The challenge of how to improve England¡¯s performance in maths education is high on the agendas of both major political parties.

The prime minister is committed to in which all young people will study maths to the age of 18. The Labour Party promises to introduce a akin to the successful introduction of phonics to teach early reading. And both parties talk about the need to address the substantial shortage of maths teachers.

Taken narrowly, none of this is wrong. Setting strong trajectories for learning maths in the early years is critically important, and England does have lower take-up of advanced maths courses than many competitor countries. And the prize for both individuals and the country is huge: a maths A Level increases your earnings by 10 per cent even compared with people in the same profession; a maths-intensive degree increases salaries further, by 30 per cent, and mathematical skills are key to employability, good health choices, financial decision making and so much more.

The problem is that a series of interventions focused on just one aspect of the maths curriculum or on only two year-groups risks overlooking the bigger picture. Yet it isn¡¯t just politicians who lack that breadth of vision. The same can apply to the academics on whose research their policies are often based.

ÖйúAƬ

As individual researchers in small teams ¨C the typical scenario ¨C we can provide good evidence to schools, teachers and politicians on how best to teach , or what type of exams best assess children¡¯s mathematical competency and at what ages, or what skills are likely to make someone a good maths teacher.

But such research ¨C often focused on specific classroom practices, interventions or trials (randomised or otherwise) ¨C risks being atomistic and disjointed. It also tends to avoid the more complex, large-scale questions that vex politicians and system leaders. Where, in our education research architecture, are we seeking to answer these bigger, systemic questions, the answer to which could impact positively on millions of children and tens of thousands of teachers across the country?

ÖйúAƬ

England is fortunate to have rich administrative datasets that education researchers can use, such as the National Pupil Database, the Longitudinal Education Outcomes Survey and the School Workforce Census. But major longitudinal cohort studies, which allow for the exploration of the interrelationship of multiple factors and can highlight where specific interventions might be best targeted, are rarely funded, even for a strategically important subject?such as maths.

Nor does it help that the relationship between education research and policy has not always been smooth in recent years. The mutual critiques and caricatures are as well known as they are unhelpful. And, nowadays, there is an increasingly complex ecosystem of organisations in the research and policy space, including those that both deliver educational programmes and engage in policy debate, such as the Ambition Institute, the National Institute of Teaching, the Teacher Development Trust and the Sutton Trust.

There is room enough for us all, but the risk of turning policy advice into a cacophony is all too real. All this research and policy insight needs to be better orchestrated to support policymakers in addressing the most pressing questions facing the educational system. And I agree wholeheartedly with Clare Brookes, professor of education?at the University of Cambridge, who recently argued in Times ÖйúAƬ that universities have a key role to play in modelling better and more informed debate.

This challenge of generating, coordinating and synthesising policy-relevant research is what the University of Nottingham¡¯s new Observatory for Mathematical Education will address. Supported with a founding grant from XTX Markets, an active supporter of maths education from primary schools to postdoctoral research, we will build a research group of more than 20 researchers, with extensive partnerships and stakeholder collaborations. This will be large enough to investigate both a range of specific issues in maths education and how they relate to one other.

ÖйúAƬ

We will explore, at scale, the interface between the quality and geographical distribution of teachers; the curriculum, assessments and resources they use; parental and wider societal attitudes to maths; the needs of employers and future citizens; the support given to young people falling behind and racing ahead; and much more. We intend to become a hub for major strategic questions related to maths education in England and beyond, channelling the expertise and insights of maths researchers and system leaders across the country ¨C rather like the University of Oxford¡¯s Migration Observatory does in its field.

Such a large-scale, programmatic approach to knowledge creation and exchange is highly unusual in the UK, and we hope to be able to show the benefit of well-orchestrated collaborations between academics, policymakers and interested external parties. Ultimately, we think this model could be replicated in a number of ways across other disciplines ¨C and truly advance the ways education research improves lives.

Andrew Noyes is the founding director of the University of Nottingham¡¯s Observatory for Mathematical Education.