Tom Sherrington

Nicky Morgan vs The Bell Curve.

Screen shot 2015-06-30 at 22.22.42
Dear Nicky, let me introduce you to the bell curve.

Dear Nicky…

I’ve just read this: https://www.gov.uk/government/news/hundreds-of-coasting-schools-to-be-transformed

In there it says this:

Schools eligible for intervention will be those which fall below a new ‘coasting’ level for 3 years.  In 2014 and 2015 that level will be set at 60% of pupils achieving 5 good GCSEs or an above-average proportion of pupils making acceptable progress.

I am now worried that you haven’t been briefed about the word ‘average’ or the new (laudable) determination by OfQual to ensure GCSE grade inflation is halted. The thing is this: by definition there are only a limited number of places on the bell-curve that can be called ‘Good GCSEs’.  You’ve decided to give a pejorative label (implicitly ‘Bad GCSEs’) to about 50% of all grades.  Now, instead of Grades 1-4 at GCSE representing any sort of achievement, they’ve been killed stone dead. Nice work. That didn’t take you long.  I have many students who by definition are unlikely to get ‘Good GCSEs’ given their starting point.  Don’t get me wrong; we’re slogging our guts out here, but the thing is this: by definition, it’s a zero sum game.  More or less. We can only get more Grade 5s and above if, on average, other students somewhere get fewer.  This is how things work now.  You may know about the reference tests being introduced to track progress between cohorts but realistically we’d only expect a slow incremental change year on year.  That’s how standards are maintained.  For this to be fair, it is imperative that grades 1-4 actually matter. However, rather than accepting this (understanding it?) you’ve decided that only the top grades count.  You appear not to appreciate – or care- that, by definition, not everyone can get them.  Do you want Glenys Stacey’s number? I suggest you give her a call.  This isn’t some kind of ‘enemy of promise’ excuse-making thing; it’s a hard-wired technicality derived from grade-setting across the examination system. You get that, right?

Once you aggregate up to whole-school data, it is then obvious that not every school can have ‘an above-average proportion of students making acceptable progress’ AND/OR above 60% of students with 5 good GCSEs. Why? Because, by definition of ‘average’ and the laws of the bell-curve, that isn’t statistically likely, especially given the link between progress and attainment.  It’s the Matthew effect – check the stats.  The ‘intervention’ concept implies that by the voodoo of academisation you will eliminate schools that are both below average for attainment and not above average for progress.  Let me explain…..?  As long as we’re holding standards to the bell curve to break the inflationary cycle, you’re on a hiding to nothing.  Actually, that’s WE are on a hiding to nothing. You’ll be fine! You’ll ‘transform’ some schools…(regression to the mean will see to that) .. but only if other schools take their place.  Do the Maths.

I know that Tough Talk is all the rage down at the DfE, but you are the Secretary of State, not the editor of the Daily Mail.  You do realise that we’re reasonably intelligent folk and we expect certain standards.  Do you really want some of my students to be told that GCSE grades that might have mattered to them are ‘Bad GCSEs’?   Do you realise how ludicrous it is to batter schools for not meeting certain standards when not everyone can reach them….by definition?  Do you ?

Seriously…. This isn’t good enough.

Yours.

58 comments

  3. Wow!! Super response. Let’s hope the people at the DfE and Nicky Morgan read this. I was shocked yesterday when I saw her being interviewed for the BBC and no one picked her up on what is fairly straightforward maths. Not really surprising, though, as the media hardly ever seem to challenge the reliability of what the government says about educational issues eg that School Direct has been a great success.

    Liked by 1 person

    Reply

    • She won’t read it. She’ll get one of her advisers to do it, who will just trot out the usual, scripted response:

      “With regard to GCSEs, the government wants all pupils to be studying subjects which give them the best start in life and give them a strong foundation for progress to further study, for work and to help keep their options open.”

      Liked by 1 person

      Reply

  4. Morgan has posited a gross outcome of all schools getting 60% A-Cs. It can be done by simply altering where the examiners put the line for the grades. Alternatively, it’s impossible but sounded good on the telly.

    Like

    Reply

  7. The statistical model they use is the “loose pack container”
    A standard articulated wagon container is divided along its length into 10 equal sections.
    The tenth nearest the cab = the top 10%, the tenth at the tail of the wagon = the bottom 10%.
    Crash test dummies or spare / former / resigned / retired Secretaries of State for Education are distributed randomly throughout the interior of the container.
    The articulated wagon is then driven at high speed by the current secretary of state for education.
    To ensure the objectivity of the methodology the current Secretary of State must not have had any experience of driving Heavy Goods Vehicles.
    When a speed of 120km per hour has been reached the Secretary of State performs and emergency stop, alternatively (preferably) she drives into a large immovable object.
    Once the dust has settled the position of the crash test dummies / former secretaries of state is examined in relation to the pre-determined sections in the interior of the container.
    As the average is clearly in the middle of the container any crash test dummies in front of this are above average.

    This methodology is highly reliable and has been tested at a variable range of speeds and variable distances, the results are largely similar. It has been found that the higher the speed and the more reckless the driving the more positive the results. A further factor that appears to improve these outcomes occurred when the current Secretary of State for Education was being interviewed by CH4 News at the same time as driving. However this may be down to the Goggle-eyed Morgan effect (the ability to stare blankly into the middle distance whilst taking verbal instructions via a dedicated Govian earpiece.

    Liked by 1 person

    Reply

  8. Depends what the point of the Policy actually is. Is it really designed to raise standards, or to provide extra impetus to ‘academise’ remaining schools?

    Like

    Reply

  9. Behind her buffoonery is a well thought out privatisation agenda. Winners? A few. Casualties many. There is no recovery from the shambles this lot have created.

