A large effect in a weak study and a small effect in a strong one are different findings. Averaging them into a single score destroys the distinction, so this reference refuses to. Every indication carries an evidence grade for how the claim was studied, and an effect signal for what was observed.
The grade is about study design
A is a randomised or controlled human trial. B is preliminary human data — pilot, open-label, case series, observational. C is mammalian in-vivo work. D is mechanistic: in vitro, cell culture, or inference from a pathway. A narrative review never lifts a grade on its own, because a secondary summary is not primary evidence.
The effect signal is about magnitude
Large, Moderate and Small describe what the cited work measured. Mixed records that studies disagree, and No effect records that something was studied and nothing was found. Those last two exist so a negative finding can be recorded.
An effect signal never renders without its grade: D · Large reads as a big effect in cell culture.
What is computed and what is authored
References are typed by study design from PubMed's own metadata; the few it cannot settle fall back to a text reading and are labelled inferred. Each compound carries the three-axis profile — preclinical depth and human evidence computed from those typed references, regulatory standing always authored with its authority cited. Indications show a grade only where their own references are linked; the rest read ungraded rather than borrowing the compound's strongest study.
The full rubric, the effect scale and the audit trail behind the links are set out on the methodology page.