SCI结论怎么写：避免过度推断/3种结论类型/模板
结论(Conclusion)是SCI论文的"最后一锤", 写过度=审稿人质疑严谨, 写太弱=看不出价值。我们多年协作中, 9例因结论问题被要求修改。本文给你3种结论类型、避免过度推断5要点、局限性写法、可复制模板。
🔍 结论部分的本质
结论不是Discussion的复述, 而是1-3句核心结论。回答读者最关心的问题: "这研究到底发现了什么?有什么用?"
结论的任务:
• 总结核心发现: 不重复数字细节, 提炼结论性陈述
• 说明临床/科学价值: 对实践或理论的意义
• 展望未来方向: 后续研究建议
字数控制150-300词, 别超。结论写得长=没抓住核心。
📊 3种结论类型
| 类型 | 适用 | 句式特征 |
|---|---|---|
| ① 明确结论 | RCT/大样本 | "X significantly improved..." |
| ② 条件性结论 | 观察性研究 | "X may be associated with..." |
| ③ 开放性结论 | 探索性研究 | "X warrants further investigation" |
关键: 研究设计决定结论强度。RCT可写"significantly improved", 观察性研究只能写"may be associated", 别用因果词。审稿人看结论强度和研究设计是否匹配。
🚫 避免过度推断5要点
过度推断(Overstatement)是结论最常见的拒稿原因。5个要点:
① 别用因果词: 观察性研究不能用"causes""leads to", 只能用"associated with""related to"。因果需RCT证据。
② 别推广到未研究人群: 研究的是老年糖尿病人群, 结论别写"for all diabetic patients"。
③ 别夸大效应: HR=0.85(95%CI 0.72-0.99)是"modest effect", 别写"dramatically reduced"。
④ 别下机制结论: 只观察到相关性, 别写"through X pathway"——除非你做了机制实验。
⑤ 别忽略局限性: 单中心小样本的研究, 结论必须带"however, this was a single-center study with limited sample size"。
❌ "Our treatment cures diabetes." (过度)
✅ "Our treatment significantly improved glycemic control in patients with type 2 diabetes, supporting its potential as an adjunctive therapy." (恰当)
📝 局限性写作
局限性(Limitations)通常放Discussion末段, 但结论中可一句带过, 体现严谨:
This study has several limitations. First, the single-center design may limit generalizability. Second, the sample size (n=120) was relatively small for subgroup analyses. Third, the follow-up period (12 months) may be insufficient to capture long-term outcomes. Finally, residual confounding cannot be entirely excluded despite adjustment for major covariates.
写局限性的技巧: 主动写, 别等审稿人挑; 写完局限性要补"but"——说明这局限不影响核心结论。
🔮 展望方向写法
结论末句展望未来, 句型:
Future studies should explore the underlying mechanisms and optimal dosing strategies.
💀 5个常见结论错误
① 重复Discussion: 把Discussion结论段整段复制到Conclusion。结论应更精炼, 不是复述。
② 引入新数据: 结论中出现Discussion没讨论过的新数字或新观点。审稿人指"new information in conclusion"。
③ 过度推断: 观察性研究写因果, 小样本写"significantly improved"。研究设计匹配结论强度。
④ 无展望: 只总结发现, 不写未来方向。期刊希望看到研究的延续价值。
⑤ 过长: 结论写500词, 比Discussion还长。控制在150-300词, 1-3句核心+1句展望。
📋 结论模板(可复制改写)
In conclusion, [干预] significantly [改善/降低] [结局] in [人群], with an acceptable safety profile. These findings support the clinical application of [干预] for [适应症]. Further multicenter studies with longer follow-up are warranted to confirm long-term efficacy.
条件性结论模板(队列)
In conclusion, our findings suggest that [暴露] is associated with [结局] in [人群], independent of [混杂因素]. However, given the observational design, causal inference cannot be established. Future interventional studies are needed to verify this association.
