Most decision support systems (DSS) based on causal models fail to analyze the empirical validity of the underlying cause-and-effect hypotheses, but instead concentrate on numerous analysis techniques within the method base. However, the soundness of these cause-and-effect-relations as well as the knowledge of the approximate shape of the functional dependencies underlying these associations turns out to be the biggest issue for the quality of the results of decision supporting procedures. Therefore this article strives towards an approach to prove the causality of nomologic cause-and-effect-hypotheses by empirical evidence as a prerequisite for the approximation of the mostly unknown causal functions. Since the latter very often show non-linear influences, it is necessary to employ universal function approximators for this purpose: consequently the proposed approach adopts artificial neural networks (ANN) as an inductive method to learn a calculational model of cause-and-effect functions from empirical time series.
Note—Participative Design of Strategic Decision Support Systems: An Empirical Assessment
