In the current paper, we compare alternative approaches to incorporating uncertainty into the statistical analysis of dichotomous choice responses. In doing so, first we employ previous modelling techniques that included uncertainty of preferences, and secondly we compare the obtained results with those coming from a novel approach here developed, a finite mixture model. The finite mixture model is a very flexible framework used to deal with preference uncertainty. Our case study uses data gathered in the Prestige oil spill valuation study from Spain.