Lesson 1.3: Synthesize Data Using Affinity Diagrams

Quality Checklist

Assess your synthesis against these indicators, and avoid what undermines it.

Signs of Effective Synthesis

Review after completing your synthesis:

graph TB
    %% ========================================
    %% START NODE
    %% ========================================

    START(["📊 <strong>YOUR SYNTHESIS</strong><br/><br/>Assess Quality Across<br/>Three Dimensions"])

    %% ========================================
    %% THREE QUALITY PILLARS
    %% ========================================

    subgraph PATTERN ["🔍 PATTERN RECOGNITION"]
        direction TB
        P1["Genuine patterns across<br/>multiple stakeholders"]
        P2["Themes that surprised<br/>or challenged assumptions"]
        P3["Actionable intervention<br/>opportunities identified"]
        P4["Both confirmatory &<br/>contradictory evidence"]
        P1 --> P2 --> P3 --> P4
    end

    subgraph VOICE ["🗣️ COMMUNITY VOICE"]
        direction TB
        V1["Stakeholder language<br/>& priorities reflected"]
        V2["Different perspectives<br/>preserved (not homogenized)"]
        V3["Cultural context &<br/>values evident"]
        V4["Power dynamics &<br/>marginalized voices acknowledged"]
        V1 --> V2 --> V3 --> V4
    end

    subgraph INTEGRATION ["🔗 INTEGRATION QUALITY"]
        direction TB
        I1["Problem Tree significantly<br/>strengthened"]
        I2["Clear traceability from<br/>insights → themes → analysis"]
        I3["Evidence base robust<br/>& community-grounded"]
        I4["Strong foundation for<br/>Theory of Change"]
        I1 --> I2 --> I3 --> I4
    end

    %% ========================================
    %% END NODE
    %% ========================================

    READY(["✅ <strong>READY</strong><br/><br/>Quality synthesis achieved<br/>Proceed to Theory of Change"])

    %% ========================================
    %% CONNECTIONS
    %% ========================================

    START --> PATTERN
    START --> VOICE
    START --> INTEGRATION
    PATTERN --> READY
    VOICE --> READY
    INTEGRATION --> READY

    %% ========================================
    %% FESTA COLORS
    %% ========================================

    style START fill:#6B7280,stroke:#4B5563,stroke-width:3px,color:#fff

    style PATTERN fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px,color:#2A2A2A
    style P1 fill:#FEF9C3,stroke:#F59E0B,stroke-width:1px,color:#2A2A2A
    style P2 fill:#FEF9C3,stroke:#F59E0B,stroke-width:1px,color:#2A2A2A
    style P3 fill:#FEF9C3,stroke:#F59E0B,stroke-width:1px,color:#2A2A2A
    style P4 fill:#FEF9C3,stroke:#F59E0B,stroke-width:1px,color:#2A2A2A

    style VOICE fill:#D9F99D,stroke:#72B043,stroke-width:2px,color:#2A2A2A
    style V1 fill:#ECFCCB,stroke:#72B043,stroke-width:1px,color:#2A2A2A
    style V2 fill:#ECFCCB,stroke:#72B043,stroke-width:1px,color:#2A2A2A
    style V3 fill:#ECFCCB,stroke:#72B043,stroke-width:1px,color:#2A2A2A
    style V4 fill:#ECFCCB,stroke:#72B043,stroke-width:1px,color:#2A2A2A

    style INTEGRATION fill:#BBF7D0,stroke:#10B981,stroke-width:2px,color:#2A2A2A
    style I1 fill:#D1FAE5,stroke:#10B981,stroke-width:1px,color:#2A2A2A
    style I2 fill:#D1FAE5,stroke:#10B981,stroke-width:1px,color:#2A2A2A
    style I3 fill:#D1FAE5,stroke:#10B981,stroke-width:1px,color:#2A2A2A
    style I4 fill:#D1FAE5,stroke:#10B981,stroke-width:1px,color:#2A2A2A

    style READY fill:#007F4E,stroke:#00b369,stroke-width:4px,color:#fff

Pattern Recognition

Community Voice Preservation

Integration Quality

Common Pitfalls to Avoid

graph TB
    %% ========================================
    %% START NODE
    %% ========================================

    DETECT(["🚩 <strong>PITFALL DETECTION</strong><br/><br/>Watch for These<br/>Warning Signs"])

    %% ========================================
    %% FOUR PITFALLS
    %% ========================================

    subgraph PIT1 ["❌ CONFIRMATION BIAS"]
        direction TB
        CB_WARN["⚠️ Warning Signs:<br/>• Clustering to confirm<br/>• Dismissing contradictions<br/>• Selecting preferred quotes"]
        CB_FIX["🔧 Fix:<br/>Seek contradictory evidence<br/>Have others review clustering"]
        CB_WARN --> CB_FIX
    end

    subgraph PIT2 ["❌ OVER-SIMPLIFICATION"]
        direction TB
        OS_WARN["⚠️ Warning Signs:<br/>• Themes too broad<br/>• Combining distinct issues<br/>• Smoothing contradictions"]
        OS_FIX["🔧 Fix:<br/>Split broad themes<br/>Embrace complexity"]
        OS_WARN --> OS_FIX
    end

