Agency to Choose Social Information Enhances Peer Influence in Adolescent Risky Decision-making
Poster No:
789
Submission Type:
Abstract Submission
Authors:
Shuhan Wang1,2, John Wang1, Natalie Melville1, Mark Orloff3, Matthew Caton1, Pearl Chiu1,2, Brooks Casas1,2
Institutions:
1Fralin Biomedical Research Institute at Virginia Tech Carilion, Roanoke, VA, 2Department of Psychology, Virginia Tech, Roanoke, VA, 3Center for Mind and Brain, University of California, Davis, CA
First Author:
Shuhan Wang
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Co-Author(s):
Pearl Chiu
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Brooks Casas
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Fralin Biomedical Research Institute at Virginia Tech Carilion|Department of Psychology, Virginia Tech
Roanoke, VA|Roanoke, VA
Introduction:
Adolescence is a unique developmental stage often identified with more engagement in risky behaviors. While evidence suggests that peer influence contributes to the spread of risk-taking among adolescents [2], little attention has been paid to how adolescents receive peer influence impacts their behavior. Given prior work showing that in non-social contexts, information acquired under agency becomes more valuable [3], we examine in adolescents whether autonomously chosen (i.e., under agency) or passively provided social information potentially leads to different extents of peer influence and attendant neural responses.
Methods:
Participants: 46 adolescents were recruited in this study around Roanoke, Virginia. After applying exclusion criteria, we analyzed data from 37 participants (22 females; age = 13.24 ± 0.44).
Experimental Design: We used a risky decision-making paradigm [1] in which participants made sequential choices between a safer and a riskier monetary gamble, either after seeing the choices of others (Info trial) or alone (Solo trial). On Info trials, participants were shown the choices of two anonymous social peers before they made their own choices. On Solo trials, participants made choices without viewing social others' choices. To evaluate the effects of agency on peer influence, these trials were shown under 'Choice' or 'No-Choice' conditions. In the 'Choice' condition, participants decided whether or not to view others' choices, and in the 'No-Choice' condition, a computer algorithm pseudorandomly selected whether social information was displayed.
Computational Modeling: To quantify the impact of social information on adolescent risky choices, we adapted a behavioral economic utility model with parameters capturing the value of social information (other-conferred-utility), risk preference, inverse temperature, and the effect of agency on each of these. All parameters were fit using hierarchical Bayesian inference. Alternative models were compared using leave-one-out cross-validation estimation.
fMRI Analysis: Functional and structural brain images were acquired in a 3T Siemens Prisma during the task. The echo-planar images were preprocessed using Statistical Parametric Mapping 12, using a standard preprocessing pipeline. Separate design matrices were constructed to evaluate the BOLD signal when viewing social others' choices on Info trials.
Experimental Design: We used a risky decision-making paradigm [1] in which participants made sequential choices between a safer and a riskier monetary gamble, either after seeing the choices of others (Info trial) or alone (Solo trial). On Info trials, participants were shown the choices of two anonymous social peers before they made their own choices. On Solo trials, participants made choices without viewing social others' choices. To evaluate the effects of agency on peer influence, these trials were shown under 'Choice' or 'No-Choice' conditions. In the 'Choice' condition, participants decided whether or not to view others' choices, and in the 'No-Choice' condition, a computer algorithm pseudorandomly selected whether social information was displayed.
Computational Modeling: To quantify the impact of social information on adolescent risky choices, we adapted a behavioral economic utility model with parameters capturing the value of social information (other-conferred-utility), risk preference, inverse temperature, and the effect of agency on each of these. All parameters were fit using hierarchical Bayesian inference. Alternative models were compared using leave-one-out cross-validation estimation.
fMRI Analysis: Functional and structural brain images were acquired in a 3T Siemens Prisma during the task. The echo-planar images were preprocessed using Statistical Parametric Mapping 12, using a standard preprocessing pipeline. Separate design matrices were constructed to evaluate the BOLD signal when viewing social others' choices on Info trials.
Results:
Adolescent choices are influenced by social others, as evidenced by greater than chance selection of the options chosen by social others (t(36) = 2.05, p = .047 for No-choice condition, t(36) = 3.34, p = .002 for Choice condition). Modeling results reveal that larger utility is conferred to the gambles chosen by social others in the Choice condition than in the No-choice condition (t(36) = 2.61, p = .013), indicating an increased value of social information under agency. Furthermore, vmPFC response tracks the utility of this social information (r = .37, p = .029 for No-choice condition; r = .37, p = .027 for Choice condition; Figure 1A). Participants with greater change in the utility of social information also show greater change in the activation of vmPFC across conditions (t(17) = 2.68, p = .016; Figure 1B). dACC is less activated in the Choice relative to the No-Choice condition when viewing social information (peak-level pFWE =.004; Figure 1C).
Conclusions:
These results support the hypothesis that in adolescents, social information acquired under agency is associated with greater value than the same information acquired passively. This increased value is associated with enhanced peer influence on choices about risky options. Regardless of agency condition, we observe valuation signals in vmPFC that encode the utility of social information. Acquiring social information with agency leads to decreased activation in the dACC, suggesting less decision conflict when adolescents make choices with agency to view social information.
Emotion, Motivation and Social Neuroscience:
Social Cognition 1
Higher Cognitive Functions:
Decision Making
Lifespan Development:
Early life, Adolescence, Aging 2
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Keywords:
FUNCTIONAL MRI
Other - Adolescence; Risk-taking; Peer Influence; Agency Effect
1|2Indicates the priority used for review
Provide references using author date format
[2] Ciranka, S. (2019). Social influence in adolescent decision-making: A formal framework. Frontiers in psychology, 10, 1915.
[3] Leotti, L. A. (2011). The inherent reward of choice. Psychological science, 22(10), 1310-1318.