QS星级评级 大学排名 指标权重 选校决策
每年排名发布以后，QS 星级评级与 QS 世界大学排名这两个名字总会被混在一起使用。前者是一套自愿参加的审计式评级，后者是一张按加权指标排序的全球榜单。2026 版排名在 2025 年 6 月 19 日发布，QS Stars 则在 2024 年 1 月换过一次方法。本文把两套工具的指标、门槛与读法拆开讲清楚，也给出一份申请期的使用顺序。
Every year, as the new tables arrive, QS Stars and the QS World University Rankings get quoted as if they were the same product. One is a voluntary audit-style rating; the other is a global league table built from weighted indicators. The 2026 ranking was published on 19 June 2025, and the QS Stars framework was revised in January 2024. This article separates their indicators, thresholds and reading habits, then sets out an order of use for application season.
一个给名次，一个给维度
QS 官网把这两套工具的分工写得很清楚：排名回答的是哪所学校表现最好，评级回答的是这所学校好在哪些方面。排名把所有入榜院校放进一条线性队列，从第 1 位排到第 1500 位左右。评级不给人排名次，它把学校放在一套统一标准下逐项打分。很多学生以为学校拿到五星就等于排名很高，其实这两件事没有换算关系。一所专科色彩很强的院校可能在星级里拿到满分，却在综合排名里排到几百位。反过来，一所综合排名前 50 的大学也可能因为类别选择与数据提交范围而只拿到四星。看清这套分工，你才知道什么时候该看名次，什么时候该看徽章。
QS draws the line between the two tools quite plainly: a ranking shows who performed best, while a rating shows what an institution is good at. Rankings squeeze every eligible university into a single order, from number 1 to roughly number 1500; a rating hands out no position at all, scoring each institution against the same fixed standards. Many students assume a five-star badge must mean a high rank, yet the two figures cannot be converted into each other; a specialist institution can collect full marks in a rating and still sit several hundred places down the world ranking. A university inside the global top 50 may hold only four stars, because of the categories it chose and the data it submitted.
2026 版排名的十个指标与权重
2026 版 QS 世界大学排名是 QS 官网公布的第 22 版，发布日期为 2025 年 6 月 19 日。它把十项指标归入五个维度，再按权重加权成总分。研究实力与发现这一维度占 50%，其中学术声誉 30%、师均引用 20%。就业与产出这一维度占 20%，其中雇主声誉 15%、就业成果 5%。学习体验就是师生比一项，占 10%；全球参与占 15%，拆成国际教师比、国际学生比与国际研究网络各 5%。可持续性指标单独占 5%，另外还有一个国际学生多样性指标，2026 版只作预览、不计权重。把这些数字记住，你就能判断一所学校名次变动到底是被哪一块数据推动的。
The 2026 edition of the QS World University Rankings is the 22nd in the series, published by QS on 19 June 2025; ten indicators are grouped into five lenses and combined into one weighted score. Research and Discovery carries half the weight, made up of Academic Reputation at 30% and Citations per Faculty at 20%; Employability and Outcomes accounts for 20%, split into Employer Reputation at 15% and Employment Outcomes at 5%. Learning Experience rests on the Faculty Student Ratio alone at 10%, while Global Engagement takes 15%, divided equally among international faculty, international students and international research networks. Sustainability holds 5%, and a new International Student Diversity measure appears as an unweighted preview rather than a scored indicator.
声誉调查与论文引用怎么进入总分
九个加权指标里，学术声誉与雇主声誉两项加起来占 45%，它们来自 QS 的全球学术调查与全球雇主调查。院校可以各提交最多 400 个联系人名单，问卷由 QS 统一发出，受访者评价的是全球院校而不只是提名自己的那一所。师均引用占 20%，用 Scopus 数据库里五年内的论文引用数除以教师人数。QS 官网说明，2026 版把纳入研究指标的院校范围扩大了六成以上，覆盖 8,000 多所，因此引用类指标的名次会出现年际位移。很多学生以为名次下滑就说明学校变差了，其实口径变化本身就会推动一批学校上下移动。判断这类变动，要看该校在具体指标上的位置，而不是只看总名次。
Reputation carries the largest single share of the weighting: Academic Reputation at 30% and Employer Reputation at 15%, both drawn from the QS global surveys. Institutions may submit up to 400 contacts for each survey, and respondents rate universities worldwide rather than only the ones that nominated them; the citations side, worth 20%, divides five years of Scopus citations by faculty numbers. QS states that the 2026 edition widened the pool of institutions assessed for research indicators by more than 60%, covering over 8,000 universities, which shifts citations-based positions from year to year. A fall in rank therefore reflects a change in measurement as often as a change in the institution itself; the indicator table is more useful than the headline number, because it shows which of the nine components moved.
