QS雇主声誉 15%权重 雇主调查 提名规则
雇主声誉是 QS 世界大学排名里唯一直接问雇主的指标。它占 15% 的权重，而 THE、软科世界大学学术排名与 U.S. News 的全球榜都没有同类问卷。这篇把这项调查的样本来路、提名上限与计分方式拆开讲清，也说明它为什么容易被误读。QS 官网的方法论页写明，世界大学排名是唯一纳入院校雇主声誉的排名。THE 官网的指标体系里只有发给学者的声誉问卷，软科官网的六项指标全部取自第三方数据库。
The employer reputation indicator is the one QS metric that asks hiring organisations directly, and it carries 15% of the weight in the QS World University Rankings, while the other major global tables run no equivalent employer survey. This piece follows the sample from the contact list to the score out of 100, and explains why the number is easier to misread than it looks. Read it before treating a five-place move as evidence that recruiters changed their minds.
雇主声誉在 QS 世界大学排名里占 15%
QS 世界大学排名把指标分成五组。研究与环境相关的一组占 50%，其中学术声誉 30%、师均引用 20%。就业与成果一组占 20%，雇主声誉 15%、就业结果 5%。也就是说，雇主问卷在总分里的位置是 15%，不是 20%。剩下的分数来自国际参与 15%、学习体验 10% 与可持续 5%。把这两组放在一起看，问卷类指标合计 45%。很多学生以为雇主声誉就是就业数据，其实它量的是雇主对这所学校的印象。
On the QS methodology page, employer reputation sits inside the Employability and Outcomes lens, which is worth 20% in total and splits into employer reputation at 15% and employment outcomes at 5%. The research side of the table carries more weight, with academic reputation at 30% and citations per faculty at 20%. Employers are judging a much smaller slice of the final score than most applicants assume.
QS 官网说自己是唯一纳入雇主声誉的排名
QS 官网的方法论页有一句直接表述。它写明世界大学排名是唯一纳入院校雇主声誉的排名。这句话可以在对比里验证。THE 官网的指标表只有教学声誉 15% 与研究声誉 18% 两项问卷，两份问卷都发给学者。软科官网的六项指标全部取自第三方数据，校友获奖 10%、教师获奖 20%、高被引科学家 20%、Nature 与 Science 论文 20%、SCIE 与 SSCI 论文 20%、人均表现 10%，没有问卷。U.S. News 的全球榜用 25% 的学术声誉，也是学者问卷。常见的错误是看到声誉两个字就以为各家都在问雇主。
QS states on its own methodology page that the World University Rankings are the only ranking to include an institution's employer reputation, and the comparison holds up. Times Higher Education asks academics rather than employers for its teaching and research reputation scores, while ARWU uses no survey at all and builds its six indicators from award records and citation databases. U.S. News spreads 25% of its global score across two academic reputation surveys, so for employers there is no equivalent question anywhere else in the field.
雇主问卷的样本来路与 400 位联系人上限
雇主问卷不是随机抽样。QS 官网写明，问卷分发到四类来源：往年回复过的雇主、院校提交的联系人名单、通过开放登记报名的雇主、以及调查合作方。院校可以为学术与雇主两份问卷各提交最多 400 位联系人，参加商科榜单的院校还能再多交 400 位。联系人名单的提交期在 9 月到次年 1 月，两份问卷的调查期在 11 月到次年 4 月。问卷开头先问受访者熟悉的地区与行业，后面的答题范围由此决定。雇主还要填写所在机构的规模与打算招聘的项目类型。很多学生以为这份问卷是匿名街访，其实是定向邀请加登记报名。
The employer sample is assembled rather than sampled at random; QS invites institutions to put forward contacts, re-contacts previous respondents, opens a sign-up route and works with survey partners. An institution may submit up to 400 contacts for each reputation survey, and business school participants may add a further 400, with contact submission running from September to January and the surveys themselves from November to April. That calendar is worth remembering whenever a ranking result is described as fresh.
