1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. Hi Guest, welcome to the TES Community!

    Connect with like-minded education professionals and have your say on the issues that matter to you.

    Don't forget to look at the how to guide.

    Dismiss Notice

How do you convert KS2 Maths SAT results to new Predicted GCSE Grades?

Discussion in 'Mathematics' started by Ezioclone, Nov 2, 2018.

  1. Ezioclone

    Ezioclone New commenter

    The new (post 2016) 'standardised scores' in KS2 SATS would seem to me to be a good/appropriate indicator of a student's potential at GCSE.

    So is anyone yet making GCSE predictions/target grades based on the pupil's KS2 standardised score?


    - should we expect a student with a standardised score of 100 to achieve Grade 5 at GCSE?

    - what GCSE grade might you expect from someone who had previously attained a KS2 standardised score of 110 (working at the min. so-called 'greater depth' level)?

    - any idea what GCSE grade a student working at 110 might attain in the new GCSE if they were to sit the Foundation paper GCSE right now (i.e. whilst 11 years old)? There'd be no reason to do this, but I'm just curious.

    Since the new KS2 SATS are so much more thorough (and problem-solving in nature) than the previous 'levels', I'd expect them to be a much more appropriate predictor of GCSE performance than before. [Unlike with the previous 'levels', there's no way at student can now attain a standardised score of 100+ without fairly solid, properly embedded arithmetic skills].

  2. adamcreen

    adamcreen Occasional commenter

    Here we go ... The new 100 "standard" scaled score is around an old Level 4a.
    Year 6 4a, Year 9 6a, Year 11 Grade B ... so new Grade 6
    Anyone on 110 is "high standard" around an old 5b
    Year 6 5b, Year 9 7b, Year 11 Grade A ... so new Grade 7/8
    Anyone on 90 is below the "standard" so around an old 3a
    Year 6 3a, Year 9 5a, Year 11 Grade C ... so new Grade 4

    Unsurprisingly I am not in charge of GCSE predictions at my school, but as a HOD with 18 years experience, this is where I would put my own personal estimates. MMT what you say?
    strawbs likes this.
  3. strawbs

    strawbs Established commenter

    Pretty similar to Adam, I would go 100-109 target 6; 110-120 target 7/8/9
    Below 100, I'd have say 95-99 target 5; 90-94 target 4; below 90 target 1/2/3
    cach9801 likes this.
  4. pi r squared

    pi r squared Occasional commenter

    We are essentially converting new scaled scores into old KS2 levels, then using previously received wisdom to make a prediction in old GCSE grades (or apply "three levels of progress") to then convert old GCSE grades to new ones. With all the variability, statistical noise, and caveats attached to each stage of the conversion, what we are doing is barely any more statistically-robust than this:

    We know that the mean score for KS2 maths was 104 and that the mean GCSE grade for maths was pretty much on the 4/5 threshold. Unless there's a massive uptick in GCSE maths grades over the next five years it seems unlikely that your "average" 100 kid is going to be walking away with a Grade 6 - more likely a high 4 or a 5 if the wind is blowing in the right direction for them.
    webby237 and cach9801 like this.
  5. adamcreen

    adamcreen Occasional commenter

    Four levels of progress for me, but we start with generally high-achieving students and set targets accordingly. Remember these are targets not predictions (for me at least).
  6. pi r squared

    pi r squared Occasional commenter

    I wasn't singling out any one particular post or user, but this goes on to show that 3LoP is a useless measure too because for top-end you need at least 4LoP. For all of its faults, Progress 8 certainly does a better job in that regard.

    So we have new GCSE grade targets/predictions that are inexactly converted from old GCSE grades, that are based on projections of 3LoP/4LoP/P8 - each with their own inaccuracies - which in turn are based on old KS2 levels, that are then inexactly converted into new KS2 scaled scores. The process lacks so much rigour it is unbelievable, and yet schools are doing it up and down the country, either directly or through proxies such as FFT (who at least have national data to work with).

    Don't get me wrong, I understand the need to target set and I understand the need for teaching staff to have an overall idea of the potential of the group they are teaching. (Individual targets to inform a group profile are fine; individual targets as predictions for individual students is pointless). But schools are spending hours, days, and £££ on this, thinking they have ended up with a method that is more statistically-grounded or more reliable than next, yet it is just snake oil. Why don't we just say: 3% of our students got Grade 9 last year, so the target for the top 3% of current Y9s will be Grade 9. 10% got 8s so the next 10% will have targets of 8, and so on. Results need to improve? Adjust upwards accordingly. Current Y9 cohort more able than last year's Y11s? Adjust upwards accordingly. It's an inexact science but it doesn't pretend to be otherwise, and it can be done in five minutes of anyone's time.
  7. Ezioclone

    Ezioclone New commenter

    Thanks everyone.
    My view was close to Adam's.
    Perhaps I'm even more optimistic at the upper end (I think 100+ SAT scores 'should' translate to clear, but low, Grade 6's (as a modal value)).
    But I'm perhaps more pessimistic at the lower end (I'm not sure most 90 SAT scores will routinely translate to Grade 4s at GCSE). You can easily get a 90 on SATs with no conception of (even informal) algebra, or pre-algebra. There may well be some '90 SAT' score students who simply can't get to grips with even basic algebra - and that makes the Grade 4 rather difficult?

