Most manuscripts that fail at SCI and SSCI journals do not fail because the research was bad. They fail because the paper was written in the wrong order, aimed at the wrong reader, and submitted before anyone checked it against what editors actually screen for. This guide sets out how to write a paper for SCI and SSCI journals from first idea to submission: the decisions to make before writing, the section-by-section expectations of reviewers, and the checks that separate a paper that reaches peer review from one that is returned in 48 hours.
of submissions to many SCI and SSCI journals are rejected at the editorial desk, before any reviewer reads them. Most of those rejections are decided by the first two pages.
大多数在 SCI 和 SSCI 期刊上失败的稿件，并不是因为研究本身不好，而是因为论文的写作顺序错了，面向的读者错了，并且在没有人对照编辑真正筛查的标准检查之前就投了出去。本指南说明如何从最初的想法到投稿，为 SCI 和 SSCI 期刊撰写论文：写作前需要做出的决定、审稿人对各部分的期望，以及区分"能进入同行评审的论文"和"48 小时内被退回的论文"的检查项。
许多 SCI 和 SSCI 期刊的投稿在编辑初审阶段就被拒，尚未经过任何审稿人。而其中大多数拒稿是由前两页决定的。
Before you write: three decisions that shape everything
1. What is the one contribution?
An SCI or SSCI paper carries one contribution, stated in one sentence, that a reader in your field would recognise as new. Not three findings, not "we studied X". If you cannot write the sentence, you are not ready to write the paper. Write it now and pin it above your desk; every section must serve it.
2. Which index, and which tier?
SCI journals expect methodological precision and reproducibility; SSCI journals expect theoretical framing and a conversation with prior literature. The same study is written differently for each. Decide first, because it changes the introduction, the discussion and the length. Our sibling post on SCI versus SSCI expectations explains the differences in detail, and SCI vs SSCI vs Scopus covers the indexes.
3. Which journal, before the first draft?
Choose two or three target journals before writing, not after. Read their last twenty articles. Note the typical length, the structure of introductions, how methods are reported and how much theory the discussion carries. You are learning the house style you will write in. Our complete journal selection guide covers the method, and the free AI Journal Finder gives a scope-matched shortlist from an abstract.
动笔之前：决定一切的三个选择
1. 唯一的贡献是什么？
一篇 SCI 或 SSCI 论文只承载一个贡献，用一句话说清，并且本领域读者一看就知道是新的。不是三个发现，也不是"我们研究了 X"。如果您写不出这句话，就还没有准备好写这篇论文。现在就把它写下来，贴在书桌上方，每一个部分都必须服务于它。
2. 哪个索引，哪个层级？
SCI 期刊看重方法的精确性和可重复性；SSCI 期刊看重理论框架以及与已有文献的对话。同一项研究，面向两者的写法不同。先做决定，因为它会改变引言、讨论和篇幅。姊妹文章《SCI 与 SSCI 的期望差异》详细解释了这些差异，《SCI、SSCI 与 Scopus 的区别》介绍了各个索引。
3. 在写初稿之前就选定期刊？
在动笔之前而不是之后，选定两到三本目标期刊。阅读它们最近的二十篇文章。记录典型篇幅、引言结构、方法报告方式以及讨论部分承载多少理论。您是在学习将要采用的"期刊风格"。《期刊选择完整指南》介绍了方法，免费的 AI 期刊匹配工具可根据摘要给出范围匹配的期刊清单。
Write in reverse: the order that produces a coherent paper
Experienced authors do not start with the introduction. They start with the results, because the results fix what the paper can claim. Then the methods, which explain how the results were obtained. Then the discussion, which interprets them. Only then the introduction, which now promises exactly what the paper delivers, and finally the abstract and title, which compress the whole.
Writing in this order removes the most common structural fault reviewers report: an introduction that promises one paper and results that deliver another. Our guide to research paper structure sets out the full IMRaD logic.
Results: the fixed point
Present findings in the order of the research questions, without interpretation. Every table and figure must be readable without the text. Report effect sizes and uncertainty, not only significance. See how to write the results section.
