In many ways I’m thinking out loud about some of the changes that we are beginning to see in our nearly post-level World. There is life after levels and it might be quite interesting.
Throughout my career two predominant class room foci seem to have been pedagogy and assessment in terms of levelling or grading. Whilst there has been a lot of work around assessment for learning, in the last decade, too much assessment still seems to focus on churning data out and up into a progress tracking system.
At a whole school level assessment invariably means a grade or level, with a need for “fine” levels or grades – for example 4a, 4b & 4c or C-, C & C+ – having evolved over the years. The main reason for this fine levelling was the system’s need to discern progress being made on an increasingly more frequent basis with a number of assessment points during the year. Yet the reality is that students aren’t likely to move a grade or level every six, nine or even twelve weeks. For the soon to be defunct levels the expectation was that one level would take about two years to attain. Whilst the whole fine levels has no real meaning a deeper issue is that it can lead to a lack of clarity about what has been learnt or needs to be learnt next. All this data inputting has also stolen valuable time off teachers.
Rethinking Data Collection
Assessment in the hands of school leaders, local authority officers and Ofsted inspectors has done a disservice to the potential power of assessment in the hands of teachers and learners. One of the key moves schools will potentially have to make, to unleash the power of assessment, is to gather aggregated or graded data far less frequently. We currently have four assessment cycles across the Trust’s academies with data gathered and input in early November, mid-January, late March and early June. That is approximately every nine school weeks and this is the same for both phases. There are only two data gathering and collection points for Key Stage 3 subjects with less than 10% curriculum time.

Photo Credit: Dafne Cholet via Flickr cc
Given the potential limits on reliability of fine aggregated grades and time required to move students a whole grade, what would be the downside of moving to two data collection points per year? What opportunities might it present? The number of summative assessments could remain the same, four times a year, but their purpose wouldn’t be about creating a grade or level for tracking. I’m not sure that my nerves could cope with this reduction in frequency in Year 6, Year 11 or the Sixth Form. Maybe I just need a bit more courage or to make this a subsequent step on the journey.
Putting Assessment Back into the Hands of Teachers & Learners
Dylan Wiliam (2014) questions whether for all the data schools have anyone reviewing it would be able to determine what a student knew, understood or could do.
“What these mark books hardly ever record is what the student can do, and what might be their next steps in the learning. It seems likely that whatever Craig Raine’s Martian did include in his postcard home about record keeping in schools, he would certainly not regard record keeping’s prime purpose as being supporting learning.”
William, D. (2014) Principled Assessment Design, p.71
This final part, “he would certainly not regard record keeping’s prime purpose as being supporting learning”, is a rather damning indictment of much of our work on assessment. If schools were to move to more infrequent collection of grades or levels how might summative assessment change?
Using Summative Assessment Diagnostically & Formatively
Assessment’s role in the learning process should be primarily about diagnosing what a student does and doesn’t know. The diagnosis can take various forms. The graphic below is a question by question analysis of some Key Stage 4 student’s performance in a GCSE Mathematics examination. The students hadn’t actually completed the course and were faced with questions on a number of areas which they hadn’t covered.
If you reimagine this analysis of the GCSE Mathematics paper as planning for your next lessons it takes on a very different and potentially more powerful purpose.
Having chatted with the teacher about the data, the indifferent performance on Q3 (producing a stem and leaf diagram) was thought to be due the topic having been taught quite some time before. Their solution was simple, include stem and leaf diagrams as part of the lesson’s “Maths Meeting” for a week or so to remind students of what was expected.
However, Q4 (forming formulae) had produced a disappointing response overall. The teacher intended to reteach this to the whole class having discussed with colleagues how they approached this. They were particularly interested in finding a teacher whose data suggested that s/he taught this area of mathematics well. It would also be useful to capture this more effective way of teaching “forming formulae” in the scheme of learning and share at a departmental meeting.
If you look at Q6 (Decimal Multiplication) and Q7 (Forming Equations) there are some interesting patterns. Some students (1, 12, & 13) have scored well in both questions. However, students 2, 5, 6 & 11 all have full marks in decimal multiplication but zero marks in forming equations. Whereas students in 3, 4, 8 & 10 have zero marks in decimal multiplication but full marks in forming equations. You need to teach to the gaps. How could you set up the class room to allow one group of students to revisit decimal multiplication and the other group forming equations? What would you do with the students who did well in both questions?
