An eye tracking index for the salience estimation in visual stimuli

Every day we face visual stimuli able to catch our attention, but this aspect becomes crucial if the visual material has the purpose to spread a message aimed at engaging the observer. In this framework, a worthy aspect is how to measure the “visual engagement” produced by visual stimuli exposure. To this purpose, in the present study, employing the eye tracking technique, an index of visual attention (VA) has been proposed, and applied to pictures belonging to antismoking public service announcements, so to investigate the saliency of health-promoting messages in a young sample. The VA index is a non-dimensional index, defined as the ratio between the percentage of the total time spent fixating an area of interest (AOI) weighted on the total time the picture is showed on the screen, and the percentage of the area occupied by the AOI weighted on the total dimension of the picture. It could be predicted that AOI reporting higher VA values will be the ones having more saliency. Three antismoking Public Service Announcements (PSAs) images have been selected for the study and for each of them were identified: i) “picture” (such as a young man with a sarcastic expression depicted while smoking a cigarette, or the image of a lady who underwent a tracheotomy) and ii) “writing” (text of the antismoking message) AOIs. Main results of the analysis revealed that writing AOIs obtained statistically significant higher VA values than visual AOIs (p=0.03), but these held true only for an ineffective PSA, probably because the text was not perceived as pertinent with the surrounding image. On the other hand, an effective PSA obtained higher VA values in response to visual than writing AOIs observation (p=0.02). The VA index appears therefore to represent a useful tool to measure the saliency of visual stimuli elements.