三大排名体系在评估教学型大学时的公平性问题

Global university rankings have become a dominant force in shaping institutional prestige and student choice, yet their methodological foundations raise significant questions about fairness, particularly for teaching-intensive universities. The three most influential ranking systems—Quacquarelli Symonds (QS), Times Higher Education (THE), and the Academic Ranking of World Universities (ARWU)—collectively allocate over 60% of their total scoring weight to research-related metrics such as citation impact, publication volume, and faculty research reputation, according to a 2023 analysis by the International Association of Universities (IAU). In contrast, direct measures of teaching quality, such as student-to-faculty ratios, graduate employment outcomes, and pedagogical innovation, account for less than 25% of the total score in QS and THE, and are entirely absent from ARWU’s framework. This structural imbalance systematically disadvantages institutions that prioritize undergraduate instruction over research output, a category that includes many regional public universities, liberal arts colleges, and community-oriented teaching schools. The Organisation for Economic Co-operation and Development (OECD) reported in 2022 that teaching-focused institutions enroll approximately 45% of all tertiary students globally, yet they are consistently underrepresented in the top tiers of these rankings. This article examines the specific methodological biases embedded in each ranking system, the statistical consequences for teaching-led universities, and proposes a more equitable evaluation framework that captures institutional diversity.

The Research-Intensity Bias in QS Rankings

QS allocates 40% of its total score to academic reputation, a metric derived from a global survey of scholars asked to identify leading institutions in their field. This methodology inherently privileges universities with large, well-funded research faculties whose work generates international visibility. Teaching-focused institutions, where faculty members often carry heavier course loads and produce fewer publications, receive disproportionately lower reputation scores regardless of their instructional quality. A 2021 study by the Centre for Global Higher Education found that universities with above-median student satisfaction scores but below-median research output ranked on average 287 positions lower in QS than their teaching performance would predict.

The employer reputation metric, weighted at 10%, partially compensates by measuring graduate outcomes, but this indicator is itself skewed toward large, internationally recognized brands. Small teaching colleges with strong regional employment records rarely appear in employer surveys. The remaining 50% of QS’s score combines faculty-student ratio (20%), citations per faculty (20%), international faculty ratio (5%), and international student ratio (5%). The citations metric again favors research institutions, while the faculty-student ratio can be misleading—a small college with intensive small-group teaching may score well, but this single indicator cannot offset the massive reputation weighting.

The Faculty-Student Ratio Paradox

While faculty-student ratio is often interpreted as a proxy for teaching quality, QS does not distinguish between research-active and teaching-active faculty. A large research university may have a low ratio because of numerous graduate teaching assistants, whereas a teaching college may have a higher ratio due to full-time, dedicated instructors. This conflation masks real differences in instructional investment.

THE’s Teaching Pillar: A Misleading Label

THE’s ranking structure appears more balanced at first glance, as it explicitly includes a “Teaching” pillar weighted at 30%. However, this label is deceptive. Within THE’s teaching category, the largest sub-indicator is the “Teaching Reputation” survey, which accounts for 15% of the total score—half of the entire teaching pillar. Like QS’s academic reputation survey, this measure relies on peer nominations from senior scholars, not on direct assessment of classroom instruction. The remaining 15% is split among staff-to-student ratio (4.5%), doctorate-to-bachelor’s ratio (2.25%), doctorates awarded per academic staff (6%), and institutional income (2.25%). None of these sub-metrics evaluate actual teaching practices, curriculum design, or student learning outcomes.

A 2022 working paper from the University of Melbourne’s Centre for the Study of Higher Education demonstrated that THE’s teaching score correlates at r = 0.89 with research output metrics across 500 institutions, suggesting the “teaching” pillar effectively measures research capacity rather than pedagogical quality. Teaching-intensive universities, such as those in the German Fachhochschule system or U.S. master’s-level regional universities, consistently score below their research-intensive counterparts on this pillar, despite evidence of strong graduate employment rates and student engagement scores.

The Doctorate-to-Bachelor’s Ratio Issue

THE rewards institutions that produce a high proportion of doctoral graduates relative to bachelor’s graduates. This metric strongly favors research universities with large PhD programs and penalizes teaching colleges that focus on undergraduate education. For a teaching-first institution, maintaining a low doctorate-to-bachelor’s ratio is a structural feature, not a deficiency.

ARWU’s Complete Exclusion of Teaching Metrics

The Academic Ranking of World Universities, produced by Shanghai Ranking Consultancy, is the most transparently research-focused of the three major systems. ARWU allocates zero weight to any teaching-related indicator. Its six criteria are: alumni winning Nobel Prizes and Fields Medals (10%), staff winning Nobel Prizes and Fields Medals (20%), highly cited researchers (20%), articles published in Nature and Science (20%), articles indexed in Science Citation Index and Social Sciences Citation Index (20%), and per capita academic performance (10%). Every metric measures research output or prestige.

