Abstract
The deployment of visual attention to faces has been studied extensively in traditional lab-based experiments over the past thirty years. Despite the breadth of knowledge gained from these studies, little is known about how people behave outside the laboratory in this domain. For example, are faces attended to in the same way during complex real-world interactions as in lab-based studies, where faces are normally static, isolated, and devoid of social context? Here, we recorded fixations made by five observers taking part in a real transaction in a social setting (e.g., buying a coffee or snack), as part of a mobile eye-tracking task on the UBC campus (Foulsham, Walker and Kingstone, 2011). We analysed fixations to the five most relevant faces, defined by proximity, distinctiveness from the background (i.e., where crowds were viewed) and task relevance, for each recording. In a second experiment, we compared these to fixations made by 16 new observers watching videos of the transactions in a lab setting. Fixations to relevant face stimuli were more frequent in lab-based viewing conditions (17%), than in the real-world condition (13%). Moreover in lab-based viewing, there was a strong bias (47% of fixations) towards the faces of those most important to the transaction (e.g., baristas or shop assistants) compared to bystanders within the scene (25%, for real world observers). This study is among the first to investigate attention to faces in realistic situations, and the findings are explored with reference to conventions of social interaction and personality traits/interactional dispositions of the lab-based observers.
Meeting abstract presented at VSS 2012