Strategies and maladaptive behaviours in complex dynamic decision making

Abstract: Research on dynamic decision making using so called microworlds has followed two differentapproaches: the German approach of "komplexes Problemlösen" where the central problem has beenseen as that of action control and the dynamic decision making approach, where the central problemhas been seen as that of task control, In this thesis, these approaches are combined in order to seehow people regulate their behaviour so that they can manage to control a complex, opaque anddynamic decision task. The general objective of the thesis is to see how people develop mentalmodels and formulate goals as a function of observability and action possibilities. The results showthat it is not sufficient to give more specific and concrete goals to the subjects. Most subjects fail to reach a satisfactory level of performance anyhow. More important, a number of maladaptivestrategies within many of the subjects seem to be an important determinant for other maladaptivebehaviours and for bad performance. As a consequence, some of the behaviours that previously havebeen regarded as consequences of failure are instead interpreted as consequences of precursors offailure. This conclusion is supported by more empirical evidence, showing that subjects who laterend up with bad performance have received the same kind and amount of information as thosesubjects who end up with better performance. Finally, it is shown that the model enhancementprocess within the subjects is encouraged by experimental instructions that reduce the experiencedopaqueness and complexity of the decision task. Instructions that emphasise a systematic andelaborate procedure affect the experienced opaqueness of the microworld. Instructions that focus ona strategy with goals and plans seem to partly reduce the experienced complexity of the task.