Jibo's User Study Complete Dataset

Social robots are increasingly expected to engage with users; however, the extent to which users accurately perceive these expressions remains uncertain, particularly across different cultural and developmental backgrounds. This study explores how users of varying ages and ethnicities interpret the emotional expressions of Jibo, a social robot that relies on abstract motion rather than facial or vocal cues. Participants (N = 36) engaged with Jibo across three modalities (physical, digital avatar, and video-based) and attempted to recognize six robot-displayed emotions. Quantitative analysis using MANOVA and Kruskal-Wallis tests revealed no significant differences in emotion recognition across ethnic groups and modalities. A modest age effect was observed in the digital condition only, with middle-aged adults outperforming younger adults. Qualitative feedback supported these findings, with most participants citing the lack of facial features, sound cues, and motion clarity as barriers to interpretation. Despite limited recognition accuracy, the physical robot was rated as the most emotionally engaging. These results suggest that while physical embodiment may enhance perceived connection, it does not guarantee improved communication and comprehension. The study highlights the limitations of motion-only expression and reinforces the importance of clearer, multimodal cues in social robot design for broader and more inclusive user understanding.These datasets were collected as part of the user study conducted to obtain results for this research. The datasets were preprocessed, cleaned, and organized using Python libraries to ensure consistency and ease of use throughout statistical analyses.