Why do people reject mixed gambles?

AbstractRejection of low-stake positive-expected-value gambles has been traditionally attributed to loss aversion (higher utility weights for losses than gains). This paper considers an additional cognitive mechanism: A predecisional bias towards rejection without the calculation of utility. Fitting drift diffusion models to experiment datasets, we show that this bias provides the largest quantitative contribution to model fits. Additionally, it successfully predicts novel choice-decision time patterns observed in our datasets. We show that high rejection rates for low-stake positive-EV gambles can be a result of multiple different cognitive mechanisms, and that a predecisional bias is the primary determinant of this behavioral tendency.

Return to previous page