Professor Rob Coe, director of Durham University's Centre for Evaluation and Monitoring and an adviser on the development of the proposed new GCSEs, writes: 

Two weeks ago, Ofqual launched its consultation on setting grade standards at GCSE. The following week they gave more details on the arrangements for reformed GCSE, AS and A-level qualifications.

This week, so far, has been mercifully free of any new announcements. But the amount of change already in the pipeline is pretty bewildering: a new national curriculum, the end of levels, the disappearance of coursework, practicals and fieldwork from many assessments, and the divergence in assessment between England and previously comparable qualifications in Wales and Northern Ireland.

Teachers returning from the Easter break are about to begin a phase of change that many feel is unprecedented in terms of its scope and speed. Will any of this revolutionary upheaval actually improve the problems it was designed to address?

In its work on grade standards, I believe exam regulator Ofqual is moving in the right direction. The consultation document brings some welcome acknowledgement of the difficulty of setting and maintaining standards. In particular, I am pleased to see an attempt to give a coherent definition of what we mean by ‘standards’, a recognition of the limitations of past and current practice in England, a useful discussion of why apparently simple solutions including norm-referencing will not work, and an understanding of the specific problems that arise when new qualifications are introduced. All this is honest and sound – a rare combination in assessment policy.

A consideration of how grades are used and interpreted is important, and allows us to focus on understanding the different ways ‘standards’ matter, rather than just assuming there is a single, all-purpose standard.

The introduction of the national reference test is also welcome. In the last few years, Ofqual’s ‘comparable outcomes’ approach has linked GCSE grades to that cohort’s previous KS2 performance. This seemed to me to be an appropriate response to a system in which grades went up every year on the back of ‘benefit of the doubt’ judgements by examiners, without independent evidence to support the resulting trends. However, I have always seen this as a stop-gap solution. The lack of alignment between what is assessed at KS2 and the GCSEs it is used to anchor, the time interval between the two assessments and the fact that the KS2 assessment itself changes each year and is hard to anchor robustly, are all limitations of this approach.

Longer term, we really need a standard setting process that allows grades to rise if and only if there are genuine improvements in students’ performance. A well-designed national reference test should help make this possible.

An interesting feature of some of the discussion of standards is a confusion between raising the bar for a particular grade threshold and actually improving national performance. Moving from a C to a 5 as the currency of a ‘pass’ will certainly make it tougher to achieve, but won’t make anyone cleverer. Hitching our threshold to something comparable within a high performing jurisdiction will not turn us into a high performing jurisdiction. And just creating more grades at the top end in the new 9-1 scale will not make more people achieve them.

At the heart of this confusion is a misunderstanding of the importance of high expectations. Policymakers believe (rightly) that having high expectations is a characteristic of the best education systems. But it is the expectations of teachers that matter, and research tells us that changing teachers’ expectations is hard. Experienced teachers pretty much know what their students are likely to achieve; we know that simply raising the pass mark, or exhorting schools to demand more, makes no real difference to this. On the other hand, intensive, sustained, focused, high-quality professional development can make a genuine difference to what teachers believe is achievable, and to their students’ learning (an example is a New Zealand study by Helen Timperley and Gwenneth Phillips).

If we really want standards to rise, we need to find ways to make this kind of professional development the norm. To do that we may need to reduce the system-wide effort that is absorbed by coping with extensive, but essentially unproductive, changes.