Deciding regulatory assessment rules in GCSEs, AS and A levels

We have developed a set a principles for when we think about rules for aspects of GCSE, AS or A level subject assessments.


Applied requirements for GCSE assessments

This file may not be suitable for users of assistive technology. Request an accessible format.

If you use assistive technology (such as a screen reader) and need a version of this document in a more accessible format, please email Please tell us what format you need. It will help us if you say what assistive technology you use.


Most of Ofqual’s regulation of GCSEs and A levels is based on the conditions that it publishes. If there was just one exam board for each GCSE or A level, Ofqual’s role would be very different.

Rather than writing multiple conditions about assessment objectives, the proportions of non-exam assessment, or tiering etc, we might just state that the exam board should design its assessment in the way that optimises validity and is the most defensible. In making those design decisions we would expect the exam board to balance what should provide the most valid assessment against threats to that validity and wider concerns.

We might want the rationale underlying the decisions it makes to be set out in an assessment strategy that we could consider and challenge.

Of course we don’t have just one board. It’s the multiplicity of boards that arguably drives us to write many of our regulatory requirements. That is because allowing boards to take different positions on the weighting and nature of assessment aspects may prevent demonstrably comparable standards being set between qualifications with the same title.

That comparability between different boards in a subject is critical to the system –recent commentaries on small variations in the level of demand of some GCSE sample assessment materials illustrates the sensitivity of this issue. The problem we have is that if we simply ask the boards to design assessments so that they optimise validity and are the most defensible, they could reasonably come up with different but plausible solutions. Alternatively they could all come up with a similar solution to an issue but one that does not command our confidence and may not be in the public interest if, for example, it provides less demanding assessments than we judged were required.

In these circumstances, even once the Department for Education has prescribed the content of a subject, we feel obliged to specify:

  • the assessment objectives and their weightings
  • the proportions of exam and non-exam assessment
  • whether the exams should be tiered and if so how

We do this because we do not believe these aspects of assessment are appropriate for competition between the boards in a system where comparability of standards is a main driver. That does however still leave some scope for the boards when designing their specifications. For example, there is freedom to choose the types of exam questions that are used, whether an exam is on paper or on screen, and the design of marking schemes for non-exam assessment components. In addition, some of the requirements we set down are expressed as minima, maxima or ranges, leaving scope for variation in designs.

So we have a high level principle:

  • Ofqual should only specify an aspect of assessment, where not to do so may prevent demonstrably comparable standards being set between qualifications with the same title over time.

Where applying that principle indicates that we should create a regulatory requirement for an aspect of assessment, our guiding principle should then be that:

  • We write a regulatory requirement in a way that optimises validity and so is the most defensible one. In making that selection we balance what ought to provide the most valid assessment against threats to that validity and wider concerns.

In considering threats to validity, we are particularly conscious of the washback effect from assessment design to teaching and learning. We look to put in place assessment arrangements that should help realise the learning outcomes described in the subject content.

Regulations should usually take the form of conditions or requirements, depending on what is being specified. There are occasions though when it might be more appropriate to produce statutory guidance instead.

The attached table describes what requirements we have applied, or could have applied, in recent developments for eight aspects of GCSE assessment. They were written when requirements for the first two phases of reform had been published. The requirements in the table do not therefore purport to be those in place before the current reforms began and are not the basis on which Ofqual took all its decisions. A similar table could be produced for AS and A levels.

The first two rows in the table raise a question about what we do when we have subjects which are offered by only one exam board. Of course, even if only one board initially expresses interest in a subject, we must still regulate in a way that would allow another board subsequently to enter that subject market. So we must always assume there will be competing boards in a subject and seek to mitigate any possible adverse effects of competition between them.

Such subjects should be suitably aligned with others, particularly similar or cognate subjects, to ensure that we are not allowing the creation of what may be perceived as ‘soft options’ through, for example, much higher weightings of non-examination assessment.

Published 25 November 2016