In amongst my posts about dead cats and serendipity I have to remember that I started this personal blog in response to a request to give some Yoda-like advice about being a successful member of a successful SLT. And I haven’t done so since my very first post so here goes with the biggie: being the Data Fiend of SLT.
First of all a little background on me, given that I’m not the typical Data Fiend on SLT. I’m an ex Head of English who failed to sit my A-Level Maths because at the end of Y12 I scored 19% in my Pure Maths internal assessment (there were no AS Levels in them dark days) and 0% in my Statistics test. Yes, you heard right. 0% for Statistics because I panicked and never ever got the concepts of quartile and standard deviation. To be completely frank I still have no real idea of what standard deviation is.
I never came late to the data party during my stint as Second in English or as Head of English either. In fact at one school I was actually sent out of a full staff meeting for having the giggles when the Head was delivering, with a completely straight face, a tirade about how badly we were doing according to the PANDA and the PICSI. It sounded like a hybrid Chinese/Irish folk story to me. Even as I aspired to Assistant Headship – a time when I tried, against the grain of my natural inclination, to keep my nose clean in most respects – I was a regular contributor to the anti-Data debate within the staffroom.
Even as a member of an SLT with responsibility for teaching and learning and student leadership I frequently mocked (gently) my colleagues on the dark side of curriculum and data: see, there is a dark side to the dark side.
And then the unthinkable happened. I completed a data task on interview for a Deputy Headship role so well that the Head offered me the post with a responsibility for data: I had become the Data Fiend and a lifetime of mockery of others had found its own sense of schadenfreude. It was time to go from being the poacher to becoming the gamekeeper.
With all that in mind, here are my top tips for being the Data Fiend (With a Heart)TM on your SLT.
Keep it simple
Data doesn’t have to be complex. It needs to be end-user oriented, and that means students. Your job is to bear the burden of the complexities and be able to answer the tough questions from colleagues and, yes, inspection teams. But your job is also to filter away those complexities without patronising staff, parents and students.
Less is definitely more
Linked to the first point. You have to be familiar with RAISEonline, FFT, Jesson and SIMs Assessment Manager and you have to keep an eye on all the groups (ethnicity, SENDA, prior attainment, blah, blah, blah) but don’t transfer that over-abundance of data to your middle leaders and class teachers. I’ve tried it and failed spectacularly. Instead know what the main data is telling you and then help middle leaders to drill down only when they have to.
Use estimates not targets
No member of SLT can set targets for students based on any benchmarking programme. Only teachers, with the full picture of a student’s performance in the classroom, can do so. This document by the Fischer Family Trust gives exemplary guidance about this. Estimates are based on chance graphs and so it follows that more students will do either better or worse than their estimate, so why on earth would we want to set it as a target? In my opinion and experience estimates often raise expectations for both teachers and students (especially those with low prior attainment) but they can also be limiting if seen as a hurdle to be cleared. Where estimates have the greatest impact is in terms of whole cohorts and particularly whole cohorts over time: if a subject or school is consistently exceeding estimates for a cohort then there’s an almost statistical certainty that they’re doing something right.
Use cognitive abilities for estimates, not prior attainment
I initially used CATs, YELLIS and ALIS for estimates out of necessity as 30% of the cohort in my highly mobile school had no KS2 data. Now I wouldn’t have it any other way. Basing estimates on cognitive abilities is fairer to students who had a bad experience at primary school, students who found exams at 11 challenging and students who have low literacy (EAL or otherwise) but strong numeracy and non-verbal skills. Our last cohort of Y11s made spectacular ‘expected progress’ and ‘better than expected progress’ without having ever had estimates set on the basis of their KS2 results. A final reason for a switch to cognitive abilities estimates is back to the pragmatic: KS2 tests are being increasingly boycotted and the new curriculum for primary is level-free beyond those tests.
Worry about individuals and let groups take care of themselves
I was massively influenced on this one by a Data Fiend I once worked with as a middle leader and Assistant Head. His premise was simple: why focus on groups of students when, if you raise the achievement and progress of every student individually, you will tackle every potentially problematic group. And what is more in the process you will avoid being reactive and creating initiative overload. This is a difficult one in the current climate of narrowing the gap and one you will have to be prepared to defend yourself on, but I still wholeheartedly believe that it is the right approach. It will help if you have evidence of impact.
Quantitative Data is not infallible
At times many teachers, and SLT most of all, tend to see numerical and alphabetical data as being entirely objective and become unbending in their approach to analysis and evaluation of it. The truth is that it does have greater (although far from complete) reliability than the qualitative information we pick up from real interactions with teachers and students, but that it has much less validity. Taking information about attendance, behaviour, effort, aptitude, family circumstances and other factors into account when seeing achievement in the round is not necessarily about making excuses or having low expectations. It can be if teachers and leaders get the balance wrong, but you always have to weigh the two appropriately and individually for each child, and where staff have previously got that balance right then you have to exercise trust in their judgement. Where they have got it wrong previously then you have to provide support, not simply bash them with statistical certainties that are far from certain.
Data should serve pedagogy
The presence of data of itself does not improve learning. In the short term it can help a school that has been coasting to raise the achievement of individual students and, in doing so, raise its performance as an institution. The problem with such an approach is that data-driven improvement, particularly when it is linked to curriculum change only, will always reach its plateau. If schools are to push forward and improve the learning of students, then they can only do so by helping teachers to improve their practice. Data has a role to play in this but not if it is used to inculcate a climate of fear across the school. Instead data should serve pedagogy by allowing more effective transition in learning between key stages, and informing more effective differentiation within learning activities.
Data only ever asks questions
I cannot underestimate how many times you need to keep telling colleagues this if you are the Data Fiend (With a Heart)TM, especially those on the Senior Leadership Team. Data is a measure of outcomes (if we’re going to get all Ofstedy about it) and so leads to questions about the quality, efficacy and impact of provision. We should always be asking questions about these things anyway in our line management processes and our job is to inform these conversations, not dictate them. Middle Leaders, with good coaching questions, can quite often find themselves holding a final piece to the jigsaw that completely undermines any hypotheses that SLT have about achievement in their subject.
I don’t claim to have got everything (or anything) right in my time as a Data Fiend (With a Heart)TM but my school has seen results improve year on year without becoming an exams factory. My final piece of advice is to those promoted to the position of Data Fiend from a Maths or Science background. Make data sing to your Music teachers. Make data tell a story to your English teachers. Make it reveal humanity to your History teachers. Do that and you will do so much for data use in your school.