Purchase this article with an account.
Daniel Brooks, Ji Dai, David Sheinberg; Scene-based Contextual Cueing in the Rhesus Macaque. Journal of Vision 2011;11(11):1329. doi: 10.1167/11.11.1329.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The neurophysiology that underlies the guidance of attention during search through naturalistic scenes remains largely unexplored. In order to understand how scene identification leads to the prioritization of certain regions of space, we have developed an animal analogue of scene-based contextual cueing in the monkey so that we might conduct physiological experiments.
Our task was adapted from the contextual cueing task of Brockmole and Henderson (2006). The monkey must search for a semi-transparent object embedded in a real-world scene. The monkey has previous experience classifying these objects into one of two arbitrary categories by pressing response buttons. In half of the scenes (non-predictive condition), objects appeared randomly in one of eight locations. In the other half of the scenes (predictive condition), scene identity cued the location of the object through consistent scene-location pairings. Along with manual reaction time, we also tracked the monkey's eye-position in order to examine how the monkey searched through each scene and how many saccades it took to locate the target.
As in humans, the monkey showed a benefit when executing searches through repeated scenes across multiple sessions, demonstrating sensitivity to scene familiarity. The monkey also showed a significant reaction time benefit when searching through predictive scenes when compared to non-predictive scenes, demonstrating scene contextual cueing. Similar to the associations formed by human participants in Chun & Jiang (2003), these scene-location associations were stable after a week-long hiatus, indicating the long-term nature of the associations. Finally, we conducted target-absent tests during which we monitored eye-movements. In these tests, the monkey saccaded to the cued location in an average of less than two saccades and did so on greater than 90% of trials.
These results demonstrate scene-based contextual cueing in the macaque, showing promise as a method for uncovering the physiological basis for attentional guidance during active search.
This PDF is available to Subscribers Only