The LORELIA Residual Test

More about the book

Focused on method comparison studies, this work introduces a novel outlier test that addresses the challenges posed by non-homogeneous sample distributions and variances. The LORELIA (LOcal RELIAbility) residual test utilizes robust linear regression to estimate outlier limits directly from the data, avoiding assumptions about error variance models. This approach enhances the identification of outliers, critical for ensuring measurement accuracy. The effectiveness of the test is illustrated through various examples and simulations, providing a comprehensive solution for data analysis in this field.