Abstract
In a comparative visual search experiment (Pomplun, Sichelschmidt, Wagner, Clermont, Rickheit, & Ritter, 2001), two halves of a display (12, 15, or 18 items per half) contained visual primitives of various shapes (circles, triangles, squares) and colors (green, blue, red). On 50% of trials, a single primitive differed between the two halves in form, color, or form/color; on the other 50% of trials, the two halves were identical. The observers' task was to make a yes/no decision to the presence of the different primitive. Eye movements, response latency, and response accuracy were measured as a function of both display size and difference type.
Display size affected responses: As display size increased, hit rate decreased. Also, with increases in display size, response latency increased in proportion to the increase in the number of fixations, but mean fixation duration was constant. Display-size effects may be described in terms of limitations on VSTM.
Difference type (form, color, or form/color) also affected hit rate, but did not affect numbers or mean durations of fixations. Detection of form mismatch was more difficult than detection of color mismatch, and there was a hit-rate redundancy gain at all display sizes: Form/shape differences yielded higher hit rates than differences in only shape or only color. The redundancy gain may be described by a horse-race model between searches for form and color differences.