QS排名 师均引用率 引用数除以教员人数
QS 世界大学排名的评分表里，师均引用率（Citations per Faculty）占 20%，是权重排第二的单项指标。它的算法一句话能说完：一所院校的论文引用总数，除以这所院校的教员人数。分子和分母来自两个不同的来源，分子取 Elsevier 的 Scopus 数据库，分母是院校层面的教员规模。很多学生以为这一项在量老师的水平，它量的是人均科研产出被引用的密度。同一份引用总量，放在四千人的院校和五百人的院校里，师均值可以差好几倍。搞清楚这次除法在解决什么问题，比记住 20% 这个数字更有用。QS 官网把这一项列在 Research and Discovery 这一组里，与学术声誉合计 50%。
Citations per Faculty carries a 20% weighting in the QS World University Rankings, second only to Academic Reputation. The calculation looks simple: the citations attracted by an institution's papers, divided by the number of academic staff it reports. Because the two figures come from different sources, the ratio measures research density rather than institutional size.
师均引用率在 QS 总分里占 20%
QS 官网的方法页把排名拆成五组 lens，Research and Discovery 这一组占 50%。这一组里只有两项指标，学术声誉 30%，师均引用率 20%。QS 官网写明权重每年复核一次，所以 20% 不是永久固定的数字。QS 2027 版在 2026 年 6 月 18 日发布，收录 1500 多所院校，评估池里有 8000 多所。很多学生以为排名只有一个总分，其实每一项指标都能单独查。看总分之外，还要看这一项在自己关心的学科里是什么位置。总分相同或接近的两所院校，科研面板可以差得很远。
The QS methodology page groups its indicators into five lenses, and Research and Discovery accounts for 50% of the total. Inside that lens there are only two indicators, with Academic Reputation at 30% and Citations per Faculty at 20%; QS states on its website that those weightings are reviewed every year. The 2027 edition was published on 18 June 2026, covers more than 1,500 institutions, and draws on an assessed pool of over 8,000.
分子取五年发表窗口，引用数看六年
QS 官网说明，师均引用率用到两套数据，一套是五年窗口内的论文数，另一套是这批论文收到的引用数。引用数按六年统计，比论文窗口更长。新发表的论文需要时间累积引用，窗口不等长就是为了给近年产出留出时间。每年第一季度，QS 从 Scopus 拿到一份数据抽取，供新一轮排名使用。引用不是论文发表当年就结算的，它会随年份往后累积。常见的错误是把这一项当成某一年度的快照，其实它一直跟着数据窗口滚动。读榜单时记住窗口长度，才能理解名次为什么逐年微调。
QS gathers two datasets for this indicator: a paper count over a five-year period, and the citations those papers collect across six years. The citation window runs longer than the publication window so that recent output has time to attract attention. An extract arrives from Scopus in the first quarter of each year and feeds the new ranking cycle, which is why the indicator rolls forward with the data rather than freezing at publication.
引用总量随院校规模上涨
论文总量和引用总量都会随院校规模上涨，这是学术出版的基本规律。一所拥有医学院、工学院和理学院的综合大学，五年里能发表上万篇论文。这些论文收到的引用层层叠加，总量很容易到几十万次。只看总量，规模小的院校几乎没有机会进入前列。不做任何处理的引用总量，等于给大院校直接加分。分母放上教员人数，就是把这个规模因素从比较里剥出来。人均之后，比的不再是谁的体量大，而是谁的产出被引用得更密集。
Raw citation totals rise with institutional scale, because more researchers publish more papers. A large multi-faculty university may produce tens of thousands of papers in five years, and the citations those papers collect accumulate into totals that a small specialist college cannot match. Comparing those totals without any adjustment would simply reward size, whatever the quality of the research behind them.
