Primena metoda zavisnosti u analizi razlika u zaradama između polova u Republici Srbiji

Abstract

Dependence methods are the special group within the method of multivariate analysis. The application of these methods comprises the procedure data analysis, in which one or more variables explicitly are considered as dependent, while all other variables are treated as independent variables. In accordance with objectives of the doctoral dissertation, the research provides a brief overview of methods of addiction, and their basic assumptions and models. Afterwards, chosen dependence methods, primarily multivariate analysis of variance and discriminant analysis, are applied to particular problems of earnings differences between genders in the Republic of Serbia, and in order to identify the intensity of factors which cause the observed differences. The empirical verification of hypotheses, within the scope of doctoral thesis, conducted a statistical processing and analysis of secondary data, respectively data collected by the Labor Force Survey (LFS) in 2016, obtained from the Republic Institute for Statistics of the Republic of Serbia. Questionnaire of the LFS is applied to the sample, which comprises 133,704 individuals, all age structure, while the realized sample makes 16,947 individuals, because the analysis included only data relating to the working age population and the population that gave the answer on the amount of net income that is generated by in the previous month. Results of verification defined hypothesis confirmed the existence of significant statistical differences between earnings of men and women, both in the Republic of Serbia and in the Member States of the European Union. Also, statistically significant differences in earnings between the gender, there are if one takes into account the ownership, autonomy of the operation and contractual relationship of respondents. Empirical schedule of employed women by level of qualification does not deviate from the expected schedule. Classification subjects into two groups, it was found that men are more often employed in technologically advanced industries and economic activities with higher earnings, while women while women are employed in less technologically advanced industries and in economic activities that are characterized by lower earnings.