This paper was presented at a symposium, "The Strategy of Multivariate Experiments in Psychological Research," at the Meetings of the American Psychological Association in September 1959. Empirical investigations in the form of the relation between a dependent variable and several independent variables constitute the central focus of this presentation. Particular attention is given to individual differences in the form of these relations and to description of observations for individuals by systematic structures. Many procedural details have been worked out for an initial class of analyses and tried out with several collections of data. The method is outlined and illustrated by material drawn from these trial applications.