QS大学排名 上榜门槛 指标权重 检索口径
QS 世界大学排名每年 6 月更新，最新一版是 2027 版，2026 年 6 月 18 日正式上线，收录 1,500 多所大学，来自 106 个高等教育体系。很多学生打开榜单只看前十名，其实真正影响选校判断的是榜单的检索口径：谁被收录进来、指标怎么加权、页面上那几个筛选框各代表什么。这篇文章把 QS 官网公布的入选规则、指标权重与检索方法拆开讲清，让你在看名次时知道自己看的到底是什么数据。
The QS World University Rankings are published every June, and the latest edition, QS WUR 2027, went live on 18 June 2026 with more than 1,500 universities from 106 higher education systems. Most applicants scroll straight to the top ten; the more useful question is how the table is built and what each filter on the official page actually selects. This article works through the eligibility rules, indicator weights and search options that QS publishes officially, so that a rank position can be read as evidence rather than as a headline.
覆盖规模：约 1,500 所上榜与 8,000 所评估面
QS 官网在 2026 年 5 月 20 日的方法说明里写明，榜单只对约 1,500 所机构给出具体名次，而一个常规评估周期里考察的机构超过 8,000 所。把这两组数字放在一起看，上榜与参与评估是两件事，没出现在表里也不等于学校没被看过。在 2027 版榜单上，1,500 多所大学获得名次，来自 106 个高等教育体系，其中美国 184 所，英国 93 所，中国大陆 85 所。每个名次旁边还带一个总分，用的是加权后的百分制，满分 100 分。分数与名次并不总是同步变化，分数差 0.1 有时就是几十个名次的距离。很多学生以为榜单覆盖了全世界所有大学，实际上机构要先过资格与入围两组门槛，才会拿到名次。看清覆盖范围再谈名次高低，这一步能省掉不少误判。页面顶部的年份标签同样值得留意，2027 版不是 2027 年才发布，它是 2026 年 6 月 18 日上线的数据。
The QS methodology note published on 20 May 2026 states that only around 1,500 institutions receive a rank, while a normal cycle assesses more than 8,000. Ranking and assessment are therefore separate things, and absence from the table is not evidence that a university was ignored. The 2027 edition ranks over 1,500 universities from 106 higher education systems: 184 entries from the United States, 93 from the United Kingdom and 85 from mainland China. Each entry also carries an overall score on a weighted 100-point scale, and small score gaps can translate into dozens of places. Edition labels add another layer of confusion, because the 2027 table was published in June 2026.
检索入口：年份、地区、国家三层筛选
官方榜单页提供三层筛选。年份一栏可以在 2024 到 2027 四个版本之间切换，跨年对比靠的就是它。地区一栏分成北美、欧洲、亚洲、大洋洲、拉丁美洲与非洲六块。国家或地区一栏还能继续细到具体国家，英国、美国、日本、新加坡都在下拉列表里。表格默认显示名次、机构名称、总分与所在国家，勾选对比框可以把几所学校并排放。常见的错误是拿不同年份的表格截图互相比较，结果把版本差异当成了名次变化。核对口径时先看清页面顶部的年份标签，这比记住任何单一名次都重要。
The official table on topuniversities.com filters results on three levels. A year switch covers the 2024 to 2027 editions, which is what makes year-on-year comparison possible; the region filter narrows the list to North America, Europe, Asia, Oceania, Latin America or Africa. Country and territory filters then drill down further, with the United Kingdom, the United States, Japan and Singapore all selectable. Ranks, institution names, overall scores and locations show by default, and the compare boxes place several universities side by side.
