Abstract
When people look through the environment their eyes are guided in part by what they have recently seen. This phenomenon, referred to as visual priming, is studied in the laboratory through manipulations of stimulus repetition. Typically, in search tasks, response times are speeded when the same target is repeated relative to when it is changed (e.g., Maljkovic and Nakayama, 1994). Although priming is thought to be based on a memory mechanism in the visual system, there is a debate in the literature as to whether such a mechanism is driven by relatively early (e.g., feature-based accounts) or later (e.g., episodic memory accounts) processing. Across three experiments, this study utilized a computational modeling framework (Systems Factorial Technology; Townsend and Nozawa, 1995) to directly compare early and later accounts of priming and determine when visual priming influences processing within the visual system in both feature and conjunctive search tasks. Specifically, the processing of priming was assessed in terms of its temporal relation (i.e., parallel or serial) to a relatively early process (conspicuity) and a relatively later process (Rewards, Experiment 1a; Word Cues, Experiments 1b and 2) in the visual system. Results suggest that priming is processed in parallel with conspicuity and word cues within both singleton (Experiments 1a and 1b) and conjunctive (Experiment 2) search. This supports accounts of priming as an early process and suggest that models of priming as a later process within feature or conjunctive search should be rejected. Further, these results also provide evidence to suggest word cues are processed at early stages of visual processing. This supports models of visual processing that suggest high-level representations can modulate the earliest levels of the visual system. Together, these findings provide some of the strongest evidence about the temporal processing characteristics of priming to date.
Meeting abstract presented at VSS 2017