IEEE Data Descriptions welcomes contributions from all areas of science and data-focused social science without constraints on scale or specificity, provided the dataset is valuable to researchers. All submissions must be technically accurate and scientifically credible, offer a novel contribution, and properly cite pertinent previous work to support facts and assertions.

Descriptor articles are detailed accounts of open research datasets designed to facilitate reuse without validating hypotheses or conclusions. Before reviewing a submitted manuscript, all reviewers should read the Author Guide to better understand the types of articles submitted.

Reviewers may request further details or analyses to justify or explain better the data being described. Occasionally, a dataset might be straightforward enough not to need a descriptor, but publishing a Descriptor article is advisable if revealing methodologies or data specifics offers any benefit. Authors must give reviewers full access to all data as described and ensure it is publicly available in a reputable data repository.

Submissions should be evaluated based on the dataset's integrity and utility rather than the impact or novelty of associated findings. Descriptor articles should focus on the dataset's creation, underline the rigour of data production, and provide necessary details for data interpretation and reuse. Preliminary or unsupported findings should not be published.

Reviewers evaluate submissions based on data quality and reusability and may suggest modifications or additional analyses. The peer review focuses on the dataset methodological rigour, data quality, completeness of description, and data integrity.

Reviewers are selected based on expertise and experience. The review should guide authors in strengthening their manuscripts to meet publication standards, data quality, and completeness. Reviewers should respect the confidentiality of the review process and respond promptly to review requests. IEEE Data Descriptions values rapid decision-making and publication to benefit both authors and the scientific community.

When crafting a review, reviewers should provide the editor with essential insights for decision-making and offer guidance to authors on how to enhance their manuscript for potential publication. Reviewers are encouraged to suggest additional detailed descriptions or analyses that substantiate the data but should refrain from demanding extensive experimentation or confirmation of specific hypotheses beyond the Descriptor/Collection/Meta article scope. Instead, the focus should remain on the data quality and applicability, not on interpreting its implications.

The review should assess the data quality and methodological soundness, the thoroughness of the data's production, the robustness of technical validations, and whether the data's depth and coverage meet the research needs outlined. Additionally, reviewers must check if the data files are complete and appropriately stored in a suitable data repository. When submitting a review, the following questions will need to be answered, and details about insufficiency need to be provided:

1. Is the data generated using robust methods?
2. Is the technical quality convincingly supported?
3. Is the data's depth and coverage adequate?
4. Are any data-processing details sufficient for replication?
5. Is all necessary information provided for data reuse?
6. Does it adhere to the journal reporting standards?
7. Is the chosen repository suitable?
8. Are the deposited data files complete?

The reviewer can make 1 of 4 recommendations: Accept, no revisions; Accept, minor revisions (2 weeks to resubmit); Major revisions needed (4 weeks to resubmit); or Reject.

Our standard practice is to exclude reviewers whom the authors have specifically ruled out for any reason. We also generally avoid referees who have recent or ongoing connections with the authors, have previously reviewed manuscript drafts, compete to publish similar findings, have past conflicts with the authors, or have a financial stake in the research outcomes. However, biases are not always apparent, so we rely on referees to inform us of potential conflicts and abstain from reviewing if they cannot remain impartial.

IEEE strictly prohibits citation manipulation, including citation stuffing, in all its publications, such as IEEE Data Descriptions. Citation stuffing involves the inclusion of irrelevant or excessive references with the intent to artificially inflate citation metrics. This unethical practice can manifest in various ways, such as authors citing their own unrelated work, reviewers requesting unnecessary citations, or editors mandating references to boost a journal’s metrics. IEEE considers such actions a serious breach of publishing ethics and relies on editors to identify and prevent citation manipulation during the peer review process. Any instances of citation manipulation should be reported to the Editor-in-Chief for appropriate action, which may include investigations or warnings to the involved parties.