Abstract
PURPOSE. To explore the effects of familiarity and repetition on the encoding of real-life scenes with emotional content into visual short-term memory.
METHODS. Stimuli were 64 color images of real-life scenes, 32 with positive and 32 with negative emotional content (matched for arousal). During each of 1024 trials per subject we used the RSVP procedure to present 6 pictures, 1 positive and 1 negative unrepeated and 1 positive and 1 negative repeated after an intervening image (counterbalanced across durations and repetition conditions), with durations per picture ranging from 8 to 1710 ms. Following each 6-picture stream subjects were shown 8 pictures singly (4 seen and 4 new, half positive and half negative images), and asked to judge whether the given picture was present in the stream they just saw. Accuracy data from 3 subjects that had extensive previous exposure to the image set were corrected for guessing and compared across conditions.
RESULTS. In all conditions, performance with negative images was better than with positive images at all durations. Performance with negative images benefited from repetition, as would be expected from the summation of independent events across successive image exposures. Performance with repeated positive images showed instead a deficit of facilitation for exposures of 100 ms and above, suggesting that responses to successive occurrences of positive images are not independent.
CONCLUSION: We previously reported (Maljkovic & Martini, VSS '03) that RSVP curves for negative images are steeper than for positive images in conditions were subjects are unfamiliar with the image set. Familiarity eliminates this difference in steepness and increases the overall speed of performance for negative images. Furthermore, familiar positive images, but not negative images, seem to be affected by interference between successive exposures, in agreement with findings of selective attentional interactions using other emotional stimuli.
NIH EY13155 to V. Maljkovic