With respect to Q17 through to Q25 these hadn’t yet been taught. The teacher has gained some very useful information about the class’ prior learning with respect to these particular areas of Mathematics. It is pretty much non-existent. This is important as it allows them to pitch the teaching at the right level.
For other subjects, there is also the potential for more time to be spent on feedback with the associated, DIRT or MAD Time with students required to improve their work to a higher standard. This may be more appropriate in some essay based subjects or those with a few larger questions. GCSE RE comes to mind, where students answer four questions. Each question has four components and the skills for each component are the same across the questions – basic recall of a definition, relational recall, explaining and discussing. The analysis of the four sub-components of the question could help target revision practice for the class or an individual on what would have greatest impact on their overall mark and subsequent grade.
The Power of Diagnosing Before Teaching

Phot Credit: Rohit Mattoo via Flickr cc
This kind of detailed diagnostic assessment could be completed following the Year 2 and Year 6 assessments. Imagine if this powerful diagnosis of what students did and didn’t know or could do was passed forward and we threw the proverbial kitchen sink at Year 3 and Year 7 in terms of closing the gap early and decisively. It would be quite a different way of closing the gap than the more usual antics that go on it Year 6 and Year 11. Closing the gap is more of a pre-emptive action, don’t allow the gap to open. It is focussed on closing the learning gaps of every student and is focussed primarily on intervention in the class room.
Assessment is our hidden gem. It has been hijacked by a high stakes accountability system, which we all have to accept our share of the blame for, in which leaders have required multiple graded or levelled assessments to feed the progress machine. The progress machine became so powerful and hungry that we had to start inventing sub-levels and +/- grades to produce the illusion of progress to feed to the machine’s insatiable appetite. Beyond the obvious issues of validity and reliability, there is the common sense question about whether we should use our finite time most effectively. What do your students actually know or are able to do? What don’t your students know or are not able to do? Teach to the gap, teach students what they don’t know.
Other Blog Posts which may be of Interest:
Assessment: We are Thinking Before Acting
References
Wiliam, D (2014) Redesigning Schooling – 8: Principled Assessment Design. SSAT (The Schools Network) Ltd
Interesting article thank you. Here’s an activity I do with this type of data http://themathsmagpie.blogspot.co.uk/2013/07/collaborative-learning-activity-to-do.html We’re also luck that at work we had a teacher who was then able to take it and print out new individualised questions based on whatever a student couldn’t do. Very powerful.
Reblogged this on The Echo Chamber.
I agree entirely with the diagnosis… But I’m not completely with you on the cure. After all, if you’re going to use a GCSE paper diagnostically, might you not as well collect and analyset the resulting grades anyway?
I’d argue that if we remove the need to feed the data machine, then our use of assessment can – and should – be much more closely related to the taught curriculum. We might select relevant questions from past papers, but the time spent testing kids on things we’ve not yet taught, and on marking them to find out that they haven’t learned what we haven’t taught them is surely time much better spent on planning and teaching the curriculum?
And while you’re right about the potential for diagnosis of KS1 and KS2 tests, the vast majority of that information is already known if assessment (rather than tracking) is serving its purpose in the run up to the tests thought the Key stage.
That said, all the time schools feel compelled to provide data regularly (and schools is categories are constantly expected to do so), the point is rather moot in too many cases sadly. So your main point is key.
Thanks for this, Michael. The spreadsheet does actually have a total for each student to the right and a grade, I just cropped it when putting it into this post as it didn’t link to the point I was making. I’ve tried to read around before devising a system with the HTs and staff in the MAT so still not yet fully formed the thinking, more a series of proposals. Next assessment blog post will look at assessment of learning (do more, linked to the taught curriculum) and assessment for grading (do less, linked to national assessments) hence less determining of grades and putting them into the data machine.
I was actually re-reading your 7 principles of assessment yesterday and they’ll be making an appearance in the blog post, much prefer them to the DfE ones. Hope this makes some sense.
I think that’s exactly it – doing more of the right things odd exactly what’s needed. Glad the principles are useful. I could probably do with revisiting them myself. Never does any harm.
They are rather good 😉
Reblogged this on rwaringatl.