Word Count = 1264Running Head: Failure to detect dynamic scene changes due to endogenous control.
ABSTRACT [148 out of 150]Change blindness is a failure to detect changes if the change occurs during a mask or distraction. Without distraction, it is assumed that the visual transients associated with the change will automatically capture attention (exogenous control) leading to detection. However, visual transients are a defining feature of naturalistic dynamic scenes. Are artificial distractions needed to hide changes to a dynamic scene? Do the temporal demands of the scene instead lead to greater endogenous control that may result in viewers missing a change in plain sight? In the present study we pitted endogenous and exogenous factors against each other during a card trick. Complete change blindness was demonstrated even when a salient highlight was inserted coincident with the change. These results indicate strong endogenous control of attention during dynamic scene viewing and its ability to override exogenous influences even when it is to the detriment of accurate scene representation.
4What we remember of a visual scene is a result of where we attend and which details of attended locations we encode in memory. One of the most striking demonstrations of the interaction between attention and memory is change blindness. Change blindness is the failure to detect an obvious change to a scene if the change occurs is relevant with the demands of the stimulus. Endogenous control has been shown to limit capture by stimulus features in simple displays (Folk, Remington & Johnson, 1992). Does a similar tempering of exogenous control occur during dynamic scenes and can it be used to hide a change in plain sight?In the present study we asked whether viewers can be made blind to a change in a dynamic scene through endogenous control and whether this can be overridden 5 by exogenous capture of attention. In our study viewers watched a video of a simple card counting task (supplementary video 1). (We encourage readers to view the video now before reading on.) The video depicted a man's hands as he unpacked a deck of blue-backed cards and then dealt them face up on the table. The audio narration instructed viewers to "count exactly how many red cards are dealt". In the reveal, the backs of all of the cards were shown to have changed color from blue to red. It is only at this point that participants realise they have been watching a card trick. The secret behind the colour change was simple: Only the first few cards had blue backs, all the rest had red backs, and it is these that the dealer turns over at the end (Figure 1). The critical feature (the change from blue to red; Figure 1a to 1b) was in clear view, was task-relevant (participants were counting "red" faced-cards,) and only 3.4 degrees of visual angle from the attended cards.In Experiment 1, we showed fifteen participants the video whilst their eye movements were recorded (with an Eyelink 1000). After the video, participants were In summary, 45 out of 45 particip...