Audiovisual detection at different intensities and delays

In divided-attention tasks with two classes of target stimuli (e.g., auditory and visual), redundancy gains are typically observed if both targets are presented simultaneously, as compared with single-target presentation. Different models explain such redundant-signals effects, including race and coactivation models. Here we generalize one such coactivation model, the superposition model, and show that restricting this evidence accumulation process to a limited time interval, that is a temporal deadline, provides the model the ability to describe the detection behavior of two monkeys across a wide range of intensities and stimulus onset asynchronies. We present closed-form solutions for the mean absorption times and probabilities for the two-stage diffusion process with drift towards a single barrier in the presence of a temporal deadline.