The Characteristics And Neural Underpinning Of Rule-based Learning Tasks In Adolescents And Adults

Since the early adolescence(10-14 years old),their performance are as good as adults in many cognitive tasks.At the same time,they experience the imbalance between a relatively overactive striatum and a relatively under-engaged prefrontal cortex,these two brain areas have weak functional connectivity.This suggests that adolescents complete a feedback-based rule learning task may differ from adults.In order to find out the possible differences between adolescents and adults in feedback processing,this study inherits the Guess and Application Task(GAT)of the previous research,in which,participants can learn rules based on the feedback in learning phase,then use them to complete the task correctly during the application phase.Based on this,we can observe the neural underpinning of the feedback processing of the participants in the guessing phase and judge their learning effect in subsequent application phase.Because the paradigm used in previous studies is simple for adults,which can lead to a ceiling effect.This study expected to examine the differences of behavioral characteristics and neural underpinning between adolescents and adults in difficult tasks.Therefore,in experiment 1,we revised the "guess-application" task.After randomly selecting one of the four cards(red/green triangle/square)in the guessing phase,participants got feedback containing the magnitude and valence information(large/small win,large/small loss).Based on this feedback,participants could infer the reward rules of the current trial.In the subsequent application phase,four candidate cards were randomly combine and there would be two kinds of tasks with same valence(such as large win and small win)and four tasks with different valence(such as large win and small loss).Participants needed to learn to choose the better one.In experiment 1,we recruited 37 adults(25 females,12 males)to examine the feasibility of this paradigm by comparing the accuracy and response time in the application phase under feedbacks in learning phase.The results showed that compared with different valence tasks,the accuracy of the same valence tasks were significantly lower and the response time was longer,which could be more difficult.And the difficulty of each subtask was quite equal,with no extremely difficult tasks appeared.Therefore,this paradigm can effectively avoid the ceiling effect and distinguish the difficulty of the task.In the second experiment,the adolescents and adult groups would be compared to examine the similarities and differences of feedback processing between the two age groups.Experiment 2 applied this paradigm and recruited 22 adolescents(11 females,11 males 13-14 years old,average age was 14.11±0.30 years)and 34 adults(23 females,11 males 18-26 years old,average age was 21.20±2.08 years old).We recorded and analysed their behavior data,at the same time,the ERP data recorded in the guessing phase were analyzed,and the feedback-related negative wave(FRN),the late positive potential and the midfrontal theta(FM?)were used as observation indicators.From the analysis of behavior and EEG/ERPs,we conclude that:1.Overall,no matter how difficult the task was,both groups showed a correct rate and accelerated response with the task,and there was no difference between the overall correct rates.The difference in age only existed after negative feedback,and the response of adolescents in the subsequent application phase was significantly longer than adults,especially when loss was large.The results suggest that adolescents are more sensitive to large losses and need to invest more cognitive processing resources than adults to facilitate the completion of learning tasks.2.Our analysis of multiple indicators of EEG showed that although the behavioral performance of adolescents and adults was comparable,the neurological processes were different in multiple indicators.Specifically:a)FRN reflected early feedback processing,adults under the negative feedback and small feedback both triggered a larger FRN amplitude in the frontal electrode region,especially large feedback showed a larger FRN valence effect,but no obvious and concentrated frontal FRN activities were observed in adolescents;b)In the later period of feedback processing,the adolescents' posterior brain regions caused a positive wave with a significantly larger amplitude than adults;c)Compared with adults,the energy of adolescents' FM? in the early period was more active in the central electrode area,and then the activation of FM? energy in adolescents was not worse than that in adults,and both showed more obvious activities in the forehead electrode area after negative feedback.This indicates some differences on the time of feedback processing between adolescents and adults.Therefore,adolescents can perform the feedback learning tasks as well as adults,but relatively slower in response time.It reflects that adults rely more on the early neural underpinning,while the adolescents more dependent on the later neuroprocessing components to integrate the feedback information.