National Science Foundation Awards Northeastern Grant to Study AI-Powered Learning Platform - Institute for Experiential AI

National Science Foundation Awards Northeastern Grant to Study AI-Powered Learning Platform


by Tyler Wells Lynch

StudyCrafter allows users to create gamified research studies involving actual human participants.

In the social and behavioral sciences, data is highly context-sensitive. Psychologists, for example, have to contend with the subconscious ways in which an experiment’s design can manipulate the inputs of participants. The sensitivity of research settings to uncontrollable variables is partly responsible for the replication crisis we see in the social sciences.

For students, the challenge is even greater. Securing actual human participants in the classroom is difficult, and without sufficient data, they may miss out on lessons about the importance of data collection and methodology.

But what if students could learn about research design by creating gamified studies involving actual human subjects? That’s the idea behind StudyCrafter, a research development platform that was recently honored with an award from the National Science Foundation (NSF).

 

Students can build engaging projects with just a few clicks.

Developed by students and faculty at the Northeastern Game Studio, StudyCrafter allows users to create gamified research studies involving actual human participants. Simple drag-and-drop tools enable them to design story-based surveys that gather relevant data automatically and anonymously. Students can use social tools to share, iterate, and replicate studies online, drawing in more human participants and data.

Northeastern University Associate Professor and Institute for Experiential AI faculty member Casper Harteveld, Ph.D., and his team received a grant from the National Science Foundation to study how AI could improve the StudyCrafter platform. Over the next four years, they will use a mix of AI, active learning, and design iteration to augment student research methods and broaden StudyCrafter’s appeal in undergraduate settings.

 

Associate Professor Casper Harteveld, Ph.D.

“We’re basically using games as a research method,” Dr. Harteveld says. “But when you’re creating games, not only do you need to design it, you need to analyze and evaluate if the game is doing what it’s meant to do.”

StudyCrafter’s capabilities go well beyond game design. An AI component can critique the students’ experimental design, and a peer-review rubric allows students to review each other’s research. Both features are critical for establishing the fundamentals of experimental design.

Beyond the Northeastern campus, the StudyCrafter project will include a pool of pilot testers from 11 different universities. Dr. Harteveld and his team will measure the effectiveness of AI-powered learning opportunities in experimental design and hopefully answer why traditional learning environments are often unable to produce instructive research experiences.

To scale the project, Harteveld eyes four key areas of development:

  1. Sophisticated data analysis tools
  2. Interactive content for students and instructors
  3. AI-assisted research coaching
  4. Automatic prompts and scaffolding methods to encourage students to critique their own research

With the recent NSF grant, Harteveld hopes to explore more of the co-creative possibilities of AI: How can AI and data analysis be used to help students design better experiments? The long-term vision is to create a comprehensive research tool that allows teachers, instructors, psychotherapists, and people with limited programming skills to create interactive games applicable in a wide range of clinical and educational settings — something akin to WordPress or Canvas.

“Games are harder to create than a quiz or a survey,” Harteveld says. “And that’s not going to change. But we can make it easier to create a game and incorporate more playfulness in different kinds of activities. With AI, it’s a lot more feasible.”

Learn more about Dr. Harteveld’s work and how the Institute for Experiential AI can help organizations adopt human-centric AI solutions to scale operations, solve problems, and gain a footing in a world of data.