Title: What drives students' online self-disclosure behaviour on social media? A hybrid SEM and artificial intelligence approach
Authors: Ibrahim Arpaci
Addresses: Department of Computer Education and Instructional Technology, Faculty of Education, Tokat Gaziosmanpasa University, 60250, Tokat, Turkey
Abstract: This study investigated drivers of the online self-disclosure behaviour on social media by employing a complementary structural equation modelling (SEM) and artificial intelligence approach. The study developed a theoretical model based on the 'theory of planned behaviour' (TPB) and 'communication privacy management' (CPM) theory. The predictive model was validated by employing a multi-analytical approach based on the data obtained from 300 undergraduate students. The model focused on the role of security, privacy, and trust perceptions in predicting the attitudes toward the selfie-posting behaviour. The results suggested that privacy and security are significantly associated with the trust, which explains a significant amount of the variance in the attitudes. Consistently, results of the machine-learning classification algorithms suggested that attributes of the security, privacy, and trust could predict the attitudes with an accuracy of more than 61%% in most cases. Further, mediation analysis results indicated that privacy has no direct effect, but an indirect effect on the attitudes. These findings suggested a trade-off between the privacy concerns and perceived benefits of the actual behaviour.
Keywords: social media; self-disclosure; trust; artificial intelligence; machine learning.
International Journal of Mobile Communications, 2020 Vol.18 No.2, pp.229 - 241
Received: 05 Dec 2017
Accepted: 18 Oct 2018
Published online: 09 Mar 2020 *