Accessibility Page Navigation
Style sheets must be enabled to view this page as it was intended.
  • Facebook
  • Twitter
  • LinkedIn
  • YouTube

The marriage of TEF and REF – can research quality be an answer to better teaching and international standing?

Thursday 29 June 2017

Professor George Saridakis

By Professor George Saridakis

The recent results of the Teaching Excellence Framework (TEF) provide some useful insights about teaching quality in United Kingdom universities.

Unquestionably teaching is a critical function of higher education institutions and an integral part of academic duties and life. Alongside the Research Excellence Framework (REF), the TEF can help universities to re-evaluate their strategies and structures in ways that will improve the quality of their academic programmes and enhance their students’ skills and abilities as well as their learning experience.

This evaluation and tailoring of how universities operate also potentially improves their competitiveness and hierarchical standing within the education sector at both national and international levels. In turn, better educational provision and outcomes have been found to have economic and social benefits that contribute to economic growth and development as well as social cohesion and a balanced development of the social fabric.

What isn’t widely appreciated, though, is that the TEF results also provide an additional opportunity – to explore the potential relationship between research quality and teaching performance. Some might say this relationship is a marriage in jeopardy while others might claim that the divorce proceedings have already been issued. There is also the optimistic view of a happy marriage, though, suggesting a positive relationship between teaching and research proficiency. It will be interesting therefore to see if the TEF results can somehow unpack the understanding of this sometimes controversial partnership.

If one looks closely at the relationship between a university’s overall, average REF research rating and its recent TEF outcome (see Figure 1), it’s interesting to note that when the REF grade point average increases, the probability of being awarded a Bronze award decreases from 39 per cent to about 9 per cent while the chance of achieving a Gold award increases from 13 to 48 per cent.

The relationship between the REF outcomes and the Times Higher Education World University Rankings is also intriguing (see Figure 2). For simplicity, the graph has been collapsed showing the probability of a high world university ranking (within the top 250 world universities) and a low one (250 and below). As the REF grade point average increases, the probability of being a low-ranked university starts to noticeably decrease – especially with a REF rating of above 2.8. In fact the probability decreases from 94 per cent to close to zero.

Interestingly, although it is not presented here, TEF results have found not to be associated with Times Higher Education world ranking positioning.

These results suggest that research performance is positively associated with teaching quality, so if universities aim to improve both their world ranking position and TEF results they should further invest in boosting research culture among staff and students and promoting research-inspired teaching.

In other words, research can benefit teaching by bringing up-to-date theories, methodologies, practices and debates together in the classroom. This is likely to alter positively students’ views on the quality of teaching and better prepare them for the labour market challenges. Finally, as academics have long asserted, teaching exposure also increases research opportunities and stimulates research ideas.

Figure 1

Figure 1 - predicted probabilities: Gold, Silver, Bronze


Figure 2

Figure 2 - predicted probabilities: High-ranked vs low-ranked universities

These two graphs – compiled by the author – plot predicted probabilities from the estimated ordered probit model of the TEF outcomes (Figure 1) and ranking groups based on Times Higher Education World university rankings (Figure 2) against overall REF scores. Ordered probit is used when the categorical dependent variables have a natural order.