A statistical procedure for the estimation of accuracy parameters in interlaboratory studies

Area of Science:

• Analytical Chemistry
• Metrology
• Statistical Modeling

Background:

• Interlaboratory studies assess laboratory measurement accuracy.
• Repeatability and reproducibility are key accuracy parameters.
• Variance components models are standard for estimating these parameters.

Purpose of the Study:

• To introduce a novel residual analysis for variance components models.
• To detect violations of model assumptions in interlaboratory studies.
• To provide a robust alternative for accuracy parameter estimation when assumptions are violated.

Main Methods:

• Developed and applied a new residual analysis technique for variance components models.
• Utilized graphical methods for residual evaluation.
• Compared estimates from standard variance components models with a robust method.

Main Results:

• The residual analysis effectively detects violations of variance components model assumptions.
• A robust estimation method provides a reliable alternative when assumptions are unmet.
• Graphical evaluation of residuals is crucial for identifying model issues.

Conclusions:

• Residual analysis is essential for reliable accuracy parameter estimation in interlaboratory studies.
• The proposed method enhances the validity of repeatability and reproducibility estimates.
• Awareness of potential hazards in residual analysis is necessary for accurate interpretation.