Copy a prompt into Claude, ChatGPT, or Gemini. Paste your document at the bottom and run.
Paste a document and get a scored quality assessment with evidence and revision priorities.
6,008 characters
You are an expert in theory of change and program design with experience reviewing causal pathways in development and humanitarian programs. Score the causal logic of the theory of change in the document I will provide using the rubric below. The document may be a standalone ToC, or a ToC embedded in a logframe, MEL plan, contribution analysis, adaptive management memo, or any document containing a causal pathway. Focus only on the causal logic, not on stakeholder engagement, presentation, or other holistic ToC dimensions.
SCORING RUBRIC - Theory of Change Causal Logic
Score each dimension 1-5 using these criteria:
DIMENSION 1: Causal Chain Plausibility
- Score 5: All four elements present. Each causal link is plausible based on evidence or programmatic logic. Links are not collapsed (e.g., outputs to outcomes are connected via a specific mechanism, not just an arrow). Mechanisms are named (HOW does X cause Y, not just THAT X causes Y). Counterfactual logic is considered (what would happen without this intervention).
- Score 4: At least three of four elements present. Causal chain is plausible and most mechanisms named; counterfactual logic or one link partial.
- Score 3: Causal chain is broadly plausible but some links collapsed. Mechanisms implicit. Counterfactual not considered.
- Score 2: Multiple links collapsed into single arrows. Mechanisms missing. Logic feels aspirational rather than grounded.
- Score 1: Causal chain is implicit or absent. Outputs and outcomes listed without connecting logic.
DIMENSION 2: Assumption Identification
- Score 5: All four elements present. Assumptions are named at each link in the causal chain. Assumptions are testable or at minimum trackable (someone could observe whether they hold). Critical assumptions are distinguished from minor ones (e.g., "critical: government budget for service delivery is sustained"). Risk to the chain if an assumption fails is acknowledged with consequence stated.
- Score 4: At least three of four elements present. Assumptions named at most links and trackable; critical-vs-minor distinction or failure consequence partial.
- Score 3: Some assumptions named but not at every link. Testability not addressed. Critical assumptions not distinguished.
- Score 2: Generic assumptions listed at the bottom of the ToC, not tied to specific links. No criticality ranking.
- Score 1: No assumptions named, or assumptions are platitudes ("stakeholders will cooperate").
DIMENSION 3: External Conditions and Risks
- Score 5: All four elements present. External conditions affecting the causal chain are identified across relevant categories (political, economic, social, environmental). Specific risks to the chain are named with likelihood and impact. Mitigation strategies for high-priority risks are documented. Boundary conditions are stated (where this ToC applies and where it does not, e.g., "this ToC assumes a stable security context").
- Score 4: At least three of four elements present. External conditions and risks identified with mitigation; boundary conditions or likelihood/impact partial.
- Score 3: External conditions mentioned generically. Risks named but without likelihood or impact. Mitigation absent.
- Score 2: Vague mention of "external factors" without specifics. No risk register.
- Score 1: External conditions and risks not addressed.
DIMENSION 4: Evidence Base for Causal Claims
- Score 5: All four elements present. Causal claims are grounded in research, prior evaluations, or programmatic experience. Sources are cited where claims are based on external evidence (studies, evaluations, sector literature). Strength of evidence is acknowledged (strong empirical, programmatic experience, plausibility argument) for each major link. Gaps in the evidence base are flagged where they exist (e.g., "this link is based on plausibility; no direct evidence in this context").
- Score 4: At least three of four elements present. Most claims grounded and sources cited; strength acknowledgment or gap flagging partial.
- Score 3: Some claims grounded but sources missing for most. Strength of evidence not acknowledged. Gaps not flagged.
- Score 2: Causal claims asserted without grounding. Few or no sources. Evidence treated as uniform.
- Score 1: No evidence base. Claims are pure assertion.
DIMENSION 5: Testability and Falsifiability
- Score 5: All four elements present. The ToC could be tested with available or feasible methods (the team can articulate how they would know it is working). Indicators or signals are identified that would confirm or refute key links. A mechanism for revising the ToC based on emerging evidence is specified (when, how, who decides). Negative evidence is considered (what would tell us the ToC is wrong, not just what would confirm it).
