Earlier in the year I blogged about the removal of levels from the national curriculum. I ended that post daunted by the prospect of designing an assessment framework to fill the gap. Now the new National Curriculum has been through its cursory consultation and is “official”, live and statutory for maintained schools. Where at the back there used to be attainment targets correlated to national curriculum levels, there is now this simple statement:

Statement from the new National Curriculum for Key Stage 3.
And that’s your lot.
This week, I met with the excellent Heads of Faculty in our Curriculum Leadership Team to decide what we were going to do with all this autonomy. I’d asked each team to look through the curriculum and work out to what extent the existing assessment framework mapped on to the new curriculum, and where there were changes needed. With some subjects – Science, Art and Design, Languages – there was little change necessary, though some might be desirable. In others – Computing for example – the new curriculum is practically unrecognisable from the old ICT which was disapplied. Drama was never a standalone subject in the old national curriculum, existing in an annex of the English documents; now it does not exist at all except in literary study.
In almost all cases, the old assessment criteria cover far more than the new, thinned-down curriculum requires. There is of course no problem with continuing to teach and assess material that is not in the curriculum. But it is essentially impossible to sensibly match the old assessment regime to the new curriculum, even if that is what primaries are still having to do with SATs. Instead we must look again at assessment, who and what it is for, and what we want to do with it.
Core principles: criteria or norm referencing?
It seems as though the current policy direction is towards norm referencing. Ofqual’s use of comparable outcomes was defended to the education select committee, and the DfE has proposed a decile ranking system for primary schools. The removal of detailed attainment targets would open the way for a percentage or rank-ordered system of assessment. It would be feasible to assess with statements like “Sarah has successfully met 78% of the requirements of the programme of study this year” or “Richard’s attainment ranks him between between the 50th and 60th percentile in his year group.” To understand why we won’t be adopting this approach, read Sue Cowley’s Dear Tom and Debra Kidd’s Pride and Prejudice. Or just be a human being.
Criteria referencing is fraught with difficulty however. Subject experts were given the full time job of designing criteria for attainment at Key Stage 3 under the National Strategies and the behemoth of APP was created. The advantage is that learners can see where they are in their learning, what they know and what they can do, and they can also see what they need to know and be able to do to improve. The disadvantage? Well, you try condensing the whole of “English” or “Science” into a ladder of statements that a twelve-year-old can access and hold in their head. In your spare time. After a full day’s teaching. And marking. And planning lessons. Yes, in the ten minutes before you actually drop off into exhausted catatonia, please condense your entire subject into a series of accessible scaled criteria. By Monday.

Criteria-referenced assessment vs teacher; round 1.
However difficult the practicalities, however, I am wedded to criteria referenced assessment simply because it’s formative; it’s assessment for teaching, not just for measuring.
What is assessment for?
Assessment in schools is multi-purpose, which may be where the problem lies. This week Michael Gove has made further steps to separate the school-accountability-purpose of GCSE assessment from the measuring-attainment-of-learners purpose with his announcement about early entries, causing panic and confusion around the land. Within school, however, especially with the removal of a national framework, we want an assessment system for the following:
- For learners: to understand how they are progressing and what they need to do to progress further
- For teachers: to understand and measure what learners know and what they can do, in order to plan their teaching to move their learning on
- For parents: to see how their children are progressing and understand how they can help and support them in improving
- For school leaders: to understand and monitor progress and attainment across the school
This was the challenge for curriculum leaders this week. Could any assessment system possibly do all that?
Measuring progress

Can progress be mapped as a flight path?
I was quite taken with @leadinglearner Stephen Tierney’s post about progress Flight Paths. I liked the neat linear and mathematical composition of the progress charts he’d constructed. However, it presupposes a uniformity that I find problematic. I would rather take a personalised approach. When we set targets (we call then Challenge Grades) I use a standardised formula, but then open them up to teachers to suggest amendments (read about the process here).
The other issue with the standardised “flight path” approach is that using levels in that way under the old system is problematic as levels don’t necessarily correlate to a set point in progress below a GCSE grade. Under the old system, I always used the following “rule of thumb” correlation based on the Levels and Grades Conversion Tables I referred to in my own post about targets:
- Level 4 = equivalent level of attainment to a GCSE grade E
- Level 5 = equivalent level of attainment to a GCSE grade D
- Level 6 = equivalent level of attainment to a GCSE grade C
and so on. As an English teacher, I could see the sense in this. However, as Tom Sherrington (@headguruteacher) points out in The Data Delusion it’s a lot more complicated than that. In Languages, for example levels are traditionally lower as the students start Year 7 from scratch, often at level 2 or 3 (a fact recognised in a footnote to the National Curriculum for English (2000) – page 54!). However, those days are gone now – Languages teachers can call what used to be “Level 4” whatever they like now. Orange? Skylark? Mountain Lion? KitKat?
What we are going to do
The curriculum leaders all felt that they knew the existing system of assessment well and that, by and large, it worked for them. We now have the freedom to change, adapt and alter those bits that don’t work, add to or subtract from the existing criteria to suit. We also agreed that levels were preferable to telling a Year 7 that they were working at a grade E, so we are going to continue to use them. Over the coming months curriculum leaders are going to:
- Review existing assessment criteria and “tweak” where necessary to fit the new national curriculum and the subject content that they are including in their own programme of study
- Look for direct lines of progress to GCSE specifications and adjust assessment criteria according to the “rule of thumb” equivalences
- Experiment with formative approaches to make “next steps” explicit to students and, if possible, to parents
The first area gives subject specialists the autonomy to tailor assessment to the needs of their curriculum content in a way which I think is liberating and empowering. The second area aims to provide a consistency and coherence to our assessment in a way that the old national curriculum levels did not, meaning that direct comparison can be made between disciplines by learners, leaders and parents. For subjects like Drama which have never had levelled criteria of their own, this is an opportunity to build something of real value; it will bring Languages and Computing into line across the curriculum.
The third area is fraught with problems as the range of criteria is so large and condensing them blunts their accuracy. The National Strategies foundered here with APP; we may do the same. However, I am hopeful that a direct link between Key Stage 3 and GCSE assessment which can be tailored to the specification used in our context will provide more meaningful and coherent progress paths for learners in our school.
The System Problem
That last sentence presents the unresolved problem – “in our school”. At our meeting on Wednesday we effectively instituted “Chew Valley Levels” – the terminology of levels may sound like old National Curriculum language, but they have a very specific meaning to our context. The criteria may not match other schools. A level 5 in French at our school may not be the same as a level 5 at a neighbouring institution. A cynic might now chime in with the observation that this was always the case under the old National Curriculum too as nobody ever standardised. (They might also go on to cite the usual point about wide variation in Year 6 teacher assessments across a variety of primary feeder schools, but before they did that I’d urge them to read Michael Tidd’s blog Dear Secondary School Teacher…) The cynic might be right in some ways, but under the old system there was at least the expectation of national consistency. That, like the expectation of a common pay structure, common terms and conditions, and common admissions, is a thing of the past.
UPDATE – MAY 2014: I have added a new post giving our progress with assessment in the new national curriculum, including refinements and amendments to the ideas above. Click here to read!