THE泰晤士高等教育排名方法论：教学、研究、引用权重分析

In 2024, Times Higher Education updated its World University Rankings methodology, and for the first time in over a decade, the weight given to citations dropped below 30 percent. The new formula, which applies to the 2025 ranking cycle, allocates 29.5 percent of a university’s final score to research influence (citations), down from 30 percent in the previous year. Meanwhile, the teaching environment now carries 29.5 percent, and the research environment carries 29 percent. This near-perfect three-way split—a deliberate move away from the old citation-heavy model—reflects a growing consensus among global higher-education analysts that no single metric should dominate a ranking. According to the OECD’s 2023 Education at a Glance report, over 4.5 million students were studying abroad across OECD and partner countries in 2021, a number that has since risen past 5 million. For a prospective applicant weighing offers from, say, the University of Melbourne versus the University of Amsterdam, understanding what these percentage points actually mean—and what they do not measure—can be the difference between a well-informed decision and a misleading one.

The Three Pillars of the THE Methodology

THE divides its assessment into five broad categories, but the three that dominate the overall score are Teaching (the learning environment), Research (volume, income, and reputation), and Citations (research influence). Together they account for 88 percent of the final ranking. The remaining 12 percent is split between International Outlook (7.5 percent) and Industry Income (4 percent). For a student choosing a university, the first three pillars are where the real strategic thinking happens.

Teaching (29.5 percent) is the most reputation-heavy component. It includes a reputational survey (15 percent), staff-to-student ratio (4.5 percent), doctorate-to-bachelor’s ratio (2.25 percent), doctorates awarded per academic staff (6 percent), and institutional income (2.25 percent). The reputational survey alone accounts for more than half of the teaching score, meaning that a university’s brand among senior academics—rather than classroom experience—drives this metric.

Research (29 percent) is built on a reputational survey (18 percent), research income (5.5 percent), and research productivity (5.5 percent). Again, the reputation component dominates. A university that publishes prolifically but lacks name recognition among surveyed scholars will score lower here than a less productive but more famous institution.

Citations (29.5 percent) measures the average number of times a university’s published work is cited by other scholars, normalized across fields. This is the only pillar that is purely quantitative—no survey, no income data. It rewards institutions whose research is actively used by the global academic community.

What the Teaching Weight Actually Captures

The teaching pillar’s reliance on a reputational survey (15 percent out of 29.5 percent) means that a university’s teaching score is, in large part, a reflection of how its research faculty are perceived by their peers. This is a well-documented limitation. A 2022 study published in Studies in Higher Education found that reputational surveys correlate strongly with institutional age and historical prestige, not with student satisfaction or graduate outcomes. For an applicant, a high teaching score from THE may signal that a university is well-regarded in academic circles, but it does not necessarily mean smaller class sizes, better advising, or more accessible professors.

The staff-to-student ratio (4.5 percent) is the only direct proxy for individual attention. A ratio of 1:10 versus 1:30 can make a tangible difference in seminar discussion depth and office-hour availability. Yet this metric is easily gamed by institutions that hire part-time or adjunct faculty, who may not be available for student mentoring. The doctorate-to-bachelor’s ratio (2.25 percent) and doctorates awarded per staff (6 percent) further tilt the score toward research-intensive universities, where most faculty hold PhDs and where doctoral students are numerous. A teaching-focused liberal arts college that employs many master’s-level instructors will lose points here, even if its undergraduate teaching is excellent.

For a student who values close mentorship and small classes, the teaching pillar is a useful but incomplete signal. Cross-referencing THE’s teaching score with national data—such as the U.S. National Survey of Student Engagement (NSSE) or the UK’s National Student Survey—can reveal whether a high THE teaching rank corresponds to real classroom quality.

Research Weight: Reputation Over Output

The research pillar (29 percent) follows the same pattern as teaching: reputation (18 percent) far outweighs quantitative measures. The research reputation survey asks senior academics to name the best universities in their field. This method has a known bias toward large, English-speaking, historically prestigious institutions. A university in Germany or Japan that publishes excellent work in its national language may be overlooked by a predominantly English-speaking survey panel. According to THE’s own 2023 methodology report, the survey pool includes over 68,000 respondents, but the geographic distribution is not uniform—North American and European scholars are overrepresented.

Research income (5.5 percent) measures how much external funding a university attracts. This can be a proxy for lab equipment, PhD stipends, and research infrastructure. But it also favors wealthy institutions in wealthy countries. A university in a developing nation that produces high-impact research on a fraction of the budget will score lower here, even if its per-dollar output is higher.

Research productivity (5.5 percent) counts the number of papers published per academic staff member. This metric rewards volume, not necessarily quality. A university that publishes many papers in low-impact journals can score as well as one that publishes fewer papers in top-tier venues. For a prospective graduate student, the research pillar is most useful when combined with field-specific rankings. THE publishes subject-level rankings that adjust the citation weighting by discipline, which can give a more accurate picture of research strength in, say, engineering versus social sciences.

Citations Weight: Influence, Not Impact on Society

The citations pillar (29.5 percent) is the most transparent and most contested. THE normalizes citation counts by field and by year, so a paper in medicine (which typically accrues many citations quickly) is not compared directly with a paper in mathematics (where citation rates are slower). The normalization makes the metric fairer across disciplines, but it does not account for the quality of the citing sources. A paper cited by 50 low-impact journals can score the same as one cited by 10 top-tier journals.

The 2025 methodology reduced the citations weight from 30 percent to 29.5 percent, a symbolic shift that reflects a broader movement in higher-education assessment toward measuring societal impact, not just academic influence. The Dutch government’s recognition of the San Francisco Declaration on Research Assessment (DORA) in 2023, for example, has led several European universities to de-emphasize citation metrics in faculty evaluations. For students, a high citations score indicates that a university’s research is widely read and used by other scholars. This is valuable if you plan to pursue an academic career, where citation counts can affect grant funding and hiring. But if your goal is industry employment, a university with strong industry partnerships or a high industry income score (4 percent) may be more relevant, even if its citation rank is lower.

International students should also note that citation patterns vary by country. Chinese universities, for instance, have seen rapid citation growth over the past decade, driven by massive state investment in research. According to the 2023 Nature Index, China surpassed the United States in the share of high-quality natural-science publications in 2022. Yet the teaching and research reputation scores for many Chinese universities remain lower than their citation scores would predict, creating a gap that can confuse applicants who rely solely on overall rank.

How the Remaining 12 Percent Can Tip a Decision

The International Outlook pillar (7.5 percent) measures the proportion of international students, international staff, and international co-authorship. For a student seeking a globally diverse campus, this is a direct metric. A university with a high international outlook score is likely to have support services for international students, such as orientation programs, visa assistance, and multicultural events. The Industry Income pillar (4 percent) measures how much research income a university earns from industry, scaled by the number of academic staff. This is a proxy for real-world application of research. A university with a high industry income score may have stronger internship pipelines, corporate partnerships, and technology transfer offices.

For a student deciding between two universities with similar overall ranks, these smaller pillars can break the tie. University A might rank 50th globally with a strong teaching score but a low international outlook. University B might rank 55th with a weaker teaching score but a high industry income score and a large international student body. If your priority is employability in the private sector, University B may be the better choice, even though its overall rank is lower.

The Limits of Weighting: What THE Does Not Measure

No ranking methodology is complete, and THE’s weighting system has three well-known blind spots. First, it does not measure graduate employment outcomes directly. The closest proxy is the reputational survey, which asks academics about the quality of graduates, but this is a vague and unreliable measure. Second, it does not account for cost of attendance or return on investment. A university that charges $60,000 per year and places 90 percent of graduates in high-paying jobs may be a better financial decision than a higher-ranked university that charges $80,000 with a 70 percent placement rate. Third, THE does not capture student satisfaction or well-being. Surveys like the UK’s National Student Survey or the U.S. College Pulse survey show that student happiness correlates only weakly with rank position.

For a 17- to 22-year-old applicant, the best use of THE’s methodology is as a starting point, not a conclusion. Compare the teaching, research, and citations scores of your shortlisted universities. If one university has a significantly higher teaching score but a lower citations score, ask why. Look at the subject-level rankings. Check the university’s own published data on class sizes, graduate salaries, and student support services. For cross-border tuition payments, some international families use channels like Flywire tuition payment to settle fees, but the financial decision should be weighed alongside academic fit.

FAQ

Q1: Should I choose a university with a high THE teaching score or a high citations score?

It depends on your career goals. If you plan to pursue a PhD and an academic career, a high citations score indicates that your future professors are producing research that is widely used by other scholars, which can help you build a strong academic network. If you are aiming for industry employment, a high teaching score may be less relevant than a university’s industry partnerships or internship placement rate. According to a 2023 survey by the Graduate Management Admission Council, 72 percent of employers rated “relevant work experience” as more important than the prestige of the applicant’s university. Cross-reference THE scores with employment data from the university’s career services office.

Q2: How much does the THE ranking change from year to year, and should I worry about a drop?

Year-to-year movements of 5 to 10 positions are common and often result from methodological tweaks rather than real changes in quality. In the 2025 ranking, THE reduced the citations weight by 0.5 percent and increased the research weight by 0.5 percent. Universities strong in research but weak in citations gained slightly. If a university drops 15 or more positions in one year, investigate the reason—it may be a data correction or a change in survey response rates. For applicants, a single year’s rank is less important than a three- to five-year trend. Use THE’s archive of past rankings to see whether a university is consistently improving or declining.

Q3: Does THE rank universities differently for different subjects, and how do I find that?

Yes, THE publishes 11 subject rankings, each with its own weightings. For example, in the Clinical and Health subject ranking, citations carry 35 percent of the score, while teaching carries 27.5 percent. In Arts and Humanities, citations carry only 15 percent, and teaching carries 37.5 percent. If you are applying for a specific major, always check the subject-level ranking rather than the overall rank. A university ranked 200th overall might be ranked 50th in Engineering, which is far more relevant to your decision. THE’s subject rankings are available on its website and are updated annually in October.

References

• Times Higher Education. 2024. World University Rankings 2025: Methodology.
• OECD. 2023. Education at a Glance 2023: OECD Indicators.
• Nature Index. 2023. 2023 Research Leaders: Country and Institutional Tables.
• Graduate Management Admission Council. 2023. Corporate Recruiters Survey Report.
• UNILINK Education. 2024. Cross-Border Tuition Payment Data (internal database).