Why is getting the subject ratings right crucial?

Duncan Ross explains how Times Higher Education explores subject ratings that suit different education systems

April 23, 2020

Source: Alamy

The history of academic subjects is a long one.

In the 1230s, the University of Paris began to categorise academic pursuits into four broad faculties: Canon Law, Theology, Medicine and Arts. But it did not stop there.

In Western universities a series of relatively consistent divisions were created throughout the Enlightenment, and these were refined and codified both in national systems and also in bibliometric systems.

Today many systems coexist.

In the UK there is the JACS system – the Joint Academic Coding of Subjects, which contains 164 principal subjects.

In the US there is the CIP – the Classification of Instructional Program system, which at its most granular level breaks down subjects into 2,342 categories.

In Oceania there is the far-reaching ANZSRC – the Australia and New Zealand Standard Research Classification, which breaks down fields of research into 157 groups and 1,238 fields.

This approach to classifying subjects of academic inquiry is not restricted to nations.

Non-government organisations also have their own schemata: for example, Scopus, the research database maintained by Elsevier, uses the All Science Journal Classification Codes (which, despite the title, extend beyond science), which have 334 distinct classes.

At Times Higher Education, we explore subjects through our rankings in only 11 broad subject areas at present.

Other organisations have chosen to assess at a finer level of subject detail, but we have always been cautious of doing so, for two main reasons.

First, we do not believe that the data that universities are able to provide for rankings are fine-grained enough to allow this in a way that we would deem acceptable, given the huge influence that the THE World University Rankings have over institutional decision-making.

For example, if a mathematician also provides lectures for the physics department, does their research output count towards the mathematics or physics subject? And if a conference proceeding involves both computing and engineering, is the author a computer scientist or an engineer?

Second, as you divide the rankings into more detailed subjects, you inevitably end up imposing a particular view of how subjects should be classified: you have to fall in line with one system or another.

This, in itself, is a problem. Why should a student, or academic, in China have to explore the data on their universities through the prism of a US-based classification of subjects? Why should an Australian student have to use UK-based divisions?

At the same time, there is an increasing need for this type of subject-level analysis.

If we return to our theoretical Chinese student – they may be interested in exploring potential universities both within China and also in the US.

It would make sense to allow them to see both US and Chinese universities from the perspective of the subjects with which they are familiar – the Chinese subject classification. Similarly, a Chinese academic might want to understand opportunities to collaborate with universities elsewhere – once more through a familiar lens of a system he or she understands.

The same logic applies elsewhere. A student in Australia who is thinking of attending Peking University will understand the Australian system of subject classification, so it would be ideal if THE could represent Chinese and Australian universities in a common and understandable way.

As an initial step towards this, THE is exploring ways to produce a new level of more detailed university subject assessment, based on a rating of performance.

The approach will draw from the data and methodology of our World University Rankings and categorise university subjects on a scale from A+ to C.

We will initially explore this using a classification of subjects according to the Chinese subject hierarchy, and we look for feedback from academics and students on our approach.

duncan.ross@timeshighereducation.com