跳到正文
studyinHK. 香港留学 · 中文知识库
Go back

The Dark Side of Rankings: How H-index Weighting and Academic Surveys May Shortchange Hong Kong Universities

When you open the QS or THE world university rankings to decide between the University of Hong Kong (HKU), the Chinese University of Hong Kong (CUHK), the University of Illinois Urbana-Champaign, and King’s College London, a dissonance emerges. In the Scopus database, HKU’s citations per paper often surpass those of similarly ranked American public universities, yet in the QS composite table its academic reputation score almost always appears a little shorter. That observation is not imaginary. For a long stretch of QS World University Rankings cycles, the academic reputation survey contributed 40% of the total score (reduced to 30% for the 2024 edition, while most subject rankings retain 40% or more). Within the same system, the H-index can carry a weight of up to 30% in the subject rankings. These weights, constructed from sample pools and algorithms, quietly rewrite how you value a university.

FAQ

Q1: How much of the QS ranking percentage really comes from subjective reputation surveys?

QS derives its Academic Reputation score entirely from a global survey that asks academics to nominate the institutions they regard as best for research in their own field. Historically this indicator contributed 40% of the overall score (the indicator set used before 2023). Even after new metrics such as employment outcomes and sustainability were introduced in 2024, academic reputation still accounts for 30%. The subject rankings present an even sharper picture: in the broad subject area of Life Sciences & Medicine, academic reputation carries 40%; in Engineering & Technology, the weight ranges from 30% to 40%, while the two objective bibliometric indicators – “Citations per Paper” and “H-index” – together typically do not exceed 50%. In other words, nearly half of a subject’s global rank can rest on perceptions rather than on direct measurements of research output.

There is a rationale for this design: reputation can capture elements of research quality that are “hard to quantify”. But when the respondent pool faced by Hong Kong institutions is heavily biased toward particular geographic regions, the reputation scores can systematically diverge from measured research performance.

Q2: How is the H-index calculated in QS subject rankings, and what scores do Hong Kong institutions achieve?

The H-index (Hirsch index) measures the balance between output and impact of a scholar’s or an institution’s entire publication record: h papers each having been cited at least h times. QS subject rankings include an institution’s H-index as part of the “research impact” component; the weight varies by discipline, usually between 20% and 30%. In the 2023 QS Engineering & Technology subject ranking, the Hong Kong University of Science and Technology (HKUST) scored 91.3 (out of 100) on the H-index, and HKU 89.7, compared with 100 for the Massachusetts Institute of Technology and around 96 for the University of California, Berkeley. The gap does not look dramatic. Yet the advantage Hong Kong institutions hold on the H-index tends to be diluted by the amplifying effect of the academic reputation weight: HKUST’s reputation score in engineering was only about 87, while Berkeley reached 100. The actual difference in research impact between the two is far smaller than the reputation score gap suggests.

Take City University of Hong Kong (CityU) materials science as another example. In the 2023 QS subject ranking, its H-index score exceeded those of several UK Russell Group universities, but its academic reputation score lagged by more than 10 points, pulling the overall rank downwards. This pattern – objective metrics close, subjective indicators wide apart – recurs repeatedly for Hong Kong institutions.

Q3: How heavily are the academic reputation survey respondents weighted toward Europe and North America?

The QS academic reputation survey reaches roughly 150,000 scholars each year. Methodological supplementary documents published by QS in 2019 and 2021 reveal a pronounced regional imbalance in the composition of the scholar database:

That means fewer than seven out of every 100 voting scholars are likely to come from Greater China and East Asia – the primary academic collaboration zone for Hong Kong universities. Most scholars are more familiar with institutions on their own continent, naturally boosting the reputation votes of local universities. The academic reputation scores of American universities generally sit above what their bibliometric performance would predict, partly because North American scholars form more than half of the sample pool. Hong Kong universities’ research networks span East Asia, Europe and the Americas, yet the voting base is concentrated precisely at the end that is least “local” to them.

Q4: Do actual citation data suggest that Hong Kong universities are undervalued?

