How AI and computer science can move social work research forward

By Matthew Biddle

Artificial intelligence is changing — or upending — nearly every industry you can think of, including social work, human services and other related professions. In a special issue of Research on Social Work Practice, two University at Buffalo scholars showcase how AI and other emerging technologies can move the field forward, challenging their fellow researchers to leverage the opportunities this technology offers to create change.

Maria Y. Rodriguez, PhD, assistant professor in UB’s Department of AI and Society, and JoAnn S. Lee, PhD, associate professor in the UB School of Social Work, co-edited the special issue, which includes seven papers that each highlight different advanced computational methods — from AI and large language models to computer simulations, geographic information systems and social network analysis.

“Notably, these methods enable us to explore social complexity more robustly, by examining interactions between many people — and other factors such as environments — as well as changes over time, more elegantly than traditional statistical methods,” Lee says. “And it’s not that these computational methods should replace our current methods, but rather, they can be a useful addition to current approaches.”

Lee says it can be challenging for researchers who are pioneering new approaches to navigate the traditional peer-review process, since their methods, by definition, are new and thus relatively less understood.

“This special issue allows us to promote the pioneering work of social work scholars who are delving into these uncharted waters while bringing together experts who can assess the quality of this work,” she explains.

For Lee, this project builds on her longtime use of agent-based modeling in her research around the transition to adulthood for youth. Agent-based modeling uses computer simulations to explore how agents — like individuals, families, even atoms — interact with each other and their environments.

Lee also holds a graduate certificate in computational social science from George Mason University. Now, at the UB School of Social Work, she is building a new Institute for Social Complexity and Computation to bring together researchers who are using advanced computational methods to study complex social systems.

“Computational methods provide alternative approaches and can actually help align our research with our values,” Lee says, providing an example from her teaching.

She says her students sometimes struggle to learn statistics, since social work values and statistical principles conflict with one another.

“Social work believes in the dignity and worth of every person, but in statistics we focus on the aggregate. In fact, we typically discard the outliers — which could easily be the client we’re working with — and calculate the average,” she explains. “In an agent-based model, we can build a computer simulation with as many ‘outliers’ as we want and see what happens. Such simulations can yield surprising results.”

Overall, Rodriguez and Lee argue that AI and other computational methods offer more opportunities than challenges for social work researchers and practitioners.

In the opening editorial to the special journal issue, they write, “Advanced computational methods require social work insight and ethics to work best for humans.” And by combining innovative methods with human-centered values, they contend, researchers can ask bigger questions — and, perhaps, develop the solutions to match.

Matthew Biddle
Director of Communications and Marketing
School of Social Work
Tel: 716-645-1226
mrbiddle@buffalo.edu