Proceedings of the Annual Meeting of the Cognitive Science Society

Attentional set shifting has been characterised using accuracy- or performance-based and reaction time–based shift costs across the literature, where "shift cost" is often treated as a unitary construct despite differences in how it is defined and measured. Using a GO/NOGO task with repeated intra- and extra-dimensional shifts between auditory and visual targets, we examined response (accuracy and reaction time, RT) dynamics for cross-modal shifts, using task difficulty to manipulate attentional load prior to shifts. Our prior work using this paradigm demonstrated that increased attentional demands prolong learning (to perform at sustained high accuracy) following modality shifts, raising the question: do RT-based shift costs show a similar sensitivity to task difficulty? The results demonstrated that RT-based shift costs were not increased following difficult tasks, but instead, reflected established sensory modality dominance effects. However, RT dynamics are not independent of learning: longer pre-shift RT predicted higher subsequent learning shift costs, revealing an interaction between processing time (attentional demand) and task difficulty in shaping learning.