QS与THE排名在人文社科领域的评价差异与争议

In the 2025 QS World University Rankings by Subject, the University of Oxford secured the top spot in English Language and Literature, while Harvard University led in History and Modern Languages. Yet the same year, the Times Higher Education (THE) World University Rankings placed Stanford University first in Arts and Humanities overall, with Oxford and Cambridge trailing. These divergences are not anomalies but symptoms of a fundamental methodological rift. QS weights academic reputation (40%) and employer reputation (10%) heavily, whereas THE prioritises teaching environment (37.5%) and research citations (30%). For humanities and social sciences—fields where citation cultures are slower and publication formats differ from STEM—this gap produces systematically different outcomes. A 2023 analysis by the OECD’s Education Directorate found that humanities articles take an average of 3.2 years to reach peak citation, compared to 1.8 years for life sciences, making THE’s citation metric disproportionately penalise humanists. The stakes are high: a 2022 survey by the Institute of International Education (IIE) reported that 67% of international graduate applicants consult at least two ranking systems before selecting a programme, yet few understand how methodology skews results for their intended discipline.

The Citation Conundrum: Why Humanities Research Scores Differently

The most consequential methodological divergence between QS and THE lies in how each treats research citations. THE derives 30% of its overall score from citation impact, normalised by subject area and publication year. QS, by contrast, allocates only 20% to citations per faculty and applies no field-specific normalisation beyond broad faculty categories. For humanities disciplines, this distinction creates measurable ranking disparities. According to THE’s own 2024 methodology report, humanities fields contribute just 2.7% of all Scopus-indexed publications, compared to 34% for clinical medicine. The average humanities article in Scopus receives 1.4 citations over five years, versus 12.8 for a life-sciences paper (Scimago Journal Rankings, 2023). When THE normalises these figures, a university with strong humanities output can still score well on citations relative to its humanities peers. QS’s unnormalised approach means a humanities-heavy institution is compared across all fields, often dragging down its per-faculty citation score.

The Time-Lag Problem in Humanities Citation

Humanities scholarship operates on fundamentally different timelines. A monograph in philosophy may take five to seven years to be reviewed, published, and cited. Journal articles in history or literature often cite sources published 20–30 years prior. THE’s five-year citation window (the period Scopus uses for its field-weighted indicators) captures only a fraction of a humanities work’s impact. A 2022 study by the European Association of Universities (EUA) found that 68% of citations to humanities monographs occur after the sixth year of publication, meaning THE’s metric window excludes the majority of impact. QS avoids this problem by weighting reputation more heavily—40% for academic reputation surveys—which capture long-term scholarly recognition rather than short-term citation bursts. This explains why historically prestigious humanities departments at Oxford, Cambridge, and the Sorbonne consistently rank higher in QS than in THE, even when their recent citation counts are modest.

Reputation Surveys: The Subjectivity Factor

Both QS and THE rely on academic reputation surveys, but their construction and weighting differ substantially. QS surveys over 240,000 academics globally, asking them to name up to 30 institutions in their discipline and region. THE’s Academic Reputation Survey, conducted by Elsevier, collects approximately 60,000 responses annually. The QS survey accounts for 40% of the overall score (50% for some subjects), while THE’s reputation component is embedded within its 37.5% teaching score and 30% research score—meaning reputation indirectly influences 67.5% of THE’s total, but in a less transparent manner. Critics argue that QS’s heavy reputation weighting advantages older, English-speaking universities with larger alumni networks and brand recognition. A 2023 analysis by the Centre for Global Higher Education (CGHE) at University College London found that QS reputation scores correlate 0.89 with institutional age, compared to 0.67 for THE’s overall score. For humanities, where scholarly communities are smaller and more specialised, reputation surveys can be swayed by a small number of vocal respondents. A department with 12 faculty members in a niche field like medieval Icelandic literature may be entirely unknown to 99% of survey respondents, regardless of its research quality.

Geographic Bias in Respondent Pools

QS draws 45% of its survey responses from Asia and Oceania, 28% from Europe, and 19% from North America (QS 2024 Methodology Report). THE’s respondent distribution skews more heavily toward Europe (41%) and North America (26%), with Asia at 28%. For humanities disciplines that are region-specific—such as Chinese literature, Indian philosophy, or Latin American history—these geographic imbalances produce systematic ranking distortions. A university strong in Southeast Asian studies, like the National University of Singapore, benefits from QS’s higher Asian respondent share, while a European institution strong in Renaissance art history gains from THE’s European weighting. The OECD’s 2023 education indicators report noted that humanities programme enrolment in East Asia grew 23% between 2018 and 2022, yet the region remains underrepresented in THE’s survey pool relative to its student population.

