Comparison of several criteria to decide on the significance of effects ina robustness test with an asymmetrical factorial design

A robustness test was performed for the reversed-phase HPLC method to deter mine the total content of triadimenol. It is based on an asymmetrical facto rial design that allows examination of simultaneously a factor (column manu facturer) at four levels and 12 factors at two levels. Different error esti mates applied in a t-test statistic to decide on the significance of the fa ctor effects were compared. Dummy factor effects and experiments at the nom inal level of the quantitative factors were used to derive critical effects . Moreover, methods that estimate the critical effects from the distributio n of the non-significant effects were also applied. These statistical metho ds are compared with a graphical evaluation by means of the half-normal plo t. (C)2000 Elsevier Science B.V. All rights reserved.