Do not search for the authors, manuscript title, preprint, repository record, or related public material for the purpose of discovering identity. Routine literature searching and legitimate source or integrity checks may be conducted only as necessary for scholarly evaluation.

If access to a suspected preprint or another public version appears necessary to assess duplicate publication, undisclosed overlap, data provenance, or another integrity concern, contact the handling editor rather than contacting the authors directly.

If identity becomes known and creates or may create a conflict of interest, prior relationship, risk of bias, or inability to remain impartial, notify the handling editor promptly. Recognition alone does not automatically invalidate the review; the editor determines whether reassignment or recusal is required.

The existence of an appropriately disclosed preprint is not, by itself, a reason to recommend rejection. Evaluate the manuscript provided by the journal and report any concern about undisclosed prior dissemination, material overlap, or version inconsistency confidentially to the editor.

The invited reviewer must personally assess the manuscript and remains accountable for every comment, criticism, and recommendation submitted to the journal.

A reviewer must not ask a student, colleague, assistant, or other person to read the manuscript or prepare the review without the editor’s prior written permission. Any approved contributor must be identified to the editor and must comply with the same confidentiality and conflict-of-interest requirements.

Manuscripts, figures, data, supplementary files, review forms, and editorial correspondence must not be uploaded to public or external generative AI systems. Such use may compromise confidentiality, intellectual property, and data protection.

Generative AI must not be used to evaluate or summarise scientific content, generate substantive criticism, draft or rewrite a review, translate confidential review content, determine a recommendation, or replace the reviewer's own expertise. Conventional local non-generative spelling or grammar checking may be used only when no confidential manuscript or review content is transmitted, retained, or exposed to an external service. The reviewer remains fully responsible for every submitted comment. The complete requirements are governed by the journal's Generative AI and AI-Assisted Technologies Policy.

Briefly state the manuscript's purpose, contribution, and overall quality.

Identify the manuscript's strongest theoretical, methodological, technical, or practical features.

Explain substantive issues that affect validity, interpretation, ethics, or contribution.

Recommend clear revisions and distinguish essential changes from optional improvements.

A substantially revised manuscript is normally returned to the original reviewers when they remain available and have no relevant conflict of interest.

A minor revision may be assessed by the handling editor or returned to one or more original reviewers when specialist confirmation is required.

Reviewers should assess the revised manuscript together with the authors’ point-by-point response and determine whether each major concern has been addressed adequately.

Reviewers should avoid unnecessary new requirements but may identify substantive concerns that emerge from the revision. Further revision or re-review may be requested, and no number of revision rounds guarantees acceptance.

Reviewers who are unavailable or can no longer provide an independent assessment should notify the editor promptly. Another qualified reviewer may be appointed when necessary.

Reviewer assessments remain advisory. The handling editor evaluates the reports, the revised manuscript, and the authors’ responses before making or recommending the editorial decision.