CAREER: Developing professional skills including critical and reflective use of generative AI in computing education

Generative artificial intelligence (GenAI) is transforming the American economy, and today's computing students will lead its integration across every sector, given the increasing use of GenAI-based tools to write code, analyze data, and do other computing-related tasks. However, computing education currently emphasizes GenAI-based tool use without developing the professional judgment, continuous learning habits, and reflective thinking that employers consistently identify as top skill gaps in new graduates. This project addresses these workforce readiness gaps by developing and evaluating scalable teaching methods that help computing students build professional skills, including the ability to thoughtfully evaluate when, why, and how to use GenAI in their work. The project centers on two complementary approaches. The first involves structured peer reflection groups where students regularly meet to share progress, give feedback, and set improvement goals. The second is a novel video-assisted reflection method where students record, review, and compare their own learning and processes during programming sessions as they experiment with using GenAI-based tools, similar to athletes watching replays of their games. By producing adoptable course materials and faculty training workshops, this project will strengthen the computing workforce pipeline and help ensure that American computing graduates are prepared to be global leaders in an AI-integrated economy. This project advances computing education research and human-computer interaction through two interrelated research tasks with longitudinal mixed-methods evaluation. The first task uses design-based research to iteratively develop and evaluate peer reflection groups for learning professional skills, including iterative improvement, lifelong learning, AI use, and psychological safety (the willingness and strength to take risks such as raising difficult issues or asking for help). The reflection groups are designed to work through peer learning with minimal instructor time, making them scalable across course types and institutions. Reflection group designs will be evaluated as adoptable drop-in assignments, as a standalone one-credit course, and when integrated into a computing ethics course, with iterative refinement informed by co-design workshops with faculty to address adoption challenges. The second task develops a novel video-assisted comparative reflection assignment, where students record programming sessions using GenAI-based tools in different ways (or not at all), then systematically review, compare, and reflect on differences in their process, productivity, and learning across those sessions. The research team will assess these instructional techniques through combining quantitative outcome measures of professional skills and effective GenAI use with qualitative thematic analysis of student reflections, interviews, and programming session videos. All designs, instruments, and open-source curricular materials will be disseminated via faculty development workshops and publications to support broad adoption. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2544192 | Program: 01003031DB NSF RESEARCH & RELATED ACTIVIT,01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Gregory Nelson | Institution: University of Maine, ORONO, ME | Award Amount: $552,200 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2544192 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2544192.html