Build a Theory of Change with AI

You are a senior M&E specialist. I need to build a Theory of Change for my program, starting with the problem analysis. Based on the program description I provide, produce: 1. **Problem statement** (2-3 sentences): What is the core problem? Who is affected? What are the consequences of inaction? 2. **Root causes** (4-6 causes): What drives this problem? Distinguish between immediate causes, underlying causes, and structural causes. Present as a simple causal chain, not a flat list. 3. **Affected population**: Who experiences this problem? How are different groups affected differently? 4. **Existing responses**: What is already being done about this problem (by government, other organizations, communities)? Where are the gaps? Here is my program description: [Describe your program, its goals, target population, and context]

Based on the problem analysis, map the causal pathway for this program. Present it as a structured results chain: **Impact** (1 statement): The long-term change this program contributes to. **Outcomes** (2-4): The medium-term changes in behavior, practice, or conditions that the program expects to achieve. These should be measurable and time-bound. **Outputs** (3-6): The direct products or services the program delivers. Each output should logically lead to at least one outcome. **Activities** (5-10): The key activities that produce the outputs. For each level, explain the causal logic: WHY does this activity produce this output? WHY does this output lead to this outcome? Make the "because" explicit. Present the pathway as a structured table with columns: Level, Statement, Causal Logic ("because...").

Review the causal pathway and identify the assumptions and risks at each level. **Assumptions table:** For each causal link (activity to output, output to outcome, outcome to impact), identify: - The assumption (what must be true for this link to hold) - Current evidence for or against this assumption - How you will test or monitor this assumption during implementation - What happens to the program logic if this assumption fails **Risk register:** Identify 5-8 risks that could prevent the program from achieving its outcomes. For each: - Risk description - Likelihood (high, medium, low) - Impact if it occurs (high, medium, low) - Mitigation strategy - Who is responsible for monitoring this risk Focus on risks that are specific to this program and context, not generic risks that apply to any development program.

For each level of the causal pathway (outputs, outcomes, impact), propose SMART indicators. Present as a table with columns: - Results level (output/outcome/impact) - Result statement - Proposed indicator - Data source - Collection frequency - Baseline status (known, needs collection, or not applicable) Include 1-2 indicators per output, 1-2 per outcome, and 1 for the impact level. Prioritize indicators that are: - Feasible to collect with typical program resources - Sensitive enough to detect change within the program timeframe - Clearly linked to the causal pathway (not generic sector indicators) Also propose 2-3 assumption-monitoring indicators: indicators that track whether key assumptions are holding, not just whether results are being achieved.

Write a Theory of Change narrative summary (800-1200 words) that synthesizes everything we have built. Structure it as: 1. **Context and problem** (150-200 words): The problem, who it affects, and why it persists. 2. **Program logic** (300-400 words): How the program intervenes, the causal pathway from activities to impact, and the key "because" logic at each transition. 3. **Assumptions and risks** (200-250 words): The critical assumptions the logic depends on, the biggest risks, and how they will be monitored. 4. **Measurement approach** (150-200 words): How the program will know whether the ToC is working, including the role of assumption-monitoring indicators. Write for a mixed audience of program staff, donors, and partners. Be direct and specific. Avoid jargon where plain language works. The narrative should be understandable to someone who has not seen the diagram.