Review criteria

Significance and originality#

Does the manuscript add something the field did not already have — a new method, a result that resolves a disagreement, a dataset or system that others can build on? Significance does not require novelty on every axis at once; a well-executed replication in a new setting, or a negative result that closes off a line of enquiry, can be significant too.

Meets: the introduction states plainly what gap the work fills and the results close that gap. Does not meet: the contribution is an incremental variation on published work with no stated reason the variation matters, or the "gap" turns out to already be filled by literature the authors did not cite.

Soundness of method#

Is the method appropriate to the question being asked, and is it applied correctly? This covers experimental design, simulation setup, instrumentation, sampling, and any assumptions the authors rely on to get from raw measurement to reported result.

Meets: the method is standard for the field or, where it departs from standard practice, the departure is justified and its effect on the result is discussed. Does not meet: a control is missing that the field would expect, an assumption is stated but never checked, or the method described could not actually produce the data shown.

Validity of the results#

Do the numbers, figures and observations reported actually follow from the method described, and do the authors' conclusions follow from those results? This is where you check that error bars, confidence intervals or uncertainty estimates are present and reasonable, and that a stated significance is not being read into noise.

Meets: every claim in the discussion can be traced to a specific figure, table or measurement in the results. Does not meet: the discussion asserts something the results do not show, uncertainty is unstated where the field would expect it, or an outlier is excluded without explanation.

Adequacy of the evaluation#

Is the evaluation thorough enough to support the paper's claims at the scope the authors claim them — one test case versus a representative range, a single operating condition versus the envelope the system is meant to work in, one dataset versus several?

Meets: the evaluation covers the conditions under which the authors say the result holds, including at least one comparison against an existing method or baseline where one exists. Does not meet: the paper claims general applicability but tests one narrow case, or omits an obvious baseline without saying why.

Clarity of presentation#

Can a reader in the field follow the argument, understand the figures without hunting through the text for a missing axis label, and locate the information needed to judge the work? Clarity is a reviewing criterion, not just a copyediting concern, because a paper that cannot be followed cannot be properly evaluated either.

Meets: figures are legible and self-explanatory with their captions, the structure follows the standard shape for the field, and technical terms are defined on first use. Does not meet: a reader has to reverse-engineer what a figure shows, or the results section reports numbers the methods section never set up.

Reproducibility#

Could someone else, given the paper, repeat the work and expect to get a comparable result? This means the parameters, materials, code or data needed to redo the analysis are either included, deposited somewhere citable, or described in enough detail that omission is not the only barrier.

Meets: the manuscript states its key parameters and either shares data/code or explains a specific, defensible reason it cannot. Does not meet: a result depends on an undisclosed parameter, proprietary step or hand-tuned value that the paper does not name.

Completeness of the literature#

Does the manuscript situate itself against the work that already exists, including work that disagrees with it? A literature review that only cites work supporting the authors' position is incomplete even if every citation in it is accurate.

Meets: the related work section engages with the closest prior results, including any that conflict, and says how this work differs. Does not meet: a directly relevant prior study is absent, or is cited only in passing without engagement.

Ethics#

Where the work involves human or animal subjects, hazardous materials, safety-critical systems, or data with privacy implications, were the appropriate approvals obtained and reported, and are risks to people or the environment handled responsibly?

Meets: approvals, consent procedures or safety measures are stated where the work requires them. Does not meet: the work clearly required an approval or disclosure that is absent — flag this to the editor regardless of your recommendation on the other criteria, since it can be grounds for decline on its own.