Abstract
Despite the common belief that stronger task-irrelevant signals more severely disrupt task performance, we previously found that task-irrelevant coherent motion just below the chance-level detection threshold (subliminal threshold) more severely disrupts a task performance than supra-threshold motion signals.
One possible explanation for this paradoxical result is that eye movements induced by task-irrelevant translating coherent motion disrupts the task performance. In order to examine this possibility, 15 subjects were instructed to perform an RSVP (Rapid Serial Visual Presentation) task at the center of the display while ignoring the background motion that contained contracting, instead of translating, dots. The ratio of contracting dots to randomly moving dots (contraction coherence ratio) was varied from trial to trial. After this main experiment, the subliminal threshold of contraction (2AFC) was measured with the same subjects. The results showed that the RSVP task performance was significantly lower when the contraction coherency ratio was just below the subliminal threshold than when it was supra-threshold. Thus, the worse performance just below the threshold cannot be attributed to eye movements. This phenomenon is possibly explained as follows: The attention system usually suppresses irrelevant signals such as coherent motion to minimize their effect on task-performance. However, the attention system fails to detect and, therefore, to suppress irrelevant signals whose strength is below the subliminal threshold. This allows the subliminal irrelevant signals to disrupt the task performance more severely than supraliminal irrelevant signals that are suppressed by the attention system.
Supported by NSF (BCS-9905914), NIH (R01EY015980-01), Human Frontier Science Program (RGP18/2004), and NSF (CELEST)