Purchase this article with an account.
Nicola Anderson, Kaitlin Laidlaw, Walter Bischof, Alan Kingstone; Recurrence Quantification Analysis of Scan Patterns. Journal of Vision 2012;12(9):544. doi: 10.1167/12.9.544.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recently, much research has been devoted to understanding how people look at scenes. However, there is relatively little research and analysis of the patterns of sequential fixations that occur during the time-course of scene viewing. In the present work we use categorical recurrence quantification analysis as a method for examining scan patterns (Zbilut, Giuliani & Webber, 1998; Dale & Spivey, 2005). This analysis allows for the visualization and quantification of the temporal dynamics of complex systems. Here, we apply it at the level of fixations in order to examine sequences of fixation patterns that recur (i.e. repeat) over time in a free-viewing task. In order to test the reliability of this method, we varied both the types of scenes presented (interiors, exteriors or landscapes) as well as the availability of the surrounding scene context using a gaze-contingent window. In the context-unrestricted condition, participants could see the entire scene, whereas in the gaze-contingent restricted-viewing condition, only the small portion of the scene where they were fixating was visible. When scene context was unrestricted, participants revisited previously inspected locations more often, with patterns of re-fixations occurring significantly later in the trial period. Participants also spent more time sequentially fixating local regions of the scene when viewing was unrestricted. With gaze-contingent viewing, re-fixations and sequential local scanning were rare. Relative to exteriors or landscapes, participants’ scan patterns of interior scenes exhibited increased concentrated local scanning and they repeated specific sequences of fixations more frequently. Collectively these data show that categorical recurrence quantification analysis is a potentially powerful method for gaining new and reliable insights into the temporal characteristics of fixation patterns across scenes, context, and individuals.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only