I will deal directly with the questions of how – and whether – we should assess outcomes such as ‘successful learners, confident individuals and responsible citizens’.
We have well-designed tests ‘on the stocks’ for virtually anything – at one end of the spectrum, tests of knowledge; at the other, tests of dispositions such as ‘externalising behaviour’, motivation, etc. For me, the question is not how to assess such outcomes as much as the wider validity questions – what’s the purpose of doing it, and how are we going to use the results? We have heard from other speakers that there won’t be tests of ‘confidence’ – and therefore, presumably, no national targets for ‘confident learners’. It is perhaps self-evident that that’s not sensible. But actually – from inside the assessment community looking out – neither is it sensible to treat current assessment data in the way in which we do. Simply adding up the outcomes of each young person and presuming that this gives an unproblematic picture of a school and of the national situation is just not in line with contemporary, state-of-the-art thinking in assessment. This leads to the first point:
1: Treat pupils, schools, and the system as presenting different measurement issues
If you are interested in maximising the attainment of individuals, then one sort of assessment is likely to be optimum. Assessment for learning points the way forward. But simply adding together the assessment results from this, and assuming that this will give good accountability measures for schools – a very necessary thing – is highly problematic. If you want to know about the performance of the education system – for example, whether the national curriculum is being adequately covered – then a different approach is needed again. It is essential to get evidence for public policy on all three levels, but ‘one size’ is very likely not to fit all these purposes.
Yet if we set ‘successful learners, confident individuals and responsible citizens’ as a policy goal, we must have some means of seeing if the policy is being delivered. I would argue that this should not be done by assessing each child on each of these dimensions. But this does not rule out the use of formal tests for monitoring purposes. Just as John Bynner and Leon Feinstein have assessed individuals’ ‘externalising behaviour’ (life just happens to me, I can’t control things…) and shown how this relates to poor educational and life outcomes, so it would be sensible to survey with precision how the system is doing on developing ‘successful learners, confident individuals and responsible citizens’ – by assessing a carefully selected group of learners at a sensible series of intervals.
And on the suitable forms of assessment for continuous assessment, while senior civil servants have suggested that ‘…we’ve tried assessment for learning and it hasn’t worked…’, they are underestimating the pervasive nature of a narrow ‘performativity’ culture in our education system – written about so effectively by Warwick Mansell and repeated almost universally in the evidence on national assessment to the Select Committee on Education and Skills. This culture almost certainly suppressed the capacity of some of the pilot work on assessment for learning to show the anticipated benefit. I will return to this issue of ‘system conditions’ in a minute.
2: Do not assess everything which moves
Development of ‘successful learners, confident individuals and responsible citizens’ should, I believe, comes from the total curriculum which they receive, not from some discrete, partial ‘bit’ of their educational experience. It would be unwise to introduce an assessment of these things and assume all will flow form this. ‘Washback’ effect from assessment is not easy to manage – unintended consequences of assessment policy and practice arise all the time. Certainly, introducing a direct assessment of these aspects of education will bring attention to them, but by itself is unlikely naturally to lead to the kind of learning and educational experience which will result in their ‘deep’ development, rather than a superficial appearance of acquisition. Curriculum enrichment activities are highly valued as a form of motivating engagement in the life of a school, and the very fact that they are uncertificated often leads to children wanting involvement and leads to greater overall engagement in learning – but more of this under ‘pupil voice’ in a moment.
3: Understand the patterns of drivers and incentives which emerge from different approaches to measurement
Different assessment models generate different patterns of drivers and incentives. It’s clear that there are significant undesirable washback effects from existing National Assessment - and it’s currently unclear what patterns will be created by the new Progress Tests. While some assessment models (combined with certain forms of accountability measures) encourage teachers to cover the whole of the curriculum, others stimulate narrowing convergence on ‘that which is assessed’. It’s not just a question of having a decent balance of assessment approaches and attending to the validity and reliability of those assessments, it’s a question of understanding how the system context causes a specific combination of forms of assessment to have specific impact on incentives and drivers. It’s a crucial issue. Get it wrong, and your policy objectives immediately become distorted.
4: Focus on curriculum coherence and removal of phase discontinuities
Schmidt’s interesting international comparative work suggests that curriculum coherence is vital – by which is meant a ‘common message’ running through all elements of curriculum content, guidance materials, teacher training and professional development, and so on. Countries which have ‘curriculum coherence’ in this sense – Asian and Nordic countries – have done well in international comparisons. Conversely, countries which do not have it – such as the USA – do badly, despite having a strong ‘accountability’ apparatus. Likewise, discontinuities in learning approach, expectations and targets regarding knowledge skills and understanding between different phases of education are both confusing to learners and dysfunctional. Curriculum innovation should be seen not merely as inserting apparently worthy elements into the national requirement but as ensuring both high utility in content and good curriculum coherence. A system lacking in curriculum coherence, however worthy the components of the curriculum, is likely to be dysfunctional.
5: Listen to pupil voice and establish consistent ‘lived experiences’ – qualitative approaches as a form of measurement
‘Successful learners, confident individuals and responsible citizens’ are likely to emerge in schools where learners feel intimately involved in the life of the school, and are involved in processes such as schools councils and anti-bullying support. There’s plenty of work on this; the ‘lived experience’ of the school should embody and re-inforce the values which the education system is seeking to see established in young people. To see whether specific arrangements are working, it is vital to collect evidence on how young people are experiencing the school and the meanings they are placing on that experience. No-one did that better than the late Jean Rudduck, whose work on ‘pupil voice’ is a beacon of excellence. Sensitively collecting evidence on ‘lived experience’ provides vital qualitative evidence of the extent to which a school is developing ‘successful learners, confident individuals and responsible citizens’.
6: The impact of overcrowding the curriculum - ‘more room please’ - personalisation and reinforcement
And finally, the history of the English National Curriculum has been a history of an over-packed curriculum slowly being thinned out, but still with subjects ‘jostling’ for space and position, and a dominant ‘rush, rush, rush’ concern by teachers for getting through the huge amount of material which is present in each key stage. Forget about ‘flexibilities’; many teachers feel under huge pressure to move through the required material for fear of not covering enough, quickly enough. Again, those countries which do well in PISA and other international comparisons tend to have slimmer national cores, which allows for careful repetition and rehearsal, sophisticated iteration of theory and practice, and careful tailoring of the presentation of material to individuals. These are crucial to consolidation of learning and ‘deep understanding’ rather than ‘surface learning’ – and are essential for delivery of genuinely personalised learning. Indeed it is difficult to imagine personalisation without these as the key elements. I remain convinced that a national curriculum based on a solid core for all is vital, but that we’ve still not got the scope and balance of content right yet.
‘Successful learners, confident individuals and responsible citizens’ seem very much the right aims – we need to use the best knowledge on curriculum and assessment to ensure that they are genuinely the outcomes of education.