Temporal learning and rhythmic responding: no reduction in the proportion easy effect with variable response-stimulus intervals

The present report further investigates the proportion easy effect, a conflict-free version of the proportion congruent effect. In the proportion easy paradigm, it is observed that the difference in performance between easy (high contrast) and hard (low contrast) items is smaller in a task with mostly hard items relative to a task with mostly easy items. This effect has been interpreted as evidence for temporal learning: participants learn a faster pace (i.e., rhythm of responding) in the mostly easy context, which boosts the contrast effect, and a slower pace in the mostly hard context. In the present experiment, intervals between trials were either fixed or randomly varied from trial to trial. Interestingly, the proportion easy effect was still present with variable intervals. These data suggest that participants do not learn the regularity in timing from one response to the next (which was highly inconsistent with variable intervals). As one alternative, participants might be learning the intervals between stimulus onset and responses, which were not manipulated. They could then use this learned timing information to prepare for responding at the anticipated time, resulting in rhythmic responding. The results further imply that variable response-stimulus intervals are insufficient for controlling for rhythmic biases