Checking the Ruggedness of Analytical Methods using Plackett-Burman Design

Abstract

When analyzing something according to a previously confirmed protocol, it is inevitable that the levels of the various operating factors may change either intentionally or accidentally. Some aspects of an analytical method are robust and changing the factors has no significant consequences, while others are sensitive and subtle changes may have a significant influence on the final result. Operators should be aware of which factors are important with respect to the method’s robustness. It is time-consuming and inefficient to evaluate the effects of the various factors which may influence analytical results using a one-at-a-time procedure. The process could be improved efficiently, however, by using a class of experimental design called Plackett-Burman design, in which only n measurements are needed to evaluate up to n-1 factors. The results of a fractional factorial design are analyzed graphically and/or statistically to examine the effects of each factor on the analytical response. The results could also be used to quantitatively calculate the non-significant intervals for significant factors. In addition, the uncertainties in the various steps of the methods are revealed by a ruggedness test.