Accurately estimating temporal discounting with bias-adjusted drift diffusion models

Drift diffusion models provide a powerful method of studying the dynamics of decision making in two-alternative forced choice tasks. Recently, these models have been extended to the study of value-based decision making, including intertemporal choice. While this is a promising development, the current paper demonstrates how, under the existing formulation, drift diffusion models of intertemporal choice systematically misestimate temporal discounting due to the bias parameter included in all drift diffusion models. A bias-adjusted formulation is proposed and shown to produce more accurate estimates of temporal discounting by compensating for the bias parameter. This bias adjustment is implemented in the open-source tempodisco R package and may allow intertemporal choice research to better harness the benefits of drift diffusion models.