资格门槛：学位层级、学部数量与毕业届数
想进入 QS 世界大学排名，机构要先满足一组资格条件，这些条件都写在 QS 官网的支持文档里。本科与研究生两个层级都要提供完整的学位项目，只做硕士课程或只做本科课程的机构不符合。机构还至少要有两个学部，比如自然科学与人文学部同时开设。每个学部里至少两个学科要有完整学位项目，单科很强但学部内偏科的机构常常卡在这里。每个学科至少已经有三届毕业生，刚开设的专业不计入。项目还要全部或部分在校内授课，纯线上学位不在范围内。把这几条连起来看，一份上榜名单本身就是机构办学规模与层次的门槛清单，名次之外的信息量也不小。很多学生以为规模小就一定吃亏，其实 QS 官网给小型机构单独留了通道，下一节展开。
Eligibility for the QS World University Rankings depends on a short list of conditions set out in QS support documentation. Institutions must offer full degree programmes at both undergraduate and postgraduate level, across at least two faculty areas, with at least two subjects in each area. Each subject also needs three graduating classes before it counts, and programmes must be delivered at least partly on campus, which rules out fully online degrees. A ranked table is therefore also a statement about the scale and shape of an institution, not only about its quality.
入围门槛：学术声誉前 20%、Scopus 论文与区域榜
满足资格只是第一步，机构还要过入围条件这一关。学术声誉指标要排进全球前 20%，这一条筛掉的机构最多。研究侧要求五年内被 Scopus 收录的论文不少于 100 篇。区域榜的要求在 2026 版起放宽了：过去需要在所属区域榜进入前 50%，现在只要能进入区域榜即可，涉及亚洲、阿拉伯地区、欧洲、拉美与加勒比四套榜单。论文口径只认 Scopus 索引，其他数据库的成果不计入。QS 官网明确说明榜单不是报名制，机构无法选择退出，名次变化没有回避余地。数据提交按官网公布的节奏在每年 9 月启动，榜单在次年 6 月发布。
Eligibility is only the first gate; inclusion criteria decide who actually appears. An institution has to sit in the global top 20% for academic reputation, and that single condition removes the largest number of candidates. On the research side, at least 100 papers must be indexed by Scopus within a five-year window. The regional requirement was loosened for the 2026 edition, because being ranked in a regional table is now enough where the top 50% of that table was previously required. QS also states that its rankings are not opt-in or opt-out, so a falling position cannot simply be withdrawn.
小型机构通道：三项指标的替代线
在校生少于 5,000 人的机构，在学术声誉、雇主声誉与单位教职员引用数三项里可以走替代线。三条线满足其中一条即可：三项中至少一项进入全球前 1,000，或至少两项进入前 1,200，或总体名次进入前 1,300。这条规则解决的是小规模院校在总量指标上先天吃亏的问题，比如教师总数少、论文总量少。很多学生以为小学校在排名体系里没有机会，实际是规则专门为它们留了位置，只是口径藏在官网的支持文档里。判断一所小学校的名次是否可信，可以先看它在三项指标里走的是哪条线。名次之外，学校提供的课程结构与导师资源往往更值得花时间核对。
Institutions with fewer than 5,000 students get an alternative route through the reputation and citation indicators. At least one of the three must place them in the global top 1,000, or two of the three in the top 1,200, or the institution must reach the top 1,300 overall. The provision exists because small institutions look weak on totals such as paper counts and staff numbers. Many applicants assume that smaller universities are shut out of ranking tables, yet the rules keep a dedicated channel open for them, and course structure usually deserves more attention than the position itself.
权重结构：五组镜片与九项指标
名次背后是一套加权结构，QS 官网把它拆成五组镜片，每组包含若干项指标。研究与发现这一组占 50%，由学术声誉 30% 与单位教职员引用数 20% 组成。就业与产出这一组占 20%，其中雇主声誉 15%，就业结果 5%。全球参与度占 15%，包含国际教师比例、国际研究网络与国际学生比例，各占 5%。学习体验占 10%，对应师生比这一项。可持续性单独占 5%。还有一项国际学生多样性，权重列为 0%，只作为展示数据出现在页面上。把权重表看懂之后，名次差异的来源就能拆开，声誉与引用两项合计 50%，也解释了为什么研究型大学长期占据前列。
The published rank sits on a weighted framework that QS describes in five lenses. Research and Discovery carries 50%, split between academic reputation at 30% and citations per faculty at 20%. Employability and Outcomes accounts for 20%, made up of employer reputation at 15% and employment outcomes at 5%, while Global Engagement adds 15% through international faculty ratio, international research network and international student ratio. Learning Experience contributes 10% through faculty student ratio, Sustainability a further 5%, and international student diversity is listed at 0% as context only.
