Abstract
We limited the exposure duration of emotional photographs to investigate whether the speed of encoding was influenced by the emotional valence of the pictures. Participants viewed slide shows in which scenes (1/3 negative, 1/3 neutral, 1/3 positive emotional images from the IAPS) were presented for 60ms and then masked. The slide show was followed by an old/new recognition memory test. Memory was best for positive, moderate for neutral, and worst for negative images. These results cannot be explained by the arousal level of the images (positive and negative scenes were matched for arousal) or response biases for different emotional images (there were equal numbers of false alarms for different emotional foils during the recognition test). Follow-up studies revealed that these emotional effects disappeared when the stimuli were inverted or the encoding time was increased to 1500ms. These data suggest that the effect is due to emotional valence rather than low-level visual differences between the images (e.g., color, brightness, spatial frequency distributions) and suggest that the time required for encoding is critical to the effect. We conclude that negatively valenced emotional pictures slow encoding time. A final study investigated whether negative images were slow to encode because of an attentional bias away from them or because they produced an overall interruption in processing. This experiment presented two images per slide, one neutral and one that was negative, neutral, or positive. An attentional bias away from negative objects predicts that low recognition rates for negative images will coincide with higher recognition rates for the neutral images that appear with them. A global interruption in ongoing processing predicts low recognition accuracy for both the negative images and their neutral pair. Preliminary data suggest that negative images produce an overall interruption in processing.