Application of Data Mining Algorithms and Statistical Hypothesis Testing to Analyze Problematic Internet Use among University Students

This study aims to understand university students' internet usage habits and whether these habits are associated with certain sociodemographic and behavioral variables. Data were obtained from the General Problematic Internet Use Scale 2 (GPIUS2) survey, administered to undergraduate students from eight programs at a state university in Türkiye, and analyzed using the Apriori Association Rule Mining Algorithm and statistical hypothesis testing. The responses from 217 students were analyzed to gain insights into their problematic internet use habits. Analysis conducted using WEKA software identified significant associations with the "Mood Regulation" dimension, particularly among male students and those with high GPAs. Additionally, communication-related smartphone usage was found to be associated with the "Negative Outcomes" dimension. The research findings also reveal that university students' problematic internet use is statistically associated with the program they are enrolled in, marital status, self-reported daily smartphone usage, primary reason for smartphone use, and self-reported addiction. This study contributes a new perspective to the application of data mining techniques in social sciences and educational research, providing valuable insights into the relationships between internet usage habits and problematic usage, thus laying a foundation for future research.