Parameter estimation for the cusp catastrophe model

Before catastrophe models can achieve scientific legitimacy, they must be subjected to empirical tests with real data. This research report provides the details of a recently developed statistical procedure for the estimation of the parameters of the canonical cusp catastrophe model. The procedure is computationally no more complicated than multiple regression, and is based on the method of moments. As an example of the use of this estimation technique, it is shown that extroversion can be used as a splitting factor in the prediction of driving speed after the ingestion of alcohol. This is an example in which catastrophe theory supplies a nontrivial and successful description of the effect of a higher-level system characteristic (extroversion-introversion) on a lower-level measure of system performance (driving speed). Of course, catastrophe models can be used in and between all system levels.