Abstract
Perceptual decision-making has been long studied using two largely independent approaches. Threshold measurement is the predominant approach in psychophysics that excels at quantifying stimulus strength required for accurate perceptual decisions. In parallel, reaction time (RT) paradigms along with associated accumulation-to-bound models have been used to estimate components of perceptual decision making (e.g., decision time, non-decision time, and drift rate). It is unknown, however, whether both approaches yield the same conclusion about the sensitivity of the sensory system. To answer this question, we conducted two experiments (total n=23) where we estimated both RTs and duration thresholds for a motion and a static discrimination task. Duration threshold (DT) is defined as the shortest stimulus presentation duration sufficient to make accurate perceptual decisions. RTs and choices were fitted by a drift-diffusion model (DDM, Wiecki et al., 2013). If the DDM is correct, there should be a close relationship between DTs and drift rates, allowing us to accurately predict DTs from RT data. In the motion task (Newsome & Pare, 1988), we found a close correspondence between the empirical DTs and the DTs predicted by the DDM across 6 levels of motion coherence (10%-100%; r=0.81, p <0.0001). Surprisingly, in the static orientation discrimination task (8 contrast levels, 2%-100%), there was very little correlation between DTs and drift rates. While DTs, as expected, improved monotonically with increasing contrast, RT drift rates saturated at 6% stimulus contrast. In summary, we show a close correspondence between duration thresholds and RT drift rate for the well-established motion coherence task. This result supports the common conceptualization of drift rate as a proxy for perceptual sensitivity. However, we do not find the same correspondence in the static orientation discrimination task, indicating a surprising limitation of the DDM and experimental approaches based on it.