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

Inaccurate KS2 data resulting in unrealistic MFL target grades???

Discussion in 'Modern foreign languages' started by manumission, Sep 13, 2019.

  1. manumission

    manumission New commenter

    Hi All
    I have recently met with the SLT to discuss this year's MFL GCSE results, which were not great. As in previous years, I have noted that students from primary school are coming in with extremely high levels and believe that the KS2 results are inaccurate. The KS2 data leaves students with extremely high target grades in MFL which cannot be met. So many of my students had a target grade of 6+, when realistically their target grade should have been 4.
    Have any other MFL departments experienced this or share my view on the KS2 data? Your thoughts would be gratefully received. Thanks in advance.
  2. cornflake

    cornflake Senior commenter

    KS2 tests are marked externally - therefore you need to look at what happens in your school rather than look to blame primary schools.
  3. Jonntyboy

    Jonntyboy Lead commenter

    I share your view, and so have my colleagues in at least two schools. I have had some minor experience in primaries over the past six years, and there are certainly ways in which children are coached to get the very best possible marks. You can't really blame the schools for doing that, as long as it's done honestly, because of course they all want to get the best OFSTED to get more pupils to get more money to get more resources ...etc. etc. - you know all that.

    But it undeniably can give a false idea of the potential abilities of some Y7s. I found in the past at one school that out of half a dozen feeder primaries, kids from a couple of them would show unrealistic scores every year. Longstanding colleagues used to say it had been thus for some time.

    I'm at a new school now, and from what I can gather we reassess before Christmas and where necessary will re-set the kids for the Spring term. I know that in some schools timetable restrictions may prevent that in some cases. Of course, that won't help much with target grades, but it will help the kids learn better at an appropriate pace for their individual ability.

    All that said, it is ludicrous that students' MFL targets should be based on English and Maths. Completely different skill sets.
    agathamorse likes this.
  4. manumission

    manumission New commenter

  5. manumission

    manumission New commenter

    You are absolutely in denial if you think teachers don't assist year 6s in their SATs. The whole system needs to be looked at so that secondary schools have accurate data. I've been a head of year for several years and can confirm that to get the results, some primary schools will assist when they certainly shouldn't.
    janjomarie likes this.
  6. manumission

    manumission New commenter

    Thanks for the reply very much appreciated.

  7. sbkrobson

    sbkrobson Star commenter

    Well there is the issue.
    Why is KS2 data used to predict something so unrelated?
    And why would you say that they are "left with" a target grade?
    Your management could take a little insight from you- find out if instead of being left with predicted grades, you could create your own predicted grades, based on a subject related assessment when they enter the school. say, at Christmas in Y7. Your management will either see sense when you patiently explain how MFL exams work, or they will have zero understanding of how languages work, in which case your mission is futile. I've seen both scenarios. Do they realise that using three tenses is not a literacy or maths skill, it is a skill of practice and repetition? Do they know the role of listening, and how, across the board, is likely to lower their final score? Do they know that 25% ie speaking is not related to any literacy skills, but to recall and imitation? Do they know that the content may not be accessible to them culturally? Do they know that EAL students with low predictors may well over achieve? Spell this out to them. Explain how a more realistic predictor needs to be generated.
    Apart from this, something does not add up in what you are asking-you query Y2 predictors against GCSE results?
    Why are you querying it now? The kids who did the GCSEs have held those predictors for five years already. Surely it would have been apparent at a far earlier stage already that those predictors were meaningless?
    Jonntyboy and Stiltskin like this.
  8. Cyberman

    Cyberman New commenter

    I am not a languages specialist (my background is maths and computing) but I thought I would add my “two penneth” to your discussion.

    1. Accuracy of KS2 SATs test data

    I have some limited experience with the accuracy of KS2 SAT data, particularly in maths. I have found that in general the KS2 maths data is accurate, I would estimate over 95% of results are “correct”, it is the other <5% that cause most problems for secondary schools. Attached is part of a presentation explaining how I arrived at this conclusion.

    2. Using KS2 SAT data for predicting KS4 outcomes

    Although not completely true, KS2 SAT scores can be thought of, and used as, an indirect measure of cognitive ability.

    This is why they are quite a good STATISTICAL predictor of EXPECTED KS4 performance for a COHORT in a GCSE subject, provided you use the APPROPRIATE KS2 level to KS4 grade conversion for the subject. The larger the cohort, the better it is as a predictor.

    You must use the correct KS2 SAT score to KS4 GCSE point scores for a subject, they are different for each. MFL has a lower KS2-4 translation than most other subjects. For example, in 2018, nationally on average learners with a fine level of 5.5 (NC 5b) achieved an average point score of around 6.5 (grade 6/7) in GCSE maths but only 5.5 (grade 5/6) in languages.

    Provided you use the correct KS2-KS4 “translation curve” the KS2 data is a good STARTING POINT for setting subject targets for individual learners and then summing these to get the EXPECTED performance for the COHORT OVERALL.

    Some of these initial individual learner targets will need to be adjusted by the learners’ teacher based on their knowledge of / performance of the learners during KS3. You would expect some learners’ target grades to be lowered and some to be raised but overall the cohort target should remain ABOUT the same.

    This does mean that if you have several classes that are “setted” then lower sets are likely to have their target grades lowered and higher set learners are likely to have their target grades raised but the overall targets for the whole cohort should remain the same.

    If you look at the free resources at the link below you will find a fuller explanation and spreadsheet tools to allow you to set targets in your subject.


    Attached Files:

    Jonntyboy and agathamorse like this.
  9. sbkrobson

    sbkrobson Star commenter

    Wow.That's a brilliant precis of what OP is up against.
    What an irrelevancy in that case, a nuisance almost, spending time and giving attention to how each child performs in each MFL lesson. Who knew?!
    Might as well put my feet up a bit more.
  10. veverett

    veverett Occasional commenter

    Also don't forget that for the op's pupils to get those predicted grades, some other pupils somewhere else would have to do worse.
    agathamorse likes this.

Share This Page