The purpose of this paper is to develop a new model for describing the cognitive process of decision making. Being of a generalization of Tversky and Koehler's Support Theory, the proposed model firstly defines the choice probability among several alternatives in terms of degree of “support.” Secondly the model defines the degree of support in terms of objective probability and representativeness. Thirdly, this model specifies the integration process by which one reaches the probability of an event utilizing the already assessed subevents. This model was applied to and tasted by, real experimental data. In the experiment, subjects were asked to evaluate the proportions (a kind of probability) and the degrees of representativeness of two objects. Compared to the estimates derived by a Bayesian approach, the subjective probabilities estimated by the proposed model were shown to be closer to those reported by the subjects.