What makes great assessment?

What makes great assessment?

Following his appearance as a panel member discussing 'What makes great assessment?', organised by Evidenced Based Education in partnership with Cambridge Assessment and the Chartered College of Teaching, Tim Oates discusses the current and foreseeable assessment challenges and uncertainties facing educators, and what action the profession can take to address these challenges.

'Great assessment' is accurate. It measures things with fidelity, and by doing this, provides valuable information. It can be depended upon, whether it is high stakes summative assessment in programmes such as medical training or in general education, or whether it is putatively 'low stakes' formative assessment designed to directly support learning. The question in the classroom which probes a child's understanding and where the answers help the teacher's insights into the way in which a child's understanding is building can be as important to the educational life of a child as the grade he or she obtains in a high stakes examination.

Accuracy is at the heart of good assessment. Some analysis explains 'accuracy' in terms of 'validity' - at it simplest, whether an assessment measures what it is intended to measure. If this is a fundamental quality, alongside it, an assessment must be fit for purpose: immune to maladministration, manageable, affordable, with timely and useful results. Of course, these qualities are easy to state in principle and difficult to deliver in practice - these features can be in tension: high accuracy can mean high cost; dependable assessments can mean high levels of bureaucracy. These real world tensions mean that fine tuning, using the idea of 'fitness for purpose' is at that heart of practice in designing, administering, and evaluating 'great assessment'.

Challenges and uncertainties.

Many nations - not just England - have begun to 'up the stakes' around assessment - to use assessment as part of public accountability, and to try to 'steer' education through assessment. Under such pressures, assessment can begin to dominate the curriculum, and incentives to 'cheat' can become overwhelming - whether this is on the part of candidates or on the part of teachers being held to account. There comes a point where good assessment design and careful requirements around administration of the assessments can only do so much to support the dependability of assessment, and incentives and drivers around the assessment begin to distort both assessment and the curriculum.

The 'Goldilocks Principle' is at centre of good assessment: assessments should be challenging (to stimulate and support worthwhile learning) but not so challenging that they are unattainable through reasonable effort. Assessments should be valuable to the states which so frequently fund their development and administration but also seen as valuable by teachers and students. They should support equity and attainment in education but still discriminate accurately between those who have attained something and those who have not. The 'sweet spot' in optimising these different tensions is not easy to maintain, but must constantly be sought.


While work on 'crowd sourced assessment' shows that good questions abound in education, there is still need for highly technical quality assurance to select and refine those questions to enable tests and examinations to meet quality thresholds. Assessment agencies thus are not some institutional anachronism but offer a vital service - a service which should be delivered in conjunction with teachers and education institutions. Building trust and mutuality of interest in dependable assessment is essential - this is one means of ensuring that assessment is not seen as something wholly external to education - to be gamed whenever possible. Rather, good education needs to be seen as the best route to assessment - and this then places assessment in its rightful place - as the servant, not the master, of curriculum aims. Technology will help with robust, unobtrusive and relatively inexpensive security around assessment, just as it will help with easier registration and test-taking.

And finally, there lurks the question of cost. High quality assessment acts as a vital bridge between so many parts of the education system; reassuring parents of children's progress; reassuring the State that schools and colleges are pursuing desired goals; linking education and the labour market by providing assurance of skills and knowledge. Good assessment should be relatively inexpensive - for example, in England the cost of all GCSEs and A Levels comes in at just under 2% of the total budget for secondary schools. It is thus not cheap, but it should always be worthwhile - and thus justify its cost. Efficiencies should always be chased, but not at the expense of compromising trust and quality.

Tim Oates
Cambridge Assessment

Follow the Twitter debate under #AssessBetter

Missed the live panel debate? Watch it all here.

Download Assessment Academy's ebook here.

Research Matters

Research Matters 32 promo image

Research Matters is our free biannual publication which allows us to share our assessment research, in a range of fields, with the wider assessment community.