Visual search asymmetry for angular figures

Abstract

The well-described phenomenon of preference for curved shapes could be explained by the fluency hypothesis, which states that curvature enables their fluent processing in the visual field. According to this hypothesis, curved figures should be faster recognized among angular, while the opposite would not be the case. In order to test the fluency hypothesis, we used a classical visual search task, with two types of the target (curved and angular), three set sizes (1, 6, 12 elements), and two set types (positive and negative). In addition, we made three experimental designs: (1) classical, where the complexity of curved and angular figures was not controlled, (2) matched, where curved and angular figures were matched by the number of changes in shape, and (3) symmetrical, consisting curved and angular figures balanced by participants’ subjective ratings of complexity. Reaction time was recorded. 31 participants took part in the experiment. The results of Generalized estimating equations analysis revealed search asymmetry for curved stimulus in classical design (Wald χ2(1)=14.96; p<.001), while the angular figure is faster recognized in both matched (Wald χ2(1)=9.54; p<.01) and symmetrical design (Wald χ2(1) 218 Perception 50(1S) =86.66; p<.001). This difference was particularly noticeable in negative and large sets. Our results indicate that the hypothesis of fluency could be rejected when curved and angular stimuli are equally complex, and that angularity should be considered as a basic stimuli feature which pops out in the visual field.