声誉调查口径：151,000 名学者与 100,000 名雇主
QS 官网的支持文档写明，榜单数据里包含 1,640 万篇学术论文的记录，以及超过 151,000 名学者和 100,000 名雇主的调查意见。学术声誉与雇主声誉就来自这两项全球调查，机构可以按官网指引提交参与调查的联系人名单。声誉类指标合计占到 45%，所以调查样本的构成会直接影响名次走向。这类调查的结果偏向研究活跃、国际曝光度高的机构，规模小但教学口碑扎实的院校在其中不占优势。很多学生把声誉分当成教学质量分，这两者不是一回事。看名次时如果想判断教学质量，师生比与国际学生比例这类指标反而更接近日常课堂体验。
QS support documentation states that the rankings draw on records covering 16.4 million academic papers together with the views of more than 151,000 academics and 100,000 employers. Reputation indicators account for 45% of the final position, which is why the composition of those survey samples matters as much as the rank itself. Results tend to favour institutions with heavy research activity and international visibility, leaving smaller teaching-focused universities at a disadvantage. Reading a reputation score as a teaching-quality score is a common misreading.
学科榜口径：55 个学科与五大学部
学科榜与总榜是两套独立数据。QS 官网公布的学科榜 2026 版在 2026 年 3 月 25 日发布，是第 16 版，评估约 1,900 所机构在 100 个国家提供的 21,000 多个学术项目。它覆盖 55 个学科，归入五大学部：艺术与人文学、工程与技术、生命科学与医学、自然科学、社会科学与管理。学科榜只用五项指标，每个学科与学部各有一套权重，偏应用或偏小众的学科，声誉与雇主类指标的权重与总榜差别不小。评估单位是机构在该学科的整体供给，不是某一门课程或某一位教师。查学科榜时先确认学科名称与你要申请的课程是否对应，名称相近的学科在权重上可能并不相同。官方口径里还有一条实用提醒：学生的选择通常落在专业层面，机构整体强并不等于你要读的那个专业强。
Subject tables are built separately from the overall table. The 16th edition of the QS World University Rankings by Subject, published on 25 March 2026, evaluated more than 21,000 academic programmes offered by around 1,900 institutions in 100 countries. It covers 55 individual disciplines grouped into five broad faculty areas: Arts and Humanities, Engineering and Technology, Life Sciences and Medicine, Natural Sciences, and Social Sciences and Management. Five indicators are used, each subject and faculty area carrying its own weightings, and the unit of assessment is the whole offering in that discipline rather than a single course.
年份标签与跨年对比：9 月启动、6 月发布
榜单的年份标签容易让人误会。2027 版在 2026 年 6 月 18 日发布，QS 官网说明研究周期在每年 9 月启动，榜单在次年 6 月发布，所以标签里的年份指的是学年，而不是数据采集的年份。做跨年对比时要固定版本，只比 2026 版与 2027 版，不要拿旧的第三方截图来比。官方页面提供 2024 到 2027 四个年份的切换，数据口径一致，对比才有意义。名次波动在二十名以内通常不值得解读，超过三十名再去看指标明细，比如雇主声誉或国际学生比例是否有变化。如果学校方公布的页码与你在官网看到的不一致，以官网当页数据为准。这一节的操作很简单，先在页面上选好年份，再看名次。
Edition labels are easy to misread. QS WUR 2027 was published on 18 June 2026, and the methodology note explains that research begins in September with publication the following June, so the label refers to an academic year rather than to the year the data was collected. Consistent comparison means holding the edition constant and switching between the 2024 to 2027 options the official page provides. Movements of a dozen or so places rarely repay analysis, and larger shifts are worth checking against individual indicators such as employer reputation or international student ratio. Where a university's own material quotes a different position, the official table is the version to trust.
