Decision making requires integrating self-gathered information with advice from others. However, the arbitration process by which one source of information is selected over the other has not been fully elucidated. In this study, we formalised arbitration as the relative precision of predictions, afforded by each learning system, using hierarchical Bayesian modelling. In a probabilistic learning task, participants predicted the outcome of a lottery using recommendations from a more informed advisor and/or self-sampled outcomes. Decision confidence, as measured by the number of points participants wagered on their predictions, varied with our relative precision definition of arbitration. Functional neuroimaging demonstrated arbitration signals that were independent of decision confidence and involved modality-specific brain regions. Arbitrating in favour of self-gathered information activated the dorsolateral prefrontal cortex and the midbrain, whereas arbitrating in favour of social information engaged the ventromedial prefrontal cortex and the amygdala. These findings indicate that relative precision captures arbitration between social and individual learning systems at both behavioural and neural levels.

This dataset was acquired as part of the Master Thesis of Madeline Stecy (Sep 2015 - Feb 2016) at the ETH, Institute for Biomedical Engineering. The study is entitled “Neural Arbitration between Social and Individual Learning Systems” and is linked to an eLife Research Article submission 28-11-2019-RA-eLife-54051. In total, 40 subjects were scanned and analyzed at the Laboratory for Social and Neural Systems Research, Department of Economics, University of Zurich.

In this study, functional magnetic resonance imaging data was acquired using a Philips Achieva 3T whole-body scanner with an 8-channel SENSE head coil (Philips Medical Systems, Best, The Netherlands) as participants performed a probabilistic learning task integrating reward and social cues. The study examined the neural representation of arbitration, the dynamic weighting of self-gathered information against advice from others. 

This dataset includes participants’ unprocessed data: behavioural data, functional and structural magnetic resonance imaging data from two runs of the experimental task, as well as physiological variables acquired during scans, including heart rate and respiration. The structural scans have been skull-stripped to reduce any subject-specific identifying features. 

This dataset can be analysed using the public GIT repository owned by Dr. Andreea Diaconescu. The routines for all analyses are available as Matlab code: https://github.com/andreeadiaconescu/arbitration. The instructions for running the code can be found in the ReadMe file linked to the repository.

Authors: Supervisors: Andreea Diaconescu and Philippe Tobler; Supervisor and Sponsor: Philippe Tobler; Contributors: Lars Kasper, Christoph Mathys, Chris Burke, Zoltan Nagy; Code Review: Lars Kasper

Acknowledgments: We are grateful for support by the Swiss National Science Foundation. We are also grateful to Prof. Klaas Enno Stephan for providing guidance and funding for the study.

The README file entitled "DIACONESCU_WAGAD_ReadMe.txt" has been uploaded along with the data