Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Statistics and Econometrics

In the framework of the general linear model, residuals are routinely used to check model assumptions, such as homoscedasticity, normality, and linearity of effects. Residuals can also be employed to detect possible outliers. Various types of residuals may be defined for linear models. Similar uses may be envisaged for three types of residuals that emerge from the fitting of linear models: Marginal Residuals, Conditional Residuals and Empirical Best Linear Unbiased Predictor (EBLUP). The main concern of the thesis is to provide an overview of approaches for residual estimate, review some of the residual analysis techniques that have been used in this context and propose a standardization of the conditional residual useful to identify outlying observations and clusters. As well as using analysis for linear illustrating this by real data set taken from the literature. To conclude areal data set is given to exercises the main these methods: Marginal Residuals, Conditional Residuals and Empirical Best Linear Unbiased Predictor (EBLUP) comparing the results. Package (Stata, ver 13) is used. We can work in the future, the application of methods of residual analysis in the case that a contrast matrix structure not be full rank. The linear n observations to least-/unconfounded residuals (r = rank [XZ) not be reduces the number of observations

Issued also as CD