Individual differences in learning behaviours in humans: Asocial exploration tendency does not predict reliance on social learning

Social learning strategies regulate the wisdom and madness of interactive crowds

“…Although the social learning weight increased with increasing group size, the overall mean value was σ̄ i ≈ 0.3 (Figure 4B; Supplementary Figure 5; Supplementary Figure 6) and it decreased on average as the task proceeded (Figure 5). This implies a weaker social than asocial influence on decision-making as reported in several other experimental studies 35;54–56 although evolutionary models tend to predict heavier reliance on social learning than experimental studies report 57;58 . Thanks to this relatively weak reliance of social learning, the kind of extreme herding that would have blindly led a group to any option regardless of its quality, such as the ‘symmetry breaking’ known in trail-laying ant collective foraging systems 2;5;26 , did not occur (Figure 2).…”

“…For the asocial influence, we used a standard reinforcement learning with ‘softmax’ choice rule 75 , widely applied in human social learning studies e.g. 35;51;55 . An individual i updates the estimated average reward associated with an option m at round t , namely Q-value ( Q i,t ( m )), according to the Rescorla-Wagner rule as follows: where α i (0 ≤ α i ≤ 1) is a learning rate parameter of individual i determining the weight given to new experience and r i , t ( m ) is the amount of monetary reward obtained from choosing the option m in round 𝟙.…”

“…First, we set to test the hypothesis that the circumstances under which collective decision making will generate ‘wisdom’ can be predicted with knowledge of the precise learning strategies individuals deploy, through a combination of experimentation and theoretical modelling. The choice of an abstract decision-making task allowed us to implement a computational modelling approach, which has been increasingly deployed in quantitative studies of animal social learning strategies 35;49–51 . In particular, computational modelling allowed us to conduct a parametric description of different information-gathering processes and to estimate the parameter values at an individual-level resolution.…”

“…Second, we added resolution to our analyses by manipulating both task uncertainty and group size in our web-based experiments with adult human subjects, predicting that these factors would induce heavier use of social information in humans, and thereby alter the balance between collective intelligence and the risk of inflexible herding. To do this, we focused on human groups exposed to a simple gambling task called a multi-player ‘multi-armed bandit’, where both asocial and social sources of information were available 35;51;52 . Through development of an interactive, web-based collective decision-making task, and use of hierarchical Bayesian statistical methods in fitting our computational model to the experimental data, we identify the individual-level learning strategies of participants as well as quantify variation in different learning parameters, allowing us to conduct an informed exploration of the population-level outcomes.…”

Digital Media Behavior of School Students: Abusive Use of the Internet

“…Generally speaking, the most active, neophylic or bolder people tend to be labeled as "explorers", while the more inactive, neophobic or more timid people tend to be labeled as "non-exploratory" (Réale et al, 2007). However, it has not been proven whether the "explorer" type actually collect more information during the learning process compared to those labeled "non-exploratory" (Toyokama, Saito & Kameda, 2017). Geffet & Blau (2016) mention that "Behavior is a strictly physical, recordable and verifiable process, which consists, precisely, of being the activity by which a living being maintains and develops its life in relation to its environment, responding to it and modifying it. "…”