Method validation is the process that provides evidence that a given analytical method, when correctly applied produces results that are fit for purpose. No matter how well a method performs elsewhere, analysts need to confirm that the method is valid when applied in their laboratory. There is now much greater emphasis on method validation in the ISO 17025 accreditation standards

The performance of analytical method is primarily measured around precision and accuracy. Laboratories are required to validate the methods they use to ensure that the quality of their tests is accurate and precise. This entails proper and effective management of testing processes meant to authenticate results that come out of such tests

Introduction to basic statistics for test laboratories

Errors vs. uncertainties

Accuracies and precision

Distributions important to test/analytical labs

Samples vs. populations

Confidence limits/intervals, level of confidence

Correlations and regression in the analytical test laboratory calibration

Standards

Number of Standards

Inclusion of the blank

Range of Standards

Linearity

Correlation coefficient

Significant linearity – ANOVA

Regression Parameters

Slope, intercept, regression line

Calibration uncertainties

Method sensitivity

Limit of detection – LOD

Limit of quantization – LOQ

Specificity vs. selectivity

Determination

Repeatability uncertainties

Regression uncertainty

Minimization of the regression uncertainty

How to handle dilutions

Method of standard additions

Using Regression to compare the results of two methods/analysts/instruments

Analysis of Variance – ANOVA

Single factor, Two factor – ANOVA

Application in providing the competency of analysts, instruments and method

Control charts

Stewart charts
Range charts
Cusum Charts

Evaluation of the uncertainty of measurement

Identification of uncertainties. Quantifying, combining and extending uncertainties

Typical uncertainties: Mass,  Volume, Purity uncertainties

Repeatability uncertainties

Method validation

Setting validation criteria
Parameters to be validated