This research focused on the application of Genetic Algorithm (GA) based methodologies to collaborative problem solving. Direct mechanisms for online asynchronous group work were compared to an indirect constrained group interaction model which was modeled after recombination used in traditional roulette wheel style GAs.;This new methodology contributes to the existing literature by extending the work in the areas of Interactive Evolutionary Computation (IEC) and Human Based Genetic Algorithms (HBGA) by creating a hybrid approach, Indirect Collaborative Evolution (ICE), employing some of the methodological features found within IEC with the collaborative human-centric features of HBGA. This contribution is important as it presents a new means of facilitating group intelligence through indirect interactions (i.e., cold collaboration) and provides new tools to collaborative rich fields, such as nursing.;Groups were compared for overall quality of solutions produced relative to tool set and group size as well as the subjective experience of the participants within each group. Cross generational ratings for solutions within the experimental condition was also examined. Findings indicate that the ICE methodology did not perform significantly different from the direct methodology and that group size played an important role in group performance. ICE outperformed direct collaboration with regards to the subjective measures of group efficacy and participant satisfaction. This finding has several implications, including support for ICE as a valid tool. Advantages for indirect collaboration are discussed.
展开▼
