io-submission

Submission Preflight (io-submission)

The last check before pressing submit on Editorial Manager. IO is double-blind, so the single most common avoidable failure is a manuscript that reveals the author through first-person self-references. Verify volatile specifics on the official page before relying on them.

When to trigger

• "Submitting tomorrow" — last pass before upload
• Unsure which files/metadata Editorial Manager expects
• Confirming the chosen article type's cap is met and the manuscript is properly anonymized

Process facts (verify volatile items on the official page)

• Owner / publisher: IO Foundation / Cambridge University Press.
• Portal: Editorial Manager (confirm the current URL).
• Review model: double-blind — anonymize the manuscript; provide a separate title page.
• Article types: Research Article (≤ 14,000 words), Research Note (≤ 8,000), Essay (≤ 10,000). Word limits include tables, figures, and notes; exclude the bibliography.
• Abstract / word count / acknowledgments: submitted separately from the manuscript (abstract word cap not numerically confirmed — 待核实).
• Self-citation: after the title page, omit identity-revealing self-references; cite your own work in the third person.
• Footnotes / style: accepted manuscripts use short author-date footnotes; consistent style.
• Supplementary material: should rarely exceed twenty pages, especially at initial submission.
• ORCID: required for the corresponding author.
• Data: none at initial submission; requested at conditional acceptance; verified by IO staff before final acceptance; Data Availability Statement before the reference list.
• Fee: no submission fee stated (待核实; confirm any post-acceptance OA APC).

Preflight checklist

Article type & length

• Type chosen and its word cap met (Article ≤ 14,000 / Note ≤ 8,000 / Essay ≤ 10,000)
• Word count includes tables/figures/notes; excludes bibliography — recomputed on that basis
• Abstract, word count, and acknowledgments prepared as separate items

Anonymity (double-blind)

• No author names, affiliations, or acknowledgments inside the manuscript
• Third-person self-citation after the title page; no "as we showed in…"
• Identifying file metadata stripped (document properties, comments)
• Anonymized pre-analysis plan / OSF link if preregistration is referenced
• Separate (non-anonymous) title page prepared as a distinct file

Format & metadata

• Short author-date footnotes; consistent citation style
• Corresponding-author ORCID ready
• Data Availability Statement placed before the reference list
• Supplementary material within the ~20-page norm; figures/tables self-contained (see io-tables-figures)

Compliance & verification-readiness

• Ethics / IRB / human-subjects compliance where applicable
• Original work; preprint status checked against Cambridge preprint policy
• Reproducibility package + formal-proof appendix staged so IO staff can verify after conditional acceptance (see io-transparency-and-data-policy)

Avoidable-rejection triage (catch these before upload)

Some failures are administrative, not intellectual, and IO's double-blind, word-counted process punishes them disproportionately. Rank your final pass by severity:

Severity	Failure	Where it bites
Fatal	author identity leaks (first-person self-cite, acknowledgments, metadata)	double-blind integrity
Fatal	word count omits tables/figures/notes; paper is actually over cap	desk return for length
Medium	no Data Availability Statement before the references	transparency-policy gap

The two fatal rows are the most common avoidable IO rejections; both are caught by a clean-room read of the anonymized PDF and a recomputed word count.

Worked preflight vignette (illustrative)

A treaty-compliance study is "ready." The recompute: body 12,400 words, but two dyadic tables and three figures add ~1,900 word-equivalents and footnotes another ~400, so the true count nears 14,700 — over the 14,000 Research Article cap (figures illustrative; recount on IO's stated basis). The fix is not prose-trimming alone: move one robustness table and the balance table into supplementary material (within the ~20-page norm). The clean-room pass then catches "as we showed in Author 2021" — a first-person self-reference — rewritten in the third person. Only then does it go to Editorial Manager. Confirm every numeric cap and the current portal against the journal's current submission guidelines before relying on them.

Anti-patterns

• First-person self-references or author identifiers left in the text/metadata (breaks double-blind)
• Counting words without tables/figures/notes (IO includes them; bibliography is excluded)
• Putting the abstract/acknowledgments inside the manuscript instead of submitting them separately
• Sending a long paper into the Research Note (≤ 8,000) type
• Supplementary material far over ~20 pages at initial submission
• Uploading before a clean-room read of the anonymized PDF (the cheapest catch for the fatal failures)

Output format

【Type】Research Article / Research Note / Essay (cap met? Y/N)
【Anonymized】third-person self-cites + metadata clean? [Y/N]
【Abstract/word-count/acknowledgments separate】[Y/N]
【Word count basis】incl. tables/figures/notes, excl. bibliography? [Y/N]
【Footnotes + ORCID + DAS】[Y/N]
【Repro + proofs staged for verification】[Y/N]
【Next】await decision → io-rebuttal on R&R

Supplementary resources

• ../../resources/external_tools.md — reference managers, anonymization, repro tooling
• ../../resources/official-source-map.md — official IO URLs behind every fact in this pack