入榜门槛与名单长度
QS 世界大学排名每年把约 1,500 所院校放进榜单，但评估范围超过 8,000 所，2026 版覆盖 100 个地区。要获得参评资格，院校必须同时开设本科与研究生完整学位，至少覆盖两个学部，每个学部下至少两个学科，每个学科至少有三届毕业生，并且课程至少部分在校面授。要真正入榜，学科声誉须进入全球前 20%，五年内 Scopus 收录论文要有 100 篇以上，位于亚洲、阿拉伯地区、欧洲与拉美加勒比的院校还必须先进入本地区排名。学生规模少于 5,000 人的院校还要在学术声誉、雇主声誉、师均引用三项里至少满足一项特殊条件。常见的错误是拿一份榜单当成全球高校全集，其实没上榜不等于学校不合格。选校时先确认你想读的课程在不在榜内，再讨论名次高低。
Around 1,500 institutions appear in each edition of the World University Rankings, although QS assesses more than 8,000 and the 2026 edition covered 100 locations. Eligibility demands full degree programmes at undergraduate and postgraduate level, at least two faculty areas with two subjects each, three graduating cohorts in every subject, and at least part of each programme delivered on campus; inclusion then requires an academic reputation inside the global top 20% and at least 100 Scopus-indexed papers over five years, with institutions in Asia, the Arab Region, Europe and Latin America needing a place in their regional ranking first. Smaller universities below 5,000 students face an extra test on at least one of three indicators: academic reputation, employer reputation or citations per faculty. A missing entry usually means an institution failed a technical threshold rather than that it teaches badly; check the course itself before reading anything into the absence.
QS Stars 是自愿申请的审计式评级
QS Stars 从 2009 年开始运行，QS 官网把它描述成一次审计，而不是一张排行榜。院校自愿申请参加，不设名额上限，想参加的学校都有机会进入评估流程。2024 版方法文件写明，全球已有 700 多所院校完成评估，QS 会为每所学校指派一名分析师，从数据收集到第三方核验全程跟进。评级结果分两层：总体徽章从 1 星到 5+ 星，每个类别另有 1 到 5 星的单类徽章。很多学生以为没有星级就是学校评级差，其实一所学校可能根本没有申请。想查某校是否参评，最直接的动作是去 QS 官网的院校页面看它有没有公布徽章。把自愿申请这一点想清楚，你就不会用有无星级来给学校排座次。
QS Stars has run since 2009, and QS describes it as an audit rather than a league table; participation is opt-in with no cap on numbers, so any university willing to take part can be assessed. The 2024 methodology document records more than 700 institutions evaluated worldwide, and every institution is assigned a QS analyst who follows the process from data gathering to third-party validation. Results arrive as an overall badge from 1 to 5+ stars plus a separate badge, scored out of five, for each category examined. The absence of a badge therefore says nothing about quality, because a university may simply never have applied; the quickest check is the institution page on the QS website, where published badges are listed.
2024 年起 QS Stars 把可持续性写进评分
QS 官网说明，从 2024 年 1 月起 QS Stars 启用新方法，这是该项评级十五年历史里改动最大的一次。最大的变化是新增可持续性维度，它占院校总分的 20% 以上，下面分良好治理、社会或环境影响、多元平等包容三个子类。原有内容被重新组织成五个维度：可持续性、就业与产出、学习体验、研究与创新、全球参与。QS 同时下调了专项项目子类的权重，并把它改成可选类别。根据 QS 官网援引的国际学生调查，41% 的受访学生把可持续性视为越来越重要的考量。数据以官方方法文件为准，不同年份的星级之间不要直接相加或比较。想评估一所学校在可持续性上的实际动作，去看它在治理与减排上的公开文件，而不是只数徽章。
From January 2024 the QS Stars methodology changed, marking the largest revision in the scheme's fifteen-year history. Sustainability arrived as a new dimension worth more than 20% of the total, split into good governance, social or environmental impact, and diversity, equity and inclusion; the remaining content was regrouped into five dimensions covering sustainability, employability and outcomes, learning experience, research and innovation, and global engagement. QS also cut the weighting of the specialist programme sub-category and made it elective. In its own student survey, QS reports that 41% of respondents saw sustainability as an increasingly important factor; because the framework moves, star totals from different years should never be added together or read as a straight comparison.
