The prevailing belief in cognitive science considers agents intelligent to the degree that they apply their knowledge to make rational choices. But fifty years of research have yielded robust phenomena in which human choice seems to violate norms of rationality. Today, the biases and shortcomings of human decision-making are standard topics in such disciples as management science, public health, and economics. As awareness has increased, so have calls to develop interventions to improve decision-making.

To answer these calls, our project adopts a rigorous and empirical approach with two main objectives. First, we aim to research training interventions designed to improve choice among options with difficult tradeoffs. We target this class of decisions because of its large size and practical implications, and because it yields a cross-species phenomenon called preference reversal, which is thought to contradict axioms of rationality. Second, we aim to advance the science of bounded rationality by testing a new model of choice. Contrary to the popular view, this model shows that preference reversals arise from rational processes that are adapted to cognitive bounds and maximize utility.

The project will achieve these aims through careful computational modeling and human empirical work. By scrutinizing the view of choice as a rational process, it will produce new insights into the science and practice of decision-making.