QS THE 卫报 UCAS 大学排名 权重与数据口径
选校时几乎每个家庭都会打开排名，但排名从来不是一份。QS 与 THE 比较的是全球研究型大学，卫报大学指南看的是英国本科的教学体验，UCAS 不发布排名，它发布申请与录取数据。同一所大学在这几处的位次可以相差几十名，原因不在学校变了，而在指标、权重与统计对象不同。这份材料把四套来源的规则拆开，说明每一张表到底在量什么。
Most families open a ranking before they open anything else, yet there is no single ranking, only several sets of rules. QS and THE compare research-intensive universities worldwide, the Guardian University Guide compares undergraduate experience in the UK, and UCAS publishes application and entry data rather than positions. A university can sit dozens of places apart across them because the indicators, the weightings and the populations differ, not because the institution changed.
QS 的五个维度与半数研究权重
QS 官网的方法论页面把得分拆成五个维度。研究与发展占 50%，其中学术声誉 30%、师均引用 20%。就业与成果占 20%，由雇主声誉 15% 与就业结果 5% 组成。全球参与占 15%，国际教师比例、国际研究网络、国际学生比例各 5%。学习体验只保留师生比一项，占 10%。可持续性占 5%。雇主声誉单独计权，这在主流全球榜里并不常见。研究与发展一项就吃掉一半权重，所以论文表现与声誉调查几乎决定了名次的基本盘。这五个维度里，没有一项直接测量课堂教学质量。
QS divides its score into five dimensions. Research and Discovery carries 50%, made up of academic reputation at 30% and citations per faculty at 20%, while Employability and Outcomes carries 20% through employer reputation at 15% and employment outcomes at 5%. Global Engagement takes 15%, Learning Experience is reduced to the faculty student ratio at 10%, and Sustainability holds 5%, so reputation and citation data decide half of the total before any other consideration.
QS 的声誉调查与入榜门槛
学术声誉这一项不是打分，而是提名。QS 官网说明，学者按学科与地区提名自己认为研究最强的机构，每个类别最多可以提交 400 个联系人。一所学校在单一年度收到超过 350 份回复之后，QS 会通知停止为它征集，名单使用两年。入榜还有门槛，学术声誉要进入全球前 20%，并且需要有 Scopus 收录的论文至少 100 篇，时间窗为五年。学生规模少于 5,000 人的院校要额外满足三项之一，单项进前 1,000、两项进前 1,200，或整体进前 1,300。2027 版一共有 1,504 所院校入榜。
Academic reputation is a nomination rather than a score. Academics name the institutions they regard as strongest in their field and region, each may submit up to 400 contacts per category, and QS stops collecting nominations for an institution once it attracts more than 350 responses in a year. Entry is gated as well, because a university must sit in the global top 20% for academic reputation and hold at least 100 Scopus-indexed papers over five years, while institutions below 5,000 students must meet one of three additional thresholds.
THE 的五大支柱与三项研究类指标
THE 官网的 2027 版方法论给出五大类，共 17 项指标。教学占 29.5%，里面有教学声誉 15%、师生比 4.5%、博士与本科比 2%、博士与教职员工比 5.5%、机构收入 2.5%。研究环境占 29%，由研究声誉 18%、研究收入 5.5%、研究产出 5.5% 组成。研究质量占 30%，是五类里最重的一块，引用影响占 15%，研究实力、研究卓越、研究影响力各占 5%。国际视野占 7.5%，国际学生、国际教职、国际合著各 2.5%。产业占 4%，产业收入与专利各 2%。教学类指标里有 15% 来自声誉调查，其余来自规模、收入与人数比例。研究环境与研究质量相加是 59%。
THE builds its 2027 table from five pillars and 17 indicators. Teaching carries 29.5%, research environment 29%, research quality 30%, international outlook 7.5% and industry 4%. Research environment and research quality together account for 59% of the score, which means publication, income and citation data shape most of the outcome, while industry income and patents carry only 4% between them.
