When does digital engagement backfire? Exploring the nonlinear relationship between digital citizenship and information literacy

Purpose This study examines the curvilinear relationship between Digital Citizenship and Digital Information Literacy in detecting deceptive advertisements. It further investigates how Online media exposure and online advertisement exposure moderate this relationship, addressing gaps in prior research that assume a strict linear association between digital engagement and literacy. Design/methodology/approach A survey was conducted with 2,072 participants to measure their digital citizenship and digital information literacy, measured by the correct identification of deceptive advertisements. An initial scatter plot analysis of the data revealed a non-linear relationship, which mandates applying non-linear regression analysis to test quadratic and cubic models. Moderator analysis was also performed to assess the influence of online media and advertisement exposure on the relationship between digital citizenship and digital information literacy. Findings The comparison of regression models confirms a curvilinear rather than linear relationship between digital citizenship and digital information literacy. The quadratic model provides the best fit, indicating that digital literacy increases digital information literacy, but this effect declines once it reaches a saturation point. Furthermore, online advertisement exposure significantly moderates this relationship, with high ad exposure weakening the positive effect of digital citizenship on digital information literacy. Research limitations/implications This study refines digital literacy theory by challenging the assumption of a linear relationship between digital engagement and literacy. It underscores the importance of moderating factors, such as online media and advertisement exposure, in shaping literacy outcomes and informs new models of digital skill acquisition that account for nonlinear information processing patterns. Practical implications Policymakers and educators should implement adaptive digital literacy programs that account for cognitive overload and overconfidence effects. AI-driven personalized learning tools, VR-based media literacy training and interactive misinformation detection systems can help balance digital engagement without inducing cognitive strain. Originality/value This study challenges traditional linear models of digital literacy development by introducing a curvilinear framework. It demonstrates that excessive digital engagement can reduce digital literacy effectiveness. Unlike previous studies, it identifies threshold effects in digital literacy growth and highlights how online exposure influences this relationship, offering a novel contribution to digital literacy research.