Gelgia Fetz Fernandes: A recent experimental study conducted in your lab reveals that adolescents are less likely to learn from punishment than from reward. Can you elaborate on the results of the study?

Sarah-Jayne Blakemore: Our study, led by Stefano Palminteri, compared how adolescents and adults learn to make choices based on the available information. 18 volunteers aged 12 to 17 and 20 volunteers aged 18 to 32 completed tasks in which they had to choose between abstract symbols. Each symbol was consistently associated with a fixed chance of a reward, punishment or no outcome. As the trial progressed, participants learnt which symbols were likely to lead to each outcome and adjusted their choices accordingly.

The results showed that adolescents and adults were equally good at learning to choose symbols associated with reward, but adolescents were less good at avoiding symbols associated with punishment. Adults also performed significantly better when they were told what would have happened if they had chosen the other symbol after each choice, whereas adolescents did not appear to take this information into account. The results suggest that adolescents and adults learn in different ways, something that might be relevant to education.

To interpret the results, we developed computational models of learning and ran simulations, applying them to the results of the study. The first model was a simple one that learnt from rewards, the second model added to this by also learning from the option that was not chosen. The third model was the most complete and took the full context into account, with equal weighting given to punishment avoidance and reward seeking.

“The fear of being excluded by peers is a major determinant of adolescent decision making.”

Comparing the experimental data to the models, we found that adolescents’ behaviour followed the simple reward-based model whereas adults’ behaviour matched the most complete, contextual model.

Our study suggests that a reward-based approach, rather than punishment, is more likely to be effective in adolescent learning.

GFF: You have shown in a series of studies that the social brain is undergoing both structural and functional changes in adolescence. What are the consequences of such changes, for example with regard to behaviour?

SJB: These findings have challenged the widely-held assumption that mentalising development is completed by mid-childhood. To investigate the possibility that mentalising develops in adolescence, in parallel with the changes in the social brain, we adapted Keysar’s communication task in which the participant has to take into account someone else’s perspective and use information about this person’s intentions to guide decisions.

Using this task in a study with a large group of children, adolescents and young adults, Iroise Dumontheil and our collaborators found that mentalising undergoes behavioural development well beyond childhood. Specifically, the ability to take into account another person’s perspective to guide decisions is still relatively poor in adolescence, and shows gradual improvement throughout adolescence and into early adulthood.

Many other social cognitive processes are also developing during adolescence in parallel with changes in the social brain, including an increased importance of interactions with peers. One key question is why adolescents often take risks especially when they are with their friends. This peer influence on risk-taking has been shown experimentally in studies looking at the effect of the presence of peers on driving risks in lab-based driving video games. We hypothesised that the tendency to take risks when with peers is associated with hypersensitivity to social exclusion in adolescence, which Catherine Sebastian showed in her study.

In a study with almost 600 participants, Lisa Knoll showed that perception of risk by young adolescents is influenced more by the risk judgements of other teenagers than the judgements of adults. In contrast, children and adults are more influenced by adults than by teenagers. Thus, young adolescents appear to be particularly sensitive to the views of their own age group. On the basis of these findings, Kate Mills and I developed a theoretical framework proposing that the fear of being excluded by peers is a major determinant of adolescent decision making. An adolescent who understands the health risks of smoking, for example, might nevertheless accept a cigarette from friends because of the potential risk of social exclusion.

GFF: You describe typical adolescent behaviour as intrinsically rooted and as meaningful from an evolutionary and social perspective. How can teenage behaviour be socially beneficial?

SJB: Adolescents are known for taking risks, being influenced by their peers, being self-conscious and easily embarrassed by their parents. These are stereotypes, of course, but there is some evidence that many teenagers do undergo substantial behavioural development in these areas. This might be for adaptive reasons: Adolescence is the time when we need to gradually become independent from our parents. We need to take more risks and explore our environments, affiliate with our peers and make new friends.

It has been eloquently argued by Howard Sercombe that risk-taking, in particular, is an important evolutionary behaviour. Adolescence is the period of life in which we develop a sense of self-identity, and particularly a social self, that is how other people see us. Being particularly self-conscious and worried about what friends think might be part of this developmental process.

“It’s highly likely that the striking behavioural changes that adolescents go through are due to substantial and important changes across the brain.”

