- Executive Summary
- Introduction
- The CIRCLS Community
- Portfolio Analysis
- Bibliographic Analysis
- Field-Driven Research Synthesis
- 1. Exploration and Discovery
- 2. Equitable Co-Design
- 3.Emergent Impact
- 4. Discussion: Key Ideas
- Practitioner Reflections
- Recommendations
- Acknowledgements and Citation
- Appendices
Field-Driven Research Synthesis: Understanding the Character of Innovative Interdisciplinary Exploratory Research
Emergent Impact Through Networked Communities
Broadening impact through innovation and expansion: Ying (Collaborative Learning and AI), Charles (Simulations), Sheryl (Accessibility and Learning), Nikolas (AI), Rebecca and Colleen (Augmented Reality), and Chad and Bill (Data Analytics and Simulations)
The initial investments that NSF made in PIs’ research projects set the stage for future impacts that a) resulted in unexpected innovations and outcomes, b) moved beyond the original proposal, and c) scaled up emerging technologies for learning.
PIs detailed the emerging impacts NSF’s initial investment afforded their projects along three dimensions: securing new funding, conducting and disseminating new research, and expanding the reach of technological tools developed through NSF funding.
For example, Ying explained that the project insights from the Collaborative Learning and AI project Building a Teacher-AI Collaborative System led to securing new funding through the GBH grant Empowering Teachers to Collaborate with Generative AI for a new PBS media project. Charles shared that the Aladdin platform developed with the NSF funding from his Simulations project Science and Engineering Education for Infrastructure Transformation and Collaborative Research: Mixed Reality Labs have been incorporated into an international cooperation grant with Ukraine’s Institute of Renewable Energy and the National Academy of Sciences to help rebuild Ukraine’s energy infrastructure.
Regarding new research, Sheryl shared that the initial NSF-funded research from the Accessibility and Learning project Designing STEM Learning Environments for Individuals with Disabilities led to a longitudinal study of disabled children’s experiences with technology into adulthood. Nikolas shared that research from the AI project Supporting Designers in Learning to Co-create with AI set the stage for the design of new interactive generative AI interfaces that will be developed into applications.
Finally, Rebecca and Colleen shared that their Augmented Reality project, Combining Smartphone Light Detection and Ranging with Augmented Reality, set the stage for expanding the reach of Physics Toolbox, an augmented reality physics sensor tool, from Rebecca’s individual classroom to thousands of classrooms across the U.S. Chad and Bill also shared that CODAP, an open-source data analysis platform resulting from their Data Analytics and Simulations projects, Inquiry Space and Data in Space and Time, has become embedded in the infrastructure of education R&D with thousands of users, including researchers, developers, and students, per year.
At the start of these projects, the PIs did not know what the future impacts of their exploratory work would be. For many, moving beyond an original idea, to something larger and unexpected, was seen as a demonstration of impact, as Charles explained:
You say you want to do a proposal. Then, at the end of the project, you create this overflow of impact that is beyond the original, right? So that’s a nice demonstration of broad impact. If you’re only impacting the demographic community that you originally [set], then that’s not broad. That’s anchored. Broad impact means that the problem, the user, is unexpected. The more unexpected, the better. And that’s the beauty of it, that’s why it’s an investment.
For Charles, moving beyond the initial project parameters toward novel insights and applications was critical to using NSF funding as seed money toward achieving broader emerging impacts.
NSF’s initial investment in PIs’ projects also served to establish use cases for future stages and applications, as Bill explained:
The early work was, as is so often true, about efficacy. Does this actually work? Does it change what kids and teachers do in classrooms? The second phase is scalability: It works in these very contained situations, but does it scale to large numbers of classrooms? These collaborations have led to proof of scalability. We don’t really know how many students use CODAP, because we don’t ask students to log in. We just know that there are 500,000 unique IP addresses every year—several hundred kids on a given day. So, we’ve gone from efficacy to scalability. The third step is longevity. How do you keep this alive? And that’s what we’re struggling with now.
For many PIs, improving scalability, broadening reach, and strengthening sustainability were seen as significant indicators of future impact and challenges. Chad described the challenge of sustaining and expanding a project’s impact:
Our work is open source by desire to lower the barriers in education. And we recognize that raises issues for sustainability. We’re trying to grapple with that as we think about this open-source ecosystem. But I’m also sure that the growth that we’re describing in terms of impact on numbers and research would not have been there if the model had been a paid business model.
For Bill, both the positive outcomes and the challenges associated with their projects’ impact ended up stretching them as researchers:
So, in a way, CODAP has extended us. We’re having to stretch to accommodate it because it has been successful in getting out into the world. And now we have to stretch to grow that community and also to figure out how to maintain it in the face of changes in the underlying technology.
Improving scalability, broadening reach, and strengthening sustainability incentivized PIs to connect with fellow researchers for community and support.
