EEG: Three armed bandit gambling task
This dataset comprises electroencephalography (EEG) recordings from 23 healthy college students performing a three-armed bandit gambling task with dynamically oscillating reward probabilities. The task was designed to investigate neural correlates of reward processing and reward prediction error during adaptive decision-making and task set generation. Each stimulus condition (blue, red, and green) had oscillating reinforcement probabilities (ranging from 20% to 90%) that moved in different phases across trials. Supplementary physiological recordings including corrugator EMG and skin conductance were acquired on most participants; however, data quality for these supplementary recordings was limited and they were not extensively analyzed. Data were collected in 2014 at the University of New Mexico's Cognitive Rhythms and Computation Lab.
- Participants
- 23
- Channels
- 64 (10-10)
- Size
- 4.72 GB
- Version
- vv1.0.0
- Updated
- Jul 10, 2026