Policy and Complex Systems
consumption trends are of increasing .
The statistical robustness is quite implicit , due to the construction of the model , being the unique source of stochastic behavior is the space localization of the agents in laboratory use cases .
In configurations where the results are related to small changes in agents ’ attitude , the soundness has been verified repeating the simulation one hundred times . Two extreme cases are reported , where the system behaves in a completely different way , changing the aptitude of a unique agent among about three hundred . In the first of the two cases , we measured 0 different outcomes ; for the other one we measured the 14 % of different outcomes .
With large actual sample of agents with their real data and real geographical location , the outcome – after one year of simulated behavior – are pretty consistent with ex-post real data .
The notion of committed agent is implemented in SAM4SN , as well as the notion of quasi-committed agent .
In our model , evangelist agents are strictly committed agents , because they are very determined in their belief . Their awareness cannot decrease , so they cannot change their type . When an agent becomes evangelist , it will be forever .
Blind agents and active agents are “ quasi committed ” agents because their belonging to a type is very strong , if compared with other types of agent , like spectators and indifferent ones .
We introduced the notions of commitment and “ quasi-commitment ” as useful notions when linked to the concept of social reinforcement . Once a committed ( evangelist ) or quasi-committed agent ( a blind or an active ) is reinforced in his belief , this reinforcement is persistent and the agent remains reinforced as it was ( positively or negatively ) while not committed agents
( spectators and indifferents ) are responsive to positive or negative reinforcements .
We have to introduce the notion of committed and quasi-committed agents . In our model , committed agents coincide with evangelists , i . e . the most influential ones , while blind or active agents are “ quasicommitted ”. The notions of “ commitment ” and “ quasi-commitment ” are linked to the concept of social reinforcement . Once a committed ( evangelist ) or quasi-committed ( blind or active ) agent is reinforced in its belief , this reinforcement becomes persistent and the agent remains forever reinforced , while non-committed agents ( in our model , spectator and indifferent agents ) remain responsive to positive or negative reinforcements . In our model , awareness is affected by social reinforcement , because the reinforcement value is added to the awareness level . When both global and individual trends are of reduction , and the first is higher than the second in absolute value is , awareness increases . When both global and individual trends are of increase , and the first is higher than the second in absolute value is , awareness decreases .
As soon as enough people with high social influence ( Christakis & Fowler , 2009 ) adopt a social norm ( Kinzig et al . 2013 ), a tipping point is reached ( Gladwell , 2000 ; Levin , Barrett , Aniyar , Baumol , Bliss , Bolin ,... Sheshinski , 1998 ). The idea of a tipping point for environmental sustainability is used by Kinzig and colleagues ( 2013 ) and derives from theoretical works ( Xie , Sreenivasan , Korniss , Zhang , Lim , & Syzmanski , 2011 ) about the role that committed agents have in reaching consensus . In particular , Xie and colleagues ( 2011 ) have introduced the value of 10 % of committed agents - as a critical value for opinion diffusion - .
57