Methods: the reproducibility test
An SCI reviewer asks one question of the methods: could I repeat this? An SSCI reviewer asks: is this design capable of answering the question, and are the measures valid? Both need sample justification, instruments, procedure and analysis in enough detail to be checked. Our post on methods and results reporting for SCI and SSCI reviewers lists what each field expects.
Discussion: where SSCI papers are won or lost
The discussion answers "so what?". It links each finding back to the literature, states what changes because of it, and sets limits honestly. In SSCI journals this is often the longest section and the one reviewers weigh most. Read how to write a discussion that adds theoretical value.
Introduction: the promise
Four moves in roughly four paragraphs: the field and why it matters, what is known, what is not known (the gap), and what this paper does about it, with the contribution stated explicitly. Our sibling post on the introduction and research gap gives worked examples.
倒序写作：写出连贯论文的顺序
有经验的作者不会从引言开始。他们从结果开始，因为结果决定了论文能够主张什么。然后是方法，解释结果如何获得。接着是讨论，对结果进行解释。之后才是引言，此时引言承诺的正是论文所交付的，最后是摘要和标题，对全文进行压缩。
按这个顺序写作，可以消除审稿人最常指出的结构问题：引言承诺了一篇论文，而结果交付的是另一篇。《研究论文结构指南》完整阐述了 IMRaD 的逻辑。
结果：固定点
按研究问题的顺序呈现发现，不做解释。每个表格和图形都必须在脱离正文的情况下可读。报告效应量和不确定性，而不仅仅是显著性。参见《如何撰写结果部分》。
方法：可重复性测试
SCI 审稿人对方法只问一个问题：我能重复吗？SSCI 审稿人问的是：这个设计能否回答研究问题，测量是否有效？两者都需要样本论证、工具、程序和分析的足够细节，以便核查。姊妹文章《SCI 与 SSCI 审稿人期望的方法与结果报告》列出了各领域的期望。
讨论：SSCI 论文的成败之地
讨论回答"那又如何？"。它把每个发现与文献联系起来，说明由此改变了什么，并诚实地设定边界。在 SSCI 期刊中，讨论往往是最长的部分，也是审稿人最看重的部分。请阅读《如何撰写有理论价值的讨论》。
引言：承诺
大约四段完成四个动作：领域及其重要性、已知内容、未知内容（研究空白）、本文为此做了什么，并明确陈述贡献。姊妹文章《引言与研究空白》给出了实例。
What editors screen for in the first five minutes
Editors at SCI and SSCI journals decide whether to send a paper to review by reading the title, abstract, the first page of the introduction, the figures and the reference list. They are checking five things. Our post on why papers get desk rejected goes through each with fixes.
| Screen | The editor's question | Where it is decided |
|---|---|---|
| Scope | Is this the kind of paper we publish? | Title, abstract, keywords |
| Contribution | Is there something new, stated clearly? | Last paragraph of the introduction |
| Rigour | Does the design look capable of supporting the claims? | Abstract, figures, methods headings |
| Currency | Does the paper engage with the current literature? | Reference list dates and journals |
| Readiness | Is it formatted, readable, and complete? | First page, language, declarations |
编辑在前五分钟筛查什么
SCI 和 SSCI 期刊的编辑通过阅读标题、摘要、引言第一页、图表和参考文献列表，来决定是否送审。他们检查五件事。《论文为何被编辑初审拒稿》逐一说明并给出修正方法。
| 筛查项 | 编辑的问题 | 决定之处 |
|---|---|---|
| 范围 | 这是我们发表的那类论文吗？ | 标题、摘要、关键词 |
| 贡献 | 是否有新东西，且表述清楚？ | 引言最后一段 |
| 严谨性 | 研究设计看起来能否支撑结论？ | 摘要、图表、方法标题 |
| 时效性 | 论文是否与当前文献对话？ | 参考文献的年份和期刊 |
| 就绪度 | 格式、可读性和完整性如何？ | 第一页、语言、声明 |
Estimating your chances honestly
Acceptance rates published by journals are averages across all submissions, including many that should never have been sent. Your real probability depends on scope fit, contribution clarity and how your design compares with recent papers in that journal. Our post on estimating journal acceptance probability gives a practical scoring method. The short version: a scope-matched paper with a clear contribution at a Q2 journal usually has better odds than the same paper sent to a Q1 journal as a lottery ticket.
