Evidence Standard v1.0

Evidence strength

How well do the study designs, consistency and precision support the claim as stated?

Applicability

Do the populations, settings and time periods match the context in which the claim is used? (Reported as the Transfer Rating, never folded into the grade.)

Source traceability

Can every source be located, verified and read by anyone? Are DOIs and versions recorded?

Human accountability

Which named reviewers assessed the evidence independently, what did they declare, and who approved the grade?

Currency

How recent is the evidence, when was the card last reviewed, and what new evidence would trigger reassessment?

• Grade A · Strong

  Consistent findings from multiple well-conducted studies whose designs can support the claim as stated, with no serious unresolved contrary evidence.

  Criteria

  • Two or more independent studies with designs appropriate to the claim type (e.g. randomised or rigorous quasi-experimental designs for causal claims)
  • Effect direction consistent across studies
  • Sources verified and traceable
  • No unresolved retraction, replication failure or serious conflict of interest

• Grade B · Moderate

  Supportive evidence with important limitations: fewer studies, weaker designs, imprecision or some inconsistency.

  Criteria

  • At least one study with an appropriate design, or several weaker designs pointing the same way
  • Limitations are material but do not reverse the conclusion
  • Sources verified and traceable

• Grade C · Limited

  Evidence is sparse, indirect, low in design quality or materially inconsistent. The claim may be true but the evidence does not establish it.

  Criteria

  • Only observational, small or indirect evidence
  • Or substantial inconsistency between studies
  • Or key sources could not be fully verified

• Grade D · Insufficient or contrary

  No usable evidence for the claim as stated, or the better evidence points the other way.

  Criteria

  • No studies that address the claim
  • Or the strongest evidence contradicts the claim
  • Or the claim rests on retracted or unverifiable sources

• High transfer

  Populations, settings and conditions in the evidence closely match the stated context of use.

• Moderate transfer

  Some differences in population, setting or time that could change the size but probably not the direction of effect.

• Low transfer

  Material differences that could change the direction of effect; local validation advised before relying on it.

• Very low transfer

  The evidence comes from contexts so different that it says little about the stated context.

Grades summarise published evidence available at the review date and can be wrong. They do not assess the quality of a product, organisation or person, and they are not advice. Transfer Ratings depend on the context stated on the card. Grades issued under this version stay attached to it; later versions do not rewrite them.

Initial draft prepared for founder and governance review. Not yet in force: no grades have been issued under it.