Annals of Statistics Submission Guide
A source-checked Annals of Statistics guide for importance, originality, proof and evidence structure, supplementary material, and submission readiness.
Readiness scan
Find out if this manuscript is ready to submit.
Run the Free Readiness Scan before you submit. Catch the issues editors reject on first read.
How to approach The Annals of Statistics
Use the submission guide like a working checklist. The goal is to make fit, package completeness, and cover-letter framing obvious before you open the portal.
Stage | What to check |
|---|---|
1. Scope | Test importance and originality |
2. Package | Prepare the manuscript and supplement |
3. Cover letter | Submit through the current IMS route |
4. Final check | Pass suitability screening |
Quick answer: Submit to the Annals of Statistics when a result is both statistically original and important: it changes theory, methodology, or understanding in a way that matters beyond one proof technique or application. Formal sophistication helps only when the reader can see the problem it resolves and the new statistical capability it creates.
Use this Annals of Statistics submission guide to test the contribution, proof architecture, empirical or computational evidence, supplementary boundary, and final PDF before entering the Institute of Mathematical Statistics submission system.
The Institute of Mathematical Statistics publishes the Annals of Statistics. Its importance-and-originality standard makes this a different decision from selecting an applied outlet or a journal that primarily rewards a domain result.
The journal describes its aim as publishing the highest-quality work across contemporary statistics and explicitly emphasizes importance and originality rather than formalism. That makes the first editorial question more demanding than “Is the theorem correct?” The paper must explain why the result changes what statisticians can know, prove, estimate, or do.
Evidence basis: We checked the current Annals of Statistics manuscript-preparation instructions, manuscript-handling procedures, and journal page on August 19, 2026. IMS facts are sourced; the readiness framework below is Manusights editorial judgment.
Source limitation: IMS publishes process and preparation guidance but not a case-level account of editorial screening. This guide cannot predict an editorial decision, acceptance, or timing for a particular manuscript.
Stop signal: Pause if the paper's importance cannot be explained without invoking formal difficulty, or if the central proof logic is intelligible only after opening the supplement.
From our manuscript review practice
For the Annals of Statistics, formal difficulty is not the contribution. The main paper must make the statistical problem, new principle, proof strategy, and practical consequence legible as one argument.
Annals decision at a glance
Gate | Evidence to inspect | Hold when |
|---|---|---|
Statistical importance | The paper resolves a consequential limitation, impossibility, or open question | The result is technically neat but changes little outside a narrow construction |
Original principle | A new idea, boundary, or method drives the result | Novelty is mostly a stronger assumption, different notation, or another special case |
Proof architecture | The main result, intuition, dependency graph, and hard step are visible | Readers must traverse pages of notation before learning what is new |
Evidence discipline | Simulations, examples, or applications test the claims and boundaries | Experiments only illustrate favorable cases or substitute for theory |
Main-paper completeness | The central argument stands without the supplement | Essential definitions, assumptions, or failure cases are exiled to an appendix |
State the contribution before the notation
A strong opening lets a statistician answer four questions before the technical machinery begins:
- What statistical problem is unresolved or inadequately understood?
- Why do existing approaches fail, and under which assumptions?
- What new principle or construction resolves the problem?
- What becomes possible, sharper, or better understood as a result?
Do not use the abstract to list every theorem. Name the main intellectual move, the conditions under which it works, and the consequence a reader should care about. Later results can then be organized as support rather than presented as competing headlines.
Build a proof map the reader can inspect
Layer | What belongs in the main paper | Common failure |
|---|---|---|
Definitions | Only the objects needed to understand the main result and its boundary | Dense notation introduced before the reader sees the problem |
Assumptions | Plain-language role plus formal statement | Conditions appear technical but quietly create the conclusion |
Main theorem | Exact claim, scope, and comparison with prior results | Several variants obscure which result carries the contribution |
Proof strategy | Dependency map and intuition for the difficult step | “By standard arguments” hides the part reviewers need to evaluate |
Consequence | Statistical meaning, usable implication, or new boundary | The discussion repeats the theorem without explaining what changes |
Move routine algebra and secondary variants to supplementary material when appropriate, but keep the logic of the central proof visible. A supplement is not a place to hide the assumption that makes the headline possible.
