The purpose of the proposed research is to demonstrate a software system that makes correctional/factor type analyses of large data bases practical on microcomputer-class machines, and is able to compute the significance of psychophysiological measurements at a speed that makes real-time feedback control mechanisms possible. The software prototype developed will access a data base composed of a wide range of measurements, analyze them to discover significant patterns of response, abstract "typical" patterns for a variety of physical states for each subject, and provide for the comparison of new measurements with the previously constructed set of "typical" patterns in order to provide input for the control of subsequent external events.