1.

During development of a multi-residue supercritical fluid extraction method for the extraction of benzimidazole residues from animal liver, Danaher et al.2 investigated the influence on recovery rates of varying several method parameters, each over two levels. This was a factorial study, but here we look at four different variants of the method, A-D, without taking account of the factorial structure. We will return to the data later and take account of the structure. The percentage recovery results for one residue, triclabendazole (TCB) are shown in Table 7.6. What can we learn from the study?

1.

The ANOVA table is given below.

The LSD = 15.27 and HSD = 20.8, so by either criterion the mean of the results for variant C (75.5) is statistically significantly bigger than the next largest mean, which is variant A (52.5). The residual plots are good, so the statistical assumptions underlying the F-test and comparisons of means are likely to be valid.

2.

During development of a multi-residue supercritical fluid extraction method for the extraction of benzimidazole residues from animal liver, Danaher et al.2 investigated the influence on recovery rates of varying several...