SPECIFY THE BEST COVARIANCE STRUCTURE FOR REPEATED MEASUREMENTS DATA WITH/WITHOUT MISSING OBSERVATIONS USING MIXED MODEL

Repeated measures ANOVA is a technique used to test the equality of means. It is performed when all the members of a random sample are tested under a number of many conditions. Repeated measures data needed special methods of statistical analysis as several types of covariance structure could be applied. Each of the regression and ANOVA methods could produce invalid results because their assumptions do not consistent with repeated measures data. There are several statistical methods used for analyzing repeated measures data such as separate analysis, univariate, multivariate and mixed model methodology. Recently, the mixed model methodology was used to analyze repeated measures data by many researches because the application of this methodology is available in many computer programs. As the growth traits represent a good example of repeated measures. This methodology was performed on growth traits of 102 Awassi lambs bred on Research station of sheep and goats in Abo -Gharib west of Baghdad to evaluate several covariance structures with/without missing data that describe the body weight (repeated measures) from birth to eight months. Results revealed that the UN covariance structure is the best in complete and missing observations data with /without covariate according to goodness of fit criterion of -2 Res Log Likelihood, AIC and AICC, whereas the TOEPH covariance structure is the best for all types of data according to BIC. In conclusion: Applying mixed model methodologies confirmed its ability to deal with various covariance structures in the repeated measures data to identify the best covariance structure.