The mission of UVA’s School of Data Science is to be a "school without walls.” In other words, this twelfth school in University history has its focus on collaborating with other schools to move research forward and approach real-world issues from a variety of domains—policy, education, business, chemistry, and more.

One of the first steps in this mission is building Collaboratories—research groups crossing disciplinary boundaries to use the power of data science to advance domains across Grounds. And the first to be established is with the UVA School of Education and Human Development.

Brian Wright, associate professor in the School of Data Science, is working from the ground up to establish this collaboration.

The word Collaboratory itself has its roots in UVA history, Wright says. William Wulf, a computer scientist in the School of Engineering and Applied Science, coined the term in 1989.

Wulf’s definition of a collaboratory was “a center without walls,” in which the nation’s researchers can perform research while interacting with colleagues, accessing instrumentation, sharing data and computational resources, [and] accessing information in digital libraries.

Wright explained that the School of Education and Human Development produces an immense amount of research each year.

“Curry is one of the top education schools in the country,” he says. “They are incredibly productive in terms of generating knowledge and research—the most productive in terms of per faculty member on UVA campus, [...] which is really an incredible outlier for an education school in general.”

So where does the School of Data Science fit into a school that already does so much?

“The level of sophistication and research quality over there [the School of Education and Human Development] is really high,” Wright says. “But what they need, especially for people that are moving more towards data science approaches, is people that are familiar with those techniques and can advise them about how to use those [techniques] and contribute to research projects and then basically work together, collaboratively, to move their research forward to take advantage of the new world of data science.”

Incubator Hubs

Wright broke down how the Collaboratory is being established. He noted that the School of Data Science is going to first start working in “hotbeds of current data science activity,” at the Curry School, calling these, “Incubator Hubs.”

There are currently two Incubator Hubs to begin that collaboration. One is a project led by Ben Castleman on economic mobility. Castleman is the Newton and Rita Meyers Professor in the Economics of Education at Curry, as well as the Founder and Director of Nudge4 Solutions Lab.

“Our lab works at the intersection of behavioral economics and data science to design and test strategies to help economically disadvantaged, underrepresented populations pursue tailored educational workforce pathways that have a higher probability of leading them to success,” Castleman said.

He continued, “I think what excites me most about the collaboratory are the opportunities for...working collaboratively on projects that Curry or the School of Data Science might not be able to achieve on their own.”

Castleman noted that Curry has partnerships with state agencies, access to vast amounts of data, and the ability to design and test interventions in the real world. He explained that the School of Data Science brings expertise in the machine learning methods and algorithms that help generate these interventions.

“Bringing those two together through the collaboratory seems to bring opportunities that really strengthen our work and provide access to and opportunities for Data Science faculty and students to engage with agency partners, and data and projects that might otherwise not be as available,” Castleman added.

The second Incubator Hub is a project led by Vivian Wong on replication studies. Wong is a research methodologist in the education field as well as an Associate Professor in Research, Statistics, and Evaluation in Curry.

“Our lab has two over-arching goals—the first is to develop the methodological foundations for what we think of as a ‘replication science’—that is, we want to formalize our understanding of what a replication study is, what are the conditions for a high-quality replication study, develop diagnostic measures for assessing the quality of a replication study, and evaluate measures of how we determine ‘replication success,’” Wong explained. “The second goal is to develop easy-to-use tools for education and social science researchers to plan, implement, and analyze replication studies.”

Wong noted that in order to achieve these goals and make the approach scalable over a long period of time, collaboration between education and data science is key. Wong is looking forward to bringing together education researchers and data scientists through this Collaboratory to face difficult problems in research methodology, specifically in the field of education.

“I am excited about this effort because it has the possibility to address some intractable methodological challenges in education. As a field, we have worried a great deal about both the replicability and generalization of results across different studies, populations, and settings,” Wong said. “We hope that the crowdsourcing platform [Collaboratory] provides opportunities to include a more diverse set of researchers in the research process, who in turn are able to collect data from more diverse participants.”

Wright’s goal is that the Incubator Hubs will serve to ignite more collaboration between the School of Education and Human Development and the School of Data Science.

“The idea is that the energy we put into those Incubator Hubs can be fed down into other potential projects inside of the school,” Wright said. “Some of the activities from those two center points can be disseminated, which could likely grow into other formal collaborations between the two schools.”

Wright added that in addition to collaborating on research projects, the Collaboratory might teach School of Education and Human Development faculty and students short courses on Python or Natural Language Processing to better understand the research from the various ongoing projects within education.

Another aspect the Collaboratory will focus on is the capstone projects of students in the Masters of Data Science program. Each student must complete a capstone project prior to graduation, and Wright plans to connect students with researchers in the Curry School.

“I’ve asked Catherine Bradshaw, the Director of Research [at the Curry School], to identify three to four projects, either from the labs or from other sources with high data science potential and propose those as projects for our capstone students to engage in,” Wright explained.

Opportunities

Working closely with the School of Education and Human Development provides immense opportunities to delve into effective education and create change in the education field as a whole.

“Educational analytics is a prime target for such a Collaboratory and we will be able to report on progress in the coming months and years as this initiative develops to better educate the next generation of teachers and students,” Bourne stated. “Our goal is to build bridges across Grounds using data science as the foundation.”

Wright emphasized that this Collaboratory has the potential to help both the School of Education and Human Development and the School of Data Science. That is his hope for the partnership between the two schools.

“The goal is to find these mutually beneficial touchpoints for everyone involved,” Wright says. “And not think about one authoritative path, but to design the relationship around how everybody can benefit.”

Filter by

School of Education and Human Development Collaboratory grows data science partnerships at UVA

Get the latest news