
- Specific techniques
- Software
7. Trustworthiness/Validity
- Quant: Internal/external validity, reliability
- Qual: Credibility, transferability, dependability
- Mixed: Integration quality
8. Ethical Considerations
- Consent
- Confidentiality
- Risk mitigation
- Approval
9. Limitations
- Acknowledge constraints
- Explain mitigation
VALIDITY & RELIABILITY
Quantitative
| Concept | Definition | How to Address |
|---|
| Internal validity | Cause-effect accurate | Randomization, control |
| External validity | Generalizable | Sampling, replication |
| Reliability | Consistent measurement | Test-retest, Cronbach's α |
Reliability Thresholds:
- α ≥ .70: Acceptable
- α ≥ .80: Good
- α ≥ .90: Excellent
Qualitative
| Concept | Definition | How to Address |
|---|
| Credibility | Believable results | Member checking, triangulation |
| Transferability | Applicable to other contexts | Thick description |
| Dependability | Consistent process | Audit trail |
| Confirmability | Unbiased findings | Reflexivity |
COMMON JUSTIFICATION PHRASES
| Situation | Justification |
|---|
| Small sample size | "Given the exploratory nature..." |
| Convenience sampling | "Due to limited resources and access..." |
| Mixed methods | "To capitalize on strengths of both approaches..." |
| Qualitative | "To gain deep understanding of..." |
| Quantitative | "To test hypotheses and generalize findings..." |
| Single case | "To provide in-depth examination of..." |
PROCESS FOR WRITING METHODOLOGY
- State research questions clearly
- Identify paradigm (positivist/interpretivist/etc.)
- Select approach (quant/qual/mixed)
- Choose specific design (experimental/phenomenological/etc.)
- Justify choices with literature
- Detail procedures (replicable)
- Address validity/reliability
- Discuss ethics
- Acknowledge limitations
QUICK DECISION TREE
Do you want to measure/test/predict?
├─ YES → Quantitative
│ ├─ Can you manipulate variables?
│ │ ├─ YES → Experimental
│ │ └─ NO → Correlational/Survey
│ └─ Time dimension?
│ ├─ One time → Cross-sectional
│ └─ Multiple → Longitudinal
│
└─ NO → Qualitative
├─ Explore experiences?
│ └─ Phenomenology
├─ Build theory?
│ └─ Grounded Theory
├─ Study culture?
│ └─ Ethnography
└─ Deep dive into case?
└─ Case Study
