You are here


Mining for Data

When the Fourth International Conference on Educational Data Mining takes place July 6-8 in Eindhoven, Netherlands, educators will explore the latest advances in the practice, an increasingly popular and effective way to discover new knowledge from large and complex data sets.

Data mining, which develops methods for exploring the unique types of data that come from different settings, has been taking place for up to 30 years in other fields, from the sciences to business. But it wasn’t recognized as an emerging discipline in K12 education until the first international conference in 2008, says Ryan S.J.d. Baker, assistant professor in social science and policy studies at Worcester Polytechnic Institute in Massachusetts.

With some Worcester (Mass.) Public Schools serving as laboratories, Baker and other WPI researchers are developing and using methods for mining data that come out of interactions between students and educational software. The purpose, Baker explains, is to understand better how students respond to the software and how their responses impact their learning.

Burncoat Middle School, for example, started this school year to implement ASSISTments, a tool developed by a WPI research team led by Neil T. Heffernan, associate professor of computer science, and his wife, Cristina, an adjunct instructor of mathematics at WPI. Burncoat is an underperforming school where seventh- and eighth-grade math, English, science and social studies teachers gather regularly for scheduled meetings to review the latest data on their students.

ASSISTments, explains Heffernan, helps students learn math by presenting them with problems, then offering structured assistance if they have difficulties solving the problems. By recording students’ attempts to answer questions and the help they request, the system also assesses which concepts they have mastered and which they still need to work on.

In addition to providing teachers and parents with immediate feedback on students’ progress, the system also gives principals real-time data on what is happening in their math classrooms. “The data drive instruction and adjust instruction,” says Principal Lisa Houlihan. “If we don’t have the data, we don’t know that the road we’ve committed to is the right road.”