Efektivitas Pendekatan Deep Learning Terhadap Kontrol Diri Remaja Dalam Menggunakan Internet

Abstract

This study aimed to test the effectiveness of the deep learning approach in improving self-control internet usage of teenagers. This research was included in pre- experimental research with one group pre-test post-test design research. The research was conducted in SMP IT Masjid Syuhada Yogyakarta and the subjects were 20 students who lacked self-control. The data were obtained from the result of the scale used to measure the self control of internet usage. The instrument validity was determined by expert judgment. The instrument reliability was determined by Alpha Cronbach's results. The data analysis technique used t-test. The results of data analysis showed a significant change in 20 students of SMP IT Masjid Syuhada. The result of pre-test was 142.1, and the post-test was 159.6. The t-test was 9.447. There is a difference if p sig <0.05 and t arithmetic> t table (N 20 = 2.093). Thus, deep learning approach effectively improved self control when students of SMP IT Masjid Syuhada Yogyakarta used internet.