Peer Review

Every article published in brazilian science communications is peer reviewed, including those commissioned by the editors. Nothing is published on the strength of the author's standing alone.

The model

Review is single-anonymous: reviewers know who the authors are; authors do not know who the reviewers are unless a reviewer chooses to sign.

We do not use double-anonymous review, and the reason is specific to this journal. A synthesis of a field is normally written by someone who has spent years in it, and the manuscript's own reference list makes the authorship obvious. Concealing a name that the text reveals in its third paragraph would be a formality, not a protection.

Reviewers may sign their reports if they wish. Roughly speaking, a signed report carries more weight with authors and costs the reviewer more; we leave the choice to them and never disclose an identity without permission.

Who reviews

Each manuscript goes to at least two reviewers. Where a synthesis crosses fields — and many do — we look for one reviewer from each side rather than two from the centre.

Reviewers are chosen for expertise in the field being synthesized, not for seniority. We avoid recent co-authors, colleagues at the same institution, and anyone in an evident competing or collaborative relationship with the authors. Authors may suggest reviewers and may ask that specific people be excluded; both requests are taken seriously, and neither is binding.

What we ask reviewers to judge

The questions are different from those a research article raises, and we state them explicitly in the invitation:

• Is the claim about the field correct? Not whether each cited study is sound, but whether the synthesis states accurately what the body of work establishes.
• Is anything important missing? A synthesis that omits the work that contradicts it is not a synthesis.
• Is the line between settled and contested drawn in the right place? Overstating consensus and manufacturing controversy are equally serious faults.
• Is the evidence for each claim traceable? A reader must be able to go from a statement to the studies behind it.
• Would this be useful to someone entering the field? That is the readership.

We do not ask reviewers to judge novelty in the sense used for research articles. The originality of a synthesis lies in the formulation, not in the data.

Confidentiality

Manuscripts under review are confidential documents. Reviewers must not share them, use them, or discuss them with third parties, and must not use unpublished material to advance their own work. A reviewer who wishes to consult a colleague should ask the editor first; the colleague is then named in the report and bound by the same terms.

Reviewers who use any artificial-intelligence tool in preparing a report must say so. Uploading a manuscript under review to a public AI service is a breach of confidentiality and is not permitted.

Conflicts of interest

Reviewers must decline when they have a financial, professional or personal interest that a reasonable reader would consider capable of affecting their judgement, and must disclose lesser relationships to the editor rather than deciding alone whether they matter.

Reports

Authors receive the full text of every report. Editors do not summarise, filter or soften them; where a report is discourteous rather than critical, the editor returns it to the reviewer for revision instead of passing it on.

The decision belongs to the editor, not to a vote count. Two positive reports do not compel acceptance and one negative report does not compel rejection — but a decision that goes against the reports is explained to the author.

Recognition

Reviewers who complete a report receive a formal acknowledgement they can cite.