Due to the significant research effort devoted to discovering whether certain psychological processes are serial or parallel, it seems important to establish the degree to which such processes are identifiable and to investigate possible ways in which such knowledge can improve our experiments. General definitions of parallel and serial systems are given, followed by a qualitative summary of indentifiability results obtained with special classes of exponential systems. Some of these results are applied to a current experimental paradigm, and possible techniques are suggested to provide stronger serial-parallel tests and acquire more temporal processing information. Finally, the possibility of S's possessing the ability to manipulate his distribution of processing energy is re-acknowledged.
Abstract