    Although I also agree with Adam that the new 100 SAT ~ old Level 4a, IMO it's much more meaningful than the old 4a ever was because it guarantees good basic arithmetic skills (including FDPR, and some problem solving ability). It used to be quite common to have students who scored (say) Level 4b, and yet had all sorts of serious mis-conceptions and ignorance about basic arithmetic at lower levels. I used to have 4b and even 4a students who quickly became quite 'Maths phobic' when they subconsciously realised they had severe gaps in their lower level knowledge, and were always trying to mask and cover-up their weaknesses, rather than being prepared to work on and correct them. These serious weaknesses were always a massive impediment to future progress, and yet are nowadays, at least for those achieving 100+ SAT scores, almost eliminated.
    For those reasons, I disagree with PiRSquared that we should only be expecting Grade 4/5's from 100+ SAT students. I just wonder if PiRSquared (and perhaps many many other Secondary teachers) have really had a chance to examine just how much KS2 SATs have changed over the last 3 years? (the teaching and assessment aspects especially).

    [My posts here 5 years ago were very critical of the way the KS2 SATS could easily be 'gamed' by Primary teachers, and the serious consequences this had on students' ultimate attainment (at GCSE).
    TBH, I'm quite astonished at just how well the KS2 SAT reforms have worked (in terms of providing a much more solid foundation for future learning).
    Perhaps 'Mastery' (and particularly Bar Modelling?) can take some credit for the improvements at KS2 too, but for me it's the syllabus and format of the KS2 SAT assessments that have really driven the improvements].

    I think the next few years should be good times for Secondary Maths because these younger cohorts of students (the scaled SAT score students) are so much better prepared, and because there'll be fewer 'wishful thinking' Grade B-type students demanding to start A-Level because there'll have seen just how hard A-Level Maths becomes (through much greater exposure to Grade 7, 8 and 9 work at GCSE)!

  8. pi r squared

    pi r squared Occasional commenter

    I work in an all-through school and worked closely with our Primary phase when the new assessments came in. We also look at the actual papers of the students entering in Y7 and use these to identify students who show early signs of underperformance, so yes I have had chance to examine the changes.

    The fact still remains that 70% of current Y9s nationally scored 100+ in Maths at KS2. You seem to be suggesting that those 70% of Y9s will go on to achieve Grade 6 or better in Summer 2021 - more than double the proportion of 16-year-olds who achieved a Grade 6 or better in 2018? I cannot see it happening. I think the new KS2 SATs do a better job of preparing our youngsters for the GCSE, for sure, and we might see a small uptick in the proportion of Grade 4 upwards being awarded, but not so much that the majority of the cohort score Grade 6 or better.
  9. Ezioclone

    Ezioclone New commenter

    I didn’t quite mean to imply that.
    The Year 9’s you currently have were the first year to have sat to new style KS2 SATS, had studied for only one or two years under the new Nat Curr, and were with teachers teaching those SATs for the very first time.
    Now, two or three years later, the students we have coming through into have experienced several years of the new Primary curriculum, Year 5 and 6 teachers have learnt how to teach it better (esp. tackling prob. solving Qs), and the result, IMO, is students with a much stronger foundation in all the core arithmetic skills.
    From what I see (and am able to teach in Years 6 and 7), the ‘middle’ students coming through now really are significantly stronger than even two years ago, and this is not because they are simply a brighter cohort. It’s because they have been taught maths better (in some sense). Their English attainment really isn’t any different from before (just some change of emphasis between comprehension, grammar, spelling and writing skills). It’s Maths specifically where there has been a very marked improvement - and where the teaching now is such that ‘gaps being filled’ is rightly prioritised over prematurely ‘pushing the kids on to new mathematical content’.

    So it’s the new students entering Year 7 from now on that I expect to be able to achieve at the level I suggested.
    The current Year 9’s won’t have had the full benefits of the changes, neither will the Secondaries have had sufficient time to react to the improved Year 7’s to fully exploit the improvements.

    I guess I am quite concerned that some Secondaries might not be capitalising upon these improvements in Primary (for example, by modifying Year 7 and subsequent SoWs and in having sufficiently high expectations).
    For example, I know of some Secondaries (one where I taught last year) following the commercial ‘Mathematics Mastery’ programme, whose YeR 7 SoW seems to have been designed years ago before the improvements in Primary took place. There’s a whole load of students unnecessarily now going through this system when they could be moved on much quicker.

    Best regards, MMT
  10. bigkid

    bigkid New commenter

    It seems unlikely that there will be a significant increase in grades nationally. If the new primary curriculum and KS2 sets mean everyone has been taught better then everyone is better prepared. Thus most students will do better. This will probably result in an increase in grade boundaries and a slight increase in the number of pupils getting a 6+

    I don't think the system is set up to allow massive shifts in the proportion of students getting each grade in a short space of time
    strawbs likes this.
  11. Lucy2711

    Lucy2711 Occasional commenter

  12. SparkMaths

    SparkMaths Occasional commenter

    Don't the exam boards have full control over the grade boundries and set them after the results are in?

    If the number of students who get a Grade 6 doubles, won't they just assume that the paper was too easy and lower the grade boundries until the usual amount get Grade 6?
  13. adamcreen

    adamcreen Occasional commenter

    That's what the National Reference Test is for. And no, they don't have overall control, Ofqual oversee it and have a very good blog where they do actually explain everything, for example how they did the grade boundaries for the first year of 1-9 or the new A levels where everything changed.
    Lucy2711 likes this.

Share This Page