Rxivist logo

Rxivist combines preprints from bioRxiv with data from Twitter to help you find the papers being discussed in your field. Currently indexing 73,182 bioRxiv papers from 318,593 authors.

Do early life non-cognitive skills matter? A systematic review and meta-analysis of early life effects on academic achievement, psychosocial, language and cognitive, and health outcomes

By Lisa G. Smithers, Alyssa C. P. Sawyer, Catherine R. Chittleborough, Neil Davies, George Davey Smith, John Lynch

Posted 10 Mar 2017
bioRxiv DOI: 10.1101/115691

Background: Success in school and the labour market is due to more than just high intelligence. Associations between traits such as attention, self-regulation, and perseverance in childhood, and later outcomes have been investigated by psychologists, economists, and epidemiologists. Such traits have been loosely referred to as non-cognitive skills. There has been no attempt to systematically assess the relative importance of non-cognitive skills in early life on later outcomes. Methods: The systematic review protocol was registered with the International Prospective Register for Systematic Reviews (PROSPERO, CRD42013006566) in December 2013. We systematically reviewed electronic databases covering psychology, education, health and economics for articles published from database conception until September 2015. Titles and abstracts were screened for eligibility, and from eligible articles data was extracted on study design, sample type and size, age of participants at exposure and outcome, loss to follow up, measurement of exposure and outcome, type of intervention and comparison group, confounding adjustment and results. Where possible we extracted a standardised effect size. We reviewed all studies and rated their evidence quality as better, weak, or poor on the basis of study design and potential for confounding, selection and measurement bias. Results: We reviewed 375 studies and provided interpretation of results from 142 (38%) better quality studies comprising randomised controlled trials, quasi-experimental, fixed effects including twin studies, longitudinal and some cross-sectional designs that made reasonable attempts to control confounding. In the academic achievement category outcomes were reported in 78 publications of better quality studies which were consistent with 0.1-0.2 SD effects. Psychosocial outcomes were reported in 65 better quality studies consistent with effects of 0.3-0.4 SD. For the language and cognitive category there were 42 publications reporting better quality studies consistent with effects of 0.3-0.4 SD. For physical health, results across only eight better quality studies were inconsistent but centred around zero. Analysis of funnel plots consistently showed asymmetric distributions, raising the potential of small study bias which may inflate these observed effects. Conclusions: The evidence under-pinning the importance of non-cognitive skills for life success is diverse and inconsistent. Nevertheless, there is tentative evidence from published studies that non-cognitive skills associate with modest improvements in academic achievement, psychosocial, and language and cognitive outcomes with effects in the range of 0.2-0.4 SD. The quality of evidence under-pinning this field is generally low with more than a third of studies making little or no attempt to control even the most basic confounding (endogeneity) bias. The evidence could be improved by adequately powering studies, and using procedures and tools that improve the conduct and reporting of RCTs and observational studies. Interventions designed to develop childrens non-cognitive skills could potentially improve opportunities, particularly for disadvantaged children. The inter-disciplinary researchers interested in these skills should take a more rigorous approach to determine which interventions are most effective.

Download data

  • Downloaded 1,224 times
  • Download rankings, all-time:
    • Site-wide: 6,178 out of 73,219
    • In epidemiology: 58 out of 1,556
  • Year to date:
    • Site-wide: 23,435 out of 73,219
  • Since beginning of last month:
    • Site-wide: 23,435 out of 73,219

Altmetric data

Downloads over time

Distribution of downloads per paper, site-wide


Sign up for the Rxivist weekly newsletter! (Click here for more details.)