MODELING SYNERGY USING MANIFEST CATEGORICAL VARIABLES

This paper discusses methods to model the concept of synergy at the le vel of manifest categorical variables. First, a classification of conc epts of synergy is presented. Additive and nonadditive concepts of syn ergy are distinguished. Most prominent among the nonadditive concepts is superadditive synergy. Examples are given from the natural sciences and the social sciences. Modelling focuses on the relationship betwee n the agents involved in a synergetic process. These relationships are expressed in form of contrasts, expressed in effect coding vectors in design matrices for nonstandard loglinear models. A method by Schuste r is used to transform design matrices such that parameters reflect th e proposed relationships. An example reanalyses data presented by Bish op, Fienberg, and Holland (1975) that describe the development of thro mboembolisms in women who differ in their patterns of contraceptive us e and smoking. Alternative methods of analysis are compared. Implicati ons for developmental research are discussed.