Culture modulates eye-movements to visual novelty

Abstract

Background: When viewing complex scenes, East Asians attend more to contexts whereas Westerners attend more to objects, reflecting cultural differences in holistic and analytic visual processing styles respectively. This eye-tracking study investigated more specific mechanisms and the robustness of these cultural biases in visual processing when salient changes in the objects and backgrounds occur in complex pictures.

Methodology/principal findings: Chinese Singaporean (East Asian) and Caucasian US (Western) participants passively viewed pictures containing selectively changing objects and background scenes that strongly captured participants' attention in a data-driven manner. We found that although participants from both groups responded to object changes in the pictures, there was still evidence for cultural divergence in eye-movements. The number of object fixations in the US participants was more affected by object change than in the Singapore participants. Additionally, despite the picture manipulations, US participants consistently maintained longer durations for both object and background fixations, with eye-movements that generally remained within the focal objects. In contrast, Singapore participants had shorter fixation durations with eye-movements that alternated more between objects and backgrounds.

Conclusions/significance: The results demonstrate a robust cultural bias in visual processing even when external stimuli draw attention in an opposite manner to the cultural bias. These findings also extend previous studies by revealing more specific, but consistent, effects of culture on the different aspects of visual attention as measured by fixation duration, number of fixations, and saccades between objects and backgrounds.

Figure 1. Schematic of the experimental design showing example stimuli from the four picture quartet conditions (left) and the display timings (right).

Figure 2. Mean object fixation data for US and Singapore participants.

Mean object fixation durations (a) and number of object fixations (b) are shown across all quartet conditions (Old and New objects/backgrounds) and picture repetitions (R0 to R3).

Figure 3. Mean background fixation data for US and Singapore participants.

Mean background fixation durations (a) and number of background fixations (b) are shown across all quartet conditions (Old and New objects/backgrounds) and picture repetitions (R0 to R3).

Figure 4. Eye gaze data for US and Singapore participants.

Mean gaze distance covered within pictures (a) and proportion of gaze saccades between objects and backgrounds (b) are shown across all quartet conditions and picture repetitions.