GroupAffect-4 Dataset Captures Collaborative Group Dynamics
A new dataset called GroupAffect-4 has been launched by researchers to examine emotions within co-located groups, focusing on individual, interpersonal, and group dynamics. It includes 40 participants organized into 10 groups of four, each engaging in four diverse collaborative tasks: information pooling, negotiation, idea generation, and a public-goods game. Participants are outfitted with wrist-worn sensors, eye-tracking glasses, and close-talk microphones. The sessions gather continuous self-reports of affect, post-task surveys, task results, and Big-Five personality assessments, all synchronized to a common clock. This dataset resolves the issue of fragmented signals across various datasets, facilitating the analysis of participant physiology, eye movements, audio, self-reports, task outcomes, and personality traits, thus advancing research in affective computing and social signal processing.
Key facts
- GroupAffect-4 includes 40 participants in 10 four-person groups.
- Four collaborative tasks: information pooling, negotiation, idea generation, public-goods game.
- Each participant has wrist-worn physiology sensor, eye-tracking glasses, close-talk microphone.
- Sessions include continuous affect self-reports, post-task questionnaires, task outcomes, Big-Five personality scores.
- All data time-aligned to a shared clock.
- Dataset supports analysis of affect as individual, interpersonal, and group-level process.
- Addresses fragmentation of signals across separate dataset traditions.
- Published on arXiv with ID 2605.19765.
Entities
Institutions
- arXiv