标准审计的必选类别与自选类别
一次标准审计要覆盖若干必选类别，院校再挑两个自选类别一起评估。2024 版方法文件列出的必选项包括教学、设施或在线学习、社会影响或环境影响、研究或学术发展、全球参与、就业力、良好治理。教学这一类看师生比、学生满意度、完成率、留存率与升学情况；就业力看雇主声誉、毕业生就业率与就业服务支持；研究与创新看论文产出、引用影响、科研经费与专利申请。设施类看体育场馆、住宿与图书馆支出，包容性看残障支持服务、弱势群体项目与性别平衡。评估类别不同，星级的可比性就不同，所以两份五星证书背后的含金量可能差很远。看到学校宣传星级时，先问它参加了哪些类别，再判断这个成绩和你关心的事有没有关系。
A standard audit covers a fixed set of mandatory categories plus two elective ones chosen by the institution. The 2024 methodology lists teaching, facilities or online learning, social or environmental impact, research or academic development, global engagement, employability and good governance as required; teaching looks at staff-student ratios, satisfaction, completion, retention and progression, while employability weighs employer reputation, graduate employment rates and careers support. Research and innovation examines publications, citation impact, research income and patents, whereas facilities covers sports provision, accommodation and library spending, and inclusion looks at disability services, under-represented groups and gender balance. Because the elective choices differ from one university to the next, two five-star certificates need not mean the same thing; when a school advertises its rating, ask which categories were assessed and whether they touch the subject you plan to study.
星级门槛与 5+ 星的前置条件
QS Stars 的总体星级由累计得分决定，2024 版方法文件给出的门槛是 1 星 15 分、2 星 35 分、3 星 60 分、4 星 85 分、5 星 105 分。想拿到最高的 5+ 星，院校必须在所有被评类别里都达到 5 星，同时满足全部 5 星前置条件，两项缺一不可。分的计算方式是阈值缩放：达到最高阈值拿满分，落在最低阈值上得零分，中间按公式折算。教学类里教师与学生比、学生满意度、完成率这些指标各有自己的阈值，逐项累加才形成类别得分。很多学生把星级当成百分比或百分制，于是看到五星就以为学校排在全球前几。真正的读法是把它当体检报告，总分告诉你整体状态，分项告诉你强在哪、弱在哪，两者要一起看。
The overall star rating is driven by accumulated points, and the 2024 methodology document sets the thresholds at 15 points for 1 star, 35 for 2 stars, 60 for 3 stars, 85 for 4 stars and 105 for 5 stars. The top 5+ star badge demands a five-star result in every category assessed as well as all the five-star prerequisites, so neither condition can be skipped. Points follow a threshold-scaling rule in which the highest threshold earns full marks, the lowest earns nothing and anything between is converted by formula; inside teaching, staff-student ratio, satisfaction and completion each carry their own thresholds and add up to the category total. Reading a star rating as a percentage is the basic mistake here, so treat it as a health check instead: the overall score summarises condition, and the category badges show which parts carry the strength.
同一所学校两个结果为什么不冲突
排名的本质是比较，评级的目标是对标准，这是两套结果看起来矛盾的原因。在排名体系里，一所学校的名次取决于它与其他院校的相对位置，别人进步就可能让你后退。在评级体系里，每所学校单独对照标准打分，理论上所有院校都能拿到 5 星，名额不受限制。这就是为什么会出现排名第 200 位却握着五星徽章的情况：它在个别类别上达到了最高标准，只是整体研究体量不及综合强校。反过来，一所研究型名校也可能因为住宿、包容性或治理材料准备不足而在类别上失分。看这两套信息时，把它们当成两个不同问题的答案，一个回答相对位置，一个回答自身表现。真正值得警惕的是只用其中一个数字下结论。
A ranking is comparative by design, while a rating measures against fixed standards, and that difference explains most apparent contradictions. In a ranking, an institution's position depends on everyone else's performance, so a rival's improvement can push it down even when nothing about its own work has changed; in a rating, each university is scored alone against the criteria, and in principle every participant can reach five stars because there is no quota. This is how a university sitting around 200th in the world can hold a five-star badge, and how a research-heavy name can lose category marks on accommodation, inclusion or governance evidence. Treat the two outputs as answers to different questions, relative standing on one side and self-assessment on the other, with neither sufficient on its own.