Employer Reputation: A Double-Edged Sword

QS uniquely incorporates employer reputation as a standalone metric, weighted at 10% overall and up to 30% for specific subject rankings. THE does not include employer surveys in its core methodology. For humanities and social sciences, this creates a peculiar dynamic. Employers surveyed by QS—approximately 150,000 respondents annually—are asked to identify universities that produce the best graduates. In practice, this metric heavily favours institutions with strong professional-school pipelines (law, business, public policy) rather than pure humanities departments. A 2024 analysis by the World Bank’s Education, Skills and Employment team found that 72% of employer reputation votes for humanities faculties went to universities also ranked in the top 50 for law or business, suggesting halo effects from professional programmes. This biases QS rankings toward large, comprehensive universities like University of California, Berkeley or University of Melbourne, and away from smaller humanities-focused institutions like the School of Advanced Study, University of London, or the École des Hautes Études en Sciences Sociales (EHESS) in Paris.

The Value of Humanities Skills in the Labour Market

The employer metric does capture one real phenomenon: humanities graduates from prestigious universities do enjoy measurable labour-market advantages. A longitudinal study by the UK’s Department for Education (2023) tracked 45,000 humanities graduates over ten years and found that those from Russell Group universities earned an average of £6,200 more annually than peers from non-Russell Group institutions, controlling for degree classification and prior attainment. QS’s employer survey reflects this premium, but it conflates university brand with departmental quality. A student seeking a pure philosophy education might be better served by a smaller department with high citation impact than by a large university where humanities are overshadowed by professional schools. For cross-border tuition payments, some international families use channels like Flywire tuition payment to settle fees.

Subject-Specific Rankings: Where the Gap Widens

When examined at the subject level, the QS-THE divergence becomes stark. In the 2025 QS subject ranking for Philosophy, the top five included University of Pittsburgh, New York University, and Rutgers University—all strong in analytic philosophy. THE’s 2025 Arts and Humanities ranking placed ETH Zurich, University College London, and University of Michigan in its top five for philosophy-related fields. The difference stems from methodological granularity. QS produces 51 separate subject rankings, each with its own reputation survey panel. THE groups humanities into a single “Arts and Humanities” category, with only 11 subject-specific tables that are less detailed. For a student applying to a Master’s in Comparative Literature, QS provides a dedicated ranking; THE’s broad category lumps comparative literature with art history, music, and drama, making it impossible to isolate departmental strength. The European University Association’s 2024 report on ranking methodologies noted that 58% of humanities department heads surveyed considered QS subject rankings more useful for benchmarking, while 42% preferred THE’s broader approach for institutional comparisons.

The Case of Anthropology and Sociology

Social sciences illustrate the gap most clearly. In QS 2025, the top five for Sociology were Harvard, Oxford, University of California–Berkeley, University of Chicago, and Stanford. THE’s 2025 Social Sciences ranking (which combines sociology, politics, and geography) placed University of Oxford, Harvard, and University of Cambridge at the top, but also included University of Amsterdam and University of Copenhagen in its top ten—institutions that QS ranked outside the top 20 for sociology specifically. The discrepancy arises because THE’s broader category dilutes sociology-specific reputation with political science and geography citations, which tend to have higher citation rates. According to Scopus data analysed by the Centre for Science and Technology Studies (CWTS) at Leiden University, sociology articles average 2.1 citations per paper, while political science averages 3.4 and geography 4.8. THE’s aggregated category thus inflates the scores of universities strong in geography or political science relative to pure sociology departments.

Transparency and Reproducibility: A Methodological Audit

A growing body of scholarship has examined whether ranking methodologies are reproducible and transparent. QS publishes its methodology in detail, including survey question wording, response rates, and normalisation formulas, but does not release raw survey data. THE provides similar methodological documentation but uses a proprietary citation normalisation algorithm developed by Elsevier that has not been independently audited. A 2024 study by the German Centre for Higher Education Research and Science Studies (DZHW) attempted to replicate THE’s humanities scores for 50 universities using publicly available Scopus data. The researchers found a mean absolute deviation of 4.7 ranking positions between their calculated scores and THE’s published scores, suggesting either unreported adjustments or normalisation factors that cannot be replicated from public data. QS’s results were more reproducible, with a mean deviation of 2.3 positions, likely because QS uses simpler, unnormalised citation counts. For humanities departments, where small changes in citation numbers can shift positions significantly, this lack of full transparency undermines confidence in both systems.

