James Croft is the founder and director of the Centre for Education Economics (CfEE). He has authored and co-authored a number of reports for the Centre and its partners, including most recently, Optimising autonomy: a blueprint for education reform.

One way or another, Brexit will soon be decided, and it will become clear how in the wider context of global trade things are going to play out economically for us. There will be a different set of priorities in every area of policy, but the need for investment in human capital is likely to figure prominently across all of them. In education, this could well lead to a renewed focus on identifying and tracking advanced learners through selective or other forms of gifted education.

The economic case for seeking to improve provision in this area is compelling. Cross-national research has found that whereas a 10 percentage point increase in the share of pupils who reach basic skills in international tests is associated with an increase in the average per capita annual growth rate of 0.3 percentage points, the same increase in the share of pupils performing at the high end of the spectrum raises average per capita annual growth rate by 1.3 percentage points. Other research indicates positive effects for social outcomes would likely follow realisation of the potential of this group – such as lower crime, improved health, and improved democracy. In other words the reward for successfully nurturing the pupils with the highest potential could be considerable indeed.

It’s a good time, then, to consider how far we can be clear about what, if anything, works in gifted education.

Unfortunately, while practitioner perspectives abound, we have depressingly little to go on empirically. In research published today, we find that the literature is generally of poor quality, making it difficult, if not impossible, to draw strong policy-relevant conclusions. Moreover, whatever little rigorous evidence we do have suggests that neither school-level streaming nor gifted education programmes, on average, make much difference as ways of generating higher performance among gifted children. In other words, the existing literature leaves the field with considerable ambiguity about effective practices.

Nevertheless, to our mind, there are a few empirical studies that offer useful avenues for further research and policy understanding. Of these few studies, a number demonstrate positive effects for enrichment programmes combined with self-directed or individualised instruction. Like other approaches we are still faced with how to identify effectively those who would benefit, but importantly such strategies are not necessarily reliant on school-level tracking or streaming.

And there are additional reasons for taking these findings particularly seriously. First the ‘discovery-based’ learning model implied by these findings is consistent with cognitive research on what expert learners need, and with the literature demonstrating that traditional, teacher-directed approaches work best for other children. So we should see whether provision for gifted children could be developed in ways that might improve their achievement, while raising the floor and overall picture of school attainment using other methods.

And second, enrichment programmes, together with discovery-based individual learning, are characteristics that feature in the educational models implemented by a number of high performing economies, mostly in Asia, with whom we compete. Much-discussed Singapore in particular looks likely to have lessons for governments wishing to develop their gifted provision.

Since none of these models have been evaluated in any way that makes causal inference valid, there’s a good case for collaboration with in-country researchers around properly conducted trials. But perhaps a greater priority is to go straight to the English context to begin randomised testing of different programmes.

There are several reasons why progress in this field has been slow. One of them, perhaps with good reason, is the way that the definition of giftedness has fragmented over time, often under pressure of determining who should qualify for limited funding. Yet consistent with a wider policy framework geared to increasing educational opportunity, we shouldn’t sacrifice the potential benefits on the altar of precision. Imprecision in identification is probably not a problem that can be overcome entirely, but its adverse effects for those disadvantaged by background, and for later developers, can certainly be reduced. To avoid this trap, a multi-pronged approach to identification is needed combining universal screening, additional metrics (not necessarily test based), and reassessments over time.

So investment is needed if we are to release the human capital of those at the high-performing / potentially high-performing end of the spectrum. And most importantly, this must extend to research. We therefore recommend the endowment of a purpose-built independent research outfit, similar to what we already have in the Education Endowment Foundation for finding out what works to improve the attainment and social mobility of disadvantaged children. In this context, the purpose of the new organisation would be to fund randomised trials investigating what works in gifted education specifically.

Given our literature review, trialling enrichment and independent-learning models to find out what could work in England specifically would be a good place to start. Since trials take time, and since the current political and economic outlook is what it is, this is a policy the government should waste no time in implementing.

What works in gifted education? A literature review, by Gabriel Heller-Sahlgren,is published today by the Centre for Education Economics (CfEE).