Cet article aborde la question de la regulation de I'action humaine dans le contrde des systemes hautement complexes. La "regulation de I'action" recouvre I'interaction de la fixation des objectifs, des activites de prevision, de I'elaboration des hypotheses, du planning, de la prise de decision et de la reflexion. Des erreurs et des fautes caracteristiques relevant du planning et de la prise de decision sont decrites et rapportkes a divers aspects du systeme humain de traitement de I'information. Beaucoup de fautes et d'erreurs renvoient A un nombre restreint de caracteristiques d e I'esprit humain. On montre en outre qu'il n'est en rien facile d'kchapper a ces fautes puisque ces processus de traitement de I'information que detient I'esprit humain sont parfaitement fonctionnels dans d'autres contextes.This article deals with human action regulation when controlling very complex systems. "Action regulation" means the interaction of goal elaboration, forecasting activities, hypothesis formation, planning, decision making, and self reflection. Typical errors and mistakes in human planning and decision making are reported and related to characteristics of the human information processing system. It is demonstrated that a lot of very different errors and mistakes are based on a few characteristics of the human mind. Additionally we demonstrate that it is in no way simple to avoid these mistakes, as these information processing characteristics of the human mind are otherwise quite functional in a number of contexts.

This article explores some of the basic shortcomings and fallacies of managerial behavior in dynamic situations. In a laboratory study, 20 groups of three participants each, all with an education in business management, were observed while trying to manage a computer-simulated industrial organization called MANUTEX. This is an interactive simulation of a small garment factory. For most groups, this problem proved to be extremely difficult. The analysis of the problem-solving process and the strategies that the participants employ show that several typical mistakes were responsible for the groups' difficulties. These mistakes are integrated into a number of generic behavior patterns and uncontested basic assumptions that guide action. The participants' difficulties are not due to insufficient managerial knowledge or cognitive limitations. Rather, they stem from an incorrect use of the available knowledge, a tendency to avoid risks and reduce uncertainty, and a motivational process directed at sheltering the subjective sense of competence.

No abstract

Recent research suggests that unconscious processing is superior to conscious processing in tasks involving many decision alternatives (Dijksterhuis et al., 2006). One explanation for these findings is the limited information processing capacity of the human working memory and the almost unlimited resources of unconscious processing. The current study further investigates this topic by using more complex tasks than previously used, i.e., two complex, dynamic, and transparent tasks. Contrary to previous findings, instructions for conscious processing led to better performance in the more complex task. Results are explained referring to methodological reasons and to literature on metacognition. Besides the theoretical relevance, findings could be relevant for training programs on dynamic decision making.

No abstract