SPECIFICS OF THE APPLICATIONS OF MULTIPLE REGRESSION MODEL IN THE ANALYSES OF THE EFFECTS OF GLOBAL FINANCIAL CRISES

• Main Authors: Željko V. Račić, Biljana T. Baraković
• Format: Article
• Language: English
• Published: Croatian Operational Research Society 2010-12-01
• Series: Croatian Operational Research Review
• Subjects: Multiple linear regression model; statistical tests; coefficient of determination; F schedule
• Online Access: http://hrcak.srce.hr/index.php?show=clanak&id_clanak_jezik=139837

This paper aims to present the specifics of the application of multiple linear regression model. The economic (financial) crisis is analyzed in terms of gross domestic product which is in a function of the foreign trade balance (on one hand) and the credit cards, i.e. indebtedness of the population on this basis (on the other hand), in the USA (from 1999. to 2008). We used the extended application model which shows how the analyst should run the whole development process of regression model. This process began with simple statistical features and the application of regression procedures, and ended with residual analysis, intended for the study of compatibility of data and model settings. This paper also analyzes the values of some standard statistics used in the selection of appropriate regression model. Testing of the model is carried out with the use of the Statistics PASW 17 program.