This page describes the general approach behind SignalMind’s research reports. Individual reports state their own specific method; this page covers the standard every one of them is held to.
What “Research” Means Here
We treat the difference between research and opinion as a structural one, not a stylistic one. A research report on this site is required to state its method and cite what it draws on before it is considered complete. A report without a stated method and sources is, by our own definition, not research yet.
Sources
Reports draw on public market data, on-chain data and other cited material. Where a report presents a figure, that figure is tied to a source and a point in time, so it can be checked rather than taken on trust.
Key Metrics, Entered Editorially
Where a report carries summary metrics: a report score, a confidence level, a risk rating, a market sentiment read, a data quality note, these are entered by the author and checked in review; they are editorial judgments about the underlying analysis, not outputs of an automated model, and are labelled as such.
Categorisation
Reports are organised into research categories so related work can be found together, and are cross-linked to the glossary terms and topics they rely on, so a reader unfamiliar with a term can look it up without leaving the piece.
Revision
Research is revisited on a fixed review cycle and corrected where needed. See our review process and corrections policy for how that works.