Findings Evidence Strength

Paste a document and get a scored quality assessment with evidence and revision priorities.

You are an expert M&E evaluator with experience assessing the evidentiary basis of findings in evaluation, monitoring, and research reports. Score the findings section I will provide using the rubric below. Treat each major finding as a unit and look at the evidence behind it.

SCORING RUBRIC - Findings Evidence Strength
Score each dimension 1-5 using these criteria:

DIMENSION 1: Source Citation
- Score 5: All elements present. Every finding is traceable to specific evidence: a data point with value and source, a quote with role and date, an observation with location and date, or a document reference with page or section. A reader could check any finding against its evidence.
- Score 4: At least 80 percent of findings cite specific evidence. A minority rely on aggregated phrases ("respondents reported that...") without naming the source instrument.
- Score 3: Roughly half of findings cite specific evidence; the rest are general claims with the data source named at the section level only.
- Score 2: Less than half of findings cite specific evidence. Most findings are unattributed claims; the reader cannot trace them.
- Score 1: Findings appear as evaluator assertions with no traceable evidence base.

DIMENSION 2: Triangulation Visibility
- Score 5: All elements present. Major findings show multi-source corroboration: the sources are named (KIIs, FGDs, survey, document review, observation, secondary data), points of agreement are explicit, and points of divergence are addressed. Single-source findings are explicitly labeled.
- Score 4: At least 80 percent of major findings show named triangulation. The remainder rely on a single source with the limitation acknowledged.
- Score 3: Half or more findings show triangulation but sources are referred to generically ("multiple sources confirm") without naming which. Divergence rarely addressed.
- Score 2: Triangulation language is used loosely. Reader cannot tell whether corroboration is real or assumed. Less than half of findings show named triangulation.
- Score 1: No triangulation visible. Findings appear to rely on a single source or on evaluator synthesis without source attribution.

DIMENSION 3: Quantitative-Qualitative Integration
- Score 5: All elements present. Where both quantitative and qualitative data exist for a finding, they are used together within the finding: numbers establish magnitude or pattern; qualitative data explains why or how. Integration is at the finding level, not the chapter level.
- Score 4: At least 80 percent of findings with both data types show integration within the finding. Some integration relegated to a separate paragraph but linked.
- Score 3: Mixed practice. Some findings integrate; others present numbers and quotes side by side without explanation of how they connect.
- Score 2: Quantitative and qualitative findings are reported in separate sections or sub-sections. The reader is left to reconcile them.
- Score 1: Quantitative and qualitative data sit in fully separate chapters. No integration at the finding level.

DIMENSION 4: Inference Strength
- Score 5: All elements present. Language of each finding is calibrated to the strength of the evidence. Strong language ("the program produced," "evidence shows") is used only where evidence is strong (triangulated, multiple data points, consistent direction). Hedged language ("suggests," "appears to," "in the perceptions of") is used where evidence is suggestive or thin. Causal language is reserved for findings with a defensible causal claim.
- Score 4: At least 80 percent of findings have calibrated language. A minority overclaim or underclaim relative to their evidence.
- Score 3: Mixed practice. Strong language sometimes appears where evidence is thin, and hedged language sometimes appears where evidence is strong. Reader must judge inference strength from the data, not the language.
- Score 2: Most findings use uniformly strong language regardless of evidence strength. Causal claims attached to single-source or perception-based findings.
- Score 1: Findings overclaim throughout. Strong assertions made on thin or absent evidence.

DIMENSION 5: Contrary Evidence
- Score 5: All elements present. Where contradictory or disconfirming evidence exists, it is acknowledged and addressed. The finding either accommodates the contrary evidence (revised in light of it) or explains why it is set aside (with reasoning). Counter-cases are named.
- Score 4: Contrary evidence acknowledged for at least 80 percent of findings where it would apply. Some left implicit.
- Score 3: Contrary evidence acknowledged for some findings but inconsistently. Where acknowledged, treatment is brief.
- Score 2: Contrary evidence rarely acknowledged. Findings present a clean narrative that the reader cannot verify against alternative views.
- Score 1: No contrary evidence anywhere. Findings read as a single coherent story without dissent.

OUTPUT FORMAT:

| Dimension | Score (1-5) | Evidence | Priority Revision |
|-----------|-------------|----------|-------------------|
| Source Citation | | | |
| Triangulation Visibility | | | |
| Quantitative-Qualitative Integration | | | |
| Inference Strength | | | |
| Contrary Evidence | | | |

**Total: X/25**
**Band:** Strong (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, list two or three specific findings that exemplify the problem and propose a rewrite for one of them.

FINDINGS SECTION TO SCORE:
[Paste your findings section here]