Helping Analytical Scientists Apply Statistics

AbstractThe development and operating characterization of an analytical method or assay requires extensive collaboration between analytical scientists and statisticians. Minimal performance requirements, validation reporting and user expectations of analytical methods rely heavily on statistical concepts, calculations and terminology, This paper provides a descriptive and illustrative recitation of commonly used, and often misunderstood, statistical principles. The goal is to provide the analytical scientist with clearer understanding and practical guidance in the use of statistics. Percent relative standard deviation was used for a measure of precision through the sections dealing with nested experiments and variance components estimation, calibration experiments and sample size estimation. A comprehensive treatment of limit of detection and quantitation was given and the case when the calibration experiment has a nested error structure was also discussed.