Initializing repository with study materials

2026-03-29 15:25:34 -06:00
commit 825d54dda7
21 changed files with 3097 additions and 0 deletions
--- a/text_coding/analysis/analysis_summary.md
+++ b/text_coding/analysis/analysis_summary.md
@@ -0,0 +1,272 @@
+# Multivariate Analysis of Coding.csv: Virtue Clustering and Associations
+
+**Date:** 2026-03-28  
+**Dataset:** coding.csv  
+**Texts Analyzed:** 134  
+**Unique Virtue Categories:** 74  
+**Average Virtues per Text:** 2.78 (range: 1-5)
+
+---
+
+## 1. Executive Summary
+
+This analysis examines 134 coded texts from two sources (AFP and PR) across 74 unique virtue categories. Using multiple multivariate techniques—clustering, network analysis, and association metrics—the study reveals:
+
+- **4 distinct text clusters** with one dominant cluster containing 86% of texts
+- **3 major virtue communities** representing different conceptual frameworks
+- **Strong ethical pairings** (e.g., Care+Consent) that nearly always co-occur
+- **Source differences** in conceptual complexity (AFP: more interconnected; PR: more focused)
+
+---
+
+## 2. Cluster Analysis of Texts
+
+Using K-means clustering on binary virtue presence/absence vectors:
+
+| Cluster | Size | Key Virtues | Sources | Interpretation |
+|---------|------|-------------|---------|----------------|
+| **1** | 5 texts | Memory, Imitation, Inheritance, Tradition | AFP, PR | *Memory-focused texts* - Historical and temporal continuity themes |
+| **2** | 4 texts | Refusal, Embodiment, Resistance, Subversion | AFP only | *Resistance discourse* - Tactical opposition to systems |
+| **3** | 115 texts | Adaptability, Tension Management, Accessibility, Design | AFP, PR | **Core protocol cluster** - Dominant protocol ethics discourse |
+| **4** | 10 texts | Authenticity, Alignment, Inheritance | AFP, PR | *Authenticity/Alignment cluster* - Self-determination and tradition |
+
+**Key Finding:** Cluster 3 represents the overwhelming majority (86%) of texts, suggesting a shared "protocol ethics" discourse across sources. Cluster 2 represents a distinct "resistance" discourse found only in AFP texts.
+
+---
+
+## 3. Strongest Virtue Associations
+
+### By Co-occurrence Count (raw frequency):
+
+| Rank | Virtue Pair | Count | Notes |
+|------|-------------|-------|-------|
+| 1 | Accessibility + Situational Awareness | 4 | Practical context-sensitivity |
+| 2 | Equity + Inclusivity | 3 | Justice framework |
+| 3 | Balance + Tension Management | 3 | Managing contradictions |
+
+### By Jaccard Similarity (normalized association strength):
+
+| Rank | Virtue Pair | Jaccard Index | Interpretation |
+|------|-------------|---------------|------------------|
+| 1 | **Care + Consent** | 0.750 | *Nearly inseparable* - Ethical foundation pair |
+| 2 | Resistance + Subversion | 0.400 | Tactical cluster |
+| 3 | Refusal + Subversion | 0.400 | Resistance tactics |
+| 4 | **Equity + Inclusivity** | 0.375 | Justice-oriented |
+| 5 | Refusal + Resistance | 0.333 | Activism tactics |
+| 6 | Embodiment + Groundedness | 0.333 | Material presence |
+| 7 | Agency + Freedom | 0.300 | Autonomy cluster |
+
+**Key Finding:** The Care+Consent pairing (Jaccard = 0.750) is exceptionally strong, appearing together in 3 out of 4 possible texts where both concepts appear. This suggests an ethical foundation where care practices are inseparable from consent frameworks.
