Participants from a Fall 2014 class at the UC Berkeley School of Information have published a first-of-its-kind dataset collected using consumer-grade brainwave-sensing headsets during an in-class group exercise, along with the software code and visual stimulus used to collect the data. The dataset and documentation are available at https://bit.ly/synchronized-brain-recordings.
The group exercise was a “Color Counting Challenge” where the 25 students in the class were presented with a visual stimulus (duration 162 seconds) consisting of a timed sequence of images with colored rectangles arranged in a 5×6 matrix. The rectangles were colored in red, green, blue, yellow, and black, with varying numbers of rectangles of each color on each image. At the start of the exercise, each participant was asked to select a color (red, green, blue, or yellow). Then each participant silently counted the number of rectangles of their chosen color as each image is displayed on the projector screen.
Brainwave signals were collected from each participant using Neurosky Mindwave Mobile devices. The data was transmitted over a Bluetooth connection to a software client installed on each participant’s notebook computer, and then forwarded to a networked server for storage. The dataset was anonymized using a four-step process before being released, to protect the confidentiality of participants.
“As a class, we talked extensively about the value of data sharing as well as the importance of data stewardship,” said professor John Chuang, who coordinated the research. “I am most excited that we have come up with a novel way to support the enthusiasm of the class to contribute a unique piece of data to science while safeguarding the confidentiality of individual contributors.”