||Many economists have tried to measure the utility of wealth and to predict human behavior based on this. In my thesis, I describe three such models - the expected utility theory, the prospect theory and the maximization of the probability of economic survival theory. Using an experiment, I try to figure out which model best represents the reality. For this purpose I developed an economic game to observe decisions of people in situations involving risk which I performed with a group of high-school students. The best representation of their behavior appears to be a convex utility function according to the expected utility theory. The implied risk-seeking is, in my opinion, caused by the fact that the experiment was conducted under laboratory conditions.