Background:

Economists and psychologists have established that non-cognitive traits, such as persistence and motivation, are more strongly associated with educational, labour market and risky behaviour outcomes than cognitive measures such as IQ (1-3). This observation has been used to motivate substantial government investments in early years interventions, such as Head Start in the US and Sure Start in the UK. These programs are partly designed to develop children's non-cognitive skills and ultimately improve their outcomes as adults (4). Yet, despite the increasing enthusiasm for these initiatives amongst policy makers, there are still major unresolved questions about whether the associations of non-cognitive skills and later outcomes are due to underlying causal relationships.

The evidence about the importance of non-cognitive skills is largely based on observational studies, both cross sectional and longitudinal. It is very difficult to infer causation in these studies, because of the challenge in sufficiently accounting for all the potential differences between individuals. This may be one reason why, in practice, experimental and quasi-experimental studies have found that the success of early interventions has been mixed (5). Additionally, we do not know the extent to which the associations between non-cognitive skills and educational and labour market outcomes are driven by common genetic and environment factors. The progress in genome-wide data and statistical tools such as GCTA mean it is now possible to estimate both the contribution of common genetic variation to these traits and the proportions of the associations of non-cognitive skills and outcomes which are due to a shared genetic aetiology. Finally, we can use these methods to show how common genetic variation influences these traits across the life course (6).

Aims:

We will use Genome-wide Complex Trait Analysis to investigate the genetic architecture underlying cognitive and non-cognitive skills.