“…For instance, no indication of the normality testing prior to a multi-way analysis of variance for blood analytes was reported in mink [ 38 , 39 , 40 , 41 ], dogs [ 42 ], cats [ 43 ], and chickens [ 44 ]. Blood analytes have also been compared between two [ 33 , 45 , 46 , 47 , 48 ] or more groups [ 49 , 50 , 51 , 52 , 53 , 54 , 55 ] with a t-test without prior normality testing. The logarithmic transformation has normalizing and variance equalizing effects on positively skewed distributions [ 36 , 56 , 57 , 58 ], and has frequently been applied to some blood analytes to achieve normality [ 42 , 59 ] or homogeneity of variances [ 45 , 60 ].…”