Abstract
Humans often select multiple objects in dynamic environments that require different modes of response (e.g., while driving, look at the clock, reach to the radio, etc.). Selection is biased towards recently attended target features; for example, participants respond to a unique red target faster if the previous target was also red (i.e., priming of popout). However, It is unknown whether the representation of this feature in memory is bound to a specific action, or whether a more broad representation of the feature alone, without a specific action association, biases selection. In a series of experiments, we examined this question by investigating the transfer of priming across different response modalities or even no overt response. In Experiment 1, cues instructed participants to respond to a uniquely colored target either by reaching to that target or by pressing a key based on the target’s properties. Repeating target color on consecutive trials reduced keyboard response time, initial reach latency, and reach curvature, even when the response mode switched from one trial to the next (e.g. from a keyboard response to a reach response). In Experiment 2, target color repetition benefits occurred even following "no-go" trials in which observers had to withhold a keyboard response. However, while initial latency of reach movements was speeded following "go" trials, "no-go" trials biased selection only at the later movement stage by reducing reach curvature. Furthermore, in both experiments, the magnitude of the repetition benefit was diminished when the response mode switched. Together, these studies lend new insight into how memory biases selection and action, suggesting the existence of a high-level mechanism that biases target selection towards previous targets regardless of the previous response modality. However, the previous action does influence the magnitude of color repetition benefits, suggesting that it is represented at some level.
Meeting abstract presented at VSS 2013