Sensorimotor decisions rely on the entanglement of evidence and motor accumulation processes

Abstract

Most contemporary theories of sensorimotor decision-making formalize the process leading up to a decision as a gradual accumulation of noisy stimulus information over time. The resulting evidence signal is thought to be continuously tested against an internal criterion representing the amount of evidence required to make the decision. In the currently prevailing view, the amount of accumulated evidence required for a decision is independent of the amount of sensory evidence presented by the stimulus, and once that level is reached, a choice is categorically communicated to the motor system to execute an overt response. Recent experimental research casts doubts on both of these assumptions. Using a Leaky Integrating Threshold model, we relax these assumptions specifying both an evidence accumulation and a motor accumulation process. The evidence accumulation signal feeds into a leaky motor accumulator, and it is on the level of the motor accumulation that the final decision criterion is set. This adaptation results in a markedly better description of choice-RT data, especially when it comes to urgency manipulations. We show that this alternative theory, which proposes that sensory evidence is doubly integrated before final evaluation, does not only describe the behavioral data better, but its neural correlates can also be readily derived from EEG signatures involving systems of both evidence and motor accumulation.