Make simulations falsify, not decorate
Computational evidence is most useful when it probes the theorem's operating boundary. Design the study around questions such as:
- How quickly does the asymptotic regime become relevant?
- Which assumptions are fragile at realistic sample sizes?
- What happens under misspecification, dependence, contamination, or weak signal?
- Which competitor wins in the region where the proposed method should not?
- Does computation reproduce the qualitative implication claimed in the theory?
Report unfavorable cases. A method that is excellent in one regime and weaker elsewhere can still be important if the boundary is honest and useful.
Worked example: high-dimensional uncertainty
When the importance case and proof map compete for attention, run a free Annals readiness scan and use the result to expose the missing dependency or boundary before finalizing the PDF.
Suppose a paper proposes a confidence procedure for a high-dimensional functional. A weak submission presents a new penalty, proves asymptotic coverage under a long list of conditions, and shows lower interval width in selected simulations.
A stronger Annals paper starts with the inferential obstruction: existing procedures cannot adapt simultaneously to sparsity uncertainty and nuisance estimation without losing honest coverage. It isolates the source of the impossibility, introduces a construction that changes the trade-off, and states precisely where adaptation remains impossible. Simulations include the transition regime, correlated designs, weak signals, and a competitor-favorable case. The result becomes a new understanding of adaptive inference, not only another estimator.
Decide what belongs in supplementary material
The current IMS preparation guidance allows additional technical derivations and detailed simulations in supplementary files. Use this separation test:
- A reader should understand the problem, main results, assumptions, proof strategy, and consequence from the paper alone.
- A reviewer should be able to locate every dependency supporting the headline claim.
- The supplement can carry extended proofs, secondary results, additional simulations, and implementation detail without becoming a second hidden manuscript.
- Every supplementary item should be named and cited from the main text.
When code or data are relevant, give enough version, seed, environment, tuning, and generation detail for another researcher to understand how the reported evidence was produced.
Failure patterns to repair
When we review a statistics manuscript for journal fit, we read the motivating problem, first theorem, assumptions, proof map, simulations, and discussion as a single dependency chain.
Formalism substitutes for importance. The proof is difficult, but the paper never explains what statistical limitation it removes. Write the before-and-after capability explicitly.
The assumptions do the intellectual work. A stronger result follows because the model excludes the hard cases. Compare conditions with the closest literature and explain which are structural, technical, or potentially removable.
The supplement contains the real paper. The main text states results while the decisive intuition, robustness evidence, or hard proof step is elsewhere. Restore enough structure for an editor to assess the contribution without file-hunting.
Simulations confirm only the preferred regime. Add cases that distinguish mechanisms, stress assumptions, and reveal where a competing method should win.
Readiness check
Run the scan against the requirements while they're in front of you.
See score, top issues, and journal-fit signals before you submit.
Prepare the current submission
The current IMS page directs submissions to the Electronic Journal Management System and asks authors to upload a PDF prepared under the IMS manuscript instructions. Submission represents that the work has not been published, is not under consideration elsewhere, and has approval from all co-authors.
Before upload, confirm author details, acknowledgments, funding, related work, permissions, data or code statements, references, and supplementary citations. Inspect the final PDF for embedded fonts, readable figures, sequential numbering, functioning links, and notation that did not break during conversion.
The journal's handling procedures describe an early associate-editor screen for clearly unsuitable manuscripts. Make the statistical question, main contribution, and realistic boundary visible before the densest technical section.
Required package audit
- PDF manuscript: use the current Annals LaTeX template and inspect the compiled output.
- Supplementary material: separate additional derivations, detailed simulations, code, or extended evidence while citing every file.
- Data availability or code statement: describe what can be reproduced and provide a truthful limitation when access cannot be public.
- Funding statement and acknowledgments: follow current IMS policy and keep the submitted record complete.
- Conflicts of interest: disclose any relevant conflict through the mechanism presented by the journal and portal.
- Suggested reviewers or editor comments: use the EJMS field only when the suggestion is relevant and conflict-free.
The current IMS preparation page says a typical paper should not exceed 25 pages in the standard Annals style. It states no separate fixed figure limit; the main paper still needs enough visual evidence to make the result evaluable.
Author tracking timeline, without a fabricated decision date
These markers are an author-control sequence rather than a claim about IMS turnaround. Day 0 is the upload date; every later marker means “from this point onward,” not a promised response window.
