Abstract
When searching, people usually scan the visual field by making saccades. From the literature we know that saccade amplitude and fixation duration are related to the amount of information to be processed during fixations. Contrast borders may affect target detectability due to lateral masking. We hypothesize that the oculomotor system uses both target-background and search area border information to adjust fixation duration and saccade amplitude, enabling search to be effective.
To test the effect of the border of the search area we designed a search stimulus in which border contrast was varied to manipulate the difficulty of the search task. Target-background contrast was kept constant. The stimulus was a target (a barely visible light gray dot) placed in a further empty search area (a randomly oriented homogenously gray triangle). The target was present in 50% of the trials. Question: How does border contrast affect search performance and saccade behavior?
Results. Percentage correct was lowest en reaction times were longest for search areas having borders of high contrast. This result indicates that high-contrast borders mask stronger than low-contrast borders. The majority of the errors were misses. Fixation density was highest near borders. Surprisingly, high fixation density was not associated with small saccades amplitudes. Near borders, saccades were large and mainly directed in parallel with the borders. High fixation densities near borders were mainly caused by re-inspections. Fixation time was nearly constant over the search area. Apparently, the oculomotor system does not adjust fixation duration and saccade amplitude to compensate for increased search task difficulty near borders, resulting in lower search performance (missed targets).
Meeting abstract presented at VSS 2012