Defining Factors of Interest for Large-scale Socio-cognitive Simulations

We examine how several environmental factors influence cognition and the emergence of social networks in artificial societies. Using a large-scale socio-cognitive simulation (VIPER), we generated social networks consisting of 20, 40, and 60 agents. We tested the impact that environmental factors such as population size, map configuration, and run time, particularly on the formation of social ties in memory. We analyzed 1,080 ego-nets across 27 conditions, measuring the number of links and average degree of each network. While all these factors influenced these network measures, our results suggest that population size has the largest influence. In addition, we examined what impact activation values and retention parameters have on the construction of social networks in memory, finding that a shift in activation values resulted in a loss of links, thus indicating a cognitive foundation for Dunbar’s Number (Dunbar, 1998) for the maximum number of social ties stored in memory.