Brooks AFB, TXA new model of real-time risky decision making is introduced that predicts tradmffs between processing and risk taking during driving. This model, called Decision-Making under Risk in a Vehicle, or DRIVE, was fitted to data from a task in which subjects decided when to cross an intersection as a car approached from the cross street. Results showed that subjects attempted to cross less often before the oncoming car when it started closer to the intersection, even though objective risk was the same regardless of starting distance. Also, when the car started closer, subjects who reported having more real-life automobile accidents were less likely to take advantage of a longer opportunity to cross first. These results, along with results from fitting DRIW to the data, suggest that risk-taking effects can be accounted for by a model of risk perception, and not by a model of risk acceptance.Two questions in research on risky real-time decision making are how a decision maker (1) integrates real-time information and (2) judges risk levels to decide on the nature and timing of risky real-time responses. For example, drivers are often faced with making maneuvers that have a small probability of producing an accident. However, there can be incentives to take such risks because of time pressures. The goal of the research presented here is to identify and model processes underlying such risk taking.Up until now, major theories have treated this issue as a problem of risk acceptance, but have lacked detailed mechanisms for assessing effects and individual differences that are due to risk perception. For example, models such as Risk Homeostasis Theory (RHT; Platenius, 1985) do not explain how, under specific conditions, information is processed and integrated to time risky responses. Lacking well-specified processing mechanisms, these theories cannot separate errors due to poor distance-and velocity-estimation abilities from those due to risk taking. Consider the case in which a driver decides when to cross an intersection. She may take some time to process information about oncoming cars, but if she takes too long, she might not arrive on time at her destination. If she crosses before sufficiently processing information about these cars, she might risk colliding with one of them. She may risk crossing quickly because they are far away, but there is a potential for error because of the tendency to underestimate distances for objects that are in motion relative to an observer (Harte, 1975) and are far from the observer (Foley, 1980). Given such errors, the driver may underestimate a car's speed and mistakenly assume that she can cross safely and quickly. The Model. The study presented here introduces a model of real-time decision making that incorporates these distortions, and accounts for how drivers trade off processing time and risk taking. This model, DecisionMaking under Risk in a Vehicle, or DRIVE, simulates how drivers make choices in real time when faced with incentives to take risks.According to DRIVE, a ...