One of the most striking educational trends of the past few decades has been the rise of “teaching to the test”. In both the UK and the US, high-stakes accountability systems have led schools to focus instruction ever more narrowly on what the tests cover. Assessment has become the driver of the curriculum.

Of course, this trend has always had its critics, but in the last year or so I think we can see the tide turning against teaching to the test. Most notably, Ofsted has intervened. In a series of speeches and publications, Amanda Spielman has criticised the culture of “grade stickers” and made it clear that Ofsted inspections are there to look at the school as a whole, not just at the exam results.

But that’s not all. The nature of the argument against teaching to the test has shifted in a subtle but crucial way. Traditionally, people would argue that teaching to the test was bad because it didn’t measure all the non-academic aspects of a school. This argument implicitly accepted that teaching to the test did succeed in measuring and improving academic outcomes. However, it’s now being recognised that this is not the case. Many of the methods used for teaching to the test may improve exam performance in the short term, but they don’t actually lead to improved academic performance in the long-term.

Teaching to the test is ‘a problem’

For example, we know that you can improve performance on reading tests in the short term by reducing the amount of time spent doing subjects like history and geography, and increasing the time spent doing past reading papers and practising abstract reading skills. But we also know that background knowledge and vocabulary are vital aspects of becoming a better reader in the long-term and that such background knowledge is often developed in other subjects.

Similarly, a common tactic at GCSE is to focus instruction very precisely on the types of question structures and content in the exam. I used to see this a lot when I taught English literature. Students were able to get good grades at GCSE by writing coursework essays on very short stories by Charles Dickens or Thomas Hardy. But when they started A level, they would be expected to read several much longer 19th-century novels independently and they’d flounder. Designing the GCSE course in the narrowest way possible consistent with the syllabus might have boosted their GCSE grade in the short term, but it hampered their chances of longer-term success in that subject.

Teaching to the test is a problem not just because it can lead to the neglect of the non-academic aims of education, but because it often compromises the academic aims too. There are signs that Ofsted is recognising this, and that it wants inspection to focus not just on the results a school achieves, but how they get them.

The challenge for schools is to make sure that the curriculum determines assessment, not the other way round. That won’t be easy, but for the first time in a long time, the accountability system seems to be encouraging such an approach, rather than obstructing it.

Daisy Christodoulou is director of education at No More Marking and the author of Making Good Progress? and Seven Myths About Education. She tweets @daisychristo