Modeling Education and Internet Usage: PCA and Linear Regression Approaches

Abstract

The frequency of Internet use is significantly influenced by the formal and non-formal education of individuals. Acquiring education enables the acquisition of basic skills in the use of digital technology and influences the reduction of the digital gap created in this way. In this research, the influence of a number of education parameters on the rate of Internet adoption in 27 countries of the European Union was modeled. The principal component analysis (PCA) technique was used to identify the parameters that contribute the most to the development of education through the formation of two factors. The statistically significant score of factor one was used to form a linear regression (LR) model. The results show that the percentage of explanation of the variability of internet users is explained with 54.5% through the value of factor one. The correlation coefficient between real and projected values (0.738) shows a strong positive correlation between these values.