    Like

    Reply

  10. How can our Headteacher associations continue to sit on their hands when we are now being asked to implement a policy which is actually impossible to deliver? They are so laid back in their responses to ill thought out policies that one wonders if they are worried about not getting their future knighthoods. It is not interfering with politics when a policy is just plain impossible to implement, and trying to do so will lead to yet more ill health amongst their members. NAHT and ASCL please get some backbone. What would it take for you to make a noise that can be heard?

    Like

    Reply

  11. Tell me if I misunderstand. The assumption is that Intervening in the schools where outcomes are at the lower or middle end should improve the results in those schools. To prevent against grade inflation (i.e. keeping a bell curve shape to the grades) will then mean that other schools who are not also improving will drop into that zone. So they would then enter into the ‘needing improvement’ zone. And so on…

    The end outcome should then be that the whole bell curve is shifted to the right, i.e. over time the standard needed to get a certain grade is increased and that is a good thing that the educational establishment must aim for?

    Like

    Reply

    • But even if the entire bell curve shifted to the right, it still wouldn’t be possible for all schools to get more than 60% “good grades”. It doesn’t matter where the curve is, there will always be a fixed number of grades that are below the benchmark.

      Like

      Reply

    • This can only happen if the exam boards and/or the Sec of State collectively allow it to happen. The Tories said that Labour had ‘inflated’ the grades so they arranged it that the bell curve should move to the left. This was nothing to do with what students actually did or did not do, it was a political move. Your scenario suggests that exam boards and ministers respond to what students actually do or not do i.e. they criterion reference the results. Do you have an evidence that this will happen with those presently in charge?

      Like

      Reply

    • They just love that bell curve. It’s been applied to civil servants’ performance for several years now, causing untold misery…

      Like

      Reply

  15. Tom your best blog ever in my humble opinion . The 60% is a disgrace . The national average by the way for 5*-C ECM is around 57% I believe ! The best bit of course is the huge number of academies who will not meet the 60% !

    I used to think t hat it was just the new Lord Chancellor who used to make policy up on the back of a beer mat !

    Like

    Reply

  18. I can’t back up this assertion – a quick Google hasn’t done the trick – but I’m fairly sure that Ofqual has a maximum allowable change to the A*-C rate between cohorts. I think this was reduced a few years ago from something like +/- 3% to +/- 1%. (Possibly this is for A-Level rather than GCSE). If I’m correct then there is a specific regulation to ensure that a certain proportion of pupils get ‘bad GCSEs’ whatever teachers and schools do.

    I suppose one could make the esoteric argument that if schools below the floor target acquired some students from a school well-above the floor target, both schools could be over 60% and making above-average progress. Perhaps multi-academy chains could do some clever trading of Y11s.

    However, that’s all academic, really. I agree wholeheartedly with the level of frustration expressed in this post. The DfE have been playing this ‘must be above average’ card for a long time now and like an earlier response, it’s enormously disappointing that a decent journalist has not properly dragged a SoS over the coals for it. Just last week Nicky Morgan was saying how important it was that a positive picture was painted of schools, and here is another punch in the mouth.

    How many teachers or SLT do you know who are ‘coasting’?

    Like

    Reply

  19. Full agree. Have been thinking this since they removed ‘satisfactory’ from the Ofsted criteria. Every child should attend a good or better school is an excellent idea in principle but how can a school be good (ie better than average) if there is no average?

    Like

    Reply

  30. […] standardised-testing (please read headteacher Tom Sherrington’s excoriating Maths 101 response). Laura McInerney provided a much more effective definition, and graphically laid out – as an […]

    Like

    Reply

  32. Great piece of writing and as they say, a picture paints a thousand words. You use the phrase ‘You get it, right?’ but regrettably I don’t think they do. They are hollow words that on the surface (and to those who like a soundbite) seem to make sense but as soon as you look into what they actually mean, they are effectively meaningless and unattainable.

    Like

    Reply

  36. Education needs (a bit of) grade inflation like the economy needs (a degree of) inflation. How can you demonstrate the efficacy of your great political project without some (easily quantifiable) measure of system improvement? Rigour may be the new black for now but in time the Minister will need to show that she/he has made a difference.

    Like

    Reply

  39. I find the phrase “Good GCSE” is extremely demotivating for my students, all of whom have SEN. The government tells them despite trying their best and forcing themselves to sit through multiple stressful exams that they haven’t got good grades. My students worked their socks off to get their F and E grades and already think they are rubbish. This government rhetoric doesn’t help at all…

    Liked by 2 people

    Reply

  46. […] So there we are.  Tough talk pays!  We’ve been getting it for years.  The prevailing school improvement paradigm has been that poor performance is the fault of feckless leaders and teachers who just need a rocket up them to get their act together.   Ed Balls was full of it introducing statistically dubious floor targets and talking tough about them; Michael Gove made it an art form on many fronts- a personal obsession to slag teachers off as often as possible; Nicky Morgan took in on too when introducing the totally dubious concept of coasting schools (see Nicky Morgan vs The Bell Curve.). […]

    Like

    Reply

  47. […] This sounds an awful lot like 93% of people thinking they’re good at multi-tasking or wanting 60% of schools to be above average The very nature of the term elite conveys an exclusive group. Second, the term ‘regional’ […]

    Liked by 1 person

    Reply

  48. […] Since then however, I have thought again. My problem is not to do with competition, but compulsion. In a compulsory system of norm (or cohort)-referenced testing, no matter how hard everyone works or how well they perform, a certain proportion of them are destined for failure before they’ve even started the course. As Tom Sherrington wrote in his amusing, if genuinely angry blog post Nicky Morgan vs the Bell Curve: […]

    Liked by 1 person

    Reply

Leave a comment