一位雇主能提名多少所学校
提名有上限。QS 官网写明，受访者最多提名本国院校 10 所、国际院校 30 所，商科部分再单独提名最多 10 所商学院。能选哪些学校取决于受访者填写的地区熟悉度，不熟悉地区的学校不会出现在列表里。学术问卷里学者不能提名自己所在的院校，雇主侧的本国提名同样排除自提名。这些上限把问卷变成偏好排序，而不是完整评价。一所学校能拿到的国际票数，与它落在多少位受访者的熟悉范围内直接相关。很多学生以为票多就等于口碑好，其实覆盖面本身就是结果的一部分。
Respondents nominate within limits: up to 10 domestic institutions, up to 30 international institutions, plus a separate list of up to 10 business schools for the business sections. The options open to a respondent depend on the regional familiarity they declare at the start, and academics cannot select their own institution, while employers' domestic nominations exclude self-nominations in the same way. Those caps turn a survey into a preference ranking rather than a full evaluation of every university in the world.
国际提名怎么算：区域与国别熟悉度加权
提名不直接计数。QS 官网写明，国际提名先做加权计数，权重来自受访者的区域与国别熟悉度，也来自回复年份。熟悉度高的受访者，其提名的分量与只泛泛了解一所学校的人不同。国别权重用来处理不同国家受访者的覆盖差异。这一步做完，各校拿到的是一个加权票数，而不是原始票数。原始票数多的学校，在加权后未必排在前面。常见的错误是拿公开的回复总量去推算某校的票数。票数与分数之间隔着权重与转换两层处理。
International nominations are not counted one for one; QS weights them by the regional and country-level familiarity a respondent declares, and by the year in which the response arrived. A nomination from someone who works closely with a university's graduates does not carry the same weight as an offhand mention from a respondent who knows the region only loosely. Because those adjustments happen before anything is published, the raw number of responses tells you very little about how a particular institution scored.
本国提名按国别规模调整
QS 官网写明，本国提名先做加权计数，并且排除自提名。之后还要按两件事调整：该国达到一定国际提名水平的院校数量，以及该国回收的问卷总量。这样做是为了反映竞争。在院校多、名气大的国家，同一批雇主票要在更多学校之间分配。体量小的国家里，少数几所学校更容易拿到集中的提名。这一层调整让跨国比较不至于被国家规模带偏。常见的错误是把不同国家之间的分数差直接读成雇主偏好差。
Domestic nominations are weighted, stripped of self-nominations and then adjusted against two national factors: how many institutions from that country reach a certain level of international nominations, and how many responses the country returned in total. The aim is to account for competition, since employers in a large higher education system are spreading the same enthusiasm across far more institutions. Without that adjustment, countries with a handful of well-known universities would look stronger than they are.
五年滚动窗口：最老一年只按四分之一计
声誉分数不是一年一算。QS 官网写明，指标使用最近五年的提名合并计算。第五年也就是最老的一年按 25% 计，第四年按 50% 计，最近三年按 100% 计。这意味着某一年的问卷结果单独看并不会立刻改变名次。一份新样本要进入窗口并被加权，才会逐步体现在分数里。这个窗口也让问卷口径调整的影响变得平滑。对读者来说，名次突然变动往往来自多年的累积，而不是当年的风声。很多学生以为榜单反映的是当年的招聘行情。
The indicator rolls five years of nominations together: the fifth and oldest year is weighted at 25%, the fourth at 50%, and the three most recent years at full weight. A single year's survey therefore cannot move an institution on its own, which is why sudden shifts usually reflect several cycles of responses rather than one season of hiring news. Treat any story about this year's employers changing their minds with that lag in mind.
分数是怎么变成 100 分的
加权计数之后还要归一化。QS 官网写明，本国与国际两个计数会各自归一化到 100 分，再按世界大学排名里的 50% 与 50% 合并。之后还会做转换处理，压低极端值的影响，并把结果缩放到该学科领域的 100 分制。这套处理让分数可比较，也让分数不容易被个别异常票拉爆。排名页给出的单项分就是处理后的结果。常见的错误是拿两所学校的单项分做小数位比较。分数的可信区间比看似精确的数字要宽。
Domestic and international counts are each normalised to a score out of 100 and then combined on a 50-50 basis in the World University Rankings, with transformation applied to limit the pull of outliers. What appears on a profile page is the finished number rather than the vote count behind it. Comparing two institutions down to a decimal place on that indicator over-reads the precision of a survey built on human recall.