When one compares citations per paper between HKU and American public universities in the QS 50–60 band (and nearby), the picture sharpens:

Look at CUHK: in the 2020 Research Assessment Exercise (RAE) conducted by the University Grants Committee (UGC), 26% of CUHK’s research outputs were rated “world-leading” (4-star) and 47% “internationally excellent” (3-star). HKUST’s 4-star proportion was even higher, at 31%. In the contemporaneous UK REF 2021, the median 4-star share for Russell Group universities was around 28%. The assessed research strength of Hong Kong institutions sits in the same tier as that of research-intensive UK universities, yet in QS academic reputation scores, UK universities with equivalent research performance typically score 5 to 10 points higher – a gap that can almost entirely be attributed to respondent distribution.

CityU provides another case. In the QS 2023 Mechanical Engineering subject ranking, its citations-per-paper and H-index scores both placed it in the global top 50, but the academic reputation score dragged its overall rank outside the top 80. The split in the indicators points to one clear conclusion: objective metrics show that the research impact of Hong Kong institutions is on a par with, or stronger than, that of peer institutions in Europe and North America; the subjective reputation survey effectively “discounts” their actual strength.

Q5: How does this “dark side of rankings” actually affect mainland and overseas students who choose to study in Hong Kong?

Many students prioritise composite rankings when selecting universities, which can lead to two direct consequences:

  1. Underestimation of the quality of postgraduate training and academic resources in Hong Kong institutions. According to UGC statistics, in the 2022/23 academic year Hong Kong’s eight UGC-funded universities enrolled about 31,000 non-local students, the majority being mainland Chinese. Many applicants still use the QS composite rank as a key decision-making factor, without looking deeper into actual research impact at the subject level. For instance, HKU Dentistry placed third globally in the QS 2023 subject rankings, but that contributes little to its composite rank. Students who rely solely on the overall table may miss out on world-leading teaching and research opportunities in particular disciplines.

  2. Influence on employment expectations and policy choices related to staying in Hong Kong. According to the Immigration Department, around 18,000 visas were approved under the Immigration Arrangements for Non-local Graduates (IANG) in 2023, an increase of about 20% year-on-year. The local graduate job market’s recognition of a Hong Kong qualification does not hinge on QS ranking fluctuations; employers place more weight on professional accreditation, internship experience, and Chinese-English proficiency. However, some overseas students, influenced by ranking perceptions, may choose other study destinations, unintentionally reducing the diversity of Hong Kong’s talent pool.

The Education Bureau (EDB) has also explicitly reminded institutions in recent years that rankings are only reference tools. In a 2023 Legislative Council reply, the EDB noted: “Different ranking agencies adopt different indicators and weightings. Universities should focus on enhancing the quality of teaching and research rather than chasing rankings.” This logic aligns with the “dark side” revealed here.

Q6: Can Hong Kong universities adjust their strategies so that rankings better reflect their real academic contributions?

Structurally, QS weight adjustments are ongoing. The introduction of the H-index was itself a corrective to a purely reputation-based survey, although the H-index also has disciplinary skews: it inherently favours life sciences and physics while reflecting less well on the humanities, social sciences and engineering practice. In theory, Hong Kong universities seeking to improve their ranking positions could take several kinds of action:

Whether these strategies can be implemented without distorting the academic mission remains debatable. A former Vice-Chancellor of CUHK once said publicly: “A university should not become a slave to rankings.” Ranking agencies, too, continue to recalibrate. QS lowered the academic reputation weight from 40% to 30% in its 2024 edition and added “sustainability” and “employment outcomes” indicators – changes that may gradually narrow the reputation bias. Yet as long as subjective surveys still command a sizeable share, the sense that Hong Kong universities are “getting a raw deal” will not go away.

Q7: What other tools are available to assess the true strength of Hong Kong universities if we look beyond composite rankings?

Students and parents can combine several information sources:


The dark side of ranking algorithms is not that they “cheat”; it is that too many of us treat them as the sole yardstick for measuring educational quality. Once you know that nearly half of the respondents in the QS academic reputation survey come from North America, and that the H-index only captures certain discipline-specific citation patterns, you can re-read the position of Hong Kong universities on the league tables. What you see may be closer to a composite image stitched together by geography and algorithm than to an accurate portrait of the institutions themselves.


分享本文到:

用微信扫一扫即可分享本页

当前页面二维码

已复制链接

Related Q&A


Previous
HKU vs NTU vs NUS: Similar Ranks, Vastly Different Costs and Returns
Next
Arts & Humanities Rankings: The Gradual Slide of HKU and CUHK — Is Hong Kong’s Cultural Edge Fading?