Editorial transparency
Sources and verification
The criteria used to select, cross-check and attribute the sources behind our content.
Source hierarchy
We prioritize scientific publications, datasets, mission documentation and original communications from responsible institutions or teams. Secondary sources provide context rather than replace available primary evidence.
Traceability and verification
We check that a source supports its associated claim, retain attribution and link to the origin when public. When sources disagree, we describe the discrepancy and the level of certainty.
How we assess a source
We consider authorship, proximity to the original evidence, publication date, methodology and whether the claim can be reproduced or corroborated. A press release may establish what an organization announced, while measurements, papers and mission archives are preferred for the scientific substance behind that announcement.
The evidence record
A useful reference identifies the author or responsible institution, document title, publication date, stable URL or identifier, and the part of the story it supports. We distinguish a source that proves a measurement from one that merely supplies background, a quotation or an announcement. Access dates and preserved metadata help us reassess the evidence if a page changes or disappears.
A list of links is not verification by itself. Editors must open the source, inspect the relevant material and confirm that the wording, time period, object and degree of certainty match the published claim. Paywalled or unavailable material may be used only when enough bibliographic information and corroboration remain to make the claim auditable.
Independence, conflicts and weak evidence
Institutions, companies and mission teams are authoritative about their own actions, but they may also have promotional, financial or reputational interests. We attribute their statements and avoid presenting them as independent validation. Social posts can document what a first party said or show a time-sensitive event, but they do not replace mission data, technical documentation or peer-reviewed research for scientific conclusions.
Anonymous claims, unattributed compilations and pages that reproduce one another are treated as weak evidence. If only one source exists, that limitation is made explicit. When credible sources disagree, we do not average incompatible values: we explain the methods, dates or definitions that may account for the difference.
Source hierarchy, versions and data reproducibility
For scientific results, the preferred chain runs from data or technical archive to paper, mission documentation and the institution’s explanatory material. For launches and operations, official mission records and operator notices establish events, while independent observation can provide corroboration. Journalism, encyclopedic references and specialist databases are valuable for orientation, but important claims are traced back as close as possible to their origin.
Versions matter: a preprint, accepted paper, catalogue release and corrected dataset may contain different values. We record identifiers, release numbers and observation dates when they affect interpretation. Calculations derived by CosmoInsider must state the inputs and method in plain language. If a reader cannot reproduce an exact result because data or software are unavailable, that limitation is disclosed instead of implying full reproducibility.