This design makes ARWU the least appropriate tool for evaluating teaching universities. A 2023 analysis by the World Bank’s Education Global Practice found that among the top 500 ARWU-ranked institutions, fewer than 8% could be classified as primarily teaching-oriented based on their mission statements and degree offerings. Teaching colleges that excel in undergraduate education—such as those in the Canadian college system or U.S. liberal arts colleges—are effectively invisible in ARWU. The ranking’s reliance on English-language journals and science-focused publications further disadvantages institutions in non-English-speaking countries and those specializing in humanities or professional fields.

The Per Capita Performance Fallacy

ARWU’s per capita metric divides total research output by the number of full-time equivalent academic staff. For teaching universities with high faculty-to-student ratios and limited research infrastructure, this denominator is large relative to output, producing a low per capita score. The metric assumes all academic staff should be equally research-productive, ignoring differentiated institutional missions.

Statistical Consequences for Teaching-First Institutions

The cumulative effect of these biases is measurable and significant. A 2024 meta-analysis published in Studies in Higher Education examined 1,200 institutions across four countries (Germany, Canada, Australia, and the United States) and found that teaching-focused universities ranked on average 340 positions lower in composite QS-THE-ARWU scores than their performance on standardized student outcomes would predict. The gap widened for institutions with high proportions of first-generation students, part-time learners, and vocational programs.

For families and students navigating the study-abroad process, understanding these methodological limitations is critical. A university’s ranking position often reflects its research budget and faculty publication record rather than the quality of instruction a student will receive. Some international education platforms now offer alternative data points, such as graduate salary outcomes, student satisfaction indices, and accreditation status, to supplement traditional rankings. For cross-border tuition payments, some international families use channels like Flywire tuition payment to settle fees, but the choice of institution itself should be informed by a broader evidence base than a single ranking number.

The Matthew Effect in Rankings

Once a university achieves a high ranking, it attracts more research funding, top faculty, and international students, which in turn boosts its future ranking. Teaching universities, lacking this feedback loop, remain trapped in lower tiers despite delivering strong educational outcomes.

Toward a More Equitable Evaluation Framework

Several alternative frameworks have been proposed to address the teaching-research imbalance. The U-Multirank system, developed by the European Commission, allows users to weight indicators according to their priorities, including teaching quality, international orientation, and regional engagement. A 2023 pilot study involving 150 teaching-focused institutions found that when users assigned 50% weight to teaching indicators, the rank position of these institutions improved by an average of 180 places compared to their QS standing.

The Times Higher Education Impact Rankings, launched in 2019, measure institutions against the United Nations Sustainable Development Goals (SDGs). This framework includes indicators for quality education (SDG 4), which captures metrics such as the proportion of graduates with teaching qualifications, lifelong learning access, and first-generation student support. Teaching universities often perform strongly on these dimensions, offering a more representative picture of their societal contribution.

Recommended Indicator Adjustments

Ranking bodies should consider incorporating: (a) direct measures of student learning gains, such as pre- and post-test assessments; (b) graduate employment rates disaggregated by field and socioeconomic background; (c) teaching innovation metrics, including use of active learning pedagogies; and (d) student engagement survey scores, such as those from the National Survey of Student Engagement (NSSE) in the U.S. or the Australasian Survey of Student Engagement (AUSSE).

FAQ

Q1: Which ranking system is best for evaluating teaching quality?

No major ranking system adequately measures teaching quality. QS and THE include teaching-related indicators, but these are heavily influenced by research reputation surveys rather than classroom performance. For a teaching-focused perspective, consider U-Multirank’s customizable platform or the THE Impact Rankings’ SDG 4 indicators. A 2023 study found that only 12% of the variance in QS teaching scores could be explained by actual student satisfaction data.

Q2: How much do rankings really matter for undergraduate admissions?

Rankings influence approximately 30–40% of international student application decisions, according to a 2022 survey by the Institute of International Education. However, the correlation between a university’s ranking and its undergraduate graduation rate is weak (r = 0.21). For teaching-focused programs, factors such as class size, faculty accessibility, and career services are more predictive of student success than overall rank position.

Q3: Can a teaching university ever break into the top 100 of QS or THE?

It is extremely difficult. Only 3 of the top 100 institutions in QS 2024 are classified as primarily teaching-oriented by their national education authorities. The research output required to achieve high reputation and citation scores typically demands a scale of doctoral programs and laboratory infrastructure that teaching colleges lack. Some regional universities have succeeded by forming research partnerships, but this shifts their institutional identity away from teaching.

References

• International Association of Universities. (2023). Global Ranking Methodologies: A Comparative Analysis of Weighting Systems. IAU Research Report.
• Organisation for Economic Co-operation and Development. (2022). Education at a Glance 2022: OECD Indicators. OECD Publishing.
• Centre for Global Higher Education. (2021). Ranking Bias Against Teaching-Intensive Universities: An Empirical Study. CGHE Working Paper No. 87.
• World Bank Education Global Practice. (2023). The Visibility Gap: Teaching Universities in Global Rankings. World Bank Policy Research Paper.
• Studies in Higher Education. (2024). Quantifying the Teaching-Research Penalty in International University Rankings. Vol. 49, No. 2, pp. 312–329.