Introduction

Adolescence is seen as the transition period between childhood and adulthood. During this period there are major changes in cognitive and social-affective reasoning. Adolescents are very sensitive to social events and the influence of peers. The cognitive changes in adolescence seem to originate from the changes in the social cognitive network in the brain during the same period. The medial prefrontal cortex, the temporoparietal junction and the superior temporal sulcus are examples of brain regions that undergo a major change in both functioning and structure during adolescence.

Several studies have shown that adolescents show more risky behaviour when they are in the presence of peers. This seems to be related to brain areas involved in the affective processing of risks and rewards. These areas are more active when peers are present, especially among adolescents.

Few studies have been conducted into the influence of peers in a more positive way. Recent studies seem to indicate that adolescents are becoming more pro-social when peers are around. Current research also builds on this notion and also wants to look at the areas of the brain involved in prosocial behaviour in adolescence. Adult research indicates that the medial prefrontal cortex and the ventral striatum play a role in peer evaluation. The medial prefrontal cortex also seems more active when a social situation is only imagined.

Current study

Research was done in two age groups (12-13 and 15-16 years). They were compared with each other regarding the effects of peer influence. Previous research suggests that young adolescents are more sensitive to this.

The Public Goods Game was used during this investigation. In this social dilemma game, the participant must distribute tokens among themselves and the group. Three conditions were used to assess peers:

• Spectators (actors of similar age) who promoted prosocial behaviour (many tokens for the group: thumbs up).

• Spectators (actors of similar age) who did nothing.

• No spectators.

The spectators and the participants met before the game started. The researchers' prediction was that there should be more activity in the medial prefrontal cortex when spectators were present in young adolescents than in middle adolescents.

Methods

Attendees

The 12-13-year-olds age group included 31 participants, 15 of whom were men. The age group 15-16-year-olds contained 30 participants, including 14 men. More than 90% of the participants were born in the Netherlands. All participants spoke and wrote Dutch fluently. An IQ measurement was done with the WISC-III and the WAIS-III. All IQ scores were within the normal range and there were no significant differences in the age groups.

Experimental research

The participants were told that they were playing the public goods game online with four anonymous other players of the same age. These were also participants in the study. The anonymity would remain guaranteed. Each round the participants received five tokens, with values ​​of 1, 1.50 or 2 euros per token. All tokens given to the group were doubled and then distributed among the group members. This means that the participant has the most money left if he keeps all tokens to himself, while the group benefits the most when each participant donates all tokens to the group. Donating to the group can be seen as prosocial behaviour. The participants could not see what the others were doing and did not get to see the results per round.

72 trials were used, divided into two rounds and three conditions:

For each block of 12 trials, the participant was shown for two seconds in which condition they were. When participants received feedback at the end of the trial, they saw 0 thumbs up at 0 and 1 donated tokens, 2 thumbs up at 2 donated tokens, 4 thumbs up at 3 donated tokens and 5 thumbs up at 4 and 5 tokens. When participants had spectators but received no feedback, the participants were told that the spectators did evaluate, but their feedback would not be visible.

The spectators consisted of 44 actors, who changed per trial session. The goal was to introduce all participants to six out of ten actors, which succeeded in 75% of the cases. During the game itself, the participants saw a photo of the actors with neutral facial expression.

Procedure

This research was part of a larger research into relationships with peers. The participants arrived at this study with their best friend, who participated in another part of the study. The participant heard that it was necessary to wait for more participants, after which the (usually 6) actors came in a little later. The participant was introduced to them. The participant would be examined in the game in an MRI scanner. He or she then heard that not all other 'participants' were there yet, but that the MRI procedures would already be explained. The participant already played five practice sessions there. It was also explicitly stated that the participant would not play against the best friend. The scanning session lasted approximately one hour. Afterwards, the participants completed questionnaires and did two sub-tests of the WISC or the WAIS. This was followed by a debriefing and a compensation of 32 euros per participant.

