Appendix A
Appendix A — Evidence Register
Keep the source and the claim together
Use with: The research sections throughout the book and Reference Chapters R1, R2, R4, and R5.
Create one entry for a claim or finding you intend to use. A source may support several entries, but a broad book title or a folder of papers should not stand in for the evidence behind a specific sentence.
Keep the source, setting, study design, finding, limitation, application, and review status connected. Check product-specific evidence separately.
A1. The evidence entry
Field | What to record |
|---|---|
Question and proposed claim | The question being addressed and the exact statement you propose to use. |
Source and location | Author or organization, title, publication date, edition where relevant, stable identifier, and the passage supporting the claim. |
Access and review | What was actually read: full text, selected sections, abstract, or another source’s account. Name the reviewer and date. |
Evidence type | Theory, professional guidance, qualitative study, observational study, trial, review, local observation, or product documentation. |
Population and setting | Who was included, where the work occurred, and relevant exclusions or differences from your audience. |
What was examined | The actual intervention or exposure, comparator, outcomes, and follow-up period. For guidance, record its scope instead. |
Finding and uncertainty | What the source found, with the relevant estimate or qualification. Include missing follow-up and conflicting findings when they affect interpretation. |
Evidence judgment | Any grade reported, the grading system, and the conclusion it applies to. Do not invent a grade for an ungraded source. |
What it does not establish | The tempting conclusion the evidence cannot support—for example, attributing a service result to software alone. |
Proposed local application | What you might change, what resources it requires, and what must be tested locally. |
Product-specific evidence? | Yes, no, or not established; identify the exact product, version, task, and evaluation when applicable. |
Approved use and maintenance | Where the claim may appear, wording approved by the relevant reviewer, conflicts still open, and triggers for revision or withdrawal. |
Keep the behavioral review’s evidence judgments separate from other systems such as GRADE. R1 and R8 explain the differences. A theory, a trial result, and a documented feature answer different questions.
A2. Check the transfer into practice
Complete three sentences before writing a benefit claim:
The source establishes: [finding within its actual scope].
Our application proposes: [the design or service change].
Before claiming that it helped here, we need: [the appropriate local evidence].
If sources disagree, retain the disagreement and refer it to the appropriate reviewer. Do not let a more attractive summary silently become the source of truth.
Decision this tool supports: Use the claim as written, narrow it, seek better evidence, or leave it out.