Uniformly most powerful tests in linear models
Abstract
In the multiple regression model we prove that the coefficient t-test for a variable of interest is uniformly most powerful unbiased, with the other parameters considered nuisance. The proof is based on the theory of tests with Neyman-structure and does not assume unbiasedness or linearity of the test statistic. We further show that the Gram-Schmidt decomposition of the design matrix leads to a family of regression model with potentially more powerful tests for the corresponding transformed regressors. Finally, we discuss interpretation and performance criteria for the Gram-Schmidt regression compared to standard multiple regression, and show how the power differential has major implications for study design.
Keywords
Cite
@article{arxiv.2411.18033,
title = {Uniformly most powerful tests in linear models},
author = {Razvan G. Romanescu},
journal= {arXiv preprint arXiv:2411.18033},
year = {2025}
}
Comments
Major changes from the previous version include: - Added a version of theorem 1 for the case of correlated predictors - Added a section on multicollinearity and model interpretation - Extended the power result to include equivalent sample sizes - New data example