+
+---
+
+## 4. Virtue Communities (Network Analysis)
+
+Using network thresholding on co-occurrence patterns, three major virtue communities were identified:
+
+### Community 1: "Protocol Mechanics" (~40 virtues)
+*Core operational virtues for protocol design and implementation*
+
+**Central Members:**
+- Adaptability, Agency, Balance, Capture Resistance
+- Care, Complex Systems Tolerance, Consent
+- Constraint, Curiosity, Design, Emergent Properties
+- Equity, Freedom, Institutional Critique, Iterative Development
+- Networked Intelligence, Plurality, Replicability, Systems Thinking
+
+**Characteristics:** 
+- Largest community spanning practical and ethical dimensions
+- High connectivity to Adaptability and Systems Thinking (hub virtues)
+- Brings together ethics (Care, Consent, Equity) with operational concepts (Design, Iterative Development)
+
+### Community 2: "Collective Intelligence" (3 virtues)
+*Focused on collaborative knowledge production*
+
+**Members:** Alignment, Collaboration, Networked Intelligence
+
+**Characteristics:**
+- Small but distinct community
+- Emphasizes distributed, collaborative approach
+- Connected to Community 1 through Networked Intelligence
+
+### Community 3: "Relational Ethics" (~9 virtues)
+*Focus on social and cultural connection*
+
+**Members:**
+- Collectivity, Cultural Awareness, Empathy, Interdependence
+- Plurality, Relationality, Respect, Spatial Awareness
+- Plus contextual concepts
+
+**Characteristics:**
+- Strong ties to Community 1 through Relationality
+- Emphasizes interpersonal and cultural dimensions
+- Includes Plurality, suggesting diversity and multiplicity
+
+---
+
+## 5. Network Centrality Analysis
+
+**"Hub" Virtues** (ranked by number of connections to other virtue types):
+
+| Rank | Virtue | Connections | Key Neighbors |
+|------|--------|-------------|---------------|
+| 1 | **Adaptability** | 25 | Agency, Resistance, Long-Term Vision, Design, Systems Thinking |
+| 2 | **Design** | 23 | Agency, Equity, Emergent Properties, Inheritance, Constraint |
+| 3 | **Agency** | 23 | Resistance, Inheritance, Refusal, Autonomy, Systems Thinking |
+| 4 | **Temporal Awareness** | 19 | Emergent Properties, Long-Term Vision, Adaptability |
+| 5 | **Systems Thinking** | 19 | Agency, Design, Long-Term Vision, Constraint |
+| 6 | **Collectivity** | 17 | Interdependence, Agency, Shared Responsibility |
+| 7 | **Transgression** | 17 | Refusal, Subversion, Care, Capture Resistance |
+| 8 | **Institutional Critique** | 16 | Refusal, Design, Subversion, Agency |
+| 9 | **Plurality** | 16 | Interdependence, Agency, Systems Thinking |
+| 10 | **Relationality** | 16 | Interdependence, Accessibility, Care, Curiosity |
+
+**Key Finding:** **Adaptability** is unequivocally the central hub of this virtue network, connecting to 25 other virtue concepts. This suggests it functions as a bridging concept across multiple ethical and practical domains.
+
+---
+
+## 6. Source Comparison (AFP vs. PR)
+
+| Metric | AFP (62 texts) | PR (72 texts) | Interpretation |
+|--------|----------------|---------------|----------------|
+| **Unique virtue pairs** | 221 | 143 | AFP texts show more conceptual diversity |
+| **Avg pairs per text** | 4.06 | 2.22 | AFP texts are more conceptually dense |
+| **Network density** | 8.2% | 5.3% | AFP has more interconnected virtue networks |
+| **Top virtues** | Adaptability (8), Temporal Awareness (7), Collectivity (7), Institutional Critique (7) | Tension Management (10), Adaptability (9), Systems Thinking (9), Infrastructural Awareness (8) | AFP: critical/social; PR: technical/systemic |
+
+### AFP Code Profile (Academic/Critical)
+- **Dominant themes:** Adaptability, Temporal Awareness, Collectivity, Institutional Critique
+- **Emphasis:** Social processes, critical engagement, collective action
+- **Pattern:** Higher virtue co-occurrence suggests more conceptually complex texts
+
+### PR Code Profile (Practical/Technical)
+- **Dominant themes:** Tension Management, Systems Thinking, Infrastructural Awareness
+- **Emphasis:** Technical complexity, managing contradictions, system design
+- **Pattern:** More focused virtue profiles, strong emphasis on Adaptability
+
+**Key Finding:** Both sources prioritize **Adaptability**, but AFP has more distributed emphasis across critical/social virtues, while PR emphasizes technical/systemic concepts. The 8.2% vs 5.3% network density difference suggests AFP texts engage with more complex conceptual interconnections.
