QS学术声誉调查 7000所院校 提名 权重 雇主
选校的时候，很多学生和家长都会看 QS 世界大学排名，却很少去查这份排名里权重最高的那一项数据是从哪里来的。学术声誉占 30%，是所有单项指标里权重最高的一项，它的来源是一份每年发给全球学者的问卷。QS 官网写明的口径是，这份问卷每年评估约 7,000 所院校的提名。这篇文章把这项调查的规模、受访者名单、提名规则与权重处理逐条讲清楚，文中数字与日期都能在 QS 官网与 QS Support 页面查到原文。
Most families read a ranking table without asking who supplied the numbers underneath it. The largest single weight in the QS World University Rankings belongs to academic reputation, and that figure comes from an annual survey of academics rather than from any direct measurement of teaching. The rules below set out the size of that survey, who takes part, and how much can be read into a gap of a few places.
学术声誉 30%：权重最高的单项指标
QS 官网把世界大学排名的指标分成五组，研究与发现占 50%，就业与成果占 20%，全球参与占 15%，学习体验占 10%，可持续性占 5%。研究与发现里有两个单项，学术声誉占 30%，师均引用占 20%。学术声誉是整份排名里权重最高的单项，比师均引用高出十个百分点。它对 QS 的其他榜单同样关键，学科榜里社科与管理类的权重是 50%，人文与艺术类高达 60%。所有指标权重每年复核一次，这也是同一所大学的名次会小幅波动的原因之一。很多学生以为名次变化一定反映学校实力变化，其实权重复核与数据更新同样会带来位移。
Academic reputation sits inside the Research and Discovery lens, which carries half of the total weight. On its own the indicator is worth 30% in the World University Rankings, and in the subject tables it rises much higher, reaching 60% in Arts and Humanities. Weightings are reviewed every year, so a small movement in a table is not automatic evidence that a university has changed.
每年约 7,000 所院校：提名池的规模
QS Support 的指标说明里有一句关键的话：通过学术问卷收集全球学者的判断，每年评估约 7,000 所院校的提名。这句话说明提名是开放的，学者可以在自己熟悉的地区与学科范围内写下任何一所院校，而不是从一份固定名单里勾选。7,000 所这个量级远大于最终出现在榜单上的院校数量，上榜名单只是被提名次数与权重综合起来靠前的一部分。一所学校被学者提名，和它是否出现在榜单上，是两件不同的事。常见的错误是把 7,000 理解成榜单院校总数，再拿它去推算上榜概率。
The wording on the QS support pages matters here: nominations for roughly seven thousand institutions are evaluated each year. Nomination is open rather than a tick-box list, so an academic can put forward any institution they know within their own subject and region. Being nominated and being ranked are separate outcomes, and the seven thousand figure describes the input rather than the published table.
受访者从哪里来：四类名单来源
学术声誉问卷不做随机抽样，它靠四类来源把问卷发出去：往年的受访者、院校自己提交的联系人名单、在 QS 报名入口自行登记的人，以及 IBIS 数据库。院校提交名单是最主要的一条通道，QS 官网给出统一的 CSV 模板，学校先上传到 QS HUB 平台，再由 QS 发出问卷。因为要符合数据保护要求，学校必须先向被推荐人取得同意，同意邮件的模板由 QS 统一规定，不能自行改写后群发。问卷也不接受 info@ 这类通用邮箱与私人邮箱，收件人应当是能判断学术卓越程度的教学或研究人员。
The survey does not sample at random. It draws on previous respondents, contact lists submitted by institutions, people who sign up themselves, and the IBIS database. Consent has to be obtained before a name goes on a list, and QS supplies the approved consent template so that a university cannot press its own network in its own words.
每校 400 个联系人：上限与 350 份回复的通知线
QS 官网规定，每所院校每份调查最多提交 400 个联系人，学术与雇主两份调查各自计算。参加商学院榜单的院校可以再提交 400 个联系人，但这批名单只用于相关的商学院榜单，不会回流到世界大学排名。提交过一次的名单会被使用两年，也就是提交当年加上后一年，学校不必每个周期都重新发一遍。当某所院校的回复超过 350 份时，QS 会发出通知，学校应当停止继续联系更多人。这条规则的意思很直接：问卷要的是更广的样本，不是某一所学校把自己人喊满。
Each institution may submit up to 400 contacts per survey, with a further 400 allowed for business school rankings and used only there. A submitted list stays in use for two years, so there is no need to send it again every cycle. QS sends a notice once an institution passes 350 responses, and the institution is expected to stop recruiting participants at that point.
