Abstract
When remembering an object at a given location, observers tend to return their gaze to this location even after the object has disappeared, known as the Looking-at-Nothing (LAN) phenomenon. However, it is unclear whether the LAN phenomenon is associated with better memory performance. Previous studies reporting beneficial effects of LAN have not systematically assessed eye movements. Here we related memory performance to eye movements during memory retrieval— saccades in a free-viewing condition and microsaccades in a fixation condition. In each trial, observers (n=19) had to remember eight images of objects shown for 5 seconds, during which observers could freely move their eyes. Object pairs were shown at 12, 3, 6 and 9 o’clock at a distance of 6 degrees from screen center. At the end of each trial, observers indicated by button press whether an auditory statement about an object’s location (e.g., “Pineapple up”) was correct, incorrect, or whether the prompted object had not been shown. Results show similar memory accuracy in free-viewing and fixation conditions (85% vs. 88%). Our eye movement analysis revealed that in only 62% of free-viewing trials observers made saccades. Yet, memory accuracy did not differ between free-viewing trials in which observers did or did not move their eyes (82% vs. 89%). Similarly, in the fixation condition we did not find a benefit of micro-saccades (88% memory accuracy with vs. 89% without microsaccades). The LAN phenomenon was observed in free-viewing trials in which observers made saccades, and performance tended to be superior in those trials. However, given the global lack of memory performance differences between eye movement conditions we conclude that eye movements are not necessary for accurate memory retrieval. These results may explain why previous literature shows contradictory findings on beneficial effects of LAN.
Acknowledgement: NSERC Discovery Grant and Accelerator Supplement to MS