+
+---
+
+## 7. Frequency Distribution
+
+**Top 30 Virtues by Frequency:**
+
+| Rank | Virtue | Count | % of Texts |
+|------|--------|-------|------------|
+| 1 | **Adaptability** | 17 | 12.7% |
+| 2 | Tension Management | 13 | 9.7% |
+| 3 | Accessibility | 13 | 9.7% |
+| 4 | Temporal Awareness | 11 | 8.2% |
+| 5 | Design | 11 | 8.2% |
+| 6 | Institutional Critique | 10 | 7.5% |
+| 7 | Agency | 10 | 7.5% |
+| 8 | Relationality | 10 | 7.5% |
+| 9 | Infrastructural Awareness | 10 | 7.5% |
+| 10 | Systems Thinking | 10 | 7.5% |
+| 11 | Plurality | 9 | 6.7% |
+| 12 | Transgression | 9 | 6.7% |
+| 13 | Collectivity | 8 | 6.0% |
+| 14 | Inheritance | 8 | 6.0% |
+| 15 | Authenticity | 7 | 5.2% |
+| 16 | Long-Term Vision | 7 | 5.2% |
+| 17 | Equity | 6 | 4.5% |
+| 18 | Capture Resistance | 6 | 4.5% |
+| 19 | Respect | 6 | 4.5% |
+| 20 | Cultural Awareness | 6 | 4.5% |
+| 21 | Spatial Awareness | 6 | 4.5% |
+| 22 | Interdependence | 6 | 4.5% |
+| 23 | Shared Responsibility | 6 | 4.5% |
+| 24 | Situational Awareness | 6 | 4.5% |
+| 25 | Memory | 5 | 3.7% |
+| 26 | Embodiment | 5 | 3.7% |
+| 27 | Inclusivity | 5 | 3.7% |
+| 28 | Balance | 5 | 3.7% |
+| 29 | Reciprocity | 5 | 3.7% |
+| 30 | Emergent Properties | 5 | 3.7% |
+
+---
+
+## 8. Key Insights and Implications
+
+### 8.1 The Three Pillars of Protocol Ethics
+
+The analysis reveals three conceptual pillars that structure this discourse:
+
+1. **Adaptive Ethics** (centered on Adaptability and Design): The capacity to adjust, learn, and evolve protocols in response to changing conditions
+
+2. **Relational Justice** (centered on Care, Consent, Equity, Inclusivity): Ethical frameworks emphasizing relationship, respect, and justice
+
+3. **Systemic Resistance** (centered on Refusal, Subversion, Institutional Critique): Tactical opposition and critique of existing systems
+
+### 8.2 The Adaptability Paradigm
+
+The overwhelming centrality of **Adaptability** (highest frequency, highest connectivity) suggests this is the core organizing concept. It bridges:
+- **Ethical dimensions:** Equity, Care, Consent
+- **Operational dimensions:** Design, Iterative Development, Systems Thinking
+- **Resistance dimensions:** Capture Resistance, Resistance, Agency
+
+### 8.3 Source Convergence and Divergence
+
+- **Convergence:** Both sources treat Adaptability as central, suggesting a shared understanding that protocols must be capable of change
+- **Divergence:** AFP emphasizes critical/social dimensions (Institutional Critique, Collectivity), while PR emphasizes technical/systemic dimensions (Tension Management, Systems Thinking)
+- **Integration:** The most conceptually dense texts (highest network density) come from AFP, suggesting critical theory provides more complex conceptual interconnections
+
+### 8.4 Unexpected Pairings
+
+Several virtue pairs show unexpected strength:
+- **Care + Consent** (0.750): Suggests an ethics of care cannot exist without consent frameworks
+- **Refusal + Subversion** (0.400): Tactical language clusters together
+- **Equity + Inclusivity** (0.375): Justice requires both fair distribution and openness
+
+### 8.5 The Resistance Cluster
+
+The small cluster of resistance-focused texts (4 texts in Cluster 2) represents a distinct discourse that:
+- Appears only in AFP texts
+- Coheres around Refusal, Resistance, Subversion, Embodiment
+- Serves as a strategic counterpoint to the dominant protocol design discourse
+- May represent the critical "edge cases" that test protocol boundaries
+
+---
+
+## 9. Methodological Notes
+
+### Analytic Techniques Used:
+1. **K-Means Clustering** (k=4): Identified text groups based on virtue profile similarity
+2. **Network Analysis**: mapped virtue co-occurrences and calculated centrality (degree = number of connections)
+3. **Jaccard Similarity**: normalized measure of virtue pair association (intersection/union)
+4. **Community Detection**: threshold-based clustering of highly connected virtue groups
+
+### Limitations:
+- Small dataset (134 texts) limits statistical power
+- K-means clustering is sensitive to initialization (used deterministic starting points)
+- Binary coding (presence/absence) doesn't capture intensity or salience
+- Limited to virtues 1-5; other dimensions not analyzed
+
+### Generated Files:
+| File | Description |
+|------|-------------|
+| `cooccurrence_matrix.csv` | 25×25 matrix of virtue co-occurrence counts |
+| `jaccard_similarity_matrix.csv` | 25×25 similarity matrix (Jaccard indices) |
+| `strong_associations.csv` | Top 50 virtue pairs with association metrics |
+| `virtue_profiles.json` | Individual virtue profiles for each text |
+
+---
+
+## 10. Recommendations for Further Analysis
+
+1. **Qualitative Deep Dive:** Examine the 4 resistance-focused texts (Cluster 2) and the 10 authenticity-focused texts (Cluster 4) to understand the distinct discourses
+
+2. **Temporal Analysis:** If dates are available, analyze how virtue frequencies change over time
+
+3. **Semantic Mapping:** The Care+Consent pairing could be explored through close reading to understand the conceptual linkage
+
+4. **Source-Specific Models:** Consider whether different theoretical frameworks might be needed for AFP vs. PR texts
+
+5. **Expand to Other Codes:** Analysis currently limited to Virtue_1 through Virtue_5; expanding to other coding categories could reveal additional patterns
+
+6. **Visualization:** Generate network graphs of virtue communities to make relationships visually explicit
+
+---
+
+*Analysis generated using Python standard library (no external packages required). All calculations are fully reproducible.*