Using Logistic Regression Model to predict the Functions with Economic categorical Dependent variables.


Abstract This research includes study the importance of using logistic regression model to predict the functions with Economic categorical Dependent variables, to get rid of the statistical and conceptual problems Facing the use of the least squares method, and for that we select 25 developing countries as a sample for research, depending on the World Bank and UN data, with the help of statistical package SPSS17. The importance of research is trying to explain that analysis tools, of a limited use now, to enrich the quantitative analytical tools are currently known in the Arab economic literature, especially after the growing need to adopt a qualitative variables as dependent variables, which requires to use appropriate analytical tools. The most important conclusion of the research is that the use of binary logistic regression model with binary dependent variables are eliminated from the three Econometrics problems, because it is passing the tests of Autocorrelation , Multicollinearity and Heteroscedasticity.