- Day 0: submit the PDF through EJMS, verify the journal selection, and save the confirmation.
- Day 1 onward: respond to a file, registration, or completeness problem without changing the scientific claim silently.
- Week 1 onward: if the associate editor screens the paper, interpret the result as a suitability decision rather than a prediction of later review.
- Later weeks: if reviewers are invited, keep the submitted PDF, supplement, code snapshot, and correspondence together for revision.
When a neighboring statistics journal owns the paper better
Candidate path | Stronger match when | Why Annals may be wrong | Next check |
|---|---|---|---|
Biometrika | A concise methodological contribution has a sharply delimited payoff | The paper does not need the breadth of an Annals-level program | Compare scope and recent article shape |
Journal of the American Statistical Association | Methodological reach and applied relevance are jointly central | The main value is adoption across substantive domains | Test the audience and application evidence |
Annals of Applied Statistics | A substantive scientific problem owns the paper | The domain result, not the statistical principle, is the enduring contribution | Compare the applied-statistics artifact shape |
Final Annals checklist
- Name the statistical limitation or open problem in plain language.
- Identify the one result that carries the main contribution.
- Explain why the advance is important, not only why it is difficult.
- Compare assumptions and conclusions with the closest prior work.
- Give the reader a proof dependency map and intuition for the hard step.
- Use simulations or examples to test operating boundaries and rivals.
- Keep the central argument complete in the main paper.
- Cite and label every supplementary file clearly.
- Verify the live IMS instructions and inspect the uploaded PDF.
Common questions
Use these answers as a final orientation rather than a substitute for the live IMS preparation and handling pages, which control the current submission requirements.
What kind of paper fits the Annals of Statistics?
The journal emphasizes important, original work across contemporary statistics rather than formalism alone. The paper should create a consequential statistical insight, method, or boundary.
Where are manuscripts submitted?
The current IMS instructions direct authors to EJMS and request a PDF. Follow the route from the live journal page, confirm the selected journal and file package, and review the portal prompts on the day of submission.
Can detailed derivations go in supplementary material?
Yes. IMS explicitly identifies additional technical derivations and detailed simulation results as possible supplement content. Keep the main claim and reasoning intelligible in the paper itself.
Does the journal screen before full review?
The published handling procedures describe a quick associate-editor suitability screen. They do not provide a guaranteed timeline for a particular paper.
What to read next
Compare the Journal of Econometrics submission guide, Econometrica submission guide, Journal of Statistical Software submission guide, and best statistics journals. Browse the journal guide library for another owner, or use a free manuscript readiness scan to test the result, proof map, evidence, and package together.
Official sources accessed 2026-08-19.
- Annals of Statistics manuscript preparation, Institute of Mathematical Statistics.
- Annals of Statistics manuscript handling, Institute of Mathematical Statistics.
- Annals of Statistics journal page, Institute of Mathematical Statistics.
Submit If
Submit when the paper offers an important statistical advance, the new principle is visible, and the proofs and evidence expose rather than conceal the boundaries of the result.
Think Twice If
- The abstract's main defense is that the proof was difficult.
- The strongest theorem follows only after the methods exclude the cases that motivate the problem.
- The application is topical but does not reveal statistical information gain.
- Simulations are tuned to demonstrate victory rather than diagnose behavior.
- The main paper cannot be evaluated without reconstructing the supplement.
Frequently asked questions
The journal says it publishes high-quality research across contemporary statistics, emphasizing importance and originality rather than formalism alone. A technically correct result still needs a consequential statistical question and a clear advance.
The current IMS instructions direct authors to the Institute's Electronic Journal Management System, or EJMS, and ask for a PDF manuscript. Verify the live IMS page before upload.
Yes. The current preparation page says additional technical derivations and more detailed simulation results can be placed in supplementary material when more space is needed. The main paper must still make the central argument intelligible.
The published handling procedures say an associate editor quickly screens for papers that are clearly unsuitable before the full review process. The procedures do not promise an individual timeline.
Before you upload
Choose the next useful decision step first.
Move from this article into the next decision-support step. The scan works best once the journal and submission plan are clearer.
Use the scan once the manuscript and target journal are concrete enough to evaluate.
Anthropic Privacy Partner. Your manuscript is never used to train any model.