Neural sensitivity to conflicting attitudes supports greater conformity toward positive over negative influence in early adolescence

Abstract

Adolescents often need to reconcile discrepancies between their own attitudes and those of their parents and peers, but the social contexts under which adolescents conform to the attitudes of others, or the neurocognitive processes underlying decisions to conform, remain unexplored. This fMRI study assessed the extent to which early adolescents (n = 39, ages 12-14) conform to their parents' and peers' conflicting attitudes toward different types of behavior (unconstructive and constructive) and in response to different types of influence (negative and positive). Overall, adolescents exhibited low rates of conformity, sticking with their pre-existing attitudes 65 % of the time. When they did conform, adolescents were more likely to conform to their peers' attitudes towards constructive than unconstructive behaviors, exhibiting decreased activation in the ventromedial prefrontal cortex, dorsal anterior cingulate cortex, insula, and inferior frontal gyrus during peer conformity toward constructive over unconstructive behaviors. Adolescents were also more likely to conform when their parents and peers endorsed relatively more positive influence than negative influence, exhibiting increased activation in the temporoparietal junction when considering conforming to negative over positive influence. These results highlight early adolescents' ability to stick with their own opinions when confronted with opposing attitudes and conform selectively based on the social context.

Keywords: Adolescence; Conformity; Parent influence; Peer influence; fMRI.

Fig. 1

Attitude Conformity task. A) Two weeks prior to the scan, adolescents rated their attitudes toward everyday behaviors (in this schematic, the adolescent rating is circled for visualization purposes but was not shown during the task). B) During fMRI, adolescents were presented with opposing attitudes from their parent and peer, which were manipulated based on the adolescent’s original rating during the behavioral session. On Parent Conflict trials, just the parent’s rating conflicted from the adolescent’s but the peer’s rating matched. On Peer Conflict trials, just the peer’s rating conflicted from the adolescent’s but the parent’s rating matched. On Mutual Conflict trials, both the parent’s and peer’s ratings conflicted from the adolescent’s. Conformity was operationalized as choosing the person whose rating conflicted with the adolescent’s original rating, whereas resistance was operationalized as choosing the person whose rating matched the adolescent’s original rating. C) Social influence was manipulated in two ways: Parents and peers (1) influenced adolescents’ attitudes toward unconstructive and constructive behaviors (i.e., type of behavior) and (2) endorsed attitudes that reflected either more positive or negative influence (i.e., type of influence).

Fig. 2

Behavioral effects of type of behavior. A) Adolescents were less likely to conform to their parents’ attitudes (i.e., chose parent) on Parent Conflict trials compared to Mutual Conflict trials, regardless of the type of behavior. B) Conformity to peers’ attitudes (i.e., chose peer) on Peer Conflict trials depended on the type of behavior, such that adolescents were more likely to conform to their peers’ attitudes toward constructive than unconstructive behaviors. Note: Error bars represent the standard error of the mean. Decisions (i.e., probability of choosing peer or parent) that occurred significantly above or below chance level (i.e., 50 % probability) are denoted by significance stars above the condition, and decisions that varied across conditions are denoted with a significance bar. ***p < .001, **p < .01.

Fig. 3

Neural responses during peer conformity to unconstructive relative to constructive behaviors. A) Whole-brain results for the Unconstructive Peer Conflict > Constructive Peer Conflict contrast. Adolescents exhibited parametric increases in the B) ventromedial prefrontal cortex (vmPFC) and C) dorsal anterior cingulate cortex (dACC), and several other regions, during peer conformity toward unconstructive behaviors, whereas they showed decreases in these regions during peer conformity toward constructive behaviors.

Fig. 4

Behavioral effects of type of influence. The mean probability of conformity is shown at -1 SD (i.e., negative influence) and +1 SD (i.e., positive influence) from the mean level of influence. Adolescents were more likely to conform when their A) parents and B) peers endorsed more positive influence than negative influence relative to what participants originally reported. Note: Error bars represent the standard error of the mean. Decisions (i.e., conform or resist) that occurred significantly above or below chance level are denoted by significance stars above the condition, and decisions that varied across conditions are denoted with a significance bar. ***p < .001.

Fig. 5

Neural responses during conformity to negative vs. positive influence. A) Whole-brain analyses revealed there were parametric increases in TPJ activation when adolescents considered conforming to relatively more negative influence and no parametric changes in TPJ activation when they considered conforming to relatively more positive influence. B) A whole-brain regression analysis with the average frequency of conformity revealed adolescents who exhibited greater posterior superior temporal sulcus (pSTS) activation when considering relatively more negative vs. positive influence had lower rates of conformity toward negative vs. positive influence.