Analytical methods

Two ANOVAs were performed to investigate responses to the stimulus and feedback onset. A 3x3 ANOVA (3 conditions and 3 token values) was used for the stimulus onset and a 3x3 ANOVA for the feedback onset. In addition, a regression analysis was performed on data across the entire brain.

Regions of Interest (ROI) analyses were also carried out to see the differences between the two age groups.

Results

Behavioural analysis

Participants donated more tokens to the group when there were spectators. Most tokens were donated to conditional viewers with feedback. The tokens that were most frequently donated were the tokens with a value of one euro. The tokens with a value of two euros were the least donated. In every condition, more was donated to the group by young adolescents than by middle adolescents.

fMRI analyses at the stimulus onset

The stimulus onset is the moment that the participant decides how he or she will distribute the tokens. The condition had a major effect on various areas of the brain, including the dorsomedial prefrontal cortex, the precuneus superior temporal sulcus and the temporoparietal junction. The value of the token had no effect on the brain and there was also no interaction effect.

When spectators were present, there was more activation in the precuneus, the bilateral temporoparietal junction and the bilateral superior temporal sulcus, compared to the condition without spectators.

When spectators who gave feedback were present, there was also increased activation in the dorsomedial prefrontal cortex, compared to the condition without spectators.

There was no significant interaction effect when looking at the age group. To get a better picture of this, we looked at the ROI analyses. This showed that the young adolescents showed a greater activation difference in the dorsomedial prefrontal cortex and the left superior temporal sulcus in the condition viewer with feedback and no viewer than the middle adolescents.

Regression looked at activation in the brain and height of the donated tokens. In the conditions with feedback-giving spectators versus no spectators, more activation appeared to be in the left temporoparietal junction with a higher donation.

FMRI analyses with the feedback onset

This concerns the analyses of brain activation when the feedback screen is displayed. The repeated ANOVA measurements on the condition and token value show the main effect of condition in many different brain regions, such as the bilateral insula and the right amygdala. There was no main effect for token value and no interaction effect. When the spectator conditions were compared with the no spectator condition, there was additional brain activation in more areas in the important areas that have to do with face recognition at the spectator conditions compared to the no spectator condition. In the condition viewer with feedback, compared to the condition viewer without feedback, increased activity was visible in different brain areas,such as the bilateral insula and the right superior parietal cortex.

There appeared to be no interaction effect between age group and condition. When ROI analyses looked at this extra closely, there appeared to be an interaction effect: the young adolescents had larger activation differences between the viewer with feedback and the viewer without feedback condition than the middle adolescents.

Discussion

The social behaviour of adolescents is characterized by a high sensitivity to the opinion of peers. Current research has shown that peer feedback has a positive effect on the degree of prosocial behaviour during early and middle adolescence. If the feedback from peers is also prosocial in nature (thumbs up), the effects are greatest. Similar results have been found in adults in other studies.

To make a choice to donate in the presence of peers, the dorsomedial prefrontal cortex (involved in social influence), the temporoparietal junction and the superior temporal sulcus become active. The current research showed that prosocial feedback had little influence on the level of activity in these brain regions, only the idea of ​​peer evaluation is sufficient for the increased activity.

Other research into the influence of peers also shows extra activation in the ventral striatum. An explanation for this is that the presence of peers makes behaviour more rewarding. This activation has not been demonstrated in the current study.

Current research has found that prosocial behaviour (donating a higher amount) entails greater brain activity in the temporoparietal junction when the participant was assessed with feedback. This brain structure appears to be involved in altruism.

From a behavioural perspective, younger participants gave a higher amount to the group than older participants. This applied to all conditions, including the condition without peers. This is in line with other studies that indicate that younger adolescents are more prosocial towards strangers. Young adolescents also seem to be the most sensitive to the opinions of peers. This was also found in the brain in the current study: there was greater brain activation among young adolescents when spectators were present.

Learning and cognitive control through the presence of peers with feedback could play a role in regulating behaviour and adaptation to others. More research is needed here.

In the current study, no distinction was made between the amount of the donation and the feedback that was given (the feedback was purely geared to the number of tokens, not to the value thereof). This has consequences for the interpretation of the results found.