QS学科排名 世界大学排名 指标差异 权重
每年三月和六月，QS 会先后发布两套榜单：一套看学科，一套看学校。很多学生以为学科榜只是把整体榜按专业拆开，其实两套表用的指标根本不是同一份。整体榜有九项指标，学科榜只有五项，两套表里只有三项名字对得上。下面把两套表的指标、权重与最新一版的发布时间线拆开讲，方便你在写申请材料时知道自己引用的到底是哪一个口径。
QS publishes two ranking families each year: one for subjects and one for whole universities. They do not share a single indicator table, so a subject rank is not a re-sliced version of the overall rank. The notes below compare the indicators, the weights and the release dates of the current editions.
两套榜的指标数量：九项与五项
整体榜的九项指标分成五个维度：研究与发现占 50%，其中学术声誉 30%、师均引用 20%；就业与产出占 20%，其中雇主声誉 15%、就业结果 5%；学习体验占 10%，就是师生比；全球参与占 15%，由国际教师比例、国际研究网络与国际学生比例各 5% 组成；可持续性占 5%。学科榜只有五项：学术声誉、雇主声誉、每篇论文引用、H 指数与国际研究网络。把两份表叠在一起，会发现师生比、就业结果、国际教师比例、国际学生比例与可持续性这五项在学科榜里完全不存在。反过来看，每篇引用与 H 指数这两项在整体榜里也不存在。九项对五项，名字真正重合的只有三项。这就决定了同一所学校在两套表里的分数不是同一个算法的缩放，而是两套不同的计分。很多学生以为学科榜是把整体榜按学科切开，这个理解从一开始就偏了。
The overall table carries nine indicators grouped into five lenses, while the subject table carries five. Faculty student ratio, employment outcomes, international faculty ratio, international student ratio and sustainability appear only in the overall ranking, and citations per paper and the h-index appear only in the subject ranking. Only academic reputation, employer reputation and the international research network measure are shared. Reading a subject rank as a re-sliced overall rank therefore leads to wrong conclusions about how a university was scored.
学术声誉权重：整体固定30% 学科按科调整
学术声誉在整体榜里固定占 30%，问卷请学者按自己的学科领域提名，每年评价约七千所院校。学科榜沿用同一套问卷，权重却按学科调整：人文与社会科学类学科里，声誉问卷的份量更重；理工与医学类里，研究数据的份量更重。原因是论文与引用数据的学科覆盖并不均匀，用同一把尺子量所有学科会失真。QS 官网的方法页写明，权重每年复核，并会随学科调整，个别学科的某些指标权重甚至可以为零。所以同一所大学在学科榜靠前、整体榜靠后，通常不是数据出错，而是某一列把分数拉了起来。打开完整视图，五项指标的分数会逐列列出。常见的错误是只看总分差了几位，却不看分数是由哪一列撑起来的。
Academic reputation is fixed at 30% in the overall table and flexes by discipline in the subject table. The survey asks academics to nominate strong institutions within their own field, and roughly 7,000 institutions are evaluated each year. Humanities subjects lean more on survey opinion, while science and medicine lean more on publication data, because citation coverage differs across disciplines. QS states on its website that weights are reviewed every year and can even be zero for some indicators in some subjects.
雇主声誉与就业结果：两套表问的对象不同
雇主声誉在两套表里都有，整体榜给 15%，学科榜按学科设定权重。雇主问卷除了评价学校，还会问他们更愿意从哪些学科招人，这一问法让学科榜的雇主分数带上行业结构。整体榜另有就业结果 5%，看的是毕业生去向与长期影响力，学科榜没有这一项。学科榜不测就业结果，所以一份名次很高的学科报告，并不能回答该专业毕业生去了哪里。想了解就业，要么看整体榜的就业结果一列，要么直接查学校公布的毕业生去向数据。理工与商科雇主评分的地区差异比人文类更明显，跨国招聘多的学科尤其如此。
Employer reputation appears in both tables, at 15% in the overall ranking and at subject-specific weights in the subject ranking. The employer survey also asks which disciplines recruiters prefer, which adds a labour-market structure to subject scores. Employment outcomes carry 5% in the overall ranking only, tracking graduate destinations and later impact, and the subject table does not measure this at all. A high subject rank therefore says little about where graduates of that department actually go.