    subgraph PIT3 ["❌ ANALYSIS PARALYSIS"]
        direction TB
        AP_WARN["⚠️ Warning Signs:<br/>• Too many micro-themes<br/>• Excessive time on clustering<br/>• Stuck on outliers"]
        AP_FIX["🔧 Fix:<br/>Set time limits<br/>Focus on 5-8 themes"]
        AP_WARN --> AP_FIX
    end

    subgraph PIT4 ["❌ VOICE ERASURE"]
        direction TB
        VE_WARN["⚠️ Warning Signs:<br/>• Technical language only<br/>• External frameworks imposed<br/>• Funder-focused themes"]
        VE_FIX["🔧 Fix:<br/>Preserve stakeholder quotes<br/>Use community language"]
        VE_WARN --> VE_FIX
    end

    %% ========================================
    %% END NODE
    %% ========================================

    CLEAR(["✅ <strong>PITFALLS AVOIDED</strong><br/><br/>Synthesis maintains rigor<br/>& community grounding"])

    %% ========================================
    %% CONNECTIONS
    %% ========================================

    DETECT --> PIT1
    DETECT --> PIT2
    DETECT --> PIT3
    DETECT --> PIT4
    PIT1 --> CLEAR
    PIT2 --> CLEAR
    PIT3 --> CLEAR
    PIT4 --> CLEAR

    %% ========================================
    %% FESTA COLORS
    %% ========================================

    style DETECT fill:#6B7280,stroke:#4B5563,stroke-width:3px,color:#fff

    style PIT1 fill:#FEE2E2,stroke:#EF4444,stroke-width:2px,color:#2A2A2A
    style CB_WARN fill:#FEF2F2,stroke:#EF4444,stroke-width:1px,color:#2A2A2A
    style CB_FIX fill:#FED7AA,stroke:#F97316,stroke-width:1px,color:#2A2A2A

    style PIT2 fill:#FEE2E2,stroke:#EF4444,stroke-width:2px,color:#2A2A2A
    style OS_WARN fill:#FEF2F2,stroke:#EF4444,stroke-width:1px,color:#2A2A2A
    style OS_FIX fill:#FED7AA,stroke:#F97316,stroke-width:1px,color:#2A2A2A

    style PIT3 fill:#FEE2E2,stroke:#EF4444,stroke-width:2px,color:#2A2A2A
    style AP_WARN fill:#FEF2F2,stroke:#EF4444,stroke-width:1px,color:#2A2A2A
    style AP_FIX fill:#FED7AA,stroke:#F97316,stroke-width:1px,color:#2A2A2A

    style PIT4 fill:#FEE2E2,stroke:#EF4444,stroke-width:2px,color:#2A2A2A
    style VE_WARN fill:#FEF2F2,stroke:#EF4444,stroke-width:1px,color:#2A2A2A
    style VE_FIX fill:#FED7AA,stroke:#F97316,stroke-width:1px,color:#2A2A2A

    style CLEAR fill:#007F4E,stroke:#00b369,stroke-width:4px,color:#fff

❌ Confirmation Bias

Warning Signs:

  • ▪ Clustering to confirm predetermined themes
  • ▪ Dismissing or minimizing contradictory evidence
  • ▪ Selecting quotes that support preferred conclusions
  • ▪ Missing patterns that challenge assumptions

How to Fix:

Seek contradictory insights; have someone else review your clustering.

❌ Over-Simplification

Warning Signs:

  • ▪ Creating themes so broad they lose specificity
  • ▪ Combining distinct issues for tidiness
  • ▪ Smoothing over contradictions instead of exploring them
  • ▪ Losing nuance in pursuit of clean patterns

How to Fix:

Split broad themes; document contradictions instead of smoothing them over.

❌ Analysis Paralysis

Warning Signs:

  • ▪ Too many micro-themes with no strategic insight
  • ▪ Too much time on clustering, not pattern recognition
  • ▪ Stuck on outliers instead of strong patterns
  • ▪ Failing to synthesize themes into implications

How to Fix:

Set time limits; aim for 5-8 themes, and accept "good enough" clustering.

❌ Community Voice Erasure

Warning Signs:

  • ▪ Translating insights into technical or academic language
  • ▪ Imposing frameworks that ignore community priorities
  • ▪ Ignoring different perspectives within the community
  • ▪ Themes that please funders but miss emphasis

How to Fix:

Preserve quotes, use community language, share findings back for validation.

Self-Assessment Questions

  1. Surprise: did any theme challenge my assumptions?
  2. Diversity: do themes reflect different stakeholder groups?
  3. Actionability: are themes specific, or too vague?
  4. Traceability: can I trace themes to quotes and tree elements?
  5. Credibility: would stakeholders recognize themselves in this?
  6. Complexity: is nuance preserved, not oversimplified?

What's Next

Next: Real-World Examples

The Nigeria Youth Livelihood project.

View Examples →
Lesson 1.3: Synthesize Data 82%
Step 9 of 11 · Quality Checklist