🤖 AI Summary
Deming regression suffers from bias under heteroscedasticity—non-constant error variances—common in wide-range measurement data. To address this, we propose a weighted errors-in-variables (EIV) regression method. Our key contribution is the first systematic integration of a generalized precision profile model into the Deming regression framework, enabling concentration-dependent adaptive weighting that accurately captures heteroscedastic error structure. The method unifies EIV modeling, precision profile estimation, and weighted least squares, implemented via a standardized R package. Empirical evaluation demonstrates substantial improvements in accuracy and robustness for method comparison, particularly in high-stakes clinical assay validation where analytical consistency is critical. By explicitly modeling heteroscedasticity through measurement-concentration–dependent weights, our approach provides a statistically principled, broadly applicable solution for measurement verification under non-constant error variance.
📝 Abstract
Errors in variables (Deming) regression of measurements spanning a wide range of values requires appropriate weighting to reflect nonconstant variance. Precision profile models, mathematical relationships between measurement variance and mean, are a route to these weights. The paper describes a methodology combining general precision profile models with Deming regression and described R routines for the resulting calculations.