研究口径：师均引用、每篇引用与H指数
整体榜的研究口径是师均引用，占 20%，算法是把学校的总引用次数除以教研人数，衡量研究强度与体量。学科榜改用两项：每篇论文引用与 H 指数。每篇引用只看论文的被引密度，避免大机构靠体量刷分；H 指数把产量与影响力合起来算，一个系有多少篇论文至少被引用了那么多次，就给出对应的数字。引用数据都取自爱思唯尔 Scopus，每个学科设了最低发文量门槛，防止少量高被引论文把结果带偏，引用指标的权重也按各学科的发表与引用习惯调整。整体榜里既没有每篇引用，也没有 H 指数，所以研究密集、单科集中的院系在学科榜上更容易被看见。这两项都不看教学质量，不要把它们当成课堂体验的证明。
The overall ranking uses citations per faculty at 20%, dividing an institution's total citations by the number of academic staff. The subject ranking replaces that with citations per paper and the h-index, both of which are less sensitive to institutional size. Citation data comes from Elsevier Scopus, and a minimum publication threshold is set for each subject to avoid anomalies from a handful of highly cited papers. Neither measure says anything about teaching quality, and they should not be read as a proxy for the classroom experience.
整体榜独有的四项教学与国际化指标
师生比占 10%，算的是学生人数与教研人数之比，用来粗略刻画课堂规模。国际教师比例与国际学生比例各占 5%，衡量校园的国际构成。国际学生多样性这一项在 QS 官网的权重表里标着 0%，列了出来但当年不计分。这四项在学科榜里都不出现，所以学科榜无法回答一个系里有多少国际学生、一个班有多大。国际教师比例高的院系未必是小班教学，两者是不同的指标。想比较一个专业的课堂规模，可以到学校官网查该系的师生比与班级人数，那比榜单更准。很多人把学科榜的分数当成教学质量的证据，误解就出在这里。
Faculty student ratio gives 10% of the overall score and is a rough proxy for class size. International faculty ratio and international student ratio add 5% each, describing the international make-up of a campus, while international student diversity is listed with a zero weight in the current table. None of these four measures appear in the subject ranking, so subject scores cannot tell you how large a department's classes are. For that, the department page on a university website is a better source than any ranking.
就业结果与可持续性：整体榜的另外十分
就业结果占 5%，看毕业生能否进入高技能岗位以及后续影响力。可持续性也占 5%，评估学校在环境与社会议题上的投入与治理。两项加起来是十分之一，学科榜都没有。对本科申请者来说，可持续性与你要读的课程关系有限，但它会实实在在影响学校的总名次。两所学校在学术声誉与研究指标上接近时，可持续性与国际化分数的差异就能决定谁排在前面。想判断某所学校的名次是不是被这些外围指标撑起来的，打开完整视图比看总名次快得多。反过来说，名次下滑也不必紧张，可能只是外围指标发生了变化。
Employment outcomes and sustainability contribute 5% each to the overall table, and both are absent from the subject ranking. Employment outcomes tracks whether graduates enter high-skill roles and go on to make an impact, while sustainability covers environmental and social governance. Together they are a tenth of the overall score, which is enough to separate two institutions with similar research profiles. Checking the full view of indicator scores is the quickest way to see whether a high overall rank rests on those columns.
国际研究网络：名字相同 测的范围不同
国际研究网络在整体榜里占 5%，在学科榜里也出现，但测的范围不一样。学科榜按五大学科领域计算，即艺术与人文、工程与技术、生命科学与医学、自然科学、社会科学与管理，不逐科出分。这项指标看的是合作院校在地理上是否分散，而不是合作数量多不多。一所学校在整体榜里的网络分与在学科榜里的网络分可以不同，因为合作关系本身按领域聚集。对想读研究型专业的学生，这一列比师生比更有参考价值。
The international research network carries 5% in the overall table and also appears in the subject table, but it is measured at broad faculty-area level rather than for each narrow subject. The five areas are arts and humanities, engineering and technology, life sciences and medicine, natural sciences, and social sciences and management. The indicator rewards geographically diverse partnerships rather than a large number of them. Its score can therefore differ between the two tables for the same university, since collaborations cluster by field.
最新两版实测：2027版整体榜与2026版学科榜
用最新两版对照最直观。QS 官网在 2026 年 6 月 18 日公布整体榜 2027 版，收录 1,504 所院校、98 所新入榜，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。MIT 连续第十五年第一，帝国理工与斯坦福并列第二，牛津第四，哈佛第五。学科榜 2026 版在 2026 年 3 月 25 日发布，覆盖 1,908 所院校、55 个学科与五大学科领域，条目超过两万一千条。艺术与人文领域由牛津居首，MIT 在十二个学科里排名第一。把两组数字放在一起看，学科榜的收录面比整体榜宽出约四百所院校。专科型学校的名字就是这样出现在学科榜上的。
The two current editions make the difference concrete. QS published the 2027 overall ranking on 18 June 2026 with 1,504 institutions, 98 new entrants and 106 higher education systems, including 184 from the United States, 93 from the United Kingdom and 85 from mainland China. The 2026 subject edition followed a different calendar, released on 25 March 2026 with 1,908 institutions, 55 subjects and more than 21,000 ranking entries. MIT holds first place overall for a fifteenth consecutive year and also tops twelve subject tables, while Oxford leads the arts and humanities faculty area.
