Research

Three bisection experiments show that the mean and variance of the whole duration set, not the short and long standards, anchor temporal comparisons, as the new ensemble-distribution account predicts.

fMRI shows that learning where salient distractors tend to appear dampens early visual cortex responses at those locations, with the link to behavioural interference holding only for distractors …

Repeated visual distractor layouts can cue the location of a tactile target, but only when the tactile target arrives 250 to 450 ms before the visual context, revealing a novel visuo-tactile …

Learning repeated search layouts with co-located tactile vibrations strengthens contextual cueing, even on trials shown to the eyes alone; drift-diffusion modeling traces the boost to faster evidence …

Learned suppression of a frequent distractor location also slows targets appearing there, but this cost fades with practice once targets are equally likely at that location, pointing to a flexible …

With Stefan Glasauer we reread Vierordt (1868) and found that his randomized interval order, not timing itself, produces the classic central tendency; a random-walk sequence yields far less bias.

A task-irrelevant auditory sequence, regular or irregular, shifts whether observers see group or element Ternus motion, and it does so through the geometric mean of its intervals, consistent with …

Modeling reaction times from three pop-out search experiments shows that response history shifts the starting decision bias, whereas repeating the target dimension speeds evidence accumulation, …