Purchase this article with an account.
Matthew S. Peterson, Melissa R. Beck, Walter R. Boot, Miroslava Vomela, Arthur F. Kramer; Little is remembered about rejected distractors in visual search. Journal of Vision 2005;5(8):416. doi: https://doi.org/10.1167/5.8.416.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
During visual search, knowledge about previously examined distractors is used to guide attention towards unexamined items (McCarley et al., 2003 Psychol Sci 14 422–426). Last year, we demonstrated that the spatial location of examined items, rather than the identity or surface features of those items, is used to guide attention towards new items (Beck, Peterson, & Vomela, VSS 2003). In the experiments reported here, we investigate whether any identity information is explicitly remembered about rejected distractors. We used two different search tasks: a conventional search task and the oculomotor contingent task of McCarley et al. (2003). In all experiments, on roughly one third of the trials, search was terminated, an examined location was circled (a place holder marked that location), and observers were quizzed about the item that had been at that location (two alternative forced choice). Although observers clearly had a memory for examined items - they avoided revisiting the last 4 items in the oculomotor contingent experiments - performance in the 2AFC recognition tasks was extremely poor. When asked to discriminate a previously seen letter from one that had not appeared in the display (foil), memory performance was near 75%, regardless of lag. However, when the foil was another item that had been fixated on that trial, results from the memory task were near chance (56%). This suggests that people do remember the identity of rejected distractors, but their explicit memory for distractors does not include their locations. Interestingly, memory performance for the penultimate item (lag 1) was contingent upon whether the search display was present when the location was probed. When the display was present (items replaced with placeholders) during the memory probe, accuracy for the last examined item improved. This suggests that scene context can help to improve memory performance.
This PDF is available to Subscribers Only