- Score 4: At least three of four elements present. Testability and indicators present; revision mechanism or negative-evidence consideration partial.
- Score 3: Some indicators identified but no clear test. Revision mechanism implicit. Negative evidence not addressed.
- Score 2: ToC presented as fixed. No indicators tied to causal links. No revision plan.
- Score 1: ToC is unfalsifiable. No way to know if it is right or wrong.
OUTPUT FORMAT:
Return your assessment as a table followed by a summary:
| Dimension | Score (1-5) | Evidence from Document | Priority Revision |
|-----------|-------------|------------------------|-------------------|
| Causal Chain Plausibility | | | |
| Assumption Identification | | | |
| External Conditions and Risks | | | |
| Evidence Base for Causal Claims | | | |
| Testability and Falsifiability | | | |
**Total: X/25**
**Band:** Sound (22-25) / Adequate (17-21) / Needs Revision (11-16) / Substantial Revision (5-10)
**Single Most Important Revision:** [One specific sentence]
For any dimension scored 1 or 2, add a brief explanation and a concrete revision example.
DOCUMENT TO SCORE:
[Paste your theory of change or the document containing the causal pathway here]
Scoring Criteria
Causal Chain Plausibility
5Excellent
All four elements present. Each link plausible. Links not collapsed. Mechanisms named (HOW, not just THAT). Counterfactual logic considered.
4Good
At least three elements. Chain plausible and most mechanisms named; counterfactual or one link partial.
3Adequate
Chain broadly plausible but some links collapsed. Mechanisms implicit. Counterfactual not considered.
All four elements present. Assumptions named at each link. Testable or trackable. Critical distinguished from minor. Failure consequences acknowledged.
4Good
At least three elements. Assumptions named at most links and trackable; critical-vs-minor or failure consequence partial.
3Adequate
Some assumptions named but not at every link. Testability not addressed. Critical not distinguished.
2Needs Improvement
Generic assumptions at bottom of ToC, not tied to links. No criticality ranking.
1Inadequate
No assumptions named or assumptions are platitudes.
External Conditions and Risks
5Excellent
All four elements present. External conditions identified across relevant categories. Risks named with likelihood and impact. Mitigation for high-priority risks documented. Boundary conditions stated.
4Good
At least three elements. Conditions and risks identified with mitigation; boundary conditions or likelihood/impact partial.
3Adequate
External conditions mentioned generically. Risks without likelihood or impact. Mitigation absent.
2Needs Improvement
Vague mention of "external factors". No risk register.
1Inadequate
External conditions and risks not addressed.
Evidence Base for Causal Claims
5Excellent
All four elements present. Claims grounded in research, evaluations, or experience. Sources cited. Strength of evidence acknowledged per link. Gaps flagged.
4Good
At least three elements. Most claims grounded and sources cited; strength or gap flagging partial.
3Adequate
Some claims grounded but sources missing for most. Strength not acknowledged. Gaps not flagged.
2Needs Improvement
Claims asserted without grounding. Few or no sources. Evidence treated as uniform.
1Inadequate
No evidence base. Pure assertion.
Testability and Falsifiability
5Excellent
All four elements present. ToC testable with feasible methods. Indicators identified for key links. Revision mechanism specified. Negative evidence considered.
4Good
At least three elements. Testability and indicators present; revision mechanism or negative evidence partial.
3Adequate
Some indicators but no clear test. Revision mechanism implicit. Negative evidence not addressed.
2Needs Improvement
ToC presented as fixed. No indicators tied to links. No revision plan.
1Inadequate
ToC unfalsifiable.
Score Interpretation
Total (out of 25)
Band
Next Step
22-25
Sound
Causal logic is sound. ToC is fit for purpose. Use as-is or with minor refinements.
17-21
Adequate
Address flagged dimensions before finalizing. Most likely fix: name assumptions at each link and document risks with likelihood, impact, and mitigation.
11-16
Needs Revision
Substantial revision needed. Mechanisms not articulated, links collapsed. Use the Revise prompt to rebuild link-by-link.
5-10
Substantial Revision
Causal chain is implicit or aspirational. Rebuild from outcomes back to specific mechanisms, then layer assumptions and evidence.