“We live in a world of big data,” said Ryan de Baker, who delivered TC’s Sachs Lecture this past winter. “My browser cookies know that I’m fluent in English and Portuguese, married with children, a college professor – that I travel internationally for work, that I stay in cheap hotels when I travel internationally for work, and that I love opera. And all for the noble purpose of selling me more stuff.”

Education is harnessing big data, too, thanks to a new generation of software that generates a wealth of information about how students think, learn and make choices.

“Students make hundreds of meaningful actions per hour, from pausing and thinking to running away from a skeleton in a game to changing a setting,” Baker said.  As organizations such as The Pittsburgh Science of Learning Center store this information, “increasingly data doesn’t have to be collected, just connected.” Once connections are made among different types of data streams, “we can answer more interesting questions than we could answer before.  For example, we can ask what attitudes are associated with positive outcomes for students.”  

Data mining already makes it possible to zero in on which topics and teaching methods work best – or don’t work at all – for different students.  The next frontier: detecting students’ emotions, particularly their levels of frustration as they grapple with different kinds of challenges.

Of course, students will always run into difficulties, Baker said. “But now we’ll be able to make adjustments in mid-semester.”  

Published Wednesday, Jun. 26, 2013