How the visual environment shapes attention: The role of context in attention guidance
In our environment, visual stimuli typically appear within the context of other stimuli, which are usually not arranged randomly but follow regularities. These regularities can be very useful for the visual system to overcome the problem of limited encoding capacity by guiding attention to stimuli w...
|PDF Full Text
No Tags, Be the first to tag this record!
|In our environment, visual stimuli typically appear within the context of other stimuli, which are usually not arranged randomly but follow regularities. These regularities can be very useful for the visual system to overcome the problem of limited encoding capacity by guiding attention to stimuli which are relevant for behavior. There is growing evidence that observers use repeated contexts for guiding attention in visual search, and there is evidence that observers adapt to dynamical changes in their visual environment. However, contexts in our natural environment often come with features predicting reward, and little is known about the influence of such reward-predicting contexts on attention guidance. In addition, it is unclear how observers adapt their behavior to context features that are not relevant for the task, and little is known about individual differences in the effects of contexts. These research gaps are addressed in the present dissertation. In five studies, the present dissertation investigates how different types of contextual regularities are integrated into behavior and how these regularities guide visual attention.
Study I showed that observers use knowledge they have acquired in former encounters with similar scenes to predict the most promising item to attend to in an upcoming scene. In a visual search task in the laboratory, participants responded faster in visual contexts that repeated compared to contexts that were novel. In addition, they also moved their eyes more efficiently to the target when they encountered repeated contexts. These results suggest that participants use repeated visual contexts to learn to predict the target location.
Study I also revealed that visual contexts are especially used for specifying promising items when they predict a high reward. Context features predicting a high reward boosted the performance advantages observed with repeated contexts. This result suggests that the prediction of reward facilitates the generation of expectations about potential target locations.
Study II demonstrated that expectations about potential target locations were quite persistent, since performance benefits were observed even after many encounters with repeated contexts. Further experiments showed that participants could use even a very limited part of the visual contexts to learn to predict the target location (Study III) and that observers used also contexts that changed dynamically for specifying promising items to attend to (Study IV). These results suggest that observers use regularities in the visual context to generate expectations about promising items in their visual environment.
Finally, the last study of this dissertation (Study V) investigated how contexts of social perception are used for specifying relevant visual information. Results showed that observers differ in how they use contexts for specifying relevant visual information and suggested that an observer’s personality might be one factor explaining these differences.
In sum, the five studies of the present dissertation demonstrate that the visual system is remarkably sensitive to regularities in the visual context. It is quite efficient in extracting repeated contexts to guide attention to relevant locations when contexts are encountered again (Studies I and II), and it only needs a very limited amount of repeating contextual information to take advantage from the contexts (Study III). It also considers rewards that are signaled by features of the contexts to prioritize processing of high reward contexts. The visual system further adapts to dynamical changes in the contexts (Study IV) and uses contexts of social perception for prioritizing information, dependent on the observer’s personality (Study V). The present dissertation thus highlights that the visual context is crucial for guiding our attention in numerous situations that we encounter every day. Fortunately, we can take advantage of the visual context, which allows our visual system to cope with its limited processing capacity.