看排名时最容易出错的四个地方
排名表里最容易被误读的是年际位移。权重与纳入范围每年调整，2026 版扩大研究指标覆盖范围后，一批院校的引用类位置出现移动，这属于口径变化而不是教学质量变化。另一个常见误读是把综合名次当成专业水平，同一所学校不同学科的位次可能相差几百名，想看专业就去看学科排名。并列与区间也常被忽略，榜单里并列名次很常见，名次差距在十名以内往往谈不上实质差异。把上榜当成办学质量认证同样站不住脚，入榜门槛包含声誉调查结果与论文数量要求，规模小的特色院校天生吃亏。常见的错误是拿一年的榜单定终局，连续看三到五年的区间比盯着某一年稳得多。核对学校真实水平时，去看它的课程结构与毕业生去向，那两项不会因为权重调整而变形。
Year-on-year movement is the first thing readers misread; weightings and coverage are reviewed annually, and once the 2026 edition widened research coverage, citations-based positions shifted for a batch of universities, which is measurement change rather than a change in teaching. A second habit is to read an overall rank as subject quality, when the same university can sit hundreds of places apart across departments, so subject tables are the right place to look. Ties and narrow gaps are ignored just as often, and a difference of ten places or fewer rarely means much. Assuming that a place in the table certifies quality fails for the same kind of reason, because entry depends on survey reputation and paper counts that penalise small specialist institutions; the reliable check is the course structure and graduate destinations, which no weighting change can rewrite.
看星级时要问的三个问题
拿到一张星级证书，先问它参加了哪些类别。必选类别每次审计都一样，自选类别由院校自己挑，所以两份证书的内容范围可能完全不同。再问每个类别的星数分别是多少，总体徽章高但某一类偏低的情况很常见，那一项往往正是你关心的部分，比如住宿、就业服务或线上教学。还要问数据是哪一年的，方法在 2024 年 1 月改过，标注旧年份的评级不能直接与新年份对照。三个问题问完，你会知道这张证书能不能回答你真正的问题。想比较两所学校，就把它们在同一类别上的星数与指标数据并排看，而不是比谁的徽章多。院系层面的实际体验还得回到课程手册与在读学生的反馈上确认。
Ask first which categories were assessed, because the mandatory ones are fixed while the electives are chosen by the institution and shape how much the certificate actually covers. Ask next for the individual scores, since a strong overall badge often hides a weak category, and that weak category is frequently the one that matters to you, whether accommodation, careers support or online delivery. Ask finally which year the data belongs to, because the framework changed in January 2024 and a rating dated before that cannot be compared with a newer one. Those three answers tell you whether the badge can address your question at all, and to compare two universities you line up their scores and indicator data in the same category rather than counting badges. Department-level experience still has to be checked in course handbooks and with current students.
申请季怎么把名次与星级用起来
选校时先定专业方向，再去看学科排名，用综合名次替代专业评估是最浪费的一种做法。确定两三所目标院校之后，用 QS Stars 的类别徽章补看设施、就业支持与包容性这些排名覆盖不到的信息，把招生网页上的课程结构、必修课与考核方式一起对照。入学要求以 UCAS 官网与学校官网为准，不要用榜单上的名次推断录取难度，两者没有对应关系。涉及学费、贷款与签证的数据，去 gov.uk 与 OfS 的公开页面核对，别用二手转载的数字。把名次、星级、课程结构与官方数据放进同一张比较表，每所学校写三行，你很快就能看出真正的差别在哪里。每年 6 月排名发布后重新跑一遍这张表，成本很低，收益是避免被单一数字牵着走。
Start with the subject, then read the subject tables, because replacing a subject assessment with a general rank wastes the most useful data available. Once two or three target universities are shortlisted, use the QS Stars category badges to fill in facilities, careers support and inclusion, none of which a ranking covers well, and set those against the course structure, compulsory modules and assessment pattern on the admissions pages. Entry requirements belong on the UCAS website and the university's own pages, and nobody should infer difficulty from a rank position, since the two are unrelated. Tuition fees, loans and visa rules are best checked on gov.uk and OfS pages rather than second-hand summaries. A three-line comparison table for each university, rebuilt after the June release, costs an hour and stops a single number from making the decision.
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