Steven Walfish

Well-designed experiments can reduce the risk of coming to an incorrect conclusion during a process characterization, assay validation, or process validation study.

Statistical methods for calculating confidence intervals, tolerance intervals, and capability analysis to reduce out-of-specification situations.

How to use hypothesis correctly, and understanding the difference between one-sample, two-sample, and z-test.

The key to a good graphical presentation is to select the method that best fits the data.

Vagueness in the ICH Q2A and Q2B guidelines necessitates effective protocol design and data analysis. For specificity (detection in the presence of interfering substances), the goal is statistical differences with meaningful implications on assay performance. Linearity (results directly proportional to concentration of analyte in the sample) is typically demonstrated via least squares regression. Accuracy (difference between measured and true values) usually is presented as a percent of nominal. Precision analysis is vital because it supports claims of accuracy and linearity. A well-designed experiment and statistically relevant methods will facilitate method validation in accordance with ICH guidelines.

Statisticians partner with technical experts to design statistically valid studies to construct the appropriate analysis.