The purpose of this study was to create an adaptive agent based simulation modeling the processes of creative collaboration. This model aided in the development of a new evolutionary based framework through which education scholars, academics, and professionals in all disciplines and industries can work to optimize their collective ability to find creative solutions to complex problems. The basic premise follows that the process of idea exchange, parallels the role sexual reproduction in biological evolution and is essential to society's collective ability to solve complex problems. The study outlined a set of assumptions used to develop a new theory of collective intelligence. These assumptions were then translated into design requirements that were designated as parameters for a computational simulation that utilizes two types of machine learning algorithms. This model was developed, and 200 simulations were run for each of 48 different combinations of four independent variables for a total of 9,600 simulations. Statistical analysis of the data revealed a number of patterns enhancing the simulation agents' collective problem solving abilities. Most notably, agents' collective problem solving abilities were optimized when idea exchange between agents was balanced with individual agent time contemplating new creative strategies. Additionally, the agents' collective problem solving abilities were optimized when simulation constraints did not force the agents to converge upon one potential solution.
展开▼
