Traditionally, research was conducted by a single scientist or a small team of scientists within a single laboratory. The scientist(s) would conduct the majority of required experiments themselves, even if they did not initially have the necessary expertise or equipment. If they could not conduct an experiment themselves, they would attempt to find a collaborator in another lab to help them by using a barter system. This barter system essentially involves one scientist asking for a favor from another scientist, with the potential upside being co-authorship on any publications that are produced by the work. This type of collaborative arrangement depends heavily on personal networks developed by scientists.