Abstract
Performance in stop-change tasks has been modeled with both interactive and independent race models: the accumulator related to the first goal (GO1) either influences (interactive) or has no impact on (independent) the processes corresponding to the second goal (i.e., STOP and GO2). Here, to lend credence towards either model, we aimed to assess whether variations in GO1 influence STOP and GO2 through manipulating primary target contrast. In the present study, we recorded eye movements in two search-step experiments. Following the appearance of a luminance pop-out target amongst three distractors, participants were instructed to make an eye movement towards either the darkest (Experiment #1) or brightest (Experiment #2) stimulus as quickly as possible. In both step and no-step trials, the primary target (GO1) had variable contrast: either low (5%), medium (30%), or high (90%). However, on step-trials, following a random delay (83- 200 ms), the target changed position (GO2) and was set to low contrast. Primary target contrast significantly modulated saccade latency in no-step trials and in step-trials, non-compensated saccade latency. Conversely, this manipulation did not significantly change compensated saccade latency. These findings support the existence of an independent architecture. The changes in no-step and non-compensated saccade latencies (GO1) are predicted by either model via faster accumulation; however, compensated saccade latencies are expected to increase with contrast in the interactive model via increased inhibition from GO1 on GO2. The estimated target step reaction times also discounted a GO1-STOP interaction: they decreased with increases in primary target contrast. In conclusion, the current paradigm provides a method to test predictions regarding the alteration of race model parameters.
Meeting abstract presented at VSS 2017