Problem Tree: Step-by-Step Guide | Setup Your Project for Success

Implementation Timeline

flowchart TB
    subgraph TopRow[" "]
        direction LR
        subgraph Phase1["📋 Phase 1<br/>Research Planning"]
            direction TB
            S1["1. Define Scope"]
            S2["2. Set Objectives"]
            S3["3. Prepare Documentation"]
            S1 --> S2 --> S3
        end

        subgraph Phase3["🌳 Phase 3<br/>Tree Construction"]
            direction TB
            S7["7. Extract Core Problem"]
            S8["8. Map Root Causes"]
            S9["9. Identify Effects"]
            S7 --> S8 --> S9
        end
    end

    subgraph BottomRow[" "]
        direction LR
        subgraph Phase2["🤖 Phase 2<br/>AI-Assisted Research"]
            direction TB
            S4["4. Customize MCP"]
            S5["5. Execute Research"]
            S6["6. Verify Quality"]
            S4 --> S5 --> S6
        end

        subgraph Phase4["💬 Phase 4<br/>Stakeholder Prep"]
            direction TB
            S10["10. Convert to Questions"]
            S11["11. Plan Validation"]
            S10 --> S11
        end
    end

    S3 -.-> S4
    S6 -.-> S7
    S9 -.-> S10

    style TopRow fill:transparent,stroke:transparent
    style BottomRow fill:transparent,stroke:transparent
    style Phase1 fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px,color:#1F2937
    style Phase2 fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px,color:#1F2937
    style Phase3 fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px,color:#1F2937
    style Phase4 fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px,color:#1F2937
    style S1 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S2 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S3 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S4 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S5 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S6 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S7 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S8 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S9 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S10 fill:#D9F99D,stroke:#72B043,color:#2A2A2A
    style S11 fill:#007F4E,stroke:#00b369,stroke-width:2px,color:#fff,font-weight:bold

Phase 1: Research Planning

Step 1: Define Your Problem Scope: Write a clear 1-2 sentence problem statement

Specify affected population and geographic area

List what you already know
Step 2: Set Research Objectives: Causes and effects you need to understand

Interventions the research should help you design

Stakeholders you'll eventually need to engage
Step 3: Prepare Documentation System: Create a folder for sources

Start from the Problem Tree template

Track evidence (E) versus assumptions (A)

Example Problem Statement

Young adults aged 18-25 in rural Nyanza region, Kenya, have limited access to stable employment.

Phase 2: AI-Assisted Research

Step 4: Customize Your MCP Prompt: Copy the MCP template

Replace placeholders with your context
Step 5: Execute and Review AI Research: Run the prompt, save outputs immediately

Scan for errors or red flags
Step 6: Quality Verification: Check credibility of 3-5 sources

Verify key statistics against original sources

Flag anything that looks like assumption

Don't Skip Verification

AI can generate plausible but incorrect citations.

Phase 3: Problem Tree Construction

Step 7: Extract and Organize Core Problem: Refine the problem statement

Specify who, what, and where

Remove any causes or solutions
Step 8: Map Root Causes by Levels: Level 1: direct causes

Level 2: underlying causes

Level 3: structural causes

Tag each cause (E) evidence-based or (A) assumption
Step 9: Identify Effects by Time Horizon: Immediate, medium-term, and long-term effects

Individual, family, community, and system levels

Tag each effect (E) or (A)

Root Cause Levels

• Skills-labor market mismatch (E)
  - Educational curricula not aligned with market needs (E)
  - Limited access to practical/vocational training (E)
  - Rapid economic transition outpacing skill development (A)

Milestone: Your Problem Tree is Complete!

Causes, problem, and effects, all tagged and ready for validation.

Phase 4: Stakeholder Preparation

Step 10: Convert Assumptions to Questions: Turn each (A) item into an open-ended question

Avoid leading questions

Aim for 8-10 questions
Step 11: Plan Validation Approach: Identify who can answer your questions

Decide how you'll reach them: focus groups, interviews, meetings

Update your tree based on what you learn

Bad Leading Question

"Isn't training the real problem here?"

Good Open-Ended Question

"What keeps young people from finding good work?"

Common Questions

Generic results?: Add more geography and population detail.
Deep enough?: Stop once you reach something actionable.
Skip Phase 2?: MCP is optional but saves hours.

Step-by-Step Implementation Guide