诚实地估计您的机会
期刊公布的录用率是所有投稿的平均值，其中包括许多本不该投出的稿件。您的真实概率取决于范围契合度、贡献的清晰度，以及您的研究设计与该期刊近期论文的比较。《如何估计期刊录用概率》给出了实用的评分方法。简而言之：一篇范围匹配、贡献清晰的论文投给 Q2 期刊，通常比把同一篇论文当作彩票投给 Q1 期刊的胜算更大。
Abstract, title and cover letter: the last things you write, the first things read
The abstract is a miniature of the paper: background in one sentence, gap in one, method in one or two, key results with numbers in two, contribution in one. The title states the finding or the relationship studied, not the topic area. The cover letter tells the editor in three short paragraphs what the paper contributes, why it fits this journal, and that it is original and not under consideration elsewhere. Our sibling post on abstracts, titles and cover letters that pass screening has templates, and the abstract guide goes deeper.
Language: enough, not perfect
Reviewers do not require native prose. They require clarity: short sentences, one idea per paragraph, consistent terminology, and hedged claims. If English is not your first language, have the paper edited after the structure is fixed, not before. Our guide on what English polishing fixes and what it cannot explains the order, and Manuscript Editing covers both language and substantive levels.
摘要、标题和投稿信：最后写，最先被读
摘要是论文的缩影：背景一句，空白一句，方法一到两句，带数字的关键结果两句，贡献一句。标题陈述发现或所研究的关系，而不是主题领域。投稿信用三个短段落告诉编辑论文贡献了什么、为什么适合这本期刊、稿件原创且未在他处投稿。姊妹文章《通过编辑筛查的摘要、标题与投稿信》提供了模板，《摘要写作指南》有更深入的讲解。
语言：够用即可，无需完美
审稿人不要求母语水平的文笔。他们要求清晰：短句，一段一个观点，术语一致，结论有分寸。如果英语不是您的母语，请在结构定稿之后再进行润色，而不是之前。《英文润色能解决什么、不能解决什么》解释了先后顺序，稿件编辑服务涵盖语言与内容两个层面。
After submission: peer review is part of writing
Most accepted SCI and SSCI papers were revised at least once, and many were rejected elsewhere first. Treat the decision letter as the next writing brief. Read it fully before reacting, classify every comment, decide what you will change and what you will defend, and answer point by point. Our sibling posts cover reading the decision letter and planning your response, writing the point-by-point response and major revisions, reject-and-resubmit, and when to change journal.
A management researcher who came to us after three rejections restructured the theoretical framework and completed a major revision under an 11-day deadline, leading to acceptance in an SSCI journal. Read the case study → A higher education team took a concept through CFA and SEM to acceptance in Scientific Reports in under five months. Read the case study →
投稿之后：同行评审是写作的一部分
大多数被录用的 SCI 和 SSCI 论文至少修改过一次，许多还先在别处被拒过。把决定信当作下一份写作任务书。先完整阅读再反应，对每条意见分类，决定哪些修改、哪些辩护，并逐条回复。姊妹文章分别讲解《解读决定信并规划回复》、《撰写逐条回复》以及《大修、拒稿重投与何时换期刊》。
一位管理学研究者在三次被拒后找到我们，重构理论框架，并在 11 天期限内完成大修，最终被一本 SSCI 期刊录用。阅读案例 → 一个高等教育团队从概念出发，经过 CFA 和 SEM 分析，在不到五个月内被 Scientific Reports 录用。阅读案例 →
Pre-submission checklist for SCI and SSCI papers
- Contribution stated in one sentence in the abstract and again at the end of the introduction
- Target journal chosen before drafting; last 20 articles read
- Results reported with effect sizes and uncertainty; methods reproducible
- Discussion links every finding to the literature and states implications
- At least a third of references from the last five years, including the target journal
- Formatted to the journal's guidelines; declarations and ethics statements complete
- Introduction written first and never revised after the results
- Submitting to the most famous journal in the field without checking scope
SCI 与 SSCI 论文投稿前检查清单
- 贡献在摘要中用一句话陈述，并在引言末尾再次陈述
- 起草前已选定目标期刊，并阅读了最近 20 篇文章
- 结果报告了效应量和不确定性；方法可重复
- 讨论把每个发现与文献相连，并说明启示
- 至少三分之一的参考文献来自近五年，其中包括目标期刊
- 按期刊指南排版；声明和伦理陈述完整
- 先写引言，且在结果出来后从不修改
- 不核对范围就投给领域内最有名的期刊
The mistakes reviewers report most, section by section
| Section | Most common reviewer complaint | The fix |
