Feeling Better but Not Thinking Better: the Effects of Scene Valence and Category on Restoration.

Indirect experiences of nature (e.g. via computer/virtual reality) are thought capable of increasing cognitive performance and reducing attentional fatigue. Natural scenes are also consistently more preferred than urban scenes, but it has been suggested that restorative benefits may occur in response to scenes high in perceived appeal. The nature of reported restorative outcomes varies across studies and it is unclear whether they encompass only subjective feelings, or extend to robust improvements in objective cognitive tasks. We employed a 2 (Scene Category: Natural, Urban) x 3 (Valence: High, Intermediate, Low) between-subjects design (N=291) to measure cognitive performance (Attention Network Test, Backwards Digit Span) and subjective restoration (self-reported fatigue and task difficulty). Each of the six experimental conditions had at least 45 participants. Image valence was determined by previous ratings of natural and urban scenes for calmness by an independent sample (N=990). Based on these ratings, 30 images were randomly selected for each category from the highest, middle and lowest-rated thirds. Participants completed a cognitive battery (ANT and BDS), then viewed and rated images in a 10-minute restoration window before the second cognitive battery. Subjective fatigue was measured at baseline, after each cognitive battery and after the restoration window. Nature images were rated as more preferred and calming than urban scenes, regardless of scene valence. Within scene category, high and intermediately valent images were also rated as more preferred and calming than low images. Irrespective of scene category, participants viewing high valence images experienced significantly greater reductions in subjective fatigue following the restoration window. However, these benefits did not correlate with increased cognitive performance. Our findings indicate that subjective restoration may be modulated by image valency more so than inherent image naturalness. Future research will benefit from examining whether changes to physiological markers (e.g. HR, cortisol) correlate with subjective restoration.