How We Rate Evidence
CAB Direct uses a simple four-level evidence label to show readers how much confidence the available research can support. The label is a concise editorial summary, not a numerical score and not a substitute for reading the study details.
The four evidence levels
Level | Typical evidence pattern | How to read it |
|---|---|---|
Strong | Systematic reviews and several randomised controlled trials point in the same direction. | The result is supported by multiple higher-quality sources with consistent findings. |
Moderate | One large randomised controlled trial or consistent cohort studies support the finding. | There is meaningful evidence, but the body of research is not as broad or as consistently high-level as for a strong rating. |
Limited | Evidence comes mainly from small studies, animal studies, in-vitro work or similarly restricted research. | The finding may be useful for forming a hypothesis, but it should not be treated as established for broader populations. |
Conflicting | Good-quality studies reach materially different conclusions. | The disagreement is part of the evidence and should be described rather than averaged away. |
What affects the rating
The rating is based on the type of studies available, the consistency of their findings and how directly they apply to the question being discussed. The same study design can support different conclusions depending on what was measured, who was studied and whether later research agrees.
- Study design: systematic reviews and randomised trials generally support stronger conclusions than uncontrolled or laboratory-only evidence.
- Consistency: agreement across independent studies increases confidence; material disagreement produces a conflicting rating.
- Applicability: evidence in animals, cells or a narrow population is not automatically generalised to people or to different settings.
- Effect size and context: relative changes should be read alongside absolute numbers and the conditions under which they were measured.
Health and nutrition
The evidence-level field is required for Health & Medicine and Food & Nutrition articles. The author gives a short justification in the article, and a medical reviewer checks health and nutrition content before publication. Animal and in-vitro results must remain clearly identified as such.
Ratings can change
An evidence label reflects the body of research available at the time of review. New trials, systematic reviews or corrected data can change the balance of evidence. When a substantive update changes the interpretation of an article, CAB Direct records the correction or update rather than silently replacing the earlier claim.
For the wider sourcing and review rules, see /editorial-policy/. For visible changes to published material, see /corrections/.