Statistical Challenges in Nursing Education and Research in the Era of AI: Revisiting the 2014 Panel

Statistics remains pervasive in nursing education and research and continues to be recognized as essential to the preparation of nurse researchers and nurse scholars. In 2014 a group of statisticians with faculty appointments in US based academic nursing programs identified five foundational domains that described major challenges at the intersection of statistics and the nursing discipline. These domains included perceptions and viewpoints about statistics, the roles and responsibilities of statisticians in academic nursing settings, interdisciplinary collaboration between statisticians and nurse investigators, statistics education for nursing students, and the use of statistics in the nursing literature. Although the broader environment in which nursing education and research take place has changed considerably over the past decade, many of the concerns described in the original panel discussion remain very much present today. In several areas the challenges have become more nuanced as expectations for methodological rigor, transparency, and reproducibility have increased while persistent difficulties with foundational statistical reasoning continue to affect both students and investigators. The purpose of this panel is to revisit the five domains identified in 2014 and to consider how the issues first articulated more than a decade ago persist in current nursing education and research settings. Drawing on their experience as educators and collaborators, panelists will describe ongoing misconceptions about statistical reasoning, the evolving expectations for statisticians working within interdisciplinary teams, and the expanding range of statistical material now included in graduate curricula. The panel will also address continued concerns in the nursing literature, where questions of appropriateness, reporting quality, and reproducibility of statistical analyses remain important. Discussion will include the tension that arises when newer analytic tools and approaches are adopted without a corresponding emphasis on the underlying principles that give statistical results meaning. The goal is to provide an updated view of the five foundational domains, highlight recent evidence related to statistical practice in nursing, and offer practical recommendations for strengthening training, collaboration, and reporting across academic nursing programs. Panelists include experienced statisticians with faculty appointments or affiliations in a academic nursing: Matthew J. Hayat (Medical University of South Carolina), Todd A. Schwartz (University of North Carolina at Chapel Hill), Sarah Schmiege (University of Colorado Anschutz), Vishal Thakkar (UT Southwestern), and Michael Schoeny (Rush University). Each panelist will reflect on one or more of the original domains and discuss how the issues identified in 2014 have evolved in their own work. A moderated discussion will follow, focusing on implications for nurse training, interdisciplinary research teams, and the development of reproducible and methodologically sound statistical practice. The session aims to extend the contribution of the original panel by situating the five domains in the contemporary environment and identifying opportunities for progress in the years ahead.