Change blindness. Influence of scene content and emotional valence on change detection performance in clinical and not clinical children

Change-blindness (CB) occurs when large changes are missed under natural viewing conditions because they occur simultaneously with a brief visual disruption, like an eye movement, a blank screen, an ocular blink, or a camera cut in a film sequence. In the typical CB paradigm, the flicker task, pictures of daily life scenes are used to assess visual search efficiency (Rensink, 2000). Two versions of a picture are presented. The pictures are identical except in a specific detail. The pictures alternate repeatedly and are separated by a brief gray screen. The observers have to search the scene for what has changed between the two pictures until they detect it. As the task uses pictures of natural scenes, participants tend to give priority to some areas of the scene than to others. Consequently, they detect changes in objects of central interest (CI) faster than changes in marginal interest (MI) objects (Rensink et al. 1997). Both perceptual and semantic characteristics of the visual scene might be taken to create a sort of priority list that determines which items are going to be attended first. Changes in objects of CI pop out from the pictures, and they are usually efficiently detected. Changes in objects of MI are more difficult to detect and require serial visual search, and therefore performance is less efficient. This study aims to evaluate the visual search ability in tipically developing children and in children with psychiatric disorders. In Experiment 1, 52 healthy children and 22 adults executed the flicker task with IAPS emotional pictures. The valence of the images interferes with the attentional performance showing slower RT in detecting changes in front of emotional images. In particular, the negative pictures interfere more the CI detection, whereas the positive ones interfere more the MI detection. The results are discussed in term of biological VS motivational aspects. We hypothesize that the evolutionary role of the negative stimuli makes to interfere the attentional performance during the more automatic CI change. In contrast, the positive images interfere it through a more voluntary mechanism; the likeness of the stimuli makes the partecipants look at the picture rather than search for the change. In experiment 2, 14 children with a disorder of the autistic spectrum and 14 controls executed the same task of Experiment 1. Our results show that autistic children are slower than controls only in the MI detection. Furthermore, they differ from controls when detecting CI changes in front of negative picture and when detecting MI changes in front of positive pictures. Our results confirm the Fletcher-Watson et al’s results suggesting an impairment in the disengagement from the most rilevant items or, in alternative, in the orienting of attention. In Experiment 3, 75 healthy children and 19 adults executed the flicker task with pictures rated as appealing or unappealing. The appealingness of the images interferes with the attentional performance showing slower RT only when detecting MI change. The results are discussed in term of motivational aspects. We hypothesize that the appealingness of the stimuli interfere with the attentional performance obstructing the execution of the task. In experiment 4, 18 children with ADHD and 18 controls executed the same task of Experiment 1. Our results show that children with are overall slower than controls. Furthermore, they specifically differ from controls when detecting MI changes. The slower RT and the poor accuracy of children with ADHD on the highest demanding condition (e.g., detection of MI changes) is consistent with a deficit in attentional resources, or with a specific impairment in using serial top-down strategies due to their limited attentional resources. Overall, our results of the present study replicate the findings consistently observed with the flicker task (Fletcher-Watson, Collis, Findlay, & Leekam, 2009), demonstrating the robust nature of change blindness. All the children showed a strong change blindness effect and a clear difference between CI and MI trials.