Systems analysis of dynamic interaction: Prediction from individual parameters

These experiments conclude a program of research aimed at developing a method bywhich a dyadic unit can be studied as a formal system, a living system at the level of the group. The dyad chosen was one in which a subject acting as “teacher” repeatedly set a goal for another subject, who performed in response to that g o d Each half of the dyad was first studied individually, with input and output variables determined by their dyadic relation. This earlier research described the judgment function by which A chose goals for B in accordance with B's prior performance and then those by which B responds to various goal levels set by A. The parameters from these experiments were used here to predict a sequence of experimental interaction between A and B. Almost all the correlations between the model and the data were both highly significant and capable of accounting for a substantial proportion of the variance.