Optimizing a nonlinear mathematical approach for the computerized analysis of mood curves.

A nonlinear mathematical model for computerized description of mood curves is presented. This model reaches a high goodness of fit to the real data. It seems superior to two other models recently proposed. Using this model in a computer program for describing the mood data of a large sample of inpatients, significant and clinically meaningful group differences between the mood curves of schizophrenic, endogenous-depressive, and neurotic-depressive inpatients could be demonstrated. The application of the methodology might be helpful, e.g. in the field of evaluative research.