Abstract
Selection history has been known to influence behavioral performance independent of top-down-goals and bottom-up-saliency. However, the precise contribution of underlying processes like target facilitation and distractor inhibition for goal-directed reaching action is unclear. To address this, we used a reach-tracking and computational modelling approach. Participants reached an odd-colored target among three homogeneous distractors, while we recorded their continuous reach trajectory. Continuous trajectory provides a behavioral measure that captures the online dynamics between target and distractor competition as the target selection unfolds. To assess the relative contributions of target facilitation and distractor inhibition on reach target selection, in addition to conventional full-repetition (repeating target and distractor colors on subsequent trials) and full-swap conditions (swapping target and distractor colors on subsequent trials), we added the partial-repetition and partial-swap conditions. In the partial-repetition conditions, the target color remained unchanged, while the distractor was in new color, or vice versa; in the partial-swap conditions, target color was the distractor color of the preceding trial while distractor was in new color, or vice versa (Eimer et al., 2010). To evaluate target selection bias modulated by history, we compared the direction and the magnitude of reach curvature across conditions: a decreased reach curvature towards a distractor indicates facilitation of target selection. In accord with prior studies (Moher & Song, 2014), reach target selection was most facilitated in the full-repeat condition while most interrupted in the full-swap condition. We also observed that more increased reach curvature towards distractors in the partial-swap conditions than in the partial-repetition conditions. This result suggests that maintaining a feature in association with its role (as target or distractor) across trials is more beneficial for subsequent target selection. To quantify the dynamic interaction between target facilitation and distractor suppression, we are currently extending the neurologically inspired CoRLEGO (Choice-reaching-with-LEGO-arm-robot) model (Strauss et al., 2015).