专科型院校为什么能在学科榜进前五十
学科榜收录的院校比整体榜多出约四百所，其中不少是体量小、没有医学院与工学院的专科院校。这类学校在整体榜上很难进入视野，因为它们在教学与国际化指标上拿不到高分。但在某一个学科里，它们的论文集中、每篇引用与 H 指数突出，名次就能进前五十甚至更前。选专业时看学科榜，比看整体名次更可能发现这类学校。很多学生以为名次靠后的学校一定差，其实那要看是哪一套表、哪一列。
The subject edition covers roughly 400 more institutions than the overall edition, including small specialist schools without medical or engineering faculties. Such institutions rarely surface in the overall table, because they score poorly on teaching and internationalisation measures. Within a single discipline their research is concentrated, and citations per paper plus the h-index can push them into the top 50. For course choice, the subject table is the more revealing of the two.
名次区间、跨年比较与并列位次
名次靠后时学科榜用区间表示，例如 151-200 或 201-250，区间内的学校没有先后之分，把它们当成精确排序会误导选校。跨年份比较要小心，权重每年复核，收录范围也在变，2027 版整体榜新增了 98 所院校，新入榜的学校会把后面一串位次往后挤。位次与分数是两件事，一所学校分数涨了名次却掉了，通常是别的学校涨得更多。还有一种情况是并列名次，2027 版整体榜的第二名就是并列，两所学校共享同一个位次。看榜单时先看分数再看名次，判断会更稳。
Below the top group, subject tables report bands such as 151-200 or 201-250, and institutions inside a band are not ordered. Year-on-year comparisons need care because weights are reviewed annually and coverage changes; the 2027 overall edition added 98 institutions, which shifts a long tail of positions downwards. A rising score with a falling rank is normal and usually means other institutions improved faster. Tied ranks are also common, as in the joint second place in the 2027 overall table.
按申请结构读榜：专业看学科榜 学校看整体榜
选专业时先看学科榜，因为它把每篇引用、H 指数与学科内的声誉问卷都算进去，与你将来要读的系更贴近。评估学校品牌、转专业空间与校园构成时看整体榜，师生比、国际学生比例、可持续性这些只在那张表里。常见的错误是只看一个总名次就定校，等入学后才发现所在院系的研究实力与那个数字感觉不符。稳妥的做法是把两套表都打开，学科榜看你要读的系，整体榜看学校能不能给你想要的环境。两个榜单都不测课程内容与考核方式，这两件事只能靠课程手册与开放日确认。如果时间有限，先查学科榜，再补整体榜的师生比一列。
Use the subject table when comparing the department you plan to join, because it weights citations, the h-index and subject-level reputation surveys. Use the overall table for brand, flexibility of switching courses and campus make-up, since class size, international ratios and sustainability live only there. Picking a university from one overall rank is a common mistake, and students often find the research profile of their department does not match that number. Neither table covers course content or assessment methods, so check the course handbook and open days for those.
发布时间线：三月看学科榜 六月看整体榜
两套榜的发布节奏比较固定。学科榜 2026 版在 2026 年 3 月 25 日上线，整体榜 2027 版在 2026 年 6 月 18 日上线，于是每年春夏之交，两份带不同年份的榜单会同时在用。写申请材料或与顾问沟通时，把版本与年份写清楚，例如写 2027 版整体榜的位次，比只写一个笼统的范围更容易核对。权重表与指标说明都挂在 QS 官网的方法页上，可以自行下载数据表核对。名次只是选校的起点，课程结构、考核方式与所在城市这些事，榜单一条都不会告诉你。
The two editions follow a predictable calendar. The 2026 subject edition went live on 25 March 2026, and the 2027 overall edition followed on 18 June 2026, so two differently dated tables are in use at the same time each spring. Always quote the edition and year when you mention a rank, because a bare figure is hard for anyone else to check. Indicator weights and definitions are published on the QS methodology pages, and ranks remain only a starting point for choosing a course.
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
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图