To level or not to level

For many of us, levels are an intrinsic part of the National Curriculum. We've come to view the notion of progression through levels from the ages of 5 to 14 as a key element of teaching and assessment. We’re also used to being held accountable for the progress made, and justifying the work we do, by identifying how the pupils we teach are now working at a higher level.

It’s a vagary of the calendar, but not one that my own children are reluctant to exploit, that I started teaching not only in the previous century but in the previous millennium (thanks guys). It’s certainly true that I taught in a pre-National Curriculum levels system; whereas there was more freedom, I’m not sure we used it well and I’m certainly not inclined to hark back to it with longing (apart from the music).


We’re now faced with the challenge of deciding how best to assess progress and outcomes through new programmes of study. There’s no expectation to use levels, but we do need to track pupils effectively.

Which way now?

I think it’s important to start with looking at the purposes of assessment. I would suggest that there are multiple outcomes and that a good system needs to provide all of these. In essence we need to use assessment to:

  • inform next steps in teaching
  • provide meaningful feedback to pupils on what they've done well and how to improve
  • provide evidence for accountability
  • communicate to parents and carers how their children are doing.

Are levels (or a reworked system thereof) the answer?  What do they offer?

I would suggest that two characteristics of levels are that they offered a standardised scale and that they recognised progress in terms of stage, rather than age. Unless there’s a staggering degree of unity amongst schools I suspect the first to now be a forlorn hope but the second is worth considering. ‘Stage not age’ is useful in that it recognises that pupils reach a certain level of mastery at different times. Different pupils get to be able to multiply, for example, 56 x 78 at various points in their schooling and recognising this is useful. However, levels require descriptors and standardisation. It only works if your Level 3 judgment is triggered by the same criteria as mine. It also works best in the context of a smooth continuum of skills. Multiplication goes from 2 x 2 to, for example, the multiplication of imaginary numbers, with all points in between. A continuous scale makes sense.

Some subjects don’t work that way, however. In science, for example, the skills of Working Scientifically can be placed on a smooth continuum with little difficulty. However the concepts are more context specific. Key Stage 2 pupils study sound in Year 4 according to the new Programme of Study. Although it is quite possible (and desirable) to develop differentiated outcomes for that topic, this doesn't really work formatively after the end of the topic. Pupils might come back to multiplication next week or next month but they won’t come back to sound until the next Key Stage.

We could therefore develop, instead, a set of age related expectations. As the programmes in the core subjects are set out by year, outcomes from a topic in a year could be identified and pupils’ success at meeting those be recorded.  The idea of using developing/meeting/exceeding is worth exploring more; meeting expectations can be equated to ‘mastery’ and exceeding expectations a trigger for further challenge (in the same way that developing is a trigger for intervention). It would obviously be important to ensure that these expectations are progressive from year to year (all that time getting to grips with APP can pay dividends here).

It can be seen then that the challenge is not so much whether to use levels or not, but rather how to retain the effective aspects of levels and develop an assessment system that works across the whole curriculum (all areas of study should surely have outcomes identified) and can be used to inform teaching as well as support accountability.


It seems to me in developing a system that there are seven key challenges to resolve:

  1. Whether to go for ‘stage not age’ or age related expectations. 

The former has merit with key skills such as sentence construction and arithmetical operations but requires a shared understanding of key characteristics which is harder to secure across the breadth of the curriculum.

  1. How to develop outcomes that represent the breadth of the curriculum.

I’m very attracted by the idea of using the question ‘What outcomes should be achieved by pupils from this area of study?’ because I think that leads to a focus on the fitness for purpose of the learning activities and the use of a wide range of evidence.

  1. Whether to develop a system within a school or across a group of schools.

This isn't an argument for trying to organise a single national system – we might have common features in some subject areas but the whole curriculum (i.e. the School Curriculum, as opposed to the National Curriculum) varies from one school to another. However there is merit in collaboration and sharing ideas about what appropriate outcomes might look like. Having to justify them is a good practice.

  1. Keeping it manageable.

As Einstein said ‘Everything should be made as simple as possible, but not simpler’. He may not have been talking about school assessment systems but it applies.

  1. Making sure the system informs teaching. 

This was a key part of the original purpose of levels, but the accountability agenda muscled in. That’s not going away, but I think we should make sure that whatever we use, it gives teachers a ‘heads up’ on what’s going well and what needs to be revisited.

  1. Ofsted need to make it clear what they expect. 

I don’t mean they should have a preferred system but rather that they should make it clear what characteristics they expect to see. Otherwise we’ll all end up trying to double guess what an inspector will be impressed by and folklore will abound.

  1. How to use digital technology effectively. 

I don’t mean setting up huge (and probably unmanageable) electronic evidence folders but rather exploiting cutting edge practice regarding information systems. As well as spreadsheets more complex than NASA used to land men on the moon, there’s value in a ‘data dashboard’ so that key indicators can be easily accessed.

Well, it’s a challenge. Early indicators are that quite a few schools aren't rushing into anything radically different as yet, though many recognise the limitations of the levels system; this is a window for doing a bit of scoping and trying ideas out. There’s a chance to come up with something that does what we need it to.

Ed Walsh is Lead Consultant for Science with Cornwall Learning and Science Consultant to Rising Stars on the Rising Stars Assessment New Curriculum Progress Tests for science.

This article is based on a session on ‘Meeting the challenge of assessing the new primary National Curriculum’ presented at the Education Show in March 2014.


key stage 1, key stage 2, new curriculum, science, standardised scores

Added to your basket: