Summer School Helps Students Build Tech Tools To Improve Learning
Ashish Aggarwal wants to help students taking online classes by creating a digital teaching assistant to answer questions.
To help turn his idea into reality, Aggarwal needed to learn about data and software. So, he enrolled in Carnegie Mellon University’s Simon Initiative LearnLab Summer School, which teaches participants about the leading tools that merge education, data and technology — all of which are developed by CMU researchers.
This July, 55 students from across the globe spent a week immersed in one of four tracks: developing intelligent tutoring systems, working on computer-supported collaborative learning, learning about educational data mining or building online courses. Each track received hands-on experience working with the tools while on small project teams.
Aggarwal, a University of Florida graduate student in computer science and education, joined the computer-supported collaborative learning track, which focused on combining insights from linguistics and artificial intelligence to create platforms that mediate student discussions. His team experimented by developing a virtual teaching assistant. They used data to create an agent that could analyze what students might type in their online chat rooms, recognize their struggles and generate assistance.
In just five days, they made enough progress that Aggarwal knew he could take this learning back to Florida.
“What we have made is a very small step toward the thing I’d like to implement in the classroom,” Aggarwal said.
LearnLab Director Ken Koedinger has spent decades developing technology-enhanced learning tools.
“Many of the grad students who are here heard the rumors that it’s quite productive from a research perspective. Many come out of it with the beginnings of a publishable paper,” said Koedinger, professor of human-computer interaction and psychology. “We do have some industry folks, and they come to enhance their skills. I think this is for those folks who hunger for innovation and who like to hear about the advances of science.”
The LearnLab Summer School, now in its 12th year, matches its students with CMU mentors, who begin emailing with their teams ahead of time and help the students as they learn the technology.
“We’re sitting right next to students with sometimes one, two, three, even four laptops open at once. Somebody is helping someone build their CTAT (Cognitive Tutor Authoring Tools) tutor or do a data analysis. It’s completely consistent with what we’re teaching — learning through feedback,” Koedinger said.
Koedinger and other LearnLab researchers have spent decades developing the tools, which are available for use by researchers and institutions.
“That’s actually part of the reason we’ve always done this — we want to build tools people use. This is a way, from our point of view, to help us improve those tools,” he said.
The summer school can benefit participants long after the week is over.
Candace Walkington, an assistant professor in teaching and learning at Southern Methodist University, attended twice to learn more about intelligent tutoring systems and educational data mining. She developed experiments using an algebra tutoring system, and she learned that students benefited when problems corresponded to their out-of-school interests, like sports, music or movies.
LearnLab helped introduce Walkington to Carnegie Learning, a Carnegie Mellon spinoff company, whose tutoring systems are in use in many schools where her idea could be tested. This launched a multi-year collaboration between Walkington, Carnegie Learning and LearnLab that provided her with software programmers, helped her gain funding to implement her system and connected her with schools.
“LearnLab is a huge network of people who supported me at all stages of my research. I honestly don’t know if I would have gotten my Ph.D. or become a professor if not for this network. They made my research possible,” Walkington said.
LearnLab, originally funded by the National Science Foundation, is the scientific arm of the Simon Initiative. It leverages cognitive theory and computational modeling to identify the instructional conditions that lead to robust student learning.
2015 LearnLab Summer School