Disclosed is a method for analyzing a computer-based data processing model of a complex system, said model comprising at least two modeling levels on which the system is represented at different degrees of abstraction. The inventive method comprises the following steps: a first partial model of the system is created on a first modeling level by defining objects and links between said objects; a second partial model of the system is created on a second modeling level by defining objects and links between said objects; associations or relations between objects of the first and the second modeling level are created; at least one first chain of events is created which represents a sequence of actions on the first modeling level and describes said sequence of actions by means of the objects of the first modeling level; at least one second chain of events is created which represents the same sequence of actions on the second modeling level and describes said sequence of actions by means of the objects of the second modeling level; and at least one of the following two steps: the partial models and associations and chains of events created according to the previous steps are automatically analyzed in order to determine and output quality-related data about the model; the partial models and associations and chains of events created according to the previous steps are automatically analyzed in order to obtain data about the modeled system and output said data.