… this is interesting rather than useful, but I’ve been getting lots of emails about assessment (I’ve just got below 150 for the first time since we set up JustMaths!!) for the new 9-1 GCSE and all I can do is tell you what we’re doing in our day jobs.

Oh I’d better warn you that you may also find it a little frustrating.

Let me start by saying that we have no intention of using the specimens or the samples that are “out there” at the minute with our current year 10’s or 9’s. By their very nature they are “exemplars” and the fact that we cannot allocate grades means that it’s a “non-starter” … I know some of you have been able to fend off your SLT and said that the best you can do is allocate percentages and I have to applaud you for that. Well done! We are however making sure our students get lots of exposure of these type of questions in our teaching … the exam questions by topic that I’ve collated have been brilliant as homework with my year 10’s!

So what are we doing? We’re using normal past papers (Edexcel if you must know!) but we are adjusting the grade boundaries slightly upwards and also adapting them so that they don’t include defunct topics (such as questionnaires) and changing say, a simple Pythagoras problem to one that is more complex from the SAM’s – the fact is that if we feck around with the papers too much we lose the robustness from them having been sat by 100’s of thousands of students in the live sitting. The fact that the papers are available as word versions on the emporium means that this was a piece of cake. This isn’t ideal because once you start messing with the make-up of assessment objectives (and remember the AOs have changed for the new GCSE too!) you start changing the basis on which the grades were awarded.

Ok … so how are we converting the grades to new numbers?? Well we have a little “look-up” table that our data guru put together based on the fact that we know a C is pegged to a 4 and two-thirds the way up a C is the new 5 in 2017. The schools policy is to grade students between grade boundaries as either a .0, .2, .5 or a .8 so for example it could be a C.8 or B.2 etc and so we’re reporting these grades alongside new estimated number grades that also have these sub-grades.

The intention is that we review what we do after year 11 leave – we do know that for the end of the year tests we are going to do probably use another past paper but split it into 3. At the minute we’re thinking, we’ll add some more questions in to raise the marks to 240 and then use the grade boundaries as percentages (if that makes sense) … still very much a work in progress. I suppose we’re also hoping that the exam boards will be able to provide more information with regards where specific boundaries are likely to fall and then we may consider using one of the mocks/practice papers that I’m aware some of the boards are producing – but we keep coming back to grade boundaries. I think you’d be stupid if you were expecting them to say that “a 5 is going to be 62%” and give an exact mark but I’d be happy with knowing say, to get a 5 on the Foundation you need to be aiming for between 70 and 85% (all of these are examples by the way … DON’T QUOTE ME!). Some schools are in situations where if they pitch this wrong could have serious implications in terms of floor standards etc.

I do know that at least two of the boards have done trials of their sample materials because we took part in the trial with Edexcel. We made a decision to only do it with our set 1 students and now I wish we’d done it with all students. They have given me permission to share our results against which I have added the students last test percentage (just before Christmas was the closest test to the trial – about 6 weeks between the 2) and put together a very quick and dirty scatter graph – like I say these are undoubtably the students that we’d enter for the Higher tier what I’d have liked to have seen is what happens further down the year group. Oh hindsight is a wonderful thing but we didn’t want to take any teaching time up doing this.

In terms of drawing conclusions from this, the sample is very small and it was a high attaining group but in my opinion, there is some relationship between their percentages on the trial papers and the percentages on the current specifications – it was lower (hence supporting our decision to raise the boundaries with years 9 and 10) … remember I know what sub-grades these students got on their Christmas mocks so can at least see whether saying X% would be a 5 etc is sensible and know that we are least in the same ball-park. We also know based on experience that the last 5 months can make a massive difference to the final outcomes. 

As I said at the start its a little frustating that we didn’t do the whole cohort, especially given that all of them apart from about 6 students will do the Higher tier this Summer. As to whether we’re any closer to looking at making tiering decisions … nope but there is no pressure.

In the meantime we’re just teaching them “stuff”.

PS: I know the scale on that graph is awful but I can’t be ar5ed to find the original data set – I have some packing to do (shhhh … don’t tell anyone but I’m away for a few days soon as I “may” have a birthday coming up)