Jeunghwan Choi¹ (), Sang Chul Chong^1,2; ¹Graduate Program in Cognitive Science, Yonsei University, ²Department of Psychology, Yonsei University

People can find a target quickly using the regularity acquired from prior search experiences. For instance, previous studies showed that repeated spatial arrangements (Chun & Jiang, 1998) and spatial probabilities of a target (Geng & Behrman, 2005) could facilitate a visual search. These studies found that people could learn cues which facilitated visual search processes. In this study, we tested whether people could learn and use the task irrelevant regularity of an ensemble property to enhance a visual search performance. The stimuli consisted of 40 differently sized circles filled with oriented sinewave gratings. The target was an outlier tilted close to vertical and was tilted either clockwise or counterclockwise from the vertical orientation. The distractors were tilted close to horizontal. The task was to find the target and to report its orientation. We manipulated the average size of the stimuli to predict the location of a target. Specifically in the training phase, the hemifield of the target location was determined by the average size of presented stimuli. For example, when the average size of presented stimuli was larger than the mean size of possible stimuli size range, the target always located within a left hemifield. When the average size of presented stimuli was smaller than the mean size of possible stimuli size range, the target always located within a right hemifield. The participants did not know about these regularities. In the subsequent test phase, there were two conditions: a congruent and incongruent conditions depending on the consistency of the learned relationship between the target locations and the average sizes. We found that participants found a target more quickly in the congruent condition than in the incongruent condition. This result suggests that people can learn and use average sizes as a cue to the target location.

Acknowledgements: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2022R1A2C3004133).