How information about probabilities and pay-offs is combined in risk judgment?

In most theories of risk judgment, it is assumed that probabilities and payoffs combine multiplicatively. However, there is empirical evidence that risk judgment is better represented by an additive combination of probability and utility. So far, this additive relation has been demonstrated only for risk rates averaged over all participants. Thus, it may be argued that the observed additive combination rule is an artifact of averaging individual strategies focused either on probabilities or payoffs. In this experiment, both the average and individual risk rates were compared with prediction of the multiplicative and additive models. The additive model provides good description of actual risk rates, both for the averages and for a majority of participants individually, whereas the multiplicative model performs poorly. Moreover, the analysis of the relative importance of payoffs and probabilities in risk judgments points to individual differences in weighting both factors but does not support concerns that the additive relation between probability and amount of loss results from averaging individual strategies, in which participants pay attention to only one of these two dimension.