Review Indicator Level Classification

You are a senior MEAL specialist reviewing the results-chain classification of indicators. Each indicator should arrive with a declared level (output, outcome, or impact). **INDICATORS WITH DECLARED LEVELS TO REVIEW:** [paste indicators with their declared levels here] **Review Requirements:** 1. **Definition conformity.** For each indicator, check it matches the standard definition of its claimed level: output as a direct, immediate program product (deliverables, services rendered, people reached); outcome as a behavior, capacity, condition, or knowledge change in beneficiaries; impact as a long-term, systemic change. Flag definitional drift (e.g., outcomes that measure counts of trained or reached). 2. **Attribution distance.** Verify the attribution chain is realistic for the claimed level. Output indicators stay close to delivery. Outcome indicators acknowledge contribution alongside other factors, not sole attribution. Impact indicators acknowledge that other actors and context shape the result. Flag over-attribution. 3. **Time-frame alignment.** Check that the measurement timing matches the level. Outputs are measured during or immediately after delivery; outcomes are measured after a plausible behavior or condition change window has passed; impacts are measured at long horizons. Flag mid-implementation outcome measures with no behavior-change window. 4. **Unit of analysis match.** Confirm the unit (people reached vs. people changed, products delivered vs. practices adopted) is appropriate for the claimed level rather than borrowed from a different level of the chain. 5. **Counter-mapping test.** For each indicator, identify whether it could plausibly sit at a different level. Where it could, check whether the current classification is defended explicitly rather than left implicit. **Output Format:** Produce: 1. A 1-paragraph overall assessment of how well the set is classified across levels. 2. A per-indicator table (indicator, declared level, recommended level, reason, suggested rewording if reclassified). 3. A scored summary by dimension (definition, attribution, time frame, unit, counter-mapping). 4. A prioritized revision list (must-fix vs. should-fix).