Tuesday, Aug 4: 12:30 PM - 1:50 PM
Roundtables – Lunch
Thomas M. Menino Convention & Exhibition Center
Room: CC-Ballroom Foyer
Presentations
The boom in accessibility of artificial intelligence and machine learning software over the last decade has generated many new challenges for collaborative biostatisticians. One such challenge is the incorporation of proprietary algorithms (often industry-sponsored) into clinical risk prediction analyses. While a proprietary approach to data analysis methods is in direct conflict with the transparency and reproducibility principles that statisticians often advocate for, modern biostatisticians consulting in clinical settings must accept that proprietary models are likely to only increase in popularity and be prepared to assess a project proposal with a proprietary component. In this roundtable, we will discuss pros and cons of proprietary methods, current academic and regulatory guidance, recommended reporting guidelines, and potential quality control actions. Participants will leave with resources outlining key considerations for determining scope of biostatistician responsibility on such collaborative projects.
Keywords
Clinical Risk Prediction
Statistical Consulting
Regression Modeling
Machine Learning
Artificial Intelligence
Transparent Reporting
Speaker
Natalie DelRocco, University of Southern California, Children's Oncology Group
Co-Author
Ying Lu, Stanford University School of Medicine
For statisticians across industry, academia, and government, the key question is no longer whether GenAI can be useful, but how it can be integrated responsibly, productively, and sustainably into everyday work, and where it falls short. As practices continue to emerge organically in this rapidly evolving medium for interacting with information, this roundtable is designed to foster collective learning through peer exchange, grounded in real-world trial-and-error experiences with GenAI tools.
Participants will share 1-2 short, concrete examples of how GenAI tools have changed their professional practice. These examples may span research, learning, collaboration, documentation, analysis, communication, or decision-making. What worked, what didn't, what saved time, what created new risks, and what required rethinking long-established habits. The emphasis will be on generalizable lessons that translate across roles, sectors, and levels of technical expertise.
The conversation will surface a sharable set of practical tips, workflows, and cautionary lessons that capture shared insights and support continued learning for a broader statistical community.
Keywords
GenAI
Responsible AI Use
Productivity and Workflow Efficiency
Peer Learning
Statistical Practice
The rapid adoption of AI tools, such as large language models and automated analytics platforms, is transforming collaborative statistical practice. While AI streamlines coding, exploratory analysis, and documentation, it does not replace statistical judgment, problem formulation, or professional accountability. Instead, AI has shifted the distribution of effort, expertise, and risk within collaborative projects. This roundtable will explore how collaborative statistical work is evolving with AI, drawing on participants' experiences in consulting, team science, and applied research. Discussion topics include changes in workflows, collaborator expectations, and project scoping; tasks that have become faster or more complex; and methods for validating, documenting, and taking responsibility for AI-assisted analyses. The session will highlight effective AI–statistician partnerships and cautionary examples where AI can mislead without proper oversight. Participants will also address managing AI-generated results from collaborators, communicating uncertainty and rigor, and defining successful statistical collaboration as AI becomes more integrated into practice.
Keywords
Collaborative statistical practice
Statistical consulting
Professional accountability
Statistical judgment
Artificial intelligence
Speaker
Li Zhang, University of California
Co-Author(s)
Margaret Stedman, Stanford University
Emily Griffith, NC State University
Heidi Spratt, University of Texas Medical Branch
Xiaoming Sheng, University of Utah
Charlotte Bolch Walsh, Midwestern University
Robert Podolsky, Medstar Health Research Institute
Eric Vance, LISA, University of Colorado-Boulder
Mario Davidson, Vanderbilt University Medical Center
David Agboola, Procter & Gamble
Study design choices made early in a project have lasting consequences for methodological rigor, feasibility, and appropriate and interpretable inference. This roundtable will explore how statisticians collaborate with investigators to identify optimal designs for health care research and navigate challenges that arise during implementation. Participants will share experiences and strategies for guiding investigators toward designs that are both rigorous and realistic.
Keywords
Study design
Health-care research
statistical consulting