Older teenagers and young adults quicker to learn maths skills than younger teenagers

24 January 2017

Older teenagers and young adults are able to improve their fundamental maths skills and reasoning abilities more rapidly than younger teens, according to research undertaken by Professor Sarah-Jayne Blakemore at UCL and funded by the Nuffield Foundation.

These findings contrast with the common assumption that 'earlier is better' for learning and highlight late adolescence and adulthood as a potential window of opportunity for educational interventions.

The results also suggest that relational reasoning, which is often a part of IQ tests, develops over the course of adolescence and is highly susceptible to training. This highlights that IQ is not necessarily a stable characteristic of the individual and calls into question the use of IQ tests for entrance exams in schools.

Education policy tends to focus on early life interventions, a practise that is partly based on the economics of early-life investment and while the researchers acknowledge that early education is undoubtedly important for many cognitive skills such as visual or language development, certain complex cognitive skills related to mathematics may be best trained relatively late in development, from 15 onwards.

How the research was conducted

The research team randomly assigned 633 adolescents and adults aged between 11 and 33 to one of three groups who underwent training in cognitive tasks for up to 20 days.

  • One group was trained to discriminate small from large numerosities - number of items there are in a given set, an important skill as we often have to compare and judge quantities in our everyday life.
  • The second group was trained in relational reasoning, which is the ability to detect abstract relationships between groups of items and is related to fluid intelligence.

Both these skills are relevant to education and correlate with mathematics.

  • The third group was trained in face perception, which is not related to mathematics. This group served as a control group.
  • Participants in the first group showed improved performance, but only those aged 15-33.
  • All age groups improved their performance when trained in relational reasoning, but older adolescents and adults showed the highest training benefits.
  • Face processing showed limited training effects and no differences between age groups.

These findings suggest that skills relating to mathematics are more efficiently learned in late adolescence and adulthood than earlier in adolescence. These findings highlight the relevance of this late developmental stage for education, and challenge the assumption that ‘earlier is always better’ for learning.

Read more: When to teach what: are there sensitive periods for learning in adolescence?