除以教员人数解决的是院校之间的可比性
QS 官网对这一步的解释很直接，引用数除以教员人数，是为了把不同院校的规模差异考虑进去。人均之后，一所只有五百名教员的研究型院校，可以和一所有四千名教员的综合大学放在同一张表上比。这个比值量的是平均水平，不是上限，也不代表每一位教员都被引用。分母数的是人，分子数的是论文收到的引用，两者之间没有一对一关系。常见的错误是拿这个比值去评价单个老师。它描述的是院校层面的科研密度，落到个人身上并不成立。看这一项的时候，把它当成院校属性，不要当成教师评语。
QS gives a plain reason for the division: the citation count is divided by the number of individuals in the faculty in order to take into account different sizes of institution. The result is an average rather than a ceiling, and it says nothing about how many citations any one academic has attracted. Read it as a property of the institution, not as a verdict on the staff who work there.
数据源是 Elsevier 的 Scopus
引用数据不是 QS 自己数的，它来自 Elsevier 的 Scopus 数据库。QS 官网把 Scopus 描述成一个来源中立的摘要与引用数据库，由独立学科专家维护。每年第一季度的抽取，是这一轮排名全部引用数据的来源。很多学生以为排名机构会自己统计论文与引用，实际背后是第三方数据库在支撑。数据库的收录范围决定了哪些论文根本没被算到。这也是为什么一所院校能否进入榜单，会先看它五年内在 Scopus 被索引的论文数量。同样一篇论文，被数据库收录与被数据库忽略，是两种完全不同的处境。
QS does not count citations by hand, and the underlying data comes from Elsevier's Scopus database. QS describes Scopus as a source-neutral abstract and citation database curated by independent subject matter experts, and an extract is supplied in the first quarter of each year for the new ranking cycle. What Scopus indexes therefore shapes what can be counted at all, and that is also why the entry threshold refers to Scopus-indexed papers.
作者层面的自引不计入
自引会被排除，QS 官网写明用的是作者层面的定义，而不是期刊层面或院校层面的定义。也就是说，一位作者引用自己作品的次数，不进入这位作者的引用计数。排除自引是为了避免院校通过内部互引把引用数做大。期刊层面的定义会把同一本刊之间的引用关系都算作自引，范围过大，作者层面更细。这一类处理细节不会出现在榜单页面上，只写进方法说明。做申请研究时如果想引用具体口径，应该回到 QS 官网的方法说明，而不是二手解读。引用口径差一点，结论就可能反过来说。
Self-citations are excluded, and QS states that the definition used here is author-level rather than journal-level or institution-level. A citation of an author's own work by that same author therefore never enters the citation count for the institution, however often it is repeated. The narrower definition matters, because a journal-level rule would sweep in a great many citations between papers that merely share a publication venue.
论文类型先筛一遍，评论与勘误类不计入
分子里的论文要先过一道类型筛选，一些非研究性的内容类型不计入。QS 官网把具体类型的定义交给 Elsevier 的 Scopus，方法页只写明会有排除。一篇东西在数据库里被归为哪种类型，会决定它是否进入计算。很多学生以为发出来的东西都会算论文，这个想法在数据库口径里不成立。跟踪这一层筛选的意义在于，它让引用密度的比较更接近研究本身。想核对一所院校的口径，先看它在 Scopus 里的论文数量，再看引用。数量与引用一起看，才能判断一所院校的科研面板是不是被少数高引论文撑起来的。
Before anything is counted, certain content types are excluded from the analysis, and QS points to Elsevier's Scopus for the definitions of those types. Whether a piece of output is treated as a paper therefore depends on how the database classifies it, not on how the author describes it. That screening is meant to keep the indicator close to genuine research output, and it is also why a long list of publications does not automatically translate into a high citation count.
五大学科领域各占 20%：先按学科归一
引用习惯在不同学科之间差别很大，生命科学与医学、自然科学的论文更容易被大量引用。QS 官网写明，如果直接用人均引用做比值，这些领域会被过度突出。为了拉平影响，QS 把研究分成五个学科领域，每个领域对这项指标贡献 20%。这一步叫学科领域归一，不是把各科论文简单加总。归一之后，人文与社会科学的院校不会被引用密度直接压下去。这也是为什么同一所院校在不同学科排名里的位置，可能和世界大学排名里的位置差得远。看排名时先问一句，这个名次是哪个口径下的名次。
Citation habits differ sharply between disciplines, and a straight ratio would place heavy emphasis on life sciences and medicine as well as on natural sciences. QS therefore equalises the influence of research across its five key faculty areas, so that each contributes 20% to the final indicator. That normalisation is the reason a strong humanities department is not automatically pushed down by a medical school with far higher citation traffic.
