IDWSDS 2025 - S121: Methods with Meaning: Celebrating The "Why" Behind Statistical Innovation

Описание: Statistical innovation is often judged by novelty — but what if we instead celebrated our methods' meaning? This session highlights the real-world motivations, challenges, and impact that drive methodological development. Rather than diving deep into equations, in this session four statisticians share the stories behind their work: why the methods were needed; what gap they fill; and how they contribute to science, policy, or public health. This session is for anyone who believes that statistical methods are not just intellectual exercises, but tools to understand and improve the world. Talks will explore bias in risk prediction, the responsible use of predictions in research, scalable validation of electronic health records, and optimizing cancer screening programs. Each project is rooted in a real-world setting-like healthcare, clinical research, or data science practice-where thoughtful methods can lead to meaningful improvements.

ORGANIZER AND CHAIR: Sarah Lotspeich, Wake Forest University
TALK TITLES AND SPEAKERS:
The Role of Congeniality in Multiple Imputation for Doubly Robust Causal Estimation, Lucy D'Agostino McGowan, Wake Forest University
Predictors of Physical Resilience in Older Adults Undergoing Total Knee Arthroplasty: Biomarker Analysis From the PRIME-KNEE Study, Marissa Ashner, Duke University
Prediction Algorithms and Persistence: The Long Road to a Simple Solution,
Rebecca Hubbard, Brown University
Do Clinical Prediction Models Perpetuate Health Disparities? Assessment of Racial and Ethnic Disparities in Suicide Prediction Models, Yates Coley, Kaiser Permanente Washington Health Research Institute