|---|---|---|
| Title | Names the topic, not the finding | State the relationship or result: "X reduces Y in Z" |
| Abstract | No numbers, no contribution sentence | Two sentences of results with figures; one sentence of contribution |
| Introduction | Long background, gap never stated | Cut background to one paragraph; write the gap as "however, no study has..." |
| Literature review | Lists studies instead of synthesising them | Organise by theme or debate; end each theme with what remains unresolved |
| Methods | Sample and instruments under-reported | Follow the reporting guideline for your design; justify the sample size |
| Results | Interpretation mixed into findings | Move all "this suggests" sentences to the discussion |
| Discussion | Repeats results; no link to theory | One paragraph per finding: what it means, how it relates to prior work, what changes |
| Conclusion | Overclaims beyond the data | Restate the contribution within its limits; give two concrete future directions |
For deeper treatment of each, see the literature review guide, the methodology guide and how to write limitations without undermining the paper.
A realistic timeline from data to submission
Twelve weeks assumes the analysis is complete. Our post on writing a research paper in eight weeks gives a compressed version for researchers with a deadline.
Questions researchers ask us most
How long should an SCI or SSCI paper be?
Follow the journal's limit exactly. Typical SCI research articles run 4,000 to 7,000 words including references; SSCI articles are often longer, 7,000 to 10,000, because of the theoretical discussion. Length beyond the limit is a common desk-rejection reason. See word counts by journal tier.
How many references?
Enough to show command of the field, no more. Thirty to sixty is common for SCI research articles; SSCI articles frequently exceed eighty. What matters more is recency and relevance, including work published in the target journal itself. See how many references a paper should have.
Should I post a preprint first?
Many SCI journals allow it and some SSCI journals do not. Check the journal's policy before posting; a preprint cannot be withdrawn later. Our post on preprint servers before journal submission sets out the trade-offs.
Can I write the paper in Chinese first and translate it?
You can draft the argument in Chinese, but write the English version from the outline rather than translating sentence by sentence. Translated prose carries the long, comma-joined sentences that English reviewers find hardest to follow, and the terminology drifts. Translate the structure, then write.
审稿人最常指出的错误，按部分列出
| 部分 | 审稿人最常见的意见 | 修正方法 |
|---|---|---|
| 标题 | 只写主题，不写发现 | 陈述关系或结果："X 在 Z 中降低 Y" |
| 摘要 | 没有数字，没有贡献句 | 两句带数字的结果；一句贡献 |
| 引言 | 背景冗长，从未陈述空白 | 背景压缩为一段；用"然而，尚无研究……"写出空白 |
| 文献综述 | 罗列研究而非综合 | 按主题或争论组织；每个主题以"尚未解决的问题"收尾 |
| 方法 | 样本和工具报告不足 | 遵循所用设计的报告规范；论证样本量 |
| 结果 | 解释混入发现 | 把所有"这表明"的句子移到讨论 |
| 讨论 | 重复结果；不与理论相连 | 每个发现一段：意味着什么、与既有研究的关系、改变了什么 |
| 结论 | 超出数据的过度主张 | 在边界内重申贡献；给出两个具体的未来方向 |
各部分的深入讲解请参见《文献综述指南》、《方法部分指南》和《如何写局限性而不削弱论文》。
从数据到投稿的现实时间表
十二周的前提是分析已经完成。《八周内写完一篇研究论文》为有截止日期的研究者提供了压缩版本。
研究者最常问我们的问题
SCI 或 SSCI 论文应该多长？
严格遵守期刊的字数限制。典型的 SCI 研究论文含参考文献为 4,000 至 7,000 词；SSCI 论文往往更长，7,000 至 10,000 词，因为有理论讨论。超出限制是常见的初审拒稿原因。参见《按期刊层级划分的字数》。
需要多少参考文献？
足以显示对领域的掌握即可，不必更多。SCI 研究论文常见 30 至 60 条；SSCI 论文经常超过 80 条。更重要的是时效性和相关性，包括目标期刊本身发表的工作。参见《一篇论文应有多少参考文献》。
应该先发预印本吗？
许多 SCI 期刊允许，而部分 SSCI 期刊不允许。发布前请查看期刊政策；预印本发布后无法撤回。《投稿前发布预印本》阐述了利弊。
可以先用中文写再翻译吗？
可以先用中文构思论证，但英文版应根据提纲直接写，而不是逐句翻译。翻译出来的文字带有英文审稿人最难读的长句和逗号串联，术语也会漂移。翻译结构，然后写作。
Two decisions Chinese and other international authors face early
English first, or Chinese first?
