MAXIMIZATION OF SUBJECTIVE EXPECTED UTILITY OR RISK CONTROL - EXPERIMENTAL TESTS OF RISK HOMEOSTASIS THEORY

Valid and meaningful studies examining the risk homeostasis theory (RHT) not only have to be experimental; they also have to be able to make predictions specific to this theory. Generally this is not true for earlier studies. Therefore two experiments were developed for which the theory of the maximization of the subjective expected utility (SEU-model) and the risk homeostasis theory (RHT) make different predictions. In the first experiment, a computer simulation of an intersection situation, forty subjects were asked to move as many vehicles as possible across an intersection making use of the gaps in the flow of traffic while avoiding collisions with the crossing traffic. As a within-subjects-factor the level of difficulty of the task was varied by changing the number of suitable gaps in the moving traffic. Subjects could react either by changing the number of crossings and/or the number of collisions. The between-subjects-factor, the variation of the relative size of the reward for a successful crossing and of the penalty for a collision, allowed discriminating behavioural predictions between RHT and the SEU-model. The results of this experiment support the predictions of RHT. In a second experiment we tried to explore whether (n = 24) subjects behave according to RHT or according to the SEU-model if they can freely choose the level of difficulty (risk) for tasks after they had already performed tasks with a mandatory level of difficulty which differed from their desired level of difficulty. In this experiment the results support the predictions of the SEU-model almost without exception, thus differing markedly from the observations of the first experiment. We conclude that 'achievement situations' and 'true risk situations' differ in characteristic ways, with the assumption that the SEU-model is more able to make valid predictions for the former, and RHT for the latter type of situation.