Duties of a Referee
SIMODS seeks to publish papers that are novel, correct, and significant.
The task of a referee is to provide a professional opinion on a paper, including justification for the evaluation. In particular, the reviewer must make a recommendation on whether the paper should be accepted (perhaps after revision) or rejected from the journal.
Furthermore, we require the referee to make an explicit determination (via a menu) of the overall significance of the paper. This assessment balances the depth of the contribution and the breadth of the impact.
A referee is expected to read the paper with sufficient care to be confident that it is mathematically sound. Nevertheless, it is not necessary to check every detail. The author has final responsibility for the content.
The referee must also be candid about his or her level of expertise on the topic of the paper, as well as the comprehensiveness of the review.
In accepting an assignment, a referee commits to return a report within the timeframe specified by the editor.
If a paper is severely flawed, the referee should send a summary report to the editor as soon as possible. This report should explain briefly and concretely why the paper is unlikely to be publishable in SIMODS.
For most papers, SIMODS expects a detailed report within six weeks. If more time is required, the reviewer should communicate with the editor about the timeline for completing the review.
SIMODS always values thoughtful and careful reviews more than rapid reviews. In particular, papers that are long or difficult may merit extra attention.
Content of the Report
The referee should summarize the contributions of the paper and then comment on the major and minor shortcomings, if any. A list of corrections (spelling, grammar, or mathematics) is generally helpful. All criticism should be constructive and free from animus.
In compiling the review, please consider the following questions:
- Who would be interested in the results of this paper?
- Is the paper technically sound?
- Does the paper represent progress?
- Are the results surprising?
- Does the paper provide context for the research, including a treatment of related work?
- Does the discussion illuminate the importance of the results/methods/experiments?
- Is the bibliography complete?
- Is the title apropos?
- Does the abstract give sufficient context and summarize the key contributions?
- Is the writing clear?
- Is the notation clear?
- Is the paper well organized?
- Does the paper include appropriate tables, figures, algorithms, or code?
- Are there any mathematical or technical errors?
- Are theoretical claims properly justified?
- Are algorithm statements sufficiently detailed?
- Are the experimental results complete enough to support the claims?
- Are the experimental results reproducible?
- Are the methods compared to the state of the art?
It is vital to the field, as well as to the reputation of SIAM and its journals, that conflicts of interest (COIs) be avoided. For all assigned papers, editors and referees are required to declare any conflict of interest or even the appearance of one. COIs include handling a paper authored by someone to whom the editor or reviewer has a close relationship (personal, mentoring, research, professional, or financial). COIs also include handling papers on a topic closely related to the editor's or referee's active research for which editorial delay or rejection of the paper could give the appearance of conflict.
Editors will do their utmost to avoid assigning papers that would present a COI, but self-reporting is also mandated due to the difficulty of tracking all affiliations of authors, referees, and editors. If there is a COI, it is SIAM policy that the manuscript be assigned to a different editor or referee. If there is an appearance of a COI but no actual conflict, then the situation should be referred to the editor-in-chief or the SIAM office for a final decision on how to proceed.
All manuscripts are submitted in confidence. Editors and referees are bound to confidentiality about manuscripts that they handle or are asked to handle. Neither the editor nor the reviewer may divulge the contents of a paper unless it is publicly available, e.g., on the arXiv. A reviewer may request the permission of the editor to discuss a paper with others in certain cases, e.g., a professor may ask a PhD student to assist with the review or vice versa. Editors and reviewers must never use the manuscript to advance their own research until the manuscript is made public by the authors.
Timeliness in reviews can also be an ethical matter. Not only does the career advancement of the members of the field depends on timely publication of their results, but the success of the field as a whole can be impacted. We strongly encourage referees to submit their reviews on time or to inform the editor of unavoidable delays.
- A situation in which an individual has competing interests or loyalties. In the context of your referee duties, please notify the editor if any of the authors falls into one of these groups, even if it's just the appearance of a conflict of interest:
- Spouse or family member (lifetime)
- Present or past PhD students and postdocs (lifetime)
- Collaborators within the last 48 months
- Authors from your home institution
- Authors from institutions where you currently seek employment
- Close personal or professional relationships
- Most important: Would a reasonable person with all the relevant facts question your impartiality? Are there any other extenuating circumstances that might interfere with your ability to be impartial? If the answer to either of these is yes, notify the editor.
- S. Krantz, "A Primer of Mathematical Writing," AMS, Providence, 1997.