Data consisting of individual records of male and female animals of purebred Bos Indicus beef cattle (Nellore, Guzerá, Gir and Indubrasil) weighted every three months from birth to 24 months of age available from National Archive of Brazilian Zebu Breeders Association (ABCZ) were used to evaluate two alternatives (co)variance analyses for body weight. In the first analysis the Box-Cox family transformation y ilambda = (ylambda-lambda)/lambda (<FONT FACE=Symbol>l ¹</FONT> 0) or y i = log y i, for lambda=0 was effective in reducing the asymmetry of the coefficients and variance heterogeneity for all weights and breeds. In the second analysis a usual standard mixed model for repeated measures was used to evaluate covariance structures which accounted for the within-individual variances using the goodness of fit criteria provided by PROC MIXED: c² distribution for the null hypothesis, Akaike's Information Criterion and Schwarz's Bayesian Criterion. The best covariance structure for all breeds was the Unstructured followed by the Factor-Analytic for Nellore, Gir and Indubrasil breeds and the Heterogeneous Composed Symmetry for the Guzera breed.
Akaike's Information Criterion; beef cattle; Box-Cox transformation; covariance structures selection; Schwarz's Bayesian Criterion
