The psychology of art and emotional experience: a theme-color-emotion framework for understanding viewer-reported...

Visual art, as a non-verbal medium for emotional expression and regulation, demonstrates unique therapeutic potential in mental health intervention. However, existing research lacks an integrated theoretical framework to systematically explain how the visual characteristics of artworks influence viewers' emotional responses. This study aims to construct and validate the "Theme-Color-Emotion" (TCE) theoretical framework, revealing the interactive mechanisms between thematic content, color characteristics, and emotional responses to visual artworks.

Based on the WikiArt Emotions dataset (n = 4,105 artworks) and the ArtEmis dataset (455 K emotion annotations), chi-square tests, correspondence analysis, and multidimensional cross-analysis methods were employed to systematically examine the association patterns between theme types, color characteristics, and emotional responses.

Artworks with natural landscape themes were significantly more associated with positive emotional responses than other theme types (67.3% vs. average 52.1%, χ ² = 234.7, p < 0.001). Cool colors (blue, green) clustered with calm and contemplative emotions, while warm colors (red, orange, and yellow) clustered with pleasant and exciting emotions. The TCE framework achieved 87.6% accuracy in classifying artworks according to viewer-reported emotional valence, a 9.3 percentage point improvement over the single-color baseline model.

Operationalized through a logistic regression model integrating thematic and chromatic features, the TCE framework reveals that theme and color jointly account for viewer-reported emotional valence in ways that neither dimension captures in isolation. These findings provide a systematic analytical perspective for understanding the associations between visual artwork characteristics and viewer emotional responses, offering evidence-based implications for therapeutic art design and art therapy practice.