Discriminant Analysis of Social Media Usage Among Students of UIN Sumatera Utara

Abstract

This study aims to identify significant differences between students who actively use social media and those who do not in the Faculty of Science and Technology at UIN Sumatera Utara, determine distinguishing variables between the two groups, develop a two-factor discrimination model, and test the model's accuracy. The study sampled 1,800 students from 220 study programs, with data collected through a questionnaire. The credibility and validity of the questionnaire were tested using SPSS with Wilk's Lambda and F-test. Discriminant analysis was employed to identify distinguishing variables, such as access, photo uploads, status updates, link sharing, and live stories, and to determine the membership proportions of each group. The results indicate significant differences in behavior between active and inactive social media users, with cross-validation showing a classification accuracy of 77.0%. These findings provide valuable insights into social media usage patterns among university students.