Mobilizing networked communities of practice: Marjorie (Simulations), Roxanne (AI), Erin (Collaborative Learning), Roxanne (AI), Lorna (Accessibility and Learning), and Ying (Collaborative Learning and AI)
PI participation in the CIRCLS networked community was essential to their engagement with NSF-funded interdisciplinary, exploratory research because it a) provided an environment that supports researchers’ professional development through mentorship and camaraderie, b) used networked infrastructures to broaden interdisciplinary collaboration and address challenges and barriers to conducting interdisciplinary, exploratory research, and c) connected researchers with practitioners to prioritize implementation.
Participation in the CIRCLS community promoted PIs’ growth as scientists by tapping into a national expert group to challenge each other with new knowledge and perspectives, as Marjorie shared:
I think it’s important that we bridge the knowledge of technology, and how it can be used and developed, with subject matter experts. I think that part of the NSF community through CIRCLS is as valuable as funded research projects because it really helps individuals grow as scientists and researchers. That’s what I’m interested in. You’re able to tap into a national group of experts to challenge each other and help each other, which is just really special. I really value the community aspects of all this.
Participating in networked research communities provided a supportive environment for PIs to learn about different disciplines and engage in interdisciplinary work. As a result, CIRCLS was seen as one of the few networked communities that offered concrete avenues for broadened participation and interdisciplinary bridging, as Erin shared below:
I think CIRCLS in particular is essential in this kind of space where people are doing such innovative things across disciplines. It feels like an important catalyzer for this kind of research. There’s been these dual goals of broadening participation and bringing people in who haven’t traditionally been part of CIRCLS. Honestly, I feel broadening participation is really important. I think that bringing the current PIs into that conversation and creating that bridge is also important.
As such, PIs felt that networked communities like CIRCLS provided mentorship, camaraderie, and support in fostering environments conducive to interdisciplinary engagement. For example, Roxanne shared her experience at a CIRCLS convening, where she discovered diverse design communities across various disciplines:
I learned a lot when I went to the CIRCLS convening. I enjoyed seeing researchers across these different disciplines. It made me realize that, as much as engineering has a design community, computing also has a design community, and there’s a whole different set of conferences and publication venues that they favor. It definitely opened my eyes to a broader range of audiences and communities and places to propose work, to publish work, and to learn from what other people are doing.
This exposure expanded Roxanne’s understanding of various knowledge communities and introduced her to the conferences, proposal opportunities, publication venues, and audiences relevant to disciplines adjacent to her field of research.
Other PIs were appreciative of CIRCLS’ supportive infrastructure for enabling collaborations with researchers in adjacent fields. For example, Lorna shared that her involvement with the 2023 CIRCLS convening made her feel part of a broader research community across diverse domains of interest to her work:
The in-person convenings have been really a great way to learn about similar work in parallel domains and see what other ways people are looking at these intersections between technology and learning. And it’s been a great way to share my work with the community. So, I’ve had collaborative calls with people. I’ve explored grant proposals with people. This new proposal that we received good news about recently is a collaboration with some colleagues from the CIRCLS community.
The resulting interdisciplinary collaborations that emerged from PIs’ participation in CIRCLS activities were viewed as positive examples of emerging and future impacts achieved through networked engagement.
However, PIs also described challenges—such as barriers to disseminating interdisciplinary exploratory research findings considered novel and unprecedented—as endemic to engaging in research that doesn’t neatly fit into established academic disciplines, as Erin explains:
I feel that is one of the issues with exploratory research: What I’m doing in this project, and in my lab, there aren’t models for. We’re not following a template for how you build these technologies. And so, you struggle at times because you have to invent this all. It means when you go to publish, then other people are looking at it like, “I don’t have a model for how to perceive this.” That’s something I’ve experienced quite a bit. So this is how I’m thinking about the impact of this work, and the opportunities that arose from it, and the ways in which I feel we’re still working to disseminate the outcomes to influence the academic community more.
In Erin’s view, established research ecosystems do not currently support and amplify interdisciplinary exploratory work. This belief was echoed by several PIs who believed that current publication structures for disseminating interdisciplinary, exploratory research are still primarily siloed along established research trajectories.
Finally, many PIs we interviewed appreciated that the CIRCLS community brought researchers back to implementation by connecting them with practitioners, centering practitioner knowledge, and enabling future collaborations between researchers and practitioners.
Ying described her positive experiences engaging with practitioners through her participation in a CIRCLS Expertise Exchange session:
I participated in an Expertise Exchange session, which is a panel where I co-hosted with two other experts in the field. Our panel was around conversational AI. I liked sharing our experiences and also hearing what other panelists did. But more importantly, we had a lot of discussions with the audience who came to our panel. I got a lot of questions that were actually very inspirational for my own research. I also really like that the panelists are not just researchers but also community members who are educational practitioners.
We’re not just researchers talking about how we develop conversational AI, we also have the perspective from educators to share what they think conversational AI might support, or might not be so helpful, in their classroom instruction. Just try to bring us back to the actual implementation piece, to how we could make our research more useful for the teachers and students. I’ve actually been in touch with the educator that was part of our panel. She has connections with schools, and I love to get her insight on how they see the practical implications of our research.
Ying enjoyed working with practitioners across disciplines and appreciated the opportunity to facilitate discussions among CIRCLS community members. For her, this work included supporting participants in developing the skills needed to better understand and translate technical language from one field to another.