Rao's simple covariance structure in growth curve models and outliers

The aim of this study is to examine Rao 's simple covariance structure in Growth Curve Models and to use the technique of outlier identification. Firstly, by using Growth Curve Models it is showed that under which conditionals Least Square Estimation and Maximum Likelihood Estimation give good results. Secondly, Rao 's simple covariance structure and many useful covariance structures which are included in the simple covariance structure as special cases arc discussed. Finally, in Growth Curve Models with simple covariance structure solution of the problem of multiple discordant outlier identification is discussed. Having given the theoretical structure, an extensive application is realized. In the application part, both girl and boy baby 's growth process is observed for each three data sets of Gazi University Medical Faculty Department Child Healty Diseases. These data sets are considered for corresponding to Growth Curve Models. Thus Outliers are identified. Science Code : 0303 Key Words : Growth Curve Models, Rao' s Simple Covariance Structure, Discordant Outliers Page Number: 110 Adviser : Prof. Dr. Müslim Ekni