THE 的声誉调查窗口与版本变化
声誉类指标在 THE 的体系里合计 33%，教学声誉 15%，研究声誉 18%。支撑 2027 版的那一轮学术声誉调查在 2025 年 11 月至 2026 年 2 月之间完成。2027 版结果在 2026 年 9 月 30 日公布，牛津第一，麻省理工第二，普林斯顿与斯坦福并列第三，剑桥第五。THE 官网同时写明，2028 版会给自引设上限，并调整研究质量类指标。很多学生以为名次是固定事实，其实它是某一版方法在某一年的产物。把发布年份与版本记住，比记住名次更有用。
Reputation accounts for 33% of the THE score, split between teaching reputation at 15% and research reputation at 18%. The survey behind the 2027 edition ran from November 2025 to February 2026, and the results were published on 30 September 2026, with Oxford first, MIT second and Princeton and Stanford joint third. THE has already confirmed that the 2028 edition will cap self-citations and adjust the research quality metrics, so a position describes one version of one method rather than a permanent property.
卫报为什么按专业而不是按学校排
卫报大学指南的做法与前两套完全不同。官网方法论写明，它把每个开课单位当作一个系，对 66 个学科分别排名，数据只覆盖全日制本科学位的学生。机构层面的总表不是独立算出来的，而是各专业表的加权平均。学生人数决定一个专业对学校总分的贡献，专业表包含的院校数量再决定该专业的影响力。一所学校要出现在八个以上专业表里，才会进入主榜。学生人数越多，这个专业越能拉动学校总分；专业表里院校越少，单个名次的波动越大。这也解释了为什么同一所大学在不同专业表里的相对位置会差很多。
The Guardian works at the level of the department rather than the whole institution. Its methodology page explains that each provider of a subject counts as a department, that 66 subjects are ranked separately, and that only full-time first-degree students enter the data. The institution table is a weighted average of those subject tables, a provider must appear in at least eight subject tables to be included, and that weighting means a university can look very different from one subject to the next.
卫报的八项指标与医学专业的例外
卫报用八项指标覆盖学生的整个学习周期。入学标准与师生比各占 15%，医学类专业分别是 24%。生均支出占 5%，医学类 10%。续读占 15%，医学类 10%，看的是学生是否读到大二。教学满意度与评估反馈满意度各占 10%，教学满意度在医学类升到 16%。增值分占 15%，医学类专业不使用。毕业去向占 15%，看毕业 15 个月后的升学与就业层级。入学标准统计的是新生实际平均分数，不是课程宣传的录取条件。
The Guardian uses eight measures that follow the whole student lifecycle. Entry standards and student-staff ratios carry 15% each, rising to 24% for medical subjects, expenditure per student carries 5%, and continuation and value added carry 15% each, with value added excluded for medicine. Teaching satisfaction and assessment feedback carry 10% each, and career prospects carry 15%, based on what graduates are doing 15 months after leaving.
UCAS 发布的不是名次而是录取口径
UCAS 官网并不发布大学排名，它发布的是申请与录取数据。UCAS Tariff 的用途写在官方指南里，把不同资格放在同一把尺子上，方便院校比较申请者。Tariff 只在 Level 3 或苏格兰 SCQF Level 6 的资格上生效，GCSE 一分也没有。院校不一定要使用 Tariff，也不一定要接受某个资格，它们可以只按等级发条件，比如 AAA。UCAS 官网还提醒，课程页写的是院校自己的录取要求，不是可以换算的名次。积分只是院校之间的一种比较工具，它不衡量资格难度，也不保证录取。很多学生以为积分够了就等于满足条件，这个理解会直接导致申请失手。
UCAS publishes application and entry data rather than a ranking. The Tariff exists so that providers can place different qualifications on one scale, and it applies only to Level 3 or SCQF Level 6 qualifications, with GCSEs carrying no points at all. Providers do not have to use Tariff points or accept a qualification simply because it is listed, so a points total never replaces the conditions printed on the course page.
UCAS Tariff 的换算方式
Tariff 的算法是档位相乘，资格的大小档乘以成绩档。A-Level 落在 4 档，A* 对应 14 档，乘出 56 分，A 对应 12 档，乘出 48 分，B 对应 10 档，乘出 40 分。往下是 C 32 分、D 24 分、E 16 分。拓展项目资格的最高分是 28 分，等于半个 A-Level。T Level 的最高档 Distinction* 是 168 分，高于三个 A* 的总和。学徒制也已纳入 Tariff，按预期时长分档，36 个月及以上的合格等级是 112 分。
The Tariff multiplies a size band by a grade band. An A Level sits in size band 4, so an A* takes grade band 14 for 56 points, an A takes band 12 for 48 and a B takes band 10 for 40, with C at 32, D at 24 and E at 16. An EPQ A* is worth 28 points, a T Level Distinction* reaches 168, and Level 3 apprenticeships are tariffed by expected duration, with a 36-month programme passing at 112 points.
