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

Composite indices data are high-dimensional; a composite index is a single number for each observation and no matter how informative it is, it cannot capture all the features of such highly-dimensional data. One purpose of this thesis is to look at these data from various angles using different techniques in an attempt to discover and extract the wealth of information that these data contain. We particularly focus our attention on two challenging issues:The high dimensionality of the data and indicator weights. We look at the data from a multivariate point of view and use existing techniques such as the Principal Components Analysis, Multidimensional Scaling to develop the following new indices from the same data:1.The Equal Weights Index (EWI) 2.Construction of three indices based on Principal Components Analysis 3. Construction of two indices based on The Multidimensional Scaling These indices can be used to complement the existing indices. We illustrate these methods using two data sets: The Ibrahim Index of African Governance (IIAG) data and Egypt{u2019}s Economic and Social Rights Fulfillment Index (ESRFI). The R programming Language for the calculations and graphs is used

Issued also as CD