Gove’s Failed Exams

This morning Simon Jenkins got me thinking about the role of data in education. Michael Gove’s plan to introduce a ‘baseline’ exam for five year olds in the UK has just been heavily criticized by a group of 120 world experts including Guy Claxton and David Whitebread. They cite the lack of any research to support the effectiveness of such testing on a child’s long-term success, and the growing evidence that in fact it does much to harm a child’s self-image at a crucial stage in life.

Gove in school

No doubt Michael Gove – or in this case Nick Clegg, who announced the reforms as part of a speech on the extension of the pupil premium – feels he is acting with the best of intentions. Centralised testing regimes are very much de rigueur in an increasingly laissez-faire social market economy, fuelled by a series of influential reports by big consultancy firms and a misunderstanding of the value of sources like the OECD’s PISA studies. Much against their de-centralising values, politicians see it as a means to retain control.

The line of argument to support testing regimes is business school 1.0. To aim for a serious improvement in any enterprise, you need to know first what the improvement is that you’re aiming for, and second whether you’re making any progress in getting there. You then gather lots of data against the resulting KPIs or MOS to see where you’re succeeding and where you’re failing, making the necessary judgements based on hard evidence and scientific certainty. So far, so simple, and who could possibly disagree?

The seam of this ideology continues to run deep throughout our schools. When I was teaching (albeit in a secondary school), I was pro-data. I found it hard to understand how I could know whether my pupils were learning if I didn’t have data from the beginning and end of the year, preferably evaluated based on GCSE grade work. But looking back I see a different story. Evidence of a white line fever, and its ill-effects, was everywhere. Kids dropped enriching subjects on the curriculum to double the time spent in English and Maths, teachers ‘cooked’ coursework to achieve class targets.

I still believe that having good data is important. But blanket testing regimes for five year olds will not produce good data, they’ll produce superficially comparable data: the kind of data that allows system leaders to wield macro incentives and design distant policies; the kind of data that obsesses and perversely incentivises the behaviour of under-resourced, highly-stressed educationalists; the kind of data that does almost nothing to improve the quality of learning for the tested child.


Data can work, but – and this is crucial – only in the hands of enlightened and well-supported professionals. A good teacher is continually gathering data on a child’s development, not according to crude once per Key Stage tests, but using all of the indicators that they can. In Finland – by some measures the best education system in the world – there are no national tests until seventeen. Instead teachers gather work samples, photos, interact with pupils and use their professional judgement to evaluate whether and where a pupil is succeeding, falling behind, or requiring some specialist intervention.

This professionalism is precisely what is undermined by the exams for five year olds that Gove is proposing. After all, why should we invest in teacher professionalism if we can simply give them a target, then beat them with it until they get there – or sack them if they don’t? Here two philosophies collide. On the one side, the belief that the market drives improvement, requiring incentives that depend on data. On the other, the conviction that we should begin reform with trust and investment in the people that do the work. As we have seen in the distortion of AFL from child learning guide to child target-setting and performance evaluation tool, the former risks undermining the latter.

And yet we know that teacher quality and professionalism is the decisive factor in child development – all of the evidence, from PISA to meta-studies by the likes of John Hattie, points this way. But as Claxton et al. point out, instead of “pursuing an enlightened approach informed by global best practice, successive ministers have prescribed an ever-earlier start to formal learning”.  It’s nonsensical to believe that negative incentives and mangerialism can bring about an increase in professional expertise. Policymakers continue to see the value of performance data, but not the cost.

At worst, such ill-conceived testing regimes reduce a child’s learning to an aggregated profit or loss business target that does actual harm.  At best, they are a well-intentioned tool badly applied. In this application, they obfuscate the true path to better schooling for our kids: investment in initial and ongoing teacher training, fostering increased creative professionalism and building a culture of research-driven practice. Good data is an important formative tool, but aggregate summative data is useful only for governments and economists. Do we want to see our young people individuals or a faceless mass of data to be crunched? On this question – as ever – we should trust the experts, not the politicians.

This entry was posted in EDUCATION, EDUCATION POLICY and tagged , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *