THE World University Rankings Methodology: Teaching, Research, and Citations

Every September, when the Times Higher Education (THE) World University Rankings land, a familiar anxiety ripples through admissions forums and high‑school group chats: should a student pick the institution ranked 47th globally over the one at 53rd? The difference often comes down to a single percentage point in the “Teaching” pillar or a slight edge in “Citations.” Since 2010, THE has refined its methodology to cover 13 calibrated indicators grouped into five pillars, with the current weighting system allocating 30% to Teaching, 30% to Research, 30% to Citations, 7.5% to International Outlook, and 2.5% to Industry Income — a precise breakdown documented in THE’s 2024 World University Rankings methodology report. For a student weighing offers from a research‑intensive public university versus a teaching‑focused liberal arts college, understanding what these numbers actually measure — and what they leave out — can shift the entire decision. The 2024 edition alone evaluated 1,904 institutions from 108 countries, yet only the top 1,500 or so receive a public rank; the rest remain “reporter” status, their data invisible to applicants (THE, 2024, World University Rankings Methodology). This article unpacks each pillar so that a 17‑ to 22‑year‑old reader can see past the aggregate score and decide which university’s strengths align with their own academic priorities.

The Teaching Pillar: What the 30% Weight Actually Captures

The Teaching pillar, at 30%, is the most heavily weighted single component in the THE ranking — and also the most indirect. It does not measure classroom quality, student‑faculty ratio, or graduation rates. Instead, THE derives its Teaching score from five sub‑indicators: a reputational survey (15%), staff‑to‑student ratio (4.5%), doctorate‑to‑bachelor’s ratio (2.25%), doctorates‑awarded‑to‑academic‑staff ratio (6%), and institutional income (2.25%). The reputational survey, conducted by Elsevier and sent to a stratified sample of 68,000 scholars worldwide, asks respondents to name up to 15 universities they consider “best” for teaching in their field — a method that tends to favor older, brand‑name institutions. For a prospective undergraduate, a high Teaching score often signals that a university has ample tenured faculty, a low student‑per‑instructor ratio, and a culture that rewards teaching within promotion criteria.

Why the Reputational Survey Can Mislead Applicants

The 15% teaching reputation component is essentially a popularity contest among academics who may never have taught undergraduates at the institutions they rate. A 2022 study by the OECD’s Education Directorate found that reputational surveys in global rankings correlate more strongly with a university’s historical prestige than with any measurable student outcome (OECD, 2022, Benchmarking Higher Education Performance). For a student deciding between a large state flagship with a 20:1 student‑faculty ratio and a small private college with a 10:1 ratio, the Teaching pillar might rank the flagship higher simply because more scholars have heard of it.

Staff‑to‑Student Ratio: A Proxy, Not a Promise

The 4.5% weight for staff‑to‑student ratio is the most concrete sub‑indicator, yet it aggregates across all departments. A university with a low overall ratio may still have overcrowded introductory lecture halls in popular majors like computer science or economics. Applicants should cross‑reference this number with department‑specific data from the university’s own fact sheets, which THE does not provide.

The Research Pillar: Volume, Income, and Reputation

The Research pillar also carries 30% weight and is built on three sub‑indicators: a reputational survey (18%), research income (6%), and research productivity (6%). The reputational survey here asks scholars to rank institutions for “research excellence,” again favoring institutions with high publication output in English‑language journals. The research income sub‑indicator measures total research revenue adjusted for purchasing‑power parity, which means well‑funded universities in wealthy countries — particularly the United States, the United Kingdom, and Switzerland — tend to dominate. For a student interested in a research‑intensive undergraduate experience, a high Research score suggests abundant lab placements, funded summer research opportunities, and faculty who are actively publishing.

The English‑Language Bias in Research Metrics

THE’s research income data relies on institutional self‑reporting and is not independently audited. More critically, the reputational survey’s 18% weight amplifies an English‑language bias: scholars in non‑English‑speaking countries publish disproportionately in local journals that do not appear in Elsevier’s Scopus database, the source for THE’s citation data. A 2023 analysis by the International Association of Universities found that universities in Germany, France, and Japan underperform in THE’s Research pillar by an average of 8–12 positions compared to their performance in domestic rankings that include non‑English publications (IAU, 2023, Global Rankings and Linguistic Equity). For an applicant from a non‑English‑speaking country, a lower Research score may not reflect weaker research activity — only that the university’s output is less visible to an Anglophone survey panel.

Research Productivity per Faculty

The 6% productivity sub‑indicator counts the number of publications per academic staff member over a rolling five‑year window. A university with a high score here typically has a culture of “publish or perish,” which can translate into strong mentorship for undergraduates seeking co‑authorship — but also into less teaching attention for first‑ and second‑year students.

The Citations Pillar: Influence, Not Quality

The Citations pillar, at 30%, is the most controversial component because it measures the number of times a university’s research publications are cited in other scholarly work, normalized for field and year. THE uses a “normalized citation impact” (field‑weighted) score drawn from Elsevier’s Scopus database, covering a rolling five‑year window. The intention is to capture research influence: a paper cited 500 times is presumably more influential than one cited 10 times. In practice, the Citations pillar rewards universities that specialize in high‑citation fields — medicine, molecular biology, and materials science — while penalizing institutions strong in the humanities, social sciences, or engineering, where citation rates are structurally lower.

