Appendix A — Templates and evaluation rubric
A.1 Claim labels
- Supported: evidence supports the claim within its stated scope.
- Incorrect: evidence or a reproducible check contradicts the claim.
- Unsupported: necessary evidence is absent.
- Uncertain: available evidence leaves meaningful ambiguity.
A.2 Review quality rubric
Score each dimension from 0 to 3. Use these scores for feedback, not as a validated certification threshold.
| Dimension | What to inspect |
|---|---|
| Question and scope | Identifies the question and appropriate limits |
| Data and design | Checks provenance, experimental units, and assumptions |
| Verification | Provides traceable checks and consistent outputs |
| Interpretation | Connects claims to evidence and acknowledges uncertainty |
| Feedback | Explains consequences and gives actionable corrections |
| Reproducibility | Records inputs, code, environment, and outputs |
0: missing or materially incorrect. 1: identifies a concern with little support. 2: mostly correct with relevant evidence but incomplete checks or explanation. 3: justified, traceable, and clear about limitations.
A total out of 18 may summarize performance, but always report dimension-level feedback. A high total cannot compensate for a major uncorrected scientific error.
A.3 Issue severity
Record separately: critical (undermines the principal conclusion), major (materially changes a result or claim), minor (limited consequence), or not applicable. Justify the level in context.
A.4 Templates
The project includes an evaluation report, evidence log, reproducibility checklist, and case authoring specifications.