受访者先写专业与地区：决定他能提名谁
问卷开头问的不是院校名字，而是受访者自己的情况：所在机构、职务层级、最熟悉的国别与地区、最熟悉的学科领域。这些答案决定后半部分问卷里他能看到并提名哪些院校。学术类受访者不能提名自己所在的院校，这一条写在 QS 官网的问卷结构说明里。提名数量也有上限，本国院校最多 10 所，国际院校最多 30 所，商学院最多 10 所。这些上限让问卷结果更接近一份专业判断的分布，而不是让某位学者把熟悉的院校全部写满。
Respondents begin by describing themselves: their organisation, seniority, the countries and regions they know best, and their subject area. Those answers determine which institutions they can nominate later in the form, and an academic cannot select their own university. Caps of ten domestic and thirty international nominations keep the exercise closer to a set of professional judgements than a list of everything a respondent happens to know.
问卷窗口：11 月开放，次年 3 月初截止
QS 官网给出的下一轮时间表很清楚，院校提交联系人的窗口从 2026 年 9 月 15 日开始，到 2027 年 1 月 31 日结束，学术声誉问卷与雇主声誉问卷都在 2026 年 11 月开放，到 2027 年 3 月初截止。名单提交得越早，联系人越早收到问卷。超出截止日期才收到的回复不会作废，它们会计入下一年的分析，所以晚交的名单只是当年用不上。QS 也写明，如果院校错过名单提交的截止日期，或者发出报名链接太晚，QS 无法保证这些联系人能在分析截止前完成作答。这条时间线解释了一件事：学术声誉数据每年更新，但更新的是上一轮收到的回复。
The published timetable for the next research cycle runs from 15 September 2026 to 31 January 2027 for contact submission, with both reputation surveys open from November 2026 until early March 2027. Replies that arrive after the cut-off are not discarded; they are counted in the following year's analysis. QS also states that it cannot guarantee a late list will be answered in time, which is why an early submission matters.
15% 与 85%：本土提名与国际提名的权重
收集到提名之后，QS 把每所院校的提名拆成本土与国际两笔计数，再按权重合并。世界大学排名里这两笔的权重是本土 15%、国际 85%。学科榜的权重不同，本土占 33%，国际占 67%。这种处理让问卷结果更接近全球学术圈的整体判断，而不是某一国学者对自己熟悉院校的集中提名。很多学生以为国际权重高就等于小国院校容易上位，实际还要看后面的地区与学科熟悉度权重，越界的提名会被降权。两笔计数都会先做归一化，再合并成一个百分制分数。
Once the replies are in, QS separates domestic from international nominations and combines them with fixed weights. In the World University Rankings the split is 15% domestic and 85% international, while the subject tables use 33% and 67%. The intent is a global reading of reputation rather than an aggregation of national preferences, and both counts are normalised before they are merged into a score out of 100.
地区与学科熟悉度：越界提名会被降权
一份提名如果来自受访者本地区与本学科之外，权重会降低，这是 QS 官网明写的处理方式。理由也写得很直白，让被过度代表的地区不至于盖住声音较小的地区。除了熟悉度，QS 还会按受访者所在国家套一层国别权重，把有国际提名的院校数量与来自该国的提名人数作比较，参与率过低的地区会被豁免，以免小样本带来噪声。世界大学排名与地区榜不使用国别权重这一层。此外还有年份权重，用来平衡不同年份回复量的差别。这几层权重叠在一起，说明问卷里单张提名并不等值。
A nomination made outside a respondent's own region or faculty carries a lower weight, a rule the QS website explains as a way of preventing over-represented regions from drowning out quieter ones. Country weights compare how many institutions in a country attract international nominations with how many nominators come from that country, and low-participation places are exempted to avoid small-sample noise. Year weights then adjust for differences in response volumes between cycles, so a single nomination is never worth the same as another.