It’s important to remember that these stereotypical adolescent behaviours are not new. Socrates and Plato wrote about them, as did Shakespeare. They are also evident across species. Mice go through a period of about 25 days of ‘adolescence’, between puberty and becoming sexually mature adults. During this period, they take more risks, explore their environment more, and are more social. A paper published in 2014 showed that adolescent mice drink more alcohol when they are with other adolescent mice, and this is not the case for adult mice. So the peer influence effect is not restricted to human adolescents!

GFF: Your research is highly relevant for teachers and parents. As a neuroscientist, are you willing to offer any advice on how to work with a teenager who is not motivated to learn?

SJB: The neuroscience research is probably too young to draw out strong implications for teachers and parents. It’s a growing field and no doubt it will at some point in the future have implications for education. The main implication at the moment, I think, is that adolescent-typical behaviour is not just down to changes in hormones at puberty or social changes in schools. It’s highly likely that the striking behavioural changes that adolescents go through are due to substantial and important changes across the brain.

One finding with possible implications for education comes from Russell Foster’s group in Oxford. He has shown that circadian rhythm, the body clock, changes at around puberty. At this time, melatonin, the sleep-inducing hormone, is produced later at night, so young people start to find it hard to go to sleep in the evening, and to get up early in the morning. This suggests that school start times might be delayed so that young people don’t have to wake up unnaturally early to go to school each morning. Perhaps later school start times would have social-emotional and educational benefits. There are trials in the US and the UK looking at this possibility.

GFF: Some of your research has been molded by questions asked by teenagers themselves. What are challenges and opportunities for a researcher to communicate with and involve the general public?

SJB: I have been actively involved in public engagement for many years – it’s highly motivating working with the general public and policy makers. There’s a lot of interest in science amongst the public, and I feel that it’s important for scientists to engage with non-specialists about their research.

Since being seconded to the UK House of Commons Select Committee for Education in 2000, I have been interested in the links between neuroscience and education. In the last 15 years, I have been involved in several public engagement projects with young people. I was lucky to be asked to advise a play about the teenage brain (Brainstorm), written by young people from schools in Islington, which was performed at the Royal Albert Hall and the National Theatre in London.

I have also worked with teenagers on a project with the goal to give them an insight into the lives and inspirations of the sort of people who become scientists or work in science related areas. Through the funding of the Royal Society Rosalind Franklin Award, I set up a website, which contains interviews with scientists based on 23 questions submitted by teenagers. The project aims to encourage more young people to consider choosing science subjects at school or university.

Working with young people is incredibly rewarding. In my lab, we often ask young people to be involved in the design of our experiments. They have ideas and insights that we simply wouldn’t have, being that much older, and having grown up in a rather different world to theirs! For example, one of our experiments included stimuli with a picture of a cassette. A young person at school pointed out that few children would recognise this object! We changed the stimulus as a result.


Sarah-Jayne Blakemore is a Royal Society University Research Fellow and Professor of Cognitive Neuroscience at University College London. She is Deputy Director of the UCL Institute of Cognitive Neuroscience and leader of the Developmental Cognitive Neuroscience Group. Her research focuses on the development of social cognition and decision making in the typically developing adolescent brain. In 2015, Sarah-Jayne Blakemore was awarded the Klaus J. Jacobs Research Prize for her exceptional work.

The Blakemore Lab at UCL


Worth knowing

  • A collaboration project led by Kate Mills, with four different cohorts from three different countries, each including child and adolescent participants who underwent MRI scanning at different time points, found strikingly similar developmental trajectories between cohorts. Cortical grey matter volume increased during childhood, peaking at some point in late childhood, and declined substantially during adolescence, stabilising into the twenties. Whole-brain white matter volume increased gradually during childhood and adolescence.
  • In an earlier study Sarah-Jayne Blakemore and colleagues used fMRI to investigate how neural activity during a mentalising task (the ability to understand intentions) differs between adolescents and adults. Compared with adults, adolescents show higher activity in medial prefrontal cortex and lower activity in temporo-parietal junction.
  • Another study by Stephanie Burnett looked at the processing of social emotions, such as guilt and embarrassment, which are emotions that involve mentalising. Thinking about social emotions was associated with higher levels of activity in medial prefrontal cortex, and lower levels of activity in the anterior temporal cortex, in adolescents compared with in adults. A meta-analysis showed that the finding of higher medial prefrontal cortex activity during mentalising in adolescents than in adults has been replicated by multiple studies using a variety of mentalising tasks. Why this is the case is not well understood.
Keep up to date with the BOLD newsletter