Data analysis of the pattern information of the collective decision-making process in subterranean termites species

Data were collected by a bioassay design to study the collective decision-making process of food searching for a subterranean termite species. The movements of termites were recorded by voltage-drop sensors, video-taping, and manual counting. Principal components analysis was applied to separate all the experimental replicates into three groups dependent upon termites final settlements. Multi-phase regression was conducted to determine the change-points in the process of food searching. The result of the data analyses indicated two general types of termite behavior. For both of these groups, termites seem to identify the food source quickly, but the movements of termites after finding the food are very different.