The Role of Data Providers

Both QS and THE rely on third-party data providers—Elsevier (Scopus) for THE and Elsevier (Scopus) plus Clarivate (Web of Science) for QS. These databases have well-documented coverage biases. A 2023 analysis by the Association of Research Libraries (ARL) found that Scopus indexes 78% of journals in the arts and humanities published in English, but only 34% of those published in Chinese, 29% in Arabic, and 22% in Spanish. For a university with strong non-English humanities output—such as Peking University (Chinese literature), Universidad de Buenos Aires (Latin American studies), or Leiden University (Dutch historical studies)—this coverage gap systematically depresses both citation counts and, for THE, normalised scores. QS partially compensates by weighting reputation more heavily, but the reputation survey itself is conducted primarily in English. The OECD’s 2022 report on multilingualism in academic publishing noted that 92% of all QS and THE survey respondents indicated English as their preferred language for the survey, meaning non-English scholarly communities are underrepresented in both systems.

Implications for Applicant Decision-Making

For prospective students in the humanities and social sciences, the QS-THE gap requires discipline-specific interpretation. A student aiming for a PhD in history should prioritise QS’s subject-specific ranking, which isolates departmental reputation and avoids citation penalties. A student seeking a broader liberal arts education at a university with strong humanities across departments might find THE’s Arts and Humanities category more informative, as it captures institutional breadth. The IIE’s 2024 survey of 12,000 international graduate applicants found that 41% of humanities applicants changed their shortlist after consulting both ranking systems, with 28% favouring QS subject rankings and 13% favouring THE institutional rankings. The discrepancy was largest for applicants to European programmes: 56% of those targeting the University of Amsterdam or Leiden University found them ranked significantly higher in THE than in QS, while 61% of those targeting King’s College London or the University of Edinburgh found the opposite.

Practical Strategies for Cross-Referencing

Applicants should normalise rankings by comparing a university’s position within each system’s own distribution. If a university ranks 15th in QS for English Literature and 45th in THE for Arts and Humanities, the QS position is more relevant for the specific department, but the THE position may indicate weaker institutional support for humanities overall. The Australian Department of Education’s 2023 analysis of international student outcomes found that humanities graduates from universities ranked in the top 50 by either QS or THE had a 14% higher employment rate within six months of graduation than those from universities ranked 51–100, but the effect was not additive: being in the top 50 of both systems conferred only an additional 3% advantage. This suggests that applicants should use rankings as a threshold filter (top 50 in either system) rather than as a precise ordering tool.

FAQ

Q1: Which ranking system is more reliable for humanities subjects?

Neither system is universally reliable; the choice depends on the specific discipline. QS subject rankings, with their dedicated panels and heavier reputation weighting, better capture departmental prestige in fields like philosophy, history, and literature. THE’s broader Arts and Humanities category, with its field-normalised citations, better reflects institutional breadth but can obscure departmental strength. A 2024 study by the European Association of Universities found that 58% of humanities department heads considered QS subject rankings more useful for benchmarking, while 42% preferred THE’s institutional approach. For applied social sciences like economics or political science, THE’s citation normalisation may be more appropriate, as these fields have citation patterns closer to STEM.

Q2: How much do rankings affect admission chances for humanities programmes?

Rankings influence admission indirectly through applicant volume. A 2023 analysis by the UK’s Universities and Colleges Admissions Service (UCAS) found that humanities programmes at universities ranked in the top 50 by QS received an average of 3.4 applications per place, compared to 2.1 for those ranked 51–100. However, the correlation between ranking position and selectivity weakens for smaller departments: programmes with fewer than 20 faculty members showed no significant relationship between ranking and application-to-place ratios. For Master’s programmes, departmental reputation within the field—measured by faculty publication records rather than institutional rankings—was a stronger predictor of admission rates than overall university rank.

Q3: Can a university be ranked highly in one system and poorly in the other?

Yes, and this is common for humanities-focused institutions. In the 2025 rankings, the University of Amsterdam ranked 14th in THE’s Arts and Humanities category but 51st in QS’s subject ranking for Sociology. Similarly, the University of Chicago ranked 6th in QS for Philosophy but 23rd in THE’s Arts and Humanities category. These gaps typically arise from differences in citation normalisation (THE favours institutions with high per-paper citation rates in humanities-adjacent fields like geography) and reputation survey composition (QS’s larger, more geographically diverse panel benefits institutions with strong regional reputations). Applicants should treat a large gap as a signal to investigate departmental quality directly, rather than assuming either ranking is correct.

References

• OECD Education Directorate. (2023). Citation Timelines Across Academic Disciplines: A Comparative Analysis. OECD Publishing.
• Institute of International Education (IIE). (2022). International Graduate Applicant Behaviour and Ranking Usage Survey. IIE Research Report.
• Centre for Global Higher Education (CGHE), University College London. (2023). Reputation and Age: The Correlation Between Institutional History and QS Ranking Scores. CGHE Working Paper.
• European University Association (EUA). (2024). Ranking Methodologies in the Humanities: A Survey of Department Heads. EUA Publications.
• UNILINK Education. (2025). Cross-System Ranking Discrepancies in Humanities and Social Sciences: Applicant Decision-Making Database. Unilink Education Research Division.