Quality Assurance Decision Flow
graph TD
START["🌳 Preliminary Problem Tree<br/>from AI-Assisted Research"]
%% Checkpoint 1: Core Problem Review
CP1{"✓ CHECKPOINT 1<br/>Core Problem<br/>Quality Check"}
CP1_PASS["✅ Specific, observable,<br/>no embedded causes"]
CP1_FAIL["❌ Generic, vague, or<br/>has embedded solutions"]
CP1_FIX["🔧 Revise Core Problem<br/>Refer to Step 2<br/>Define clear scope"]
%% Checkpoint 2: Root Causes Review
CP2{"✓ CHECKPOINT 2<br/>Root Causes<br/>Quality Check"}
CP2_PASS["✅ 2-3 levels deep,<br/>(E) and (A) tagged,<br/>multi-dimensional"]
CP2_FAIL["❌ Shallow, untagged,<br/>single-dimension"]
CP2_FIX["🔧 Deepen Cause Analysis<br/>Refer to Step 8<br/>Add (E) and (A) tags"]
%% Checkpoint 3: Effects & Research Review
CP3{"✓ CHECKPOINT 3<br/>Effects & Research<br/>Quality Check"}
CP3_PASS["✅ Time horizons mapped,<br/>sources verified,<br/>gaps acknowledged"]
CP3_FAIL["❌ No time dimension,<br/>unverified sources,<br/>no gaps identified"]
CP3_FIX["🔧 Strengthen Effects<br/>& Research<br/>Refer to Steps 6 & 9"]
%% Final Outputs
READY["🎯 READY FOR<br/>STAKEHOLDER<br/>VALIDATION<br/>(Lesson 1.2)"]
QUESTIONS["📝 8-10 Validation<br/>Questions Prepared<br/>from (A) items"]
%% Flow
START --> CP1
CP1 -->|Pass| CP1_PASS --> CP2
CP1 -->|Fail| CP1_FAIL --> CP1_FIX --> CP1
CP2 -->|Pass| CP2_PASS --> CP3
CP2 -->|Fail| CP2_FAIL --> CP2_FIX --> CP2
CP3 -->|Pass| CP3_PASS --> QUESTIONS
CP3 -->|Fail| CP3_FAIL --> CP3_FIX --> CP3
QUESTIONS --> READY
%% Festa Colors
style START fill:#F59E0B,stroke:#D97706,stroke-width:2px,color:#1F2937
style CP1 fill:#10B981,stroke:#059669,stroke-width:2px,color:#fff
style CP2 fill:#10B981,stroke:#059669,stroke-width:2px,color:#fff
style CP3 fill:#10B981,stroke:#059669,stroke-width:2px,color:#fff
style CP1_PASS fill:#72B043,stroke:#5A8F36,stroke-width:1px,color:#fff
style CP2_PASS fill:#72B043,stroke:#5A8F36,stroke-width:1px,color:#fff
style CP3_PASS fill:#72B043,stroke:#5A8F36,stroke-width:1px,color:#fff
style CP1_FAIL fill:#E12729,stroke:#B91C1C,stroke-width:1px,color:#fff
style CP2_FAIL fill:#E12729,stroke:#B91C1C,stroke-width:1px,color:#fff
style CP3_FAIL fill:#E12729,stroke:#B91C1C,stroke-width:1px,color:#fff
style CP1_FIX fill:#F37324,stroke:#E05C1B,stroke-width:1px,color:#fff
style CP2_FIX fill:#F37324,stroke:#E05C1B,stroke-width:1px,color:#fff
style CP3_FIX fill:#F37324,stroke:#E05C1B,stroke-width:1px,color:#fff
style QUESTIONS fill:#72B043,stroke:#5A8F36,stroke-width:2px,color:#fff
style READY fill:#10B981,stroke:#059669,stroke-width:3px,color:#fff
Pass all three checkpoints and you're ready to convert assumptions into stakeholder questions.
Core Problem Quality
- Specific population and geographic scope defined Name the population and place.
- Problem stated without embedded causes or solutions Keep causes separate from the problem statement.
- Observable, measurable, and clearly urgent Verifiable through data or direct observation.
- Connects to your organization's mission and capacity Within your mission and capacity.
Root Causes Analysis
- Causes answer why the problem exists Would the problem exist without it?
- 2-3 levels of depth with sub-causes Direct, underlying, and structural causes.
- Mix of evidence-based (E) and assumption (A) items At least 3-5 items marked (A).
- Covers multiple cause categories Problems are rarely one-dimensional.
- Potentially addressable through interventions Some should be within an NGO's reach.
Effects Mapping
- Realistic consequences across time horizons Immediate, medium-term, and long-term.
- Individual, community, and system-level impacts considered Not just personal; families and systems too.
- Clear connection between problem and effects You can explain the causal link.
- Demonstrates why solving this problem matters Urgent enough to justify the intervention.
Research Quality
- Major claims supported by credible, recent sources UN agencies, government statistics, or established NGOs.
- Evidence clearly separated from assumptions Nothing left unmarked.
- Knowledge gaps and uncertainties acknowledged Honesty, not weakness.
- 8-10 stakeholder validation questions prepared Open-ended, non-leading, targeting your assumptions.
- Ready to learn from community insights Go in to learn, not to confirm.
Red Flags: Pause Before Moving Forward
- ❌ No (A) tags: claiming perfect knowledge isn't credible
- ❌ Generic problem: could describe any country or population
- ❌ Single-cause thinking: problems are rarely that simple
- ❌ No source verification
- ❌ Leading validation questions: "Don't you think..."
Scoring Your Problem Tree
- 16-18
- Ready for validation.
- 13-15
- Good foundation, address gaps first.
- 10-12
- Revisit the Step-by-Step Guide.
- Below 10
- Return to Phase 2 or 3.
Common Questions
- How many causes and effects?
- 4-6 each. Clarity beats quantity.
- Can't verify a source?
- Tag it (A).
- Validate everything first?
- No, be honest about uncertainty.
Next Steps
- Ready
- Move to Lesson 1.2.
- Gaps remain
- Revisit Step-by-Step Guide.
- Want inspiration
- Check Examples.