学科榜的额外规则：专科院校与跨榜比较
学科榜不直接使用总体学术声誉，而是把某学科的声誉与该学科所在学部的声誉作比较，学部榜则把学部声誉与总体声誉作比较。这样设计是为了不让整体声誉强的大学在每一个学科里都占便宜，也让专科型院校更容易被看见。被识别为专科型、并且在相关学科或学部开设课程的院校，可能获得额外加成。只在一个具体学科内作答的学者，其提名权重更高。地区权重同样会按受访者自称熟悉的地区套用。看学科榜时不能把它的分数直接当作总体声誉的换算结果。
Subject rankings compare reputation in a narrow subject against reputation in the wider faculty, while faculty tables compare the faculty against overall reputation. The design gives specialist institutions a fairer hearing and stops broad universities from trading on brand alone. Respondents who work in a single discipline carry more weight in those tables, and regional weighting is applied again on top.
雇主问卷与其他指标：学术声誉只占一部分
声誉数据不只来自学者。雇主声誉问卷占世界大学排名 15%，受访者是民营或公共机构里参与招聘决策的人，高等教育与研究机构的人员不在邀请范围。其余指标各有分工，师均引用占 20%，师均学生比占 10%，就业成果占 5%，国际教师比 5%，国际研究网络 5%，国际学生比 5%，可持续性 5%，国际学生多样性目前为 0%。把 30% 与 15% 相加，声誉类指标一共 45%，接近总分的一半。这意味着排名里接近一半的分数来自问卷判断，而不是可以独立复核的运行数据。
Reputation data also comes from employers: the employer survey is worth 15%, and its respondents must be involved in hiring rather than working in a university or research institute. Citations per faculty carry 20%, faculty student ratio 10%, employment outcomes 5%, international faculty ratio 5%, international research network 5%, international student ratio 5% and sustainability 5%. The two reputation indicators together account for 45% of the total, which is worth remembering before a ranking is treated as a neutral measurement.
数据清洗与处罚：为什么核查清单不公布
QS 官网写明，问卷汇总后会做一系列校验，包括清理自我引用与高度关联的作者数据、对异常值做抑制、剔除可疑数据。清洗过程以非人工方式完成，并且明确不公布完整的核查清单，理由是防止有人针对规则做游戏。规则的另一面是处罚：如果一所大学被认为在引导联系人如何作答，例如不按 QS 规定的模板发信，或者有其他不端做法，可能被处以一个周期内不参与排名、撤销当年提名等处分。这些条款都写在 QS Support 的问卷指引页上。
After collection, QS runs automated checks that strip self-citations and excessive affiliations, damp outliers and reject spurious data, and it deliberately keeps the full list of validations out of public view so that the process cannot be gamed. Universities that appear to coach respondents, or that ignore the prescribed consent templates, can face sanctions ranging from exclusion for a cycle to the removal of that year's nominations.
这份调查对选校的实际用法
知道调查规模之后，用法会不一样。7,000 所院校的提名池说明这是一份覆盖面很广的声誉问卷，但学术声誉反映的是学者对研究与合作声誉的判断，它不衡量课程设置、班级规模或本科生的上课体验。想比较具体专业，优先看学科榜，因为学科榜给学术声誉的权重在社科与管理类是 50%，人文与艺术类是 60%。同一档次里相差几位，不必当成决定性差异，权重每年复核，名次本来就会小幅移动。想看教学相关的信息，可以回到学习体验与师生比这类指标，或者直接查课程结构与课时安排。
A nomination pool of about seven thousand institutions is a broad reputation exercise, but it measures how academics judge research standing rather than how a course is taught. For one subject, the subject tables matter more because academic reputation carries 50% or 60% of the weight there. A gap of a few places is a weak signal on its own, since weightings are reviewed each year and the underlying data changes with every survey window.
常见误区：把调查规模当成录取难度
常见的错误是把排名位次直接当成录取难度或教学质量。声誉调查问的是学者愿意提名哪些院校，与某门课的录取比例、一位导师带多少学生没有直接关系。还有一个误区是只看总榜不看方法论，把不同口径的数字混着比较，学科榜的学术声誉权重明显高过总榜。想核对一份榜单，最省事的入口是 QS 官网的方法论页面与 Support 页面的指标说明，两页都写明了权重、数据来源与问卷窗口。看完这两页再回头看名次，判断会稳得多。
A ranking position is not an admission difficulty and not a teaching-quality score, yet it is often read as both. The survey simply records which institutions academics are willing to nominate. Anyone checking a table should start with the methodology page and the indicator notes, where the weights, data sources and survey windows are set out in full.
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