学科榜里的雇主声誉权重不一样
QS 官网的指标表列出各榜单的权重。世界大学排名是 15%，区域榜多为 20%，欧洲区域榜 15%，可持续榜只占 3%。学科榜按学部给权重，艺术与人文 20%、工程与技术 30%、生命科学与医学 10%、自然科学 20%、社会科学与管理 30%。商科榜单更看重雇主，全球 MBA 榜 35%、商业硕士榜 30%、在线 MBA 榜 30%。学科榜还有两条规则：某学科的雇主声誉会与该机构整体雇主声誉对比，被识别为专科强校的学校可能获得加成。回复者熟悉的学科越少，其提名权重越高。这一步是为了让专科强校不被综合声誉盖住。
Employer reputation is not a fixed 15% across the portfolio; QS applies different weights by ranking, from 3% in the sustainability table to 35% in the Global MBA table, and the subject rankings range from 10% in life sciences and medicine to 30% in engineering and technology and in social sciences and management. In subject tables a university's employer score for that field is also compared with its overall employer score, and specialist institutions may receive a boost. Responses from employers who claim knowledge in fewer disciplines are weighted more heavily, an adjustment designed to stop broad brand strength from crowding out focused departments.
雇主声誉和就业结果不是一回事
就业与成果这一组里有两个指标。雇主声誉 15% 来自问卷，量的是雇主的印象。就业结果 5% 来自院校提交的数据，量的是毕业生的实际去向与收入表现。印象与实际数据可以背离。一所学校可能雇主口碑很好，但毕业生留在本地的比例高、起薪一般。另一所学校可能在雇主问卷里低调，却在收入数据上表现突出。QS 在商科榜单里把这种背离单独标出，称为雇主认知不足画像。看单项分时先分清这两项，判断会稳很多。
Inside the same lens sit two different measurements: employer reputation at 15%, which is survey-based and measures how recruiters see a university, and employment outcomes at 5%, which comes from institutional data and tracks what graduates actually do. The two can pull in opposite directions, and QS flags the gap explicitly in its business school analysis when career outcomes outrun employer recognition. Reading the single indicator without that distinction leads to conclusions about graduate pay that the survey never supported.
样本规模给排名划出的边界
QS 官网公布的整体规模是 2000 所院校入榜、8000 所被评估、151000 份以上学术回复与 100000 份以上雇主回复。雇主声誉数据集另有一组数字：82000 份以上雇主问卷回复，覆盖 150 多个行业与 140 个学科，共 220 万个雇主印象数据点。问卷回复分布不均，行业与地区之间差别明显。入榜还有门槛：学生少于 5000 人的院校，必须在学术声誉、雇主声誉与师均引用三项中至少一项进前 1000，或两项进前 1200，或整体进前 1300。规模小的专科院校因此常被挡在榜外。看名次前先确认这所学校是否真的参与了这项指标。
QS publishes its own scale figures: around 2,000 institutions ranked, more than 8,000 assessed, over 151,000 academic responses and more than 100,000 employer responses in a typical cycle. Its employer reputation dataset cites over 82,000 employer survey responses across more than 150 industries and 140 disciplines, and institutions with fewer than 5,000 students must clear an additional threshold in academic reputation, employer reputation or citations per faculty to be ranked at all. The indicator is shaped by who answered and who was eligible, not only by who hires well.
读榜单的用法：先看口径再看名次
雇主声誉是一份偏好调查，不是招聘统计。它的价值在于告诉你，招聘方在想起某所学校时会先想到谁。它不能说明某个专业的课程强度，也不能替代官网上的入学要求与毕业生去向数据。想判断就业前景，可以把它与就业结果指标、学校官网的毕业生去向页放在一起看。想判断自己是否适合，课程结构、考核方式与往届学生的真实反馈更直接。榜单每年发布，口径也每年复核，读完方法论再读名次是省时间的顺序。常见的错误是只记名次，不记这一年用了哪套口径。
Employer reputation is a survey of preference rather than a recruitment record, and its value lies in showing which universities come to mind first when employers think about hiring. It says nothing about how demanding a specific course is, and it should be read alongside employment outcomes, published graduate destination data and the entry requirements on a university's own pages. Read the methodology before the table, because the weightings behind the indicator are reviewed every year.
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