Abstract
Humans heavily rely on vision to evaluate safe versus dangerous environments. It is debated whether color information contributes to the perception of affective scenes (Cano et al., 2009; Codispoti et al., 2011). Previous studies compared scenes with strong emotional valence and, as a control condition, neutral scenes. However, we argue that if color plays a significant role in affective scene perception, the effect should be most pronounced in stimuli with weak or moderate emotional valence, since obviously fearful scenes may be perceived as such with and without color. A set of 157 colorful images devoid of human faces but with a wide range of other contents (e.g., fires, spiders, car crashes) was collected from the Internet. We tested the effect of color and its interaction with novelty of the images by asking participants to rate the fearfulness of each stimulus. Participants were divided into two groups. One group saw the images in color and then in gray-scale while the other group saw the images in gray-scale first. A mixed-model ANOVA revealed significant main effects of both color and observing order. The interaction between the effects of color and observing order is also significant with the difference between color and gray-scale images being greater when participants saw the images the first time rather than the second time, suggesting perhaps a color priming effect. Moreover, the effect of color was found to be the strongest for images that were rated as low or moderately fearful, and not significant for images that were rated as highly fearful, despite the fact that the ratings did not reach the maximum suggesting that the results were not caused by a ceiling effect. These results are consistent with our prediction, highlighting the importance of using stimuli across a full spectrum of valence to study affective scene perception.
Meeting abstract presented at VSS 2012