四套来源为什么给出不一样的名次
四套来源的统计对象不同，名次对不上是设计使然。QS 与 THE 比较的是全球研究型大学，指标里含论文、引用与声誉调查。卫报只看英国院校的全日制本科，按专业排名，还包含生均支出与续读。UCAS 完全没有名次，只有课程层级的录取要求与积分换算。一所研究实力强但本科班额偏大的大学，可能在 THE 靠前，却在卫报的师生比一项吃亏。三套表的收录门槛也不同，QS 要求学术声誉进入全球前 20%，卫报要求学校至少出现在八个专业表里。常见的错误是把 QS 名次直接当作录取难度，QS 的指标里没有任何一项测量录取竞争度。
The four sources measure different populations, so their orders disagree by design. QS and THE compare research-intensive universities worldwide using publications, citations and reputation, the Guardian looks only at full-time undergraduate provision in UK departments, and UCAS offers no position at all, only course-level requirements and a points scale. A research-heavy university with large undergraduate classes can rank well on one table and lose ground on the Guardian student-staff ratio, and treating a QS place as a measure of admission difficulty is a mistake, because nothing in that framework measures competition for places.
读表时真正要看的四项数据
拿到一张表，先看年份与统计对象，再看与自己专业对应的那张表。卫报官网的方法论页写明，表里的数据主要来自 2024/25 学年，续读与就业两项涉及更早的队列，毕业去向取自毕业 15 个月后的调查。想知道能否被录取，回到 UCAS 官网的课程页逐条读要求。想知道课程支持力度，看续读与增值分。想知道毕业去向，看就业层级的口径，卫报把专业、管理与技术类职位以及继续深造都算作正向结果。
Before trusting any table, check its year and its population, then find the subject table that matches your own course. The Guardian's methodology states that its data comes mainly from 2024/25, that continuation and career metrics refer to earlier cohorts, and that career prospects come from a survey taken 15 months after graduation. Admission conditions live on the UCAS course page, support shows up in continuation and value added, and destinations sit in the career metric, which counts professional, managerial and technical occupations and further study as positive.
版本与年份带来的比较陷阱
排名每年更新，规则也会跟着动。THE 官网已公布 2028 版要给自引设上限，并调整研究质量类指标。QS 的权重每年复核，指标构成近年有过增删。卫报的方法论写明，结构自 2008 年以来大体稳定，但数据来源机构近年出现过中断，某些指标因此改用两年平均。一份表如果不知道它的年份与版本，比较就失去意义。常见的错误是拿 2023 年的名次去解释 2026 年的申请结果。
Tables are refreshed every year and their rules move with them. THE has announced a self-citation cap and adjustments to research quality metrics for 2028, QS reviews its weightings annually, and the Guardian records that while its structure has been broadly stable since 2008, disruption at the data providers has forced two-year averages in some metrics. Comparing two positions without knowing their editions is meaningless, and quoting a 2023 placing to explain a 2026 application outcome is a common error.
选校前的四步核对
定下课程之前，可以做四步核对。先确认表格的年份与统计对象，再找出与自己专业对应的那张表，因为卫报按专业发布，机构总表只是加权平均。接着回到 UCAS 官网的课程页，逐条读出等级与科目要求，注意条件里可能写着某个指定科目。然后把目标院校的续读、增值与毕业去向看一遍，医学类专业的权重与普通专业不同。如果条件里写着指定科目，积分再高也替代不了那一门。做完这四步，排名就从一个个名次变成可以追问的依据。
Before shortlisting, run a short check. Confirm the edition and the population, then find the subject table that matches your course, since the Guardian publishes by subject and the institution table is only a weighted average. Read the entry requirements on the UCAS course page line by line, including any named subject, and finish with the continuation, value added and destination data for that department, remembering that medical subjects use different weights.
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