Analisis Faktor Perkembangan Interaksi Anak terhadap Penggunaan Gawai Pada Usia Dini Menggunakan Kombinasi Regresi Logistik Biner dan Bayesian Network

Abstract

The rapid advancement of digital technology has had a significant impact on the way young children interact and socialize. If not managed properly, the use of digital devices (gadgets) has the potential to hinder children’s communication skills and social adaptation to their environment. This research was conducted with the primary objective of identifying and analyzing the determinants of children’s social interaction skills related to gadget use. The method used was a combination of Binary Logistic Regression analysis and Bayesian Network analysis to obtain comprehensive results. This field research took place at MIS Al-Hasbi, Nogo Rejo Village, Galang District. The sample population consisted of 112 first- and second-grade students, determined through a Stratified Random Sampling procedure. Data collection was conducted by distributing questionnaires covering ten measurement dimensions. These dimensions include: the length of time children spend using gadgets, the variety of content accessed, the presence of parental supervision, restriction strategies, the content selection process, and the children’s speech level. Validity test results showed that 9 out of 10 items were valid, and all variables were reliable with a Cronbach’s value > 0.60. Logistic regression analysis showed that the independent variables simultaneously had a significant effect on children’s interaction skills, while the partial effect was insignificant. Bayesian network analysis revealed a complex cause-and-effect relationship, where educational content served as a connecting node between content type, selection, and restrictions, which influenced speaking responses. The results of the study emphasized the importance of mentoring and selecting educational content to support social interactions in early childhood.