Methodological Framework
Research Framework
Tracing the journey from raw data to civic insight
PHASES: 6
•
SAMPLE: n=450
•
CONFIDENCE: 95%
•
METHOD: MIXED
α1
n=450, CI=95%
Sampling Theory
Coverage: 12 regions
Data Discovery
Phase 1/6We listen where others don't.
1.1
Forum Scouting
1.2
Ethical Clearance
1.3
Regional Sampling
1.4
Crawl Design
1.5
Metadata Verification
NODES: 5
•
DEPTH: 1
•
STATUS: ACTIVE
β2
λ=0.85, κ=0.92
Inter-rater Reliability
Anonymization: 100%
Data Collection
Phase 2/6Silence becomes data.
2.1
Forum Scraping
2.2
Anonymization
2.3
Storage & Tagging
2.4
Community Validation
NODES: 4
•
DEPTH: 2
•
STATUS: ACTIVE
γ3
TF-IDF × BERT
Semantic Analysis
Clusters: 28 topics
Language Processing
Phase 3/6Words reveal worlds.
3.1
Tokenization
3.2
Semantic Clustering
3.3
Sentiment Calibration
3.4
Topic Embedding
NODES: 4
•
DEPTH: 3
•
STATUS: ACTIVE
δ4
r=0.78, p<0.001
Correlation Analysis
Cross-regional: 89%
Thematic Mapping
Phase 4/6Patterns emerge from chaos.
4.1
Cluster Labelling
4.2
Context Comparison
4.3
Temporal Dynamics
4.4
Cross-Regional Synthesis
NODES: 4
•
DEPTH: 4
•
STATUS: ACTIVE
ε5
UX Score: 8.4/10
Usability Testing
Engagement: +340%
Visualization & Interaction
Phase 5/6Data becomes narrative.
5.1
Node Mapping
5.2
Interactive Layers
5.3
Geo-Temporal Interface
5.4
Insight Narratives
NODES: 4
•
DEPTH: 5
•
STATUS: ACTIVE
ζ6
Impact Factor: 2.8
Citation Analysis
Policy refs: 14
Civic Impact
Phase 6/6Voices shape policy.
6.1
Policy Reflection
6.2
Media Outreach
6.3
Educational Use
6.4
Open Archive
NODES: 4
•
DEPTH: 6
•
STATUS: ACTIVE
Main Stage
1.1
Sub-ProcessData Flow
Research Methodology Notation
n: Sample Size•CI: Confidence Interval•λ: Cronbach's Alpha•κ: Cohen's Kappa•r: Correlation Coefficient•p: Significance Level
Silence speaks.