Research

Higher education presents one of the most consequential sites of AI integration, but is simultaneously one of the least understood. Students, instructors, researchers, and administrators use and think about AI differently, hindering educators’ ability to obtain a clear picture of unfolding AI adoption. Interdisciplinary research is required to paint this picture across different contexts and populations. My research does exactly that using mixed-methods traditions across education, engineering, sociology, psychology, religious studies, philosophy, science and technology studies (STS), and history. I investigate how students, faculty, institutions, and researchers internationally conceptualize AI, infuse it into their practices, and — in doing so — restructure the educational systems they constitute. Importantly, underscoring all these approaches is a dedication to translate research findings into educational practice, whether faculty development programs, curricular frameworks, academic integrity systems, or student coursework. Application is central.

Collaborations

Methods Contributions

I contribute to discourse about how researchers themselves use AI and what responsible AI-infused methodology looks like.

Dissertation: Accountable by Design

If AI is changing how expertise is developed and trusted, researchers are not exempt; my dissertation studies that change in qualitative research. It argues that accountability for AI-infused qualitative analysis has to be designed into the analysis rather than assumed of the analysts, and develops that claim across three manuscripts: a critical review of how automatic speech recognition deferred and obscured transcription’s interpretive choices, a vocabulary and visual language for refutable reflexivity that makes any AI-infused claim traceable and contestable, and a test of that standard comparing human, prompted-LLM, and decision-model labels on one codebook. Read more about the dissertation →

Other Methods Work

My published article in the International Journal of Qualitative Methods used LLMs to analyze qualitative educational data, interrogating LLM-infused workflows as a methodological intervention compared to human-based analysis. This work asks how LLMs afford and constrain research processes, and what constitutes “meaningful” and “trustworthy” findings when AI is part of the analytic chain.

An in-press book chapter for Sage’s 2027 International Handbook of Engineering Education Research Methods provides a conceptual landscape for researchers navigating these systems, and a companion piece under review at Studies in Engineering Education reframes AI use in research as inherently interpretive work requiring appropriate technical and reflexive practices. The Problem With Prompting (in preparation, Organizational Research Methods) argues for decision models over prompting alone as a route to traceable and justifiable GenAI-infused qualitative data analysis. A further in-progress article analyzes researcher conceptualizations of AI through conference publications, finding an overwhelming treatment of AI systems as instrumental “tools” rather than as transformational sociotechnical systems. See Publications for the full list.

Research Program

My research agenda operates across four organizational levels: students, instructors, institutions, and nationalities.

Students

My master’s thesis in Computer Science has surveyed over 370 STEM graduate students from six R1 U.S. institutions, constituting one of the largest empirical datasets on this population’s AI habits to date. Findings reveal significant polarization in how STEM graduate students relate to AI — with measurable effects on disciplinary understanding, advising relationships, and perceived legitimacy with peers. See Publications for student-focused work from this project, including an in-preparation study on dependency-based AI adoption patterns targeting Computers & Education: AI.

Instructors

Through the NSF CAREER project Minds and Machines (#2339702, PI: Katz), I interview engineering instructors about how they conceptualize GenAI and how those conceptions reach students as course policy. Two instructors in the same department may use, think, and talk about GenAI entirely differently, and that variation is what makes instructors a critically under-examined population in AI-in-education discourse. Work from this strand includes an analysis of how engineering faculty use metaphors to construct AI understanding, a study of when and how faculty permit AI use in their courses, and work with NEASTE on engineering internships examining how students notice differences between how practitioners and instructors orient toward AI.

Institutions

Studying individual students and instructors is necessary but insufficient — institutional structures shape what is possible. I actively collaborate with AI governance leads at Virginia Tech and Cornell University. In March 2026, I initiated a cross-institutional workshop at Virginia Tech bringing together graduate instructors, educational technologists, and academic integrity organizations around effective AI course policy. Earlier work with Cornell’s discipline-based education research community explored how LLMs are reshaping qualitative research practice.

International

AI adoption is not only a national question. Through an ongoing Fulbright Brazil collaboration, I study and support AI governance coordination at transnational scales, working with institutional leaders making decisions that impact real students. I participated in multi-day workshops in São Paulo in September 2025 and Arlington, Virginia in May 2026. See the Fulbright Brazil workshop for details.

Governance & Outreach

My research connects me with central decision-makers overseeing AI adoption in STEM higher education. At the K-12 level, early conversations with the New York State School Boards Association are exploring governance systems that protect students against AI harms while supporting responsible exploration across school districts. My collaborations at Virginia Tech, Cornell, and abroad have sharpened strategies for spanning disciplinary and institutional boundaries.

One such strategy requires “AI translation” to bridge different vocabularies across technical and educational communities. A 2026 ASEE conference paper documents the wide-ranging language engineering instructors use to describe AI: from “algorithms” to “assistants” to “tsunamis.” My prior industry experience as a software engineer at two Fortune 100 companies and my M.S. in Computer Science provide a direct link with practitioners driving AI change and demand.

Browse my Talks and Teaching pages for invited presentations and workshops across these governance and outreach efforts.

Vision

Higher education is where the next generations of AI researchers, engineers, and academics are trained, so getting AI adoption “right” here matters enormously. My work is fundamentally about occupying and fortifying the bridge between communities that design AI systems, those that use them, and those that study these human-AI relationships. I see my research contributing frameworks that develop competency, trust, and expertise around AI practice in STEM higher education, particularly for graduate students, instructors, and researchers navigating this critical transition. Looking ahead, that means building AI systems compatible with how students, researchers, and instructors actually work: transcript engineering as the anchor, alongside an ethnography of software engineers’ AI workflows, design-based research with instructors, and graduate research training.