Affective Response Dataset for Virtual Workspaces: Based on Color Stimuli and Multimodal Physiological Signals

Abstract

Highlights: What are the main findings? In terms of emotional valence, the red workspace significantly elicited “anxious”; the yellow space readily induced feelings of “happy”; whereas the blue space was closely associated with a higher sense of “calm”. In terms of emotional arousal, the red and yellow workspaces were associated with a higher state of arousal, whereas the blue and green ones resulted in a lower level of arousal. The physiological data solidly supported this subjective arousal reports. What are the implications of the main findings? On the theoretical level, this study bridges a gap in the affective dataset of virtual reality workspaces, providing support for the development of “affective-intelligent” virtual workspaces to enhance user experience. On the practical level, the distinct emotional impacts of red, yellow, and blue color schemes offer valuable insights for optimizing traditional workspace designs, thereby contributing to the improvement of the participants’ mental health. In the context of post-pandemic remote work normalization and the emergence of the metaverse, virtual workspaces have attracted significant attention as critical digital infrastructure with promising application prospects. While virtual workspaces enable efficient task performance, compared with traditional ones, the lack of emotional connection between humans and machines adversely affects participants’ mental health. The emergence of affective computing has made it possible to endow virtual workspaces with “affective intelligence”. Therefore, this study aims to clarify the relationship between color and participants’ emotions in virtual workspaces through an experiment involving 48 participants, and eight virtual workspaces were constructed, incorporating four color conditions (red, blue, yellow, and green) and two workspace types (shared and single). Data were synchronously collected using the Positive and Negative Affect Schedule (PANAS), a questionnaire item on arousal, electrodermal activity (EDA), and heart rate variability (HRV). The results successfully established specific associations between colors and emotions: red with “anxious”, yellow with “happy”, and blue with “calm”. Although no specific emotion word was identified for green, this study successfully achieved the emotion classification of virtual workspaces and constructed a corresponding dataset. These findings provide a theoretical foundation for the development of affective computing models.