Ethical Standards

Scholarly publishing rests on the assumption that a reader can trust what a paper claims and can tell whose work it reports. These standards describe the conduct that assumption requires.

They apply at every stage — inquiry, submission, review, revision, acceptance, and publication — and they continue to apply to work already published. Section 2 binds everyone. Sections 3 to 6 set out what each role must do.

College Mathematics publishes advanced high school students, undergraduates, and educators alongside established mathematicians, and many of our authors are writing for publication for the first time. Where a lapse is minor and unintentional, the editorial office will normally resolve it through correction and guidance rather than sanction. This page exists so that expectations are clear before a manuscript is written.

Copyright ownership, licensing, and permission to reuse material are governed separately by the Copyright & Licensing page. Where a matter touches both, both apply. These standards are reviewed periodically; the date above shows the most recent revision.

Find what applies to you

Authors, referees, editors, and AIAS are each bound by the following.

• Be truthful about the work. Nothing may be fabricated, falsified, or presented so as to misrepresent who did what.
• Treat unpublished manuscripts as confidential. Their content and their existence alike are not to be shared, posted, or discussed outside the editorial process.
• Do not submit a manuscript under review to a public AI service. Doing so discloses unpublished work to a third party.
• Disclose any interest that could appear to sway your judgment, and step aside where it is serious. Disclosure resolves most such situations; concealment turns a manageable one into a grave matter.
• Judge the mathematics, not the person. Nationality, institution, gender, race, religion, political belief, and career stage have no bearing on any decision.
• Report a significant error or suspected breach promptly, whether it is yours or another’s, and cooperate with any inquiry under section 8.

By submitting a manuscript, every named author affirms that:

• the work is their own, and every source relied upon is cited — including preprints, lecture notes, online postings, software, and unpublished communications;
• it is not under consideration elsewhere and has not been published in substantially the same form;
• permission has been obtained for any material owned by others, as set out in the Copyright & Licensing page;
• every author has seen and agreed to the version submitted; and
• any AI assistance and any competing interest has been disclosed.

Where a manuscript develops the author’s own earlier writing — a thesis chapter, preprint, competition report, or course project — cite it and explain the relationship in the cover letter. Prior circulation of a preprint does not disqualify a submission.

Authorship

Authorship requires a substantive intellectual contribution. A named author must have done at least one of the following:

• conceived the problem or the approach taken to it;
• developed or verified the mathematical argument;
• performed or validated the computations the paper reports; or
• written or substantially revised the exposition.

Every named author should be able to explain the paper as a whole and to say which parts are theirs. Help that falls short of that belongs in the acknowledgments, with the person’s consent.

Naming someone who did not contribute, and omitting someone who did, are misconduct in either direction.

Where work is mentored or supervised, settle the mentor’s role openly before submission rather than assuming it. A mentor who shaped the mathematics is an author; one who advised on presentation is acknowledged. The editorial office will discuss the question in confidence with either party.

After submission, any change to the author list requires all three of:

• a request from the corresponding author;
• a stated reason for the change; and
• the written agreement of everyone affected, including any person being added or removed.

The Editor-in-Chief decides. After acceptance the list is final except by published correction.

Plagiarism and duplicate publication

Plagiarism is presenting another person’s words, arguments, proofs, figures, data, or ideas so that a reader may believe they originated with the author. It takes three common forms:

• Verbatim copying — text, figures, tables, or code reproduced without quotation or credit.
• Close paraphrase — a source followed so closely that only the surface wording differs.
• Omitted citation — a result or technique used without pointing to where it came from. This is the most common form, and the hardest to judge.

Freely available material — a course page, a forum answer, a preprint, a piece of published software — requires attribution exactly as a printed journal does.

In mathematics the line takes judgment. An argument belonging to the common vocabulary of a field needs no particular source; a technique a reader would recognize as belonging to a specific paper should be cited even if you reconstructed it independently. Where the two are hard to tell apart, cite.

Most disputes reaching the editorial office concern a missing citation rather than copied text, and most are oversights. An editor weighs three questions:

• Would a reader come away crediting the author with someone else’s result?
• Does that result carry the weight of the paper’s contribution?
• Was the author placed to know of the source?

Where the answer to all three is yes, the omission is a serious matter. Short of that, it is a defect to repair in revision, not a finding of misconduct.

Duplicate publication is the reappearance of your own published work, in whole or in substantial part, without saying so.

Submissions may be compared against the literature at any time, before or after publication. Where a concern arises, the corresponding author receives it in writing and a reasonable period to reply before any decision is taken.

Artificial intelligence tools

We do not prohibit AI tools. We require disclosure, and we require that responsibility for the work stay with its human authors. How much you must say depends on what the tool did:

Where the manuscript must describe the use, that description should:

• identify the tools and versions used;
• describe the prompts or lines of inquiry pursued; and
• explain how the authors verified the output.

See Human Orchestration on the For Authors page. Two further rules apply throughout:

• An AI system cannot be named as an author: it cannot vouch for the work, answer questions about it, or correct it.
• A fabricated citation, theorem, or figure is the authors’ error, not the tool’s.

