Examine The Internet Addiction Scale Of Students In Turkey And Iraq Comparatively With The Multivariate Adaptive Regression Splines (MARS) Method
##plugins.themes.bootstrap3.article.main##
Abstract
In recent years, the Internet has altered how modern life is lived. Internet addiction is considered a current condition that could affect mental health. The aim of the study is to develop a scale for internet addiction and examine the factors that can affect the addiction status of individuals. The literature uses the novel non-parametric method known as MARS. MARS has the ability to provide accurate parametric forecasts as well as readable model curves. The sample of the study is at the international level, and cosmopolitan universities from each country (Turkey and Iraq), and the target audience was determined as university students.
The data for the study came from a questionnaire designated, which was meant to evaluate a person's level of addiction. The Internet addiction scale is a 5-point Likert scale, with the lowest score that each participant can get from the scale is 35 and the highest score being 175. The population of the study consists of 2235; 1220 students (427 males and 793 females) from the Turkish sample and 1015 students (461 males and 550 females) from the Iraqi sample were assigned by random sampling method. For MARS data mining analysis, the default values of the SPM 8.2 program were taken as a basis, and operations were carried out considering the entire data set. MARS obtained nine base functions of the model for Turkey and twenty base functions of the model for Iraq. As a result of the MARS analysis method, it was proven that the predictors of Daily Internet Usage may be seen as the primary initiating factor for Internet dependence for Turkish and Iraqi students following the research model identified by MARS for both datasets. These results have been supported in the literature, which eventually resulted in the conclusion that using the Internet leads to Internet dependence.