论文按 Elsevier 的 ASJC 学科码归入领域
论文归入哪个学科领域，不是靠人工评审，而是按 Elsevier 的 ASJC 学科码分类。QS 官网写明分类依据是这套学科码，它把期刊与论文映射到固定的学科目录。一所院校的引用总量先按科目分开，再在五个领域之间做平衡。这里有一个容易忽略的点，跨学科论文只会落在它的主分类里。常见的错误是以为归一化会按论文的实际内容重新判断。实际做法是数据库怎么归类，就按那个归类进入计算。了解这一层，就不会把学科归一误读成主观打分。
Papers are assigned to faculty areas through Elsevier's ASJC subject codes rather than by panel judgement. That mapping decides which of the five areas a paper counts towards, and the balance between the areas is then applied to the institution's citation totals. Nobody re-reads the papers to decide where they belong, so the classification follows the database rather than an editorial view of each piece of work.
一个算例：总引用量大的院校师均值可能更低
假设甲校五年内的论文共收到四十万次引用，全校教员四千人。人均引用就是一百次。假设乙校同年限收到八万次引用，教员五百人，人均引用是一百六十次。乙校的引用总量只有甲校的五分之一，人均值反而高出六成。这个例子说明总量排名与人均排名可以给出相反的结论。如果甲校的引用集中在生命科学，归一之后人均值还会再变一次。所以师均引用率适合回答科研密度这类问题，不适合直接回答谁的教学更好。把它放进选校清单时，先想清楚你要问的是哪一个问题。
Suppose one university receives 400,000 citations over five years with 4,000 academic staff, which works out at 100 citations per faculty member. A second institution with 80,000 citations and 500 staff reaches 160 per member, so it sits ahead despite attracting a fifth of the citations. Totals and averages can therefore point in opposite directions, and the normalisation step can change the gap again.
这一项量的是科研密度，不是本科教学
QS 官网把师均引用率定义为院校科研的相对强度与体量。官方给出的判断是，引用多通常意味着教员在受认可的期刊上发表、有较强的合作、选题有较广的读者。这些判断都落在科研一侧，没有涉及课堂、作业批改和课程设计。教学体验在 QS 的指标体系里有另一个代理项，就是权重 10% 的师生比。很多学生以为师均引用高的学校课上得更好，这两个结论之间没有直接链路。选校时把这一项当成科研面板读，再去看教学相关的官方口径，判断会稳一些。两项指标各答各的问题，混着用只会让判断更糊。
QS describes the indicator as a measure of the relative intensity and volume of research at an institution. A higher volume of citations suggests that staff are publishing in respected journals, engaging in strong collaborations and working on topics that merit a wide readership; none of that is evidence about teaching. QS handles the learning side through a separate indicator, the Faculty Student Ratio, weighted at 10%.
门槛与读法：至少 100 篇五年内 Scopus 论文
进入 QS 世界大学排名要先满足几条门槛，学术声誉进入全球前 20%，五年内在 Scopus 被索引的论文至少一百篇，并且进入所在区域的区域排名。规模不足五千名学生的院校还要额外满足指标条件。这些门槛意味着纯教学型院校和很小的专业院校，可能根本不出现在这项指标里。读这一项时有两个动作值得做，一是看学科排名里的位置，二是看五年趋势而不是单年名次。QS 官网写明权重每年复核，口径调整会带来名次移动。选校时把这一项和英国学生办公室的 Discover Uni 官方数据放在一起看，信息更完整。排名是入口，课程页上的官方数据才是你要读的那一页。
Eligibility comes first, because an institution must be in the top 20% globally for Academic Reputation, have at least 100 papers indexed by Scopus over five years, and appear in a regional ranking. Teaching-led colleges and very small specialist providers can therefore be absent from the indicator altogether. Reading subject-level results and five-year trends tells you more than a single year's position, especially since QS reviews the weightings annually.
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