Disclosure is never held against a submission. Method honestly reported is itself a contribution.

Data and computations

• Retain the notes, code, and datasets supporting published claims, ordinarily for at least five years.
• Supply them, or access to them, on reasonable request from the editorial office.
• Describe computations well enough to be repeated — software, version, parameters — and deposit code publicly where feasible.
• State the provenance of any dataset you did not generate and confirm its terms permit your use.

Where a result rests on a computation a referee cannot reasonably repeat, we may ask for the supporting materials as a condition of review.

During review

• Do not attempt to identify or contact referees. Concerns about a review go to the editorial office, which will address them.
• Suggesting referees, or asking that someone be excluded for a stated reason, is welcome — though the editor is not bound by either.
• A review is not a favor to be traded or bought. Any attempt to influence one through payment, pressure, or personal appeal ends consideration of the manuscript.

Review is single-blind: you are told who the authors are, and they are not told who you are.

• Return the review by the agreed date, or tell the editor promptly that you cannot.
• Assess the mathematics and be specific, courteous, and useful, distinguishing a defect in the argument from a matter of taste.
• Decline, or disclose and let the editor decide, where you:
  • have recently collaborated or co-authored with an author;
  • share an institution or department with an author;
  • supervised an author, or were supervised by one;
  • have a personal relationship with an author; or
  • are pursuing work that competes with the manuscript.
• Do not retain or copy the manuscript once the review is done, and do not circulate, post, or upload it anywhere.
• Do not use what you learn from an unpublished manuscript to advance your own work, and do not delay a review to gain an advantage.
• Tell the editor if the manuscript closely resembles work you know to be published or under review.

Obtain the editor’s agreement before consulting a colleague or involving a student in a review. Anyone so involved is bound by the same confidentiality and should be named to the editor.

• Decide on merit alone — correctness, significance, clarity, and fit with the journal’s scope.
• Hold invited contributions, special collections, and submissions from board members to the same standard as any other manuscript.
• Take no part in handling a submission you have authored or hold a conflict in. It passes to another editor, and you receive no access to the reports or the decision.
• Neither use nor disclose the content of an unpublished manuscript for any purpose of your own.
• Give reasons for a decision sufficient for the author to understand it.
• Give an author a fair opportunity to answer any complaint made about their work.
• Act on a well-founded concern whether it arrives before or after publication.

These standards bind the journal as well as those who submit to it. AIAS undertakes to:

• maintain and publish these standards, and review them periodically;
• keep editorial decisions with the editors, so that no donor, sponsor, partner, or institution influences whether a manuscript is accepted;
• support editors and referees facing an ethical question, and bear the cost of investigating one properly;
• investigate allegations against referees and editors on the same terms as those against authors; and
• keep published articles openly accessible and preserve the record, including every notice of correction, retraction, and concern.

AIAS is an independent nonprofit publisher. College Mathematics charges no submission fee and no article processing charge, and no payment of any kind bears on an editorial decision.

Retraction corrects the record; it is not a punishment. A retraction the authors themselves request on discovering their own error is responsible conduct, and the notice will say so.

Silent alteration of a published article is not permitted. Typesetting, metadata, and formatting may be changed without notice; anything affecting content carries one.

Your message should identify:

• the article or manuscript concerned;
• the nature of the concern;
• the specific passages, results, or figures involved;
• the prior work or source relied upon, with citations or links; and
• whether you are willing to be identified to the authors.

Concerns are examined on their substance rather than on who raises them. Anonymous reports are accepted, though one offering no verifiable detail may be impossible to act upon.

How a concern is handled

• Acknowledgment. The editorial office confirms receipt and records the concern.
• Initial assessment. The Editor-in-Chief considers whether it falls within these standards and has an evidential basis. If not, the matter is closed and the reporter told why.
• The authors’ response. Otherwise the corresponding author receives the substance of the concern and a reasonable period to reply. Authors are never asked to answer an allegation they have not been shown.
• Independent assessment. Where specialist judgment is needed, the Editor-in-Chief consults mathematicians unconnected to any party.
• Determination. The possible outcomes are:
  • no further action;
  • a published correction;
  • a required revision before consideration continues;
  • rejection, or withdrawal from consideration; or
  • retraction, or an expression of concern.
• Communication. The outcome goes to the authors and, where identifiable, to the person who raised the concern.

AIAS may contact an author’s institution, funder, or another publisher where a concern is serious and cannot be resolved through the journal alone, and will cooperate with an inquiry properly brought by another journal, publisher, or institution.

Career stage is no defense to fabrication, but it bears on how an inadvertent lapse is handled. Where a student author has erred through inexperience, the editorial office will normally address it instructively, and a concern resolved that way leaves no lasting mark. Records are retained by the editorial office, and a concern is examined however long ago the work was published.

Questions about these standards, requests for guidance before submission, and reports of a possible breach may be directed to:

An author unsure whether something must be disclosed, cited, or acknowledged is encouraged to write before submitting. A question asked in advance is always easier to settle than a concern raised afterwards.