The distinct roles of dedication-based and constraint-based mechanisms in social networking sites

Purpose - The purpose of this paper is to investigate the effects of dedication-and constraint-based mechanisms on users' post-adoption behavior in the social networking site (SNS) context. Design/methodology/approach - The proposed framework uses user satisfaction and trust belief to capture the dedication-based mechanism and perceived switching costs and social norms to capture the constraint-based mechanism. Hypotheses were tested by applying partial least squares to data from 250 experienced Facebook users. A structural equation modeling was used to test the validity of the proposed research models. Findings - The analysis results show that SNS users' continuance intention is jointly affected by two distinctive mechanisms: a dedication-based one and a constraint-based one, the former playing a more critical role. The findings indicate that both perceived relative benefits and perceived enjoyment significantly influence user satisfaction. Learning and network size were found to be the key predictors of perceived switching costs. Research limitations/implications - This study applies the dedication-and constraint-based models by incorporating numerous sets of antecedents. The framework provides a theoretical lens of how two distinctive mechanisms influence SNS users' post-adoption behaviors. Practical implications - The analysis results provide several insights that can aid SNS providers understand SNS users post-adoption behaviors. Moreover, the findings will help SNS providers effectively facilitate dedication-and constraint-based mechanisms by enhancing the key antecedents of two distinctive mechanisms. Originality/value - SNSs have become an important component of individuals lives. However, few systematic works investigate the fundamental mechanisms leading to SNS users' continued usage. In an attempt to extend the horizons of SNS research, this study incorporates a set of antecedents to dedication-and constraint-based models.