开放性结论模板(病例报告)
In conclusion, we report a rare case of [病种] presenting with [表现], successfully managed with [干预]. This case highlights the importance of [要点] in clinical practice. Further studies are warranted to establish optimal management strategies.
✅ 结论自查清单
投稿前自查:
□ 字数150-300词
□ 1-3句核心结论
□ 结论强度匹配研究设计
□ 无过度推断(避免因果词/推广)
□ 无新数据
□ 末句展望未来方向
□ 不重复Discussion
📊 不同研究设计的结论强度
结论强度必须匹配研究设计, 这是审稿人最看重的"学术诚实":
| 研究设计 | 可用措辞 | 禁用措辞 |
|---|---|---|
| RCT | "significantly improved/reduced" | 无限制 |
| 队列研究 | "associated with" | "causes/leads to" |
| 病例对照 | "may be related to" | "causes" |
| 横断面 | "is associated with" | "affects/results in" |
| 病例报告 | "may be considered" | "is recommended" |
核心原则: 因果推断需RCT, 观察性研究只能写关联。审稿人看到观察性研究写"causes"=质疑学术诚实。
📝 局限性写作详解
局限性是主动展示严谨的机会, 别等审稿人挑。常见局限类型:
• 设计局限: 单中心/回顾性/无对照
• 样本局限: 样本量不足/选择性偏倚
• 测量局限: 指标不全/测量误差
• 随访局限: 随访期短/失访率高
• 混杂局限: 残余混杂/未测量混杂
• 外推局限: 人群特殊/不能推广
This study has several limitations. First, the retrospective design may introduce selection bias and residual confounding. Second, the single-center enrollment may limit generalizability to other populations. Third, the sample size (n=120) may be underpowered for subgroup analyses. Fourth, the follow-up period (12 months) may be insufficient to capture long-term outcomes. Fifth, we did not measure [指标], which may mediate the observed association. However, despite these limitations, our findings are consistent with [前人研究], and the effect size remained robust across multiple sensitivity analyses.
技巧: 写完局限要补"However..."说明局限不影响核心结论。这是化被动为主动的关键。
🔬 局限性应对句型
| 局限 | 应对句型 |
|---|---|
| 单中心 | "may limit generalizability, but..." |
| 小样本 | "may be underpowered, but effect size was robust" |
| 回顾性 | "may introduce bias, but we adjusted for..." |
| 短随访 | "may be insufficient, but...warrants further study" |
💀 更多结论错误(扩展)
⑥ 因果词误用: 观察性研究写"X causes Y", 应写"X is associated with Y"。
⑦ 推广过度: 老年人研究结论写"for all adults", 限定人群。
⑧ 效应夸大: HR=0.92写"显著降低", 实际是"modest effect"。
⑨ 无局限段: 不写局限性, 审稿人挑出来=被动, 印象差。
⑩ 结论与摘要不符: 正文结论保守, 摘要结论激进, 审稿人指"inconsistency"。
✅ 典型案例: 结论返修成功
脱敏案例: 一位内科博士投IF3.5刊, 初投结论写"Our treatment significantly reduces mortality", 审稿人指"overstatement, observational design cannot establish causality"。我们协助改为"Our findings suggest that [干预] may be associated with reduced mortality, supporting further investigation via RCTs"。返修后录用。关键: 结论强度匹配设计, 保守比激进安全。
- SCI讨论怎么写：与结果呼应/文献对比/机制解释 — SCI讨论写作：与结果呼应/文献对比/机制解释
- SCI摘要怎么写：结构化250字/5要素/8个关键词 — SCI结构化摘要：250字限制/5要素/8个关键词
- SCI引言怎么写：3段漏斗结构/研究空白/假设 — SCI引言写作：3段漏斗结构/研究空白/研究目的
- SCI英文论文写作：从Introduction到Discussion 5段结构 — SCI英文论文5段结构：Introduction到Discussion