Want to give your statistical test a boost?
One effective strategy is to reduce its dependence on ancillary statistics-it's like clearing the noise
to sharpen the signal.
We develop a decision-theoretic framework that improves test statistics by leveraging the principle of independence from ancillary structures.
Preprint: https://arxiv.org/abs/2507.11816
I would greatly appreciate any critical feedback, suggestions, or ideas for possible applications.
Thank you,
Albert Vexler
Professor, Department of Biostatistics
University at Buffalo
avexler@buffalo.edu
------------------------------
Albert Vexler
Professor
State University of New York
------------------------------