Human visual attention is exquisitely specialized for and captured by social information in naturalistic scenes, and this “social preference” begins as early as infancy. Recent visual perception and neuroimaging research suggest that people also have high visual interest in interacting dyads compared to non-interactors, supporting the idea that social interactions might provide unique social information, above and beyond the mere presence of two people. Thus, interactions might capture visual attention above other social information, a ‘preference’ that could contribute, across development, to social learning processes. However, very little work has directly examined if attention is more biased to social interactions than to other social information in complex scenes, nor how and when such a bias may arise or change across development.
In this work, across three free-viewing eye-tracking experiments, we investigate the development of attention to social information in naturalistic, complex scenarios, in order to better understand the role of social interactions in such processes. In experiment 1, scenes contained dyads who were either interacting or not, while in experiment 2, dyads were presented together with one or two additional non-interactors to evaluate developmental changes in attention to social information when social interactions compete for attention with other social information. In experiment 3, the pictures contained ambiguous social dyads and, after the free viewing session, participants were asked to indicate whether they perceived the two agents as interacting or not. Our aim in that experiment was to investigate age differences in the way pre-existing knowledge about social events (e.g. the concept of what it means to be a social interaction) might influence social attention.
In all three experiments, we compared attentional engagement and capture by social areas of interest (i.e., human information) with non-social information (other scene elements), further contrasted by whether scenes were interactive or not. Results revealed both children and adults manifested a strong human attentional bias in the first two experiments, but a weaker bias in the third, when the social information was ambiguous. This social bias towards human information was moderated by the presence of a social interaction only in the first experiment, but not the second or the third, and was moderated in a similar way across development. In experiment 2, when interacting people were contrasted with non-interacting people in the same scene, interacting people capture attention more quickly and engage it more strongly than other social targets when there is one other agent in the scene for both adults and children. However, this effect is smaller (and not significant) in children than for adults when an interaction competes with a pair of non-interactors. This suggests interactions can take attentional priority, but that this “interaction bias” increases across development, especially as scenes become more crowded and complex. Finally in the third experiment, we find that adults were more likely to see ambiguous scenarios as interactive compared to children. However, this difference was not reflected in the way attention was oriented to social information, as the social bias was similar in the scenes categorized as interactive or not and not different across development.
The results in this thesis are consistent with the idea that social interactions carry additional information, compared to isolated humans, and even more importantly, that this ‘bias’ to attend to social interactions is present as early as six. Finally, the implications of these findings for social attention and social development are discussed, followed by a discussion of future theoretical and experimental questions left to explore.