Cultural and Developmental Influences on Overt Visual Attention to Videos

Omid Kardan,Laura Shneidman,Sheila Krogh-Jespersen,Marc G Berman,Amanda Woodward,Suzanne Gaskins

doi:10.1038/s41598-017-11570-w

Abstract

Top-down influences on observers’ overt attention and how they interact with the features of the visual environment have been extensively investigated, but the cultural and developmental aspects of these modulations have been understudied. In this study we investigated these effects for US and Yucatec Mayan infants, children, and adults. Mayan and US participants viewed videos of two actors performing daily Mayan and US tasks in the foreground and the background while their eyes were tracked. Our region of interest analysis showed that viewers from the US looked significantly less at the foreground activity and spent more time attending to the ‘contextual’ information (static background) compared to Mayans. To investigate how and what visual features of videos were attended to in a comprehensive manner, we used multivariate methods which showed that visual features are attended to differentially by each culture. Additionally, we found that Mayan and US infants utilize the same eye-movement patterns in which fixation duration and saccade amplitude are altered in response to the visual stimuli independently. However, a bifurcation happens by age 6, at which US participants diverge and engage in eye-movement patterns where fixation durations and saccade amplitudes are altered simultaneously.

Highlights

On the one hand researchers have made strong assumptions on the universality of basic visual attentional mechanisms
One of the proposed reasons for this difference coming from developmental psychology has been that Mayan individuals demonstrate patterns of visual attention that are based on their regular experience as 3rd party observers during their childhood[12, 22,23,24,25]
If saccade amplitudes and fixation durations co-vary together in important and different ways based on age or culture, those differences will be missed with the univariate analyses, as those analyses are unable to detect

Summary

Introduction

On the one hand researchers have made strong assumptions on the universality of basic visual attentional mechanisms (e.g. refs 9 and 10). G., colors) and high- level features (e.g., actions) in the visual environment guide the allocation of attention as measured by eye-movement behavior. We investigated whether these differences followed the patterns of attention allocation that have been described at more global levels of analysis We assessed these patterns of attention across development since bifurcations due to cultural experiences are likely to develop over time. One of the proposed reasons for this difference coming from developmental psychology has been that Mayan individuals demonstrate patterns of visual attention that are based on their regular experience as 3rd party observers during their childhood[12, 22,23,24,25]. Third-party attention is relevant for indicating what visual content might be more salient for Mayan participants compared to their US peers. Does one high priority region receive a long fixation, and the region receives another long fixation, or are the fixations rapidly exchanged between the two? The hummingbird flight pattern hypothesis is relevant for this, since it implies larger/faster saccadic eye-movements and a higher frequency of shifting fixation points for Mayans when compared to the US viewers

Methods

Results

Conclusion