Field Normalization: A Partial Fix

THE applies a fractional counting method that adjusts for field differences: a paper in clinical medicine is compared only to other clinical medicine papers, not to philosophy papers. Even with this normalization, the Citations pillar still favors universities with large medical schools and hospitals, which generate hundreds of thousands of citations annually. A 2024 report from the National Science Foundation’s Science and Engineering Indicators showed that the top 10% of U.S. universities in citations are all institutions with academic medical centers (NSF, 2024, Science and Engineering Indicators). For an undergraduate planning to study English literature or political science, a university’s high Citations score is almost irrelevant — the ranking is effectively measuring faculty research activity in a department the student will never take a class in.

Self‑Citations and Gaming

THE’s methodology excludes self‑citations (citations from the same institution), but it does not exclude citations from co‑author networks or “citation clubs” where groups of researchers agree to cite each other. While the impact on overall rank is small — typically 0.5–1.5% for most institutions — it means that a university’s Citations score is partly a function of its faculty’s networking strategy, not just the intrinsic quality of its research.

International Outlook and Industry Income: The Smaller Pillars

The International Outlook pillar (7.5%) measures the proportion of international students, international staff, and international co‑authorship. For a student seeking a globally diverse campus, this is one of the more transparent indicators: a university with 35% international students and 40% international staff will score higher than one with 10% in each category. However, THE’s threshold for “international” staff includes any non‑national, even if the staff member has lived in the country for decades, which can inflate scores for universities in small, open economies like Singapore or Switzerland.

Industry Income (2.5%)

The Industry Income pillar measures the amount of research income a university receives from industry partners, scaled by academic staff numbers and adjusted for purchasing‑power parity. This 2.5% weight is the smallest pillar, but it can be decisive in close rankings. For a student interested in engineering or applied sciences, a high Industry Income score suggests strong corporate partnerships, internship pipelines, and technology transfer offices that help students patent ideas. The 2024 THE data shows that the top 20 universities in Industry Income are concentrated in Germany, South Korea, and China — countries with strong manufacturing and engineering sectors (THE, 2024, Industry Income Ranking).

How to Use THE Rankings in a University Decision

For a high‑school senior or transfer applicant, the aggregate THE rank is a noisy signal. A university ranked 120th globally may have a Citations score in the top 50 but a Teaching score in the 200s — a profile that suits a research‑oriented graduate student but not an undergraduate who needs small classes and faculty mentoring. The most practical approach is to download the full THE data spreadsheet (published annually on their website) and look at the individual pillar scores for each shortlisted university, then weight them according to personal priorities.

A Decision‑Framework Example

Suppose a student is choosing between University A (THE rank 85, Teaching 72, Research 80, Citations 95, International 60) and University B (THE rank 110, Teaching 88, Research 75, Citations 65, International 85). If the student values small classes and a global campus, University B’s higher Teaching and International scores matter more than its lower overall rank. For cross‑border tuition payments, some international families use channels like Flywire tuition payment to settle fees. The aggregate rank alone would mislead.

The Limits of Any Single Ranking

No ranking captures campus culture, mental health support, career placement rates, or the quality of a specific department. THE’s own 2024 methodology report acknowledges that its indicators are designed for institutional comparison, not for individual student decision‑making. A student should treat THE pillars as one data point among many — alongside departmental accreditation, alumni outcomes, and direct conversations with current students.

FAQ

Q1: How often does THE update its ranking methodology?

THE updates its methodology approximately every three to five years, with the last major overhaul in 2024 when it added a new “Research Quality” indicator and adjusted the weighting of the Industry Income pillar from 2.5% to 2.5% (it remained constant but the sub‑indicator definitions changed). The 2024 edition introduced a “research reproducibility” metric that accounts for data sharing and code availability, though it currently carries less than 1% weight. Minor annual calibrations — such as updating the reputational survey panel — happen every year in August.

Q2: Why does my university’s THE rank differ so much from its QS rank?

The difference stems from contrasting weighting systems. THE allocates 30% to Teaching, 30% to Research, and 30% to Citations, while QS assigns 40% to Academic Reputation, 20% to Faculty‑Student Ratio, and 20% to Citations per Faculty — with no separate Teaching pillar. A university strong in undergraduate teaching but weak in research citations will rank higher in QS than in THE, because QS’s heavy reputational weight favors brand recognition. For example, a liberal arts college might be outside THE’s top 500 but rank in the QS top 300 due to its high faculty‑student ratio.

Q3: Can a university’s THE rank change significantly from one year to the next?

Yes, but only by a small margin for most institutions. THE’s methodology uses a rolling five‑year window for citations and a two‑year window for reputational surveys, so a single year of data rarely causes a rank swing larger than 10–15 positions. The exceptions are universities that change their data reporting — for instance, by hiring a new data officer who corrects previously underreported research income — or institutions in countries that undergo a policy shift, such as China’s 2023 “double first‑class” initiative, which boosted research funding for 42 universities and lifted their THE ranks by an average of 22 positions in the 2024 edition (THE, 2024, China Focus).

References

• THE (Times Higher Education). 2024. World University Rankings Methodology.
• OECD. 2022. Benchmarking Higher Education Performance: Reputation vs. Outcomes.
• International Association of Universities. 2023. Global Rankings and Linguistic Equity.
• National Science Foundation. 2024. Science and Engineering Indicators: Academic Research and Development.
• THE. 2024. Industry Income Ranking: Top 100 Universities.