Mixed Feelings and Realities: Joyful to Nauseating Sentiments about VR on Twitter

This study examines how Twitter users publicly express emotions about virtual reality (VR) and play, with the aim of understanding how perceived affordances and limitations of the technology shape emotional responses. Motivated by skepticism about VR's future and by theories of interactivity, presence, and emotional arousal, the authors ask which emotions dominate tweets about VR and play and which positive and negative themes emerge from those emotions. The paper is framed as preliminary work within a larger National Science Foundation project on the future of VR.

The data consist of 100,363 tweets referencing "VR" or "play" from 2021, drawn through the Academic Research tract of the Twitter API. After filtering for first-person singular pronouns, 27,271 tweets were scored against the NRC emotion lexicon, which covers eight emotions and two sentiment polarities. An ANOVA and Tukey HSD test showed that anticipation was the most frequently expressed emotion, significantly higher than trust, while joy and trust did not differ statistically, and both exceeded fear. The three leading emotions, anticipation (13,660 tweets), joy (10,738), and trust (11,121), were then subjected to qualitative inductive thematic analysis.

The qualitative findings indicate that positive emotions were consistently tied to game content, such as anticipation of streaming or replaying titles and enthusiasm for specific games, but these feelings were frequently counterbalanced by hardware and physical limitations, particularly motion sickness and cybersickness. Trust was undermined by technical difficulties, including streaming problems, leading some users to abandon VR entirely. The authors conclude that VR enjoyment appears to hinge on traditional gaming content, and that breakdowns in presence and immersion caused by physical discomfort are a principal limit on emotional engagement. The authors note that the analysis is preliminary and that a fuller typology of themes remains planned, and that the study's reliance on a single year of English-language Twitter data limits generalizability.