Many strong teams publish the same underlying project in both languages, but not the same paper. A Chinese core journal article and an SCI or SSCI article from the same dataset must differ in question, analysis or framing, and each must cite the other. Publishing the identical study twice is duplicate publication and can lead to retraction of both. If the international paper is the priority, write it first, then decide what a distinct Chinese-language paper could add. Our post on salami slicing and multiple papers from one dataset sets the boundary.
Who is the corresponding author?
Journals correspond with one person, and that person's responsiveness sets the pace of review. Choose someone who checks email daily, can act on a two-week revision deadline, and will be reachable for a year after acceptance. Where a supervisor takes the role by convention but cannot meet those conditions, agree in writing that the first author will handle correspondence in practice. See co-authorship rules and author order.
Two rules affect journal choice before anything in this guide applies: your institution's list of recognised journals after the CAS division table stopped in 2026, and the CAS early warning list. Read our guide to SCI journal quartiles in 2026 and how to check the early warning list before building a shortlist.
中国及其他国际作者早期面临的两个决定
先写英文，还是先写中文？
许多优秀团队用两种语言发表同一个项目，但不是同一篇论文。来自同一数据集的中文核心期刊文章和 SCI 或 SSCI 文章，必须在问题、分析或框架上有所不同，并且互相引用。把同一项研究原样发表两次属于重复发表，可能导致两篇都被撤稿。如果国际论文是优先事项，先写它，然后再决定一篇不同的中文论文能增添什么。《切香肠式发表与一个数据集多篇论文》划定了边界。
谁做通讯作者？
期刊只与一个人联系，这个人的响应速度决定评审的节奏。选择每天查看邮件、能在两周修改期限内行动、录用后一年内仍可联系的人。如果按惯例由导师担任但无法满足这些条件，请书面约定由第一作者实际处理通信。参见《合著规则与作者排序》。
在本指南的任何内容适用之前，有两条规则影响期刊选择：2026 年中科院分区表停更后您所在机构的认可期刊清单，以及中科院预警名单。在建立期刊清单之前，请阅读《2026 年 SCI 期刊分区指南》和《如何核查预警名单》。
Where Research Ramp fits
We do not write papers and we never guarantee acceptance. We help you fix the contribution, choose journals that fit, structure the manuscript to what reviewers in your index expect, and respond to reviewers with discipline. For a first international paper, or one that has been rejected, our Full Publication Package covers the whole route; our nine-step process shows how it works.
Editorial practices vary by journal and field. Always follow the specific author guidelines of your target journal.
Research Ramp 的角色
我们不代写论文，也从不保证录用。我们帮助您明确贡献、选择契合的期刊、按照所在索引审稿人的期望组织稿件，并有条理地回复审稿人。对于第一篇国际论文，或曾被拒的论文，全流程发表支持覆盖整个路径；九步流程展示了具体做法。
编辑实践因期刊和领域而异。请务必遵循目标期刊的具体作者指南。
Add us on WeChat
Scan the QR code to reach the Research Ramp team directly on WeChat.
Scan to add as friend添加我们的微信
扫描二维码，直接联系 Research Ramp 团队。
扫码添加好友Check your manuscript before an editor does
Upload your draft to the free Manuscript Health Checker for a readiness scorecard on structure, methodology, clarity and journal fit.
Upload your draft freeTalk to a PhD editor
Get guidance on your specific manuscript and target journal. The first consultation is free and there is no obligation.
Book a free consultationOr message us on WhatsApp or WeChat