Fig 2. Hybrid social decision-making model visualization and explanation.

The hybrid model uses two weights (ω_follow and ω_oppose) to represent utility–independent contributions of social information to participant decisions, indexing the degree to which participants follow or oppose the choices of social others, independent from utility-based risk processing. (a) Effectively, the ‘follow’ and ‘oppose’ weights act to constrain the probability of choosing a given option. On Info: ‘safe’ trials, the maximum probability of choosing the safe option (regardless of any gamble information) is equivalent to 1 –ω_oppose (shown as the dotted red lines for different ω_oppose values) while the minimum probability of choosing the safe option (regardless of any gamble information) is equivalent to ω_follow (shown as the dotted green line). Likewise, on Info: ‘risky’ trials (not depicted), the maximum probability of choosing the safe option is 1 –ω_follow, and the minimum probability of choosing the safe option is ω_oppose. The black lines represent the other-conferred utility process, setting the probability of choosing the safe option as a function of gamble information transformed by one’s risk preference (according to a power utility function), sensitivity to utility (softmax choice rule), and subjective value conferred by others choosing a particular option (other-conferred utility; OCU). (b) In constructing this model, we remained agnostic as to whether oppose and follow represent dissociable processes. Nonetheless, after model estimation, we further tested if there was a relationship between these oppose and follow processes. We find no significant correlation between ω_follow and ω_oppose when measured across all participants (Pearson’s r = –0.11, P = 0.47, robust correlation, BF_null = 6.52), suggesting dissociable processes for individuals’ decisions to follow or oppose social others’ choices. Each point represents an individual participant, and lines are the regressions between the indicated parameters.