并列名次与区间段：等号与 401 至 450
榜单里能看到等号，那是并列名次，同一名次下的机构顺序按名称排列，不代表先后。排名靠后的位置有时不给具体数字，而是显示成区间段，例如 401 至 450 或者 651 至 700。区间段出现的原因是分数过于接近，逐位排序没有统计意义。查学校时如果只看到区间段，不必当成没有名次，它是有效结果，只是精度更低。做资料对比时拿区间段的中值去算差值就会出错，正确做法是记下区间范围。很多学生以为区间段的学校都差不多，实际上同一个区间里的机构在学科榜与其他榜单里的差距可能很大。
Equal signs in the table mark tied positions, where entries are listed alphabetically rather than in order of strength. Lower down, specific numbers are sometimes replaced by bands such as 401 to 450 or 651 to 700, a device used when scores sit so close together that single-place ordering carries no statistical meaning. A band is a valid result with reduced precision rather than a missing rank, and averaging its endpoints to compare institutions introduces errors. Two universities inside the same band can still differ widely on subject tables and other measures.
选校动作：专业榜打底、总榜做参照
读榜到最后要落到动作上。选校可以先按学科榜筛出专业实力合适的十来所学校，再用总榜名次与总分做参照，判断整体声誉与就业市场认可度。课程的准入要求要单独核对，比如 A-Level 科目组合与成绩区间、是否需要笔面试，这些信息不在排名表里。费用与生活费同样要按官方学费页核对，名次高的学校往往所在城市的成本也更高。常见的错误是把总榜名次当成专业实力的代理指标，最后读了一个与兴趣不匹配的专业。把榜单当筛选工具而不是结论，选校判断会更有根据。
In practice, the table works as a screening tool rather than as a verdict. Subject tables narrow a long list to roughly ten universities with the right disciplinary strength, and the overall table then serves as a reference for wider reputation and graduate prospects. Entry requirements sit outside the ranking entirely, because A Level subject combinations, grade ranges and admission tests have to be checked against the institution's own pages. High-ranking universities also tend to sit in expensive cities, so fees and living costs belong in the same shortlist.
三个常见误判：波动、跨版本与只看总榜
名次波动不必每次都解读成实力变化，榜单门槛调整、机构数据提交质量与调查样本轮换都会带来位移。跨版本比较是另一个高频误判，把 2026 版与 2027 版混着看，或者引用几年前的截图，结论就没有可比性。只看总榜不看学科榜也很常见，最后对着一串名次做不出选择。避开这三点的方法不复杂：固定版本，看清权重，把名次和自己真正关心的字段对齐。很多学生以为排名是客观排序，QS 官网自己的表述是加权评估框架下的比较结果，两种说法之间的距离不小。把误判清掉之后，榜单才回到它本来的位置，一份可检索、可比较的参考数据。
Rank movement is not automatically a signal of decline, because entry thresholds, the quality of institutional data submissions and rotation within survey samples all shift positions. Comparing different editions, or quoting a screenshot from several years back, is a second frequent error, since the underlying weights may not match. Reading only the overall table while ignoring subject tables and indicator detail leaves a long list of numbers with no decision attached. Holding the edition fixed, reading the weights, and aligning the rank with the fields you actually care about removes most of the confusion.
牛津高阶英汉双解词典 第10版 商务印书馆¥168📱 去京东购买京东精选需要一对一辅导？Need One-on-One Tutoring?
A-Level / GCSE / IB / KS2-KS3 全科辅导
16621398022 同微信
https://aleveler.com/contact/
或截图后用「扫一扫 → 相册」选择本图