An investigation on disciriminant analysis

By obtaining sample data about a research topic on the basis of statistics, generalization is made according to the results of the analysis by applying a number of statistical analyzes developed on these data. In this way, it can predict the whole universe of the subject covered by the study on the selected sample. However, it is seen that unvariate analyzes are not sufficient in the analysis of the subject studied in scientific research. Because of this limitation, multivariate statistical analysis methods have been developed. One of these methods is discriminant analysis. In the first part of this study, multivariate analysis methods are introduced. The various assumptions underlying these methods are explained. In the second part; discriminant analysis, definition, aim, basis of assumptions, discriminant analysis in case of two or more groups, the acquisition and meaning of separation functions and theoretical background of discriminant analysis are examined. In the third part, discriminant analysis was applied by using SPSS 22 package program to divide countries into very high, middle and low human developed countries by using the 2018 Human Development Index data published by